The Interactive Fly
Genes involved in tissue and organ development
The insect brain is traditionally subdivided into the trito-, deuto- and protocerebrum. However, both the neuromeric status and the course of the borders between these regions are unclear. The Drosophila embryonic brain develops from the procephalic neurogenic region of the ectoderm, which gives rise to a bilaterally symmetrical array of about 100 neuronal precursor cells, called neuroblasts. Based on a detailed description of the spatiotemporal development of the entire population of embryonic brain neuroblasts, a comprehensive analysis was carried out of the expression of segment polarity genes (engrailed, wingless, hedgehog, gooseberry distal, mirror) and DV patterning genes (muscle segment homeobox, intermediate neuroblast defective, ventral nervous system defective) in the procephalic neuroectoderm and the neuroblast layer (until stage 11, when all neuroblasts are formed). The data provide new insight into the segmental organization of the procephalic neuroectodem and evolving brain. The expression patterns allow the drawing of clear demarcations between trito-, deuto- and protocerebrum at the level of identified neuroblasts. Furthermore, evidence is provided indicating that the protocerebrum (most anterior part of the brain) is composed of two neuromeres that belong to the ocular and labral segment, respectively. These protocerebral neuromeres are much more derived compared with the trito- and deuto-cerebrum. The labral neuromere is confined to the posterior segmental compartment. Finally, similarities in the expression of DV patterning genes between the Drosophila and vertebrate brains are discussed (Urbach, 2003a).
In the trunk neuroectoderm, segment-polarity genes are expressed in stereotypical segmental stripes, and in NBs that delaminate from these domains, subdividing each neuromere along the AP axis. In the pregnathal head region the expression domains of segment polarity genes are less obvious, but analysis of engrailed and wingless expression in the head peripheral ectoderm, and of PNS mutant phenotypes, support the existence of four pregnathal segments in Drosophila: the intercalary, antennal, ocular and labral segments (from posterior to anterior). However, the identity and organization of brain structures deriving from these segments is still obscure. In order to obtain evidence concerning the number and extent of the brain neuromeres, and to map the position of their boundaries, the expression of segment polarity genes, including wingless, hedgehog, gooseberry-distal, engrailed, invected and mirror, was analyzed. The spatiotemporal pattern of their expression was traced in the neuroectoderm and in the NB-layer until stage 11, when all brain NBs are formed. The data show that segmental expression is retained for most of the investigated segment polarity genes in both the developing head ectoderm (procephalon) and brain NBs, providing landmarks for the definition of segmental domains within the developing brain NB pattern (Urbach, 2003a).
engrailed (en) expression domains in the trunk define the posterior segmental compartments, from which NBs of row 6 and 7 and NB1-2 derive. In the pregnathal head en expression was found as follows: from late stage 8 in the posterior ectoderm of the antennal segment (en antennal stripe; en as) from which four deutocerebral NBs (Dv8, Dd5, Dd9, Dd13) delaminate; from stage 9 in a small ectodermal domain in the posterior part of the ocular segment, the en head spot (en hs), from which two protocerebral NBs (Ppd5, Ppd8) evolve; and from stage 10 in an ectodermal stripe in the posterior intercalary segment (en intercalary stripe; en is), which gives rise to three to four tritocerebral NBs (Tv4, Tv5, Td3, Td5). Furthermore, from stage 11 onwards, En is weakly detected in the anteriormost ectoderm of the procephalon corresponding to the region of the 'anterior dorsal hemispheres' (en dh). About 10 weakly En-positive NBs were identified that delaminate from the en dh. Thus all four pregnathal head segments contribute to the early embryonic brain. The spatial distribution of the En-positive NBs closely corresponds to the en domains of their origin in the ectoderm. This suggests they demarcate the posterior borders of the respective brain neuromeres (Urbach, 2003a).
In the trunk, hedgehog (hh) matches en expression. This is also the case for the intercalary segment in the pregnathal head ectoderm. By contrast, the En-positive antennal stripe and head spot are only subfractions of the large hh-lacZ domain, which, between stages 9 and 10, encompasses the antennal segment and the posterior part of the ocular segment. All NBs delaminating from this domain express hh-lacZ. From stage 10 onwards, en expressing NBs maintain a strong hh-lacZ signal, whereas hh-lacZ subsequently diminishes in the neuroectoderm and in NBs between the en antennal stripe and head spot. Additionally, hh-lacZ-expressing NBs positioned dorsally to the en/hh-lacZ-co-expressing Ppd5 and Ppd8 (both NBs demarcating part of the posterior border of the ocular neuromere), appear to prolong the boundary between the deuto- and proto-cerebrum in the dorsal direction (Urbach, 2003a).
From late stage 8 onwards, Wingless (Wg) protein is expressed in a neuroectodermal domain spanning a broad area of the ocular and the anterior antennal segment (and in the invaginating foregut). This becomes clearer in En/Wg double labelling at stage 9, revealing that the en hs is localized within this Wg domain. At that stage, Wg is already detectable in about 4-5 protocerebral NBs (Pcd6, Pcd15, Pcd7, Ppd3), derived from the region with strongest Wg expression (which later corresponds to the wg head blob). Furthermore, Wg is faintly expressed in the deutocerebral Dd7 emerging from the antennal part of the Wg domain, which corresponds to the later wg antennal stripe. By stage 10, when the wg head blob is clearly distinguishable from the wg antennal stripe, about 10-12 Wg-positive NBs have emerged from this domain. In addition, a small, spot-like wg domain was found in the intercalary segment from which a single NB (Td4) delaminates. Thus, all three wg domains, the intercalary, antennal and ocular (head blob), contribute to the anlage of the brain. From late stage 9 an additional wg domain is visible in the ectodermal anlage of the clypeolabrum, which is the wg counterpart to the En/Inv-positive region in the 'dorsal hemispheres'. Upon double labelling for either asense or deadpan (both are general markers for neural precursor cells) and wg, in embryos between stage 9 and 11 no NB emerging from the wg labral spot could be detected. By stage 11 the number of wg expressing NBs originating from the ocular head blob has increased to about 16-20, which is more than 25% of the total number of identified protocerebral NBs. Three Wg-positive NBs are identified in the deutocerebrum and one in the tritocerebrum (Urbach, 2003a). P>The gooseberry (gsb) locus encodes two closely related proteins, Gsb-distal (Gsb-d) and Gsb-proximal, which are both expressed in the developing ventral nerve cord. Gsb-d is segmentally expressed at high levels in all row 5 and 6 NBs, as well as in a median row 7 NB (NB 7-1). The expression of gsb-d was analyzed during early neurogenesis in the head region; segmental expression of Gsb-d was found to be conserved in parts of the pregnathal head ectoderm and deriving NBs. Gsb-d/En double labelling shows that the gsb-d intercalary and antennal stripes are expressed anterior to the corresponding en stripes, and are partly overlapping with the en stripes. Consequently, NBs from the posterior part of the gsb-d stripe in the tritocerebrum and deutocerebrum co-express en (Td3, Dd5), and those from the anterior part co-express wg (Td4, Dd1 and Dd7; as seen in Gsb-d/Wg double labelling), resembling the situation in the ventral nerve cord. However, Dd8 and all Wg-positive protocerebral NBs do not co-express Gsb-d (except for Ppd3 which, like Ppd10, transiently expresses gsb-d during stage 10. Gsb-d can also be detected at a low level in ganglion mother cells of the respective NBs, but fades away in NBs and their progeny during germ band retraction. Expression of the protein in the brain is completely downregulated at stage 13 (Urbach, 2003a). <>In the trunk, mirror (mirr)-lacZ is expressed in segmental ectodermal stripes giving rise to mirr-lacZ-positive NBs of row 2 and several NBs that flank row 2 at stage 11. The pattern of mirr-lacZ expression in the procephalic neuroectoderm and brain NBs differs significantly from the trunk. No evidence is found of a segmental arrangement of mirr-lacZ expression in the procephalon. Interestingly, regarding the DV axis, mirr-lacZ is mainly limited to the ventral part of the procephalic neuroectodermal region (pNR) and corresponding NBs (as confirmed by mirr-lacZ/Vnd double staining, although there is a faint dorsal mirr-lacZ expression, in the region of the later invaginating optic lobe anlage, and is, at stage 9/10, roughly complementary to en, wg and gsb-d expression, the domains of which are mainly confined to intermediate and dorsal regions of the pNR. At stage 11, expression extends towards the dorsal part of the antennal neuroectoderm and is observed in all NBs of the ventral deutocerebrum, as well as in two tritocerebral (Tv5, Td8) and four ventral, protocerebral NBs (Pad1, Pcv1, Pcv2, Pcv3). Although expression is also found in the clypeolabrum, no mirr-lacZ-positive labral NBs were identified (Urbach, 2003a). p>In addition to the segment polarity genes, the dorsoventral patterning genes ventral nervous system defective (vnd), intermediate neuroblast defective (ind) and muscle segment homeobox (msh) have been shown to confer positional information to the truncal neuroectoderm, which also contributes to the specification of NBs. For the head and brain, a detailed analysis of the expression of these genes has not yet been undertaken. In order to elucidate their putative role in patterning the head and brain, the expression of vnd, ind and msh was analyzed in the procephalic ectoderm and NBs in the early embryo (until stage 11). Although the data are consistent with their role in dorsoventral patterning being principally conserved in the procephalon, significant differences are found in their patterns of expression compared with the trunk (Urbach, 2003a).
the blastodermal stage, Ventral nervous system defective protein (Vnd) is expressed in bilateral longitudinal stripes corresponding to the most ventral neuroectodermal column, and is by stage 11 detected in all ventral and two intermediate NBs of the ventral nerve cord. Interestingly, the latter co-express en and are located in the posterior compartment of each truncal neuromere. At gastrulation the ventral longitudinal vnd domain reaches anteriorly across the cephalic furrow into the procephalic neuroectoderm. By stage 9, vnd maps in the ventral neuroectoderm of the prospective intercalary, antennal and ocular segment and is observed in ventral NBs of the antennal (Dv2, Dv3, Dv6) and ocular neuromere (Pcv1, Pcv3, Pcv6, Ppv2). It appears as if the dorsal part of the Vnd-positive antennal neuroectoderm partly co-expresses ind at that stage, but the NB Dd1, which emerges from this ectodermal region expresses only ind and not vnd. This is possibly due to the transient expression of vnd in most parts of both the ventral antennal ectoderm and corresponding NBs: by stage 10 Vnd is detected in the ventral Dv2, Dv4 and Dd5, but is already downregulated in Dv3 and Dv6, and by stage 11 it is confined to Dd5 and the new Dv8. As a consequence of the downregulation of vnd, some ventral deutocerebral NBs, which delaminate between stage 9 and 11 from this domain were not observed to express vnd (e.g. Dv1, Dv5, Dv7). By stage 11 Vnd is seen in four tritocerebral NBs (Tv2, Tv3, Tv4, Tv5), in two deutocerebral NBs (Dd5, Dv8), and in a cluster of about 13 protocerebral NBs. Interestingly, vnd expression expands along the posterior border of the en intercalary stripe (en is), and is also significantly extended dorsally into the en antennal stripe; the NBs delaminating from there. The fact that vnd and en are co-expressed in Tv5 and in Dd5, Dv8 is in agreement with findings in the ventral nerve cord, where these genes are co-expressed in two intermediate NBs. This indicates that vnd demarcates the ventral part of the posterior border in trunk as well as in brain neuromeres. Furthermore, the posterior border of the ocular vnd domain (including the NBs Pcv1, Pcv2, Pcv3, Ppv1, Ppv2, Ppv3) abuts dorsally the En-positive NBs Ppd5 and Ppd8 (deriving from the en head spot), supporting the view that these NBs demarcate the posterior border of the ocular neuromere (Urbach, 2003a).
mscle segment homeobox (msh) expression is first detected at the blastoderm stage in discontinuous patches in the dorsolateral part of the neuroectoderm that later extend and form a bilateral longitudinal stripe; this domain gives rise to the lateral NBs of the ventral nerve cord. At stage 7 msh expression is detected anterior to the cephalic furrow, which expands until stage 9 to cover, as a broad domain, the dorsal ectoderm of the intercalary and the antennal segment. As evidenced by Msh/Inv double labelling during stage 9 and stage 11, the anterior border of the msh domain coincides with the posterior border of the en hs. This suggests that msh expression in the pregnathal region is restricted to the intercalary and antennal segments, and matches the border between the antennal and ocular segment. This is further supported by Msh/hh-lacZ double labelling in stage 11 embryos, using hh as a marker for the posterior border of the ocular segment. All identified brain NBs delaminating from the dorsal intercalary and antennal neuroectoderm express msh. This suggests that during early neurogenesis, msh controls dorsal identities of the procephalic neuroectoderm and brain NBs, as was shown for the ventral nerve cord. In the ventral nerve cord, most glial precursor cells (glioblasts and neuroglioblasts) derive from the dorsal neuroectoderm, and express msh. In the intercalary segment of the early brain, two glial precursors (Td4 and Td7) were identified. Interestingly, both precursors are also located dorsally and express msh. At least until stage 11 no msh expression is found in the preantennal segments (Urbach, 2003a).
Compaing the expression of DV patterning genes in the trunk and
procephalic region, the following significant differences were observed (Urbach, 2003a):
This implies that other, still unknown factors might be involved in the DV
patterning of the anterior head neuroectoderm and protocerebrum (Urbach, 2003a).
With regard to the expression of the segment polarity genes en, hh,
wg and gsb-d, as well as the DV patterning genes msh
and vnd, it is proposed that the procephalic (pregnathal) neuroectoderm gives rise to four brain neuromeres: the tritocerebrum, the deutocerebrum, the ocular and the labral neuromere (from posterior to anterior). These tightly fused neuromeres form a supraoesophageal brain hemisphere on either side. The ocular and labral neuromeres represent the most prominent part of the brain which is traditionally referred to as the protocerebrum (Urbach, 2003a).
The detailed analysis of the dynamic expression of these genes in the procephalic neuroectoderm and in the identified brain NBs allows mapping of the boundaries of the brain neuromeres. The posterior border of the tritocerebrum is clearly represented by the en- and hh co-expressing NBs Tv4, Tv5, Td3, Td5. In the antennal and preantennal neuroectoderm the expression of en, hh, wg and gsb-d is largely restricted to intermediate and dorsal regions, and NBs deriving from there. Thus, regarding segment polarity genes, a clear demarcation of the antennal and preantennal neuromeres is only possible for the intermediate and dorsal, but not for the ventral domains. vnd is observed to be co-expressed with en in some tritocerebral (Tv5) and deutocerebral NBs (Dv8 and Dd5), located at intermediate DV positions. This is consistent with observations in the trunk, where vnd expression is dorsally expanded into each en domain in the neuroectoderm, as well as at the level of NBs. It is therefore suggested that the (transiently) vnd expressing NBs Dv2 and Dv4, which follow Dd5 and Dv8 ventrally, demarcate the ventral part of the posterior border of the deutocerebrum. The intermediate part of this border is defined by the en/hh/vnd-co-expressing Dv8, Dd5, and the dorsal part by the en- and hh-co-expressing Dd9 and Dd13. For the posterior border of the ocular neuromere, the following is proposed. Under the assumption that vnd expression also marks the posterior compartment in this neuromere, the vnd expressing NBs Pcv1, Pcv2, Pcv3, Ppv1, Ppv2 and Ppv3 would demarcate the ventral part of this border. The intermediate part is defined by the en/hh-co-expressing Ppd5 and Ppd8, and the dorsal part by the Hh-lacZ-positive NBs Ppd10, Ppd11, Ppd15 and Ppd16. Interestingly, the anterior border of the msh domain abuts exactly on the posterior ocular segmental border, indicating that msh expression is confined to the trito- and deuto-cerebrum. inv expression is observed in about 10 NBs deriving from the most anterior part of the protocerebral anlage, a region that corresponds to the En-positive 'dorsal hemispheres' (en dh). It is suggested that these NBs represent the neural correlate of the labral segment. This fourth brain neuromere seems to be of rudimentary character since it is confined to the posterior segmental compartment (considering that en/inv is normally expressed in the posterior compartment), and no NBs anterior to en dh are found. Thus, the wg domain in the clypeolabral ectoderm, which is located immediately anterior to the en dh does not give rise to brain NBs. The existence of four brain neuromeres, in the spatial orientation shown, is furthermore substantiated by the segmental expression of other genes like gsb-d, sloppy paired 1 and ladybird (Urbach, 2003a).
Thus,ese data clearly support the view that the pregnathal head consists of four segments (antennal, intercalary, ocular and labral). Furthermore, it was possible to attribute to each of the four pregnathal head segments a corresponding neuromere. All segment polarity genes are segmentally expressed in the pNR as well as in brain NBs, except mirr, the segmental expression of which is not overt. wg and gsb-d are partly overlapping, and are expressed anterior to the respective en domains, which are colocalized with hh. The expression of these genes is either mainly confined to intermediate and dorsal regions of the antennal and ocular segment (in case of en, wg and gsb-d) or is at least stronger (hh) in these parts of the pNR. Consequently, with regard to segment polarity genes there is a clear segmental demarcation, which is limited to intermediate and dorsal parts of the respective neuromeres, but it remains unclear in their ventral parts (except in the tritocerebrum). Surprisingly, the DV patterning genes vnd and msh endorse a separation of brain neuromeres in AP axis. vnd expression demarcates the ventral part of the posterior border of the tritocerebrum, deutocerebrum and ocular neuromere, and msh the dorsal anterior border of the deutocerebrum. Thus, based on the expression of segment polarity genes (en/inv, hh) and DV patterning genes (vnd, msh) a reconstruction is provided of segmental boundaries in the developing brain on the level of identified cells (Urbach, 2003a).
The segmetal organization of the anterior head, in particular the origin of the labrum, the existence of a corresponding segment and its position at the anterior pole, are central issues of a long-lasting debate concerning head segmentation. Consequently, the segmental origin of the protocerebrum, the largest and most anterior portion of the brain, has been a matter of debate and there is disagreement about whether it can be assigned to the labral and/or the ocular segment (equivalent to the acron). en expression in the en dh has been attributed to the labral segment, the existence of which is further substantiated by PNS phenotypes in head gap mutants. About 10 NBs have been identified that derive from this domain and weakly express en. Immediately anterior to the en dh, within the clypeolabral ectoderm, the genes wg, gsb-d, lbe and slp1 are found to be expressed, but these domains do not contribute to the brain. The spatial pattern of expression of these genes confirms the following: the anteroposterior orientation of a labral segment, and a parasegmental character of the border between the en dh and the labral wg domain, supporting the view that the en dh is the en-expressing part of the labral segment. It is therefore concluded that the protocerebrum consists of two neuromeres, a large ocular neuromere (comprising more than 60 NBs) and a smaller labral neuromere (comprising about 10 NBs). Since en expression delimits the posterior compartment of each segment, the labral neuromere appears to be confined to the posterior compartment (Urbach, 2003a).
The protocebrum develops prominent neuropile structures such as the central complex and the mushroom bodies. On comparative morphological grounds, the protocerebrum in arthropods has been subdivided into the archicerebrum and prosocerebrum. Accordingly, the archicerebrum, which bears the optic lobes and mushroom bodies, belongs to the acron (or ocular segment), and the prosocerebrum, which comprises the remainder of the protocerebrum (including the central complex and the neurosecretory cells of the pars intercerebralis) belongs to the labral segment. The progenitor cells of the mushroom bodies are part of the ocular neuromere, supporting the view that the mushroom bodies are indeed neuropil structures of the ocular segment or archicerebrum. Consequently, the identified labral NBs would be progenitors of neurons of the pars intercerebralis. This appears likely because during further embryogenesis the en dh becomes displaced in a brain region corresponding to the pars intercerebralis of postembryonic stages. In Drosophila, little is known about the embryonic origin of the central complex. In the grasshopper, NBs in the pars intercerebralis contribute neurons to the central complex. Taking into consideration that the identified labral NBs presumably represent the progenitors of cells of the pars intercerebralis and that the fundamental 'bauplan' of the brain is believed to be conserved among insects, it is suggested that in Drosophila, progeny cells of labral NBs participate in the formation of the central complex (Urbach, 2003a).
In the trunk the neuroectoderm and NB pattern of each hemisegment is subdivided by the activity of segment polarity genes into transverse rows and by the activity of DV patterning genes into longitudinal columns. This orthogonal expression of segment polarity and DV patterning genes is principally conserved in the posterior part of the pregnathal head neuroectoderm and corresponding regions of the early brain, but becomes obscure towards anterior sites. The intercalary neuroectoderm and neuromere are subdivided by en, hh, wg and gsb-d expression into transverse-like rows and by msh, ind and vnd into longitudinal columns. Analysis of other genes that are segmentally expressed in the trunk CNS (e.g., slp1 and lbe) provides further support for the notion that the tritocerebrum behaves like a reduced trunk neuromere. Similarly, this orthogonal pattern of segment polarity and DV patterning gene expression appears to be essentially retained in the antennal neuroectoderm and deutocerebrum. However, it appears less conserved compared with the tritocerebrum because en, wg and gsb-d (and slp1) expression is confined to intermediate/dorsal sites, ind is restricted to one NB and vnd is only transiently expressed. The orthogonal expression pattern of both gene groups is to a minor extent, if at all, conserved in the posterior half of the ocular neuromere. Owing to the lack of msh expression, a dorsoventral polarity is less obvious and most ocular NBs do not express any DV patterning gene. Finally, conservation of this pattern is not evident in the labral segment. Although some segment polarity genes are expressed in the labral ectoderm, expression of DV patterning genes is missing (except for the two vnd-positive NBs, Pav1 and Pcv4, at the border to the ocular neuromere) (Urbach, 2003a).
In this contet, it is interesting to note that the head has been claimed to be composed of two distinct domains, an anterior terminal domain and a segmented region. Both domains require high levels of Bicoid protein as an anterior determinant, but the anterior terminal domain, which encompasses the labral segment and the acron (which is equivalent to the ocular segment) is primarily specified by a signalling pathway mediated by the receptor tyrosine kinase Torso. Zygotic target genes which become activated by this signalling pathway are the gap genes hkb and tll. For tll, it has been shown that (part of) its anterior, blastodermal expression is necessary for the development of the protocerebrum, which is missing in tll mutants. tll represses hb and ftz and may thus function in the head as an 'anti-segmentation' gene. tll expression, which covers the ocular and labral neuroectoderm (the latter coincides with the region of the en dh) and emerging NBs, closely corresponds to that part of the early brain where segmental features are largely obscure. A coordinated, orthogonal expression of segment polarity and DV patterning genes within the ocular and labral neuroectoderm is not obvious, and the existence of putative serially homologous NBs in those regions of the brain is less evident. This implies that tll might be a component crucial for the suppression of segmental characteristics in the ocular and labral neuromere. Furthermore, crossregulatory interactions among the segment polarity genes in the pregnathal head differ from those in the trunk and are unique for each pregnathal segment (Urbach, 2003a).
For a part of e segmented head (mandibular, intercalary and antennal) it has been proposed that a combinatorial expression of the cephalic gap genes otd, ems and buttonhead mediates metamerization by acting directly on segment polarity genes, thereby omitting the intermediate function of pair rule genes. Recent data indicate that, in the segmental patterning of this head region, other (intermediate) regulators are involved. One of these is collier, which is already expressed in the blastoderm and is required for the formation of the intercalary segment. It is controlled by the combined activity of ems and buttonhead, and the pair rule gene even skipped, thus integrating inputs from both the head and trunk segmentation system. Such factors might help to explain that trunk specific segmental characteristics are more conserved in the intercalary and antennal neuroectoderm and NBs, when compared to the ocular and labral neuroectoderm and NBs (Urbach, 2003a).
In Drosophila tDV patterning genes subdivide the trunk neuroectoderm into longitudinal columns; vnd is required for the specification of the ventral neuroectodermal column and NBs; ind and msh have analogous functions in the intermediate and dorsal neuroectodermal columns and NBs, respectively. Remarkably, homologous genes are found to be expressed in the vertebrate neural plate and subsequently in the neural tube. In the neural tube the order of expression along the DV axis is analogous to that of Drosophila: like vnd, the vertebrate homologs of the Nkx family are expressed in the ventral region; the ind homologs, Gsh-1/2, are expressed in the intermediate region; and the msh homologs, Msx-1/2/3, are expressed in the dorsal region of the neural tube (Urbach, 2003a).
These DV patterninenes are expressed in the procephalic neuroectoderm and developing brain. Furthermore, it has been observed that in the anterior, the extent of expression is specific for each gene: msh is confined to more posterior regions, and vnd expression extends into anterior regions of the brain. Moreover, the expression border of msh and vnd coincide with neuromeric borders. A comparison of the anteroposterior sequence of DV patterning gene expression in the early brain of Drosophila, with that published for the early mouse brain, reveals striking similarities. Msx3, which presumably represents the ancestral msh/Msx gene, becomes restricted to the dorsal neural tube during later development (in contrast to Msx1/2). The anterior border of the Msx3 domain is positioned within the rostral region of the dorsal rhombencephalon, thus showing the shortest rostral extension of all vertebrate DV patterning genes. This displays analogy to msh, the expression domain of which coincides with the anterior border of the dorsal deutocerebrum, thus representing the shortest anterior extension of DV patterning genes in Drosophila. Mouse Nkx2.2 extends ventrally into the most rostral areas of the forebrain. vnd is expressed ventrally in anterior parts of the ocular and labral protocerebrum. Thus, the expression of the respective homologs in both species displays the most anterior extension among DV patterning genes. Moreover, Nkx2.2 expression in the mouse forebrain suggests that Nkx2.2 may be involved in specifying diencephalic neuromeric boundaries. Similarly, in Drosophila, dorsal expansions of the vnd domain appear to correspond to the tritocerebral and deutocerebral neuromeric boundaries (Urbach, 2003a).
Furthermore, Drosophil ind and its mouse homologue Gsh1 show similarities in their expression in the early brain. In the posterior parts of the Drosophila brain, ind is expressed in intermediate positions between vnd and msh. Likewise, in the posterior part of the mouse brain, Gsh1 appears to be expressed in intermediate positions, dorsally to Nkx2.2, and in the hindbrain ventrally to Msx3. Gsh1 has been shown to be expressed in discrete domains within the mouse hindbrain, midbrain (mesencephalon) and the most anterior domain in the posterior forebrain (diencephalon). Correspondingly, in Drosophila ind expression in restricted domains within the gnathocerebrum, the tritocerebrum, deutocerebrum and ocular part of the protocerebrum, demonstrating that the anteriormost extension of ind (and Gsh1) expression lies between that of msh and vnd (Urbach, 2003a).
Taken together, consideng these similarities, it is suggested that in the Drosophila and vertebrate early brain the expression of DV patterning genes is to some extent conserved, both along the DV axis (as suggested for the truncal parts of the Drosophila and mouse CNS) and along the AP axis. Furthermore, in Drosophila large parts of the anterodorsal procephalic neuroectoderm and NBs (more than 50% of all identified brain NBs) lack DV patterning gene expression. Likewise, in the vertebrate neural tube, gaps between the expression domains of DV patterning genes have been described, raising the possibility that other genes might fill in these gaps. How DV fate is specified in the anterior and dorsal part of the Drosophila procephalic neuroectoderm, and if other genes are involved, remains to be clarified (Urbach, 2003a).
The Drosophila brain develops from the procephalic neurogenic region of the ectoderm. About 100 neural precursor cells (neuroblasts) delaminate from this region on either side in a reproducible spatiotemporal pattern. Neuroblast maps have been prepared from different stages of the early embryo (stages 9, 10 and 11, when the entire population of neuroblasts has formed), in which about 40 molecular markers representing the expression patterns of 34 different genes are linked to individual neuroblasts. In particular, a detailed description is presented of the spatiotemporal patterns of expression in the procephalic neuroectoderm and in the neuroblast layer of the gap genes empty spiracles, hunchback, huckebein, sloppy paired 1 and tailless; the homeotic gene labial; the early eye genes dachshund, eyeless and twin of eyeless; and several other marker genes (including castor, pdm1, fasciclin 2, klumpfuss, ladybird, runt and unplugged). Based on the combination of genes expressed, each brain neuroblast acquires a unique identity, and it is possible to follow the fate of individual neuroblasts through early neurogenesis. Furthermore, despite the highly derived patterns of expression in the procephalic segments, the co-expression of specific molecular markers discloses the existence of serially homologous neuroblasts in neuromeres of the ventral nerve cord and the brain. Taking into consideration that all brain neuroblasts are now assigned to particular neuromeres and individually identified by their unique gene expression, and that the genes found to be expressed are likely candidates for controlling the development of the respective neuroblasts, these data provide a basic framework for studying the mechanisms leading to pattern and cell diversity in the Drosophila brain, and for addressing those mechanisms that make the brain different from the truncal CNS (Urbach, 2003b).
The cephalic gap genes arepressed in large domains of the procephalon and play a crucial role not only in the patterning of the peripheral ectoderm, but also in regionalizing the brain primordium. The segmental organization of the Drosophila brain is based on the expression pattern of segment polarity and DV patterning genes. To see whether the cephalic gap genes respect the neuromeric boundaries segment polarity and DV patterning genes, and to provide a basis for studying their potential role in the formation or specification of brain precursor cells, the expression was studied of orthodenticle, empty spiracles, sloppy paired 1, tailless, huckebein, and hunchback in the developing head ectoderm, as well as in the entire population of identified NBs during stages 9-11 (Urbach, 2003b).
In the cellular blastoderm tailless (tll) haseen shown to be expressed in an anterior horseshoe-shaped stripe in the cellular blastoderm, which after gastrulation shows a region of high ('HL domain') and a region of low level of tll expression ('LL domain'), and at stage 9 covers most of the protocerebral neuroectoderm. Using a tll-lacZ line at stage 9 tll expression has been found in the developing brain in most protocerebral NBs (except the dorsoposterior ones). During stages 9-11 tll-lacZ expression expands in the protocerebral neuroectoderm beyond the En-positive head spot (hs). By stage 11 it is detectable in all protocerebral NBs. In addition, tll-lacZ is found in some ventral and dorsal deutocerebral NBs, indicating that tll is not exclusively confined to protocerebral progenitors (Urbach, 2003b).
During early neurogenesis in the unk, empty spiracles (ems) is metamerically expressed in lateral ectodermal patches. In the head, it acts as a gap gene, which is expressed in a circumferential procephalon domain in the early cellular blastoderm. During gastrulation this circumferential stripe dissolves into three smaller ectodermal domains between the anterior part of the mandibular segment and the posterior part of the ocular segment; these domains are not in segmental register. During further development, the third domain splits into a mandibular/intercalary and an antennal component. All these domains contribute NBs to the brain. In addition to ems expression in the intercalary and antennal segments, and the corresponding trito- and deutocerebral neuromers, ems expression was also detected in a small neuroectodermal region. Finally, a further ems patch is located in the dorsoanterior procephalic ectoderm ('dorsal patch'), which becomes part of the labral ectoderm and does not appear to give rise to brain NBs. Thus, from stage 9 onwards, part of the antennal/ocular ems domain overlaps with the En-positive hs, and from stage 10/11 onwards these genes are also found to be co-expressed in the en hs-derived protocerebral NBs Ppd5 and Ppd8 (although en and ems expression also partly overlaps in the trunk ectoderm, a co-expression of both genes in trunk NBs is never seen). In contrast to earlier observations, showing that most of the tritocerebral NBs are included in the ems-expressing domain, only the dorsal Td6 was identified as Ems-positive. Ems protein is detectable in clusters of brain cells until the end of embryogenesis (Urbach, 2003b).
The sloppy paired (slp)locus contains the two related genes slp1 and slp2. slp1, which acts as a head gap gene, plays a predominant role in head formation, while slp2 is largely dispensable. In the trunk neuroectoderm, where slp1 has a function as a pair-rule and segment polarity gene, it is segmentally expressed in neuroectodermal stripes as well as in NBs of row 4 and 5. This segmental appearance of slp1 expression is found to be conserved in parts of the procephalon. In the blastoderm, Slp1 protein is detected in a large domain of the procephalon anlage, which subsequently diminishes in its anterior/ventral part. As a result, only the posterior half of the original slp1 domain remains as a circumferential ring ('head stripe') and gets separated from the anterodorsal part ('head cap'). To follow the dynamics in the Slp1 expression pattern, Slp1/En double labelling was examined during stages 8-11. The 'head stripe' corresponds to the slp1 stripe of the prospective mandibular segment, and the posterior part of the 'head cap' to the Slp1-positive stripe in the prospective antennal segment (slp1 as). At the beginning of gastrulation, a new Slp1 ectodermal spot in the anterodorsal procephalon is observed; this spot later becomes part of the labral ectoderm. In addition, at stage 9, three new ectodermal domains become detectable: one stripe anterior to the en intercalary stripe belonging to the intercalary segment, and two small spots in the region of the ocular segment (anterior to the en head spot). Except for the labral domain, the slp1 domains contribute NBs to the brain. Thus, slp1 is segmentally expressed in the procephalic neuroectoderm and subsets of brain NBs, resembling the situation in the trunk. At stage 11 patchy expression of Slp1 becomes detectable within the ocular and labral ectoderm and in some underlying ocular and labral NBs. Some of these NBs initiate slp1 expression after delamination; e.g. Pcv6 and Pcd2 delaminate at stage 9 and do not express slp1 before stage 11. Slp1 expression is observed in the brain until the end of embryogenesis (Urbach, 2003b).
huckebein (hkb), a termial gap gene, is first expressed at the anterior and posterior blastodermal poles, where it is required for the specification of the endodermal anlagen, and later for the invagination of the stomodeum. After gastrulation, hkb becomes transiently expressed in a repetitive pattern in the trunk neuroectoderm and in eight, mainly intermediate, NBs per hemineuromere. In the procephalic region at the cellular blastoderm stage, hkb (Urbach, 2003b).
Expression is detected
in a centrally ocated stripe and a dorsal etodermal spot. hkb in situ hybridization combined with anti-Inv antibody staining reveals that during stage 9/10 the hkb stripe covers most of the antennal ectoderm and reaches into the anterior region of the intercalary segment, and the hkb spot covers part of the ocular ectoderm. During stage 9 hkb transcript in the ocular spot becomes progressively restricted to the delaminating protocerebral NBs, Pcv7 and Pcd2, and remains strongly expressed in both NBs until stage 11. In the antennal domain during stage 10/11 the transcript becomes confined to three to five deutocerebral NBs. However, using a hkb-lacZ line (5953) the marker is expressed in all deutocerebral NBs at stage 10. At stage 11, hkb-lacZ was not detectable in Dd8 and Dd11, indicating that hkb is not a general deutocerebral NB marker. In the tritocerebrum, hkb is expressed only in Td6 (stage 10) and in Tv1, Td8. Thus, although expressed in a few trito- and protocerebral NBs, hkb expression appears to be mainly confined to the antennal neuroectoderm and NBs. Compared with the transcript, which becomes restricted to the NBs during stage 9-11, hkb-lacZ expression has a longer perdurance in the peripheral ectoderm and corresponding NBs. By stage 14, most of the hkb transcript has disappeared and is confined to some deutocerebral cells; hkb-lacZ is strongly expressed until the end of embryogenesis in deutocerebral, and at a lower level, in protocerebral cells, the putative progeny of the identified Hkb-positive brain NBs (Urbach, 2003b).
hunchback (hb) expression i the anterior half of the embryo falls below the limit of detection at the beginning of germ band extension, but accumulates during the extended germ band stage in the CNS, where it is transiently expressed in early, fully delaminated, trunk NBs (S1 and S2) and their progeny. Antibody staining reveals that, from stage 8 onwards, Hb protein is not detected in the head neuroectoderm, but is very dynamically expressed in brain NBs. At stage 9, only about half of the identified deuto- and protocerebral NBs show Hb protein at a detectable level, suggesting that Hb is not a general marker for early NBs. Correspondingly, it is found that Hb protein is also lacking in particular S1 and S2 NBs of the trunk. In some of the early brain NBs, Hb first becomes detectable at stage 10, after their delamination. For example, the early NBs Pcv9 and Pcd6 delaminate at late stage 8 but do not start Hb expression before stage 10. By stage 10, Hb is expressed in about 26 brain NBs, most of which delaminate between stage 9 and 10. In most of these NBs, Hb expression is progressively lost, but is observed in an increasing amount of progeny cells. At stage 11, it is confined to a small subpopulation of about five tritocerebral and four to six protocerebral NBs. Thus, as opposed to the trunk, hb expression in the brain is not limited to early NBs. Hb is expressed in the brain until stage 15, when it is detected in a few cells of the protocerebrum (Urbach, 2003b).
Taken together, among the cephalic gap gens, slp1 appears to respect segmental boundaries during early neurogenesis of the brain. By contrast, in the considered period of development (stage 9-11), the expression of ems, otd and tll does not seem to respect these borders, contradicting claims in previous reports. All three genes are expressed in NBs deriving from ectodermal domains that are part of two or three neighboring segments. For example, ems is expressed in a small number of NBs comprising about six posterior ocular and four anterior deutocerebral NBs (all of which derive from the same ems domain, except Dv3 and Pcv5), and one tritocerebral NB. Accordingly, ems mutants show defects in the intercalary, antennal, and the ocular segment (e.g., the en hs is missing). Considering that ems is expressed in only a few trito- and deuto-cerebral NBs it is remarkable that ems mutants show a deletion of the trito- and deuto-cerebrum. An explanation for this could be that ems expression, which during earlier development covers the neuroectoderm of the respective segments, possibly confers specific identities to the arising trito- and deuto-cerebral NBs. The lack of these NBs might be responsible for the loss of NB-specific gene expression, and (secondarily) for the gross morphological defects seen in the ems mutant brain. A similar proposal has been made to explain the brain defects that occur in buttonhead (btd) mutants, although btd is not expressed in NBs of the corresponding brain regions (Urbach, 2003b).
The expression was analyzed of the homeogenes proboscipedia (pb) and labial (lab), both members of the Antennapedia complex and known to be expressed in the head ectoderm and in the brain after mid-embryogenesis. Antibody staining against Pb reveals that at stage 11, the protein is restricted to internal cells of the mandibular segment (presumably mesodermal cells) and to dorsal ectoderm of the maxillary and labial appendages. No Pb protein was detected in brain NBs (Urbach, 2003b).
lab has been described as being express in the posterior tritocerebrum at stage 14. Using an antibody, the expression of Lab protein during early neurogenesis was investigated. From stage 9 onwards, Lab is detected in the ectoderm of the intercalary segment, and presumably in a small part of the posteroventral antennal segment. At that stage, the only NB expressing Lab protein is Dv2. Double labelling against En reveals that at stage 11 Lab is expressed throughout the ectoderm of the intercalary segment. The Lab domain overlaps posteriorly with the en intercalary stripe (en is), indicating that posterior borders of lab expression and of the intercalary segment coincide. The character of the anterior border of the lab domain is less clear. Dorsally, it runs along the posterior border of the en antennal stripe (en as); ventrally, however, it reaches the anterior border of the en as. This suggests, that the anterior border of the lab domain is segmental in the dorsal region and parasegmental in the ventral region. Interestingly, also for scr and dfd, which are other members of the ANT-C, it has been reported that they initiate expression in a jagged stripe resolving into a pattern that is dorsally segmental and ventrally parasegmental. All NBs arising from the Lab-positive neuroectoderm express lab, among them all tritocerebral NBs and two ventral NBs, which are attributed to the deutocerebrum (Dv2 and Dv4) because they are located on the same anteroposterior level as the en-expressing Dv8 and Dd5 (Urbach, 2003b).
dachshund (dac) is involved in the evelopment of the eye and the mushroom bodies where it is expressed already in the progenitor cells. Using an antibody, Dac expression was found in the trunk CNS not before stage 12; it is expressed in only two or three cells (not NBs) per neuromer. In the procephalon, Dac is already detected by stage 9 in a small area of the dorsal ocular neuroectoderm from which four Dac-positive NBs (Pcd4, Pcd8, Pcd9, Pcv9) delaminate. It has been suggested that the NBs delaminating from this Dac domain represent the progenitors of the mushroom body and co-express eyeless (ey). In disagreement with this, it was found that, at that stage, the co-expression of both genes is confined only to a small region of the Dac-positive neuroectoderm and to only one of the four identified Dac-positive NBs. As evidenced by Dac/Ey antibody double labelling, this NB (Pcv9) is one of the five Ey-positive brain NBs identified at stage 9. Until stage 11, the Dac-expressing ocular domain expands into the antennal segment and into the optic lobe anlage (now encompassing also the ectodermal region called 'para-MB neuroectoderm'), and a further spot appears in the clypeolabral ectoderm. At this stage, Dac protein can be observed in 13 protocerebral NBs and in the tritocerebral Tv2, but in no deutocerebral NBs. From stage 12 onwards, Dac becomes expressed in an increasing number of scattered cell clusters in the brain and ventral nerve cord (Urbach, 2003b).
eyeless (ey), which encodes a mof the Pax6 family of transcription factors, is a crucial regulator for eye development. It is expressed in the embryonic ventral nerve cord and brain and has been shown to be involved in the development of the mushroom bodies. Although it has been suggested that ey is expressed in the progenitor cells of the mushroom bodies, ey expression in the evolving NB pattern of the ventral nerve cord and brain has not been described to date. Using an Ey antibody (which principally shows the same pattern as ey mRNA in situ hybridization), Ey in the trunk is found to be expressed in segmentally reiterated ectodermal stripes and, at stage 11, in six NBs per hemineuromere. In the posterior pregnathal head segments (intercalary and antennal), a segmental expression of Ey appears to be fundamentally conserved. Ey/En double labelling reveals that the Ey-positive spots (and emerging NBs) in both segments are localized just anterior to their En-positive counterparts. Ey protein is detected, by stage 9, in five NBs: Pcv6, Pcv7, Pcv9, Pcd2 (which derive from an ocular ectodermal Ey domain) and Dv6 (which develops from a small ectodermal Ey spot in the antennal segment). By stage 10, ey becomes expressed in an intercalary ectodermal spot, from which, by that stage, the Ey-positive Td2 and (slightly later) Td1 delaminate, and in a second ocular ectodermal spot from which Ppd12 develops. In contrast to the intercalary and antennal ectodermal spot, Ey protein becomes depleted during stage 11 in the ocular domain. At late stage 11, Ey is found in 15 brain NBs, including two intercalary, three deutocerebral, nine ocular and one labral NB (Pad7). During subsequent development of the brain, the Ey expression pattern becomes complex, especially in the preantennal segments, but it appears to be mainly confined to the progeny of the identified NBs (Urbach, 2003b).
The second Drosophila Pax6 gene, twin of eyeless
(toy), has been shown to be expressed in the blastoderm in an
anterodorsal patch that represents the posterior region of the procephalon anlage, and later in the embryonic brain (including the mushroom body and visual system). At stage 9, Toy expression encompasses the dorsal ocular and the anterodorsal part of the antennal ectoderm. All NBs that delaminate from this part of the neuroectoderm express Toy. By stage 11, these include about 40 protocerebral NBs, all of which derive from the ocular ectoderm, except the putative labral Pav1, and two deutocerebral NBs. This pattern of expression in NBs closely matches the pattern that has been observed for otd. In some of these NBs expression of Toy protein is transient and ceases during stage 11 (e.g., Dd3). To determine whether both Pax6 genes, ey and toy, are co-expressed in identified NBs, toy in situ hybridization was performed combined with an Ey antibody staining. Interestingly, toy appears to be expressed only in those Ey-positive NBs deriving from the ocular segment; a co-expression was not seen in Ey-positive NBs of the trito- and deuto-cerebrum, or in Pad7. During later embryogenesis, Toy protein is expressed in cells of the proto- and deutocerebrum, as well as in the tritocerebrum, although no Toy-positive tritocerebral NBs were identified (Urbach, 2003b). In order to establish further molecular markers that are specifically expressed in subsets of brain NBs, the expression of castor, Fasciclin 2, klumpfuss, ladybird early, POU-domain 1 gene, runt and unplugged were investigated for details of the spatiotemporal expression pattern of these genes in the neuroectoderm and brain NBs (stages 9, 10 and 11) (Urbach, 2003b).
In the trunk, the pair-rule gene runt is expre in segmental domains of the ventral neuroectoderm and in five NBs of row 2 and 3 and two NBs of row 5. Runt has also been shown to be expressed in an anterodorsal region of the blastoderm, corresponding to the presumptive head region. En/Runt antibody co-labelling reveals that this Runt domain contributes to the ocular segment. In addition to the ocular segment, patches of runt expression are in the intercalary, antennal and clypeolabral ectoderm, and in subsets of protocerebral and deutocerebral NBs. At stage 11, the protein is expressed in a total of 23 brain NBs, some of which initiate Runt expression after delamination from Runt-negative ectoderm, and in a large number of postmitotic cells until the end of embryogenesis (Urbach, 2003b).
In the ventral nerve cord castor (cas), encodng a zinc-finger protein, has been shown to be expressed in 18 NBs per hemineuromere, including early (S1-S2) and late delaminating (S3-S5) NBs, and to be involved in cell fate control within NB lineages. In the procephalon, cas expression is not detectable before stage 10. It is dynamically expressed in the central and dorsal neuroectoderm of the ocular segment, in the median antennal segment, and, by stage 11, in the labral segment, which is surprising since cas is not expressed in the neuroectoderm of the trunk. A proportion of Cas-positive protocerebral and deutocerebral NBs are derive from these domains. Most NBs appear to delaminate from Cas-negative neuroectoderm, and start to express cas at the time of formation, or show a reproducible delay in the onset of cas expression. The latter may already have produced a part of their lineage, which likewise has been proposed for early trunk NBs (e.g. NB7-4). At late stage 11, Cas is expressed in about 60% of the total number of identified brain NBs (Urbach, 2003b).
Using an antibody against the cell membrane glycoprotein Fasclin 2 (Fas2), it has been found that in the procephalic region Fas2 is first expressed by late stage 10 in an ectodermal patch at the border between the intercalary and antennal segment. Later it also covers the posterodorsal ocular neuroectoderm (including the optic lobe anlage) and part of the labral ectoderm. Fas2 is also detected in brain NBs emerging from the antennal and intercalary neuroectoderm, and at a low level in a few dorsal ocular NBs. It has been found that Fas2 controls proneural gene activity in the eye/antennal imaginal disc, raising the possibility that it functions likewise in the procephalic neuroectoderm. However, Fas2 expression in almost all identified brain NBs is initiated after delamination from Fas2-negative neuroectoderm, suggesting that Fas2 in the procephalic neuroectoderm is not involved in the regulation of proneural genes. It has been shown that Fas2 appears on the surface of neural somata prior to axon outgrowth; these neurons belong to 'fiber tract founder clusters' that pioneer the main axonal tracts in the brain. Considering position and time point of development, it is suggested that the identified Fas2-positive deuto- and trito-cerebral NBs (Tv1, Tv2, Td1, Td2, Td6, Td8; Dv2, Dd9, Dd11) are the precursors of the 'D/T fiber tract founder cluster' (Urbach, 2003b).
In the trunk, the zinc-finger transcription factor Klumpfuss (u) is expressed from stage 10 onwards in an increasing number of NBs, and at stage 11, almost all NBs (except NB2-3 and NB6-4) show nuclear Klu staining. The expression of Klu in the procephalon was analyzed using an antibody against Klu and the P-lacZ enhancer trap strain klu P212 which basically shows an identical expression pattern. Klu is not expressed in the neuroectoderm. Similar to the situation in the trunk CNS, Klu protein is first found at a detectable level at stage 9, in a subset of (about 17) brain NBs and at late stage 11 in almost all brain NBs. For most NBs, there is a significant delay between birth and onset of klu expression. Klu also appears to be expressed in ganglion mother cells, as was shown for the trunk (Urbach, 2003b). ladybird (lb), a tandem of the homeobox genes ladybird early (lbe) and ladybird late (lbl), both of which encode transcription factors, show a similar expression pattern, with lbe activity slightly preceding that of lbl. At stage 11, both genes are expressed in segmental repetitive patches in the laterodorsal trunk ectoderm and specifically in one NB per hemineuromere, the lateral NB 5-6. Using an antibody against Lbe the protein is first observved by stage 10 in three small procephalic patches in the labral, ocular and antennal ectoderm, and at stage 11 in an additional patch of the intercalary ectoderm. Lbe is selectively expressed in only four brain NBs on either side: one in the tritocerebrum (Td4), one in the deutocerebrum (Dd7) and two in the protocerebrum (Ppv3, Pcv8). Wg/Lbe double labelling demonstrates that Lbe and Wg expression are colocalized in the intercalary, antennal and labral ectoderm, and in Td4 and Dd7; remarkably, the ocular Lbe-positive domain and corresponding NBs (Ppv3 and Pcv8) are Wg negative. Lbe protein is detected in the progeny of the identified brain NBs until the end of embryogenesis (Urbach, 2003b).
The two closely related Drosophila POU-domain genes, pdm1nubbin) and pdm2, are co-expressed in the developing CNS (before stage 13) and have been shown (at least with respect to the specification of the first ganglion mother cell of the truncal NB4-2) to be functionally redundant. pdm1 is expressed in the trunk neuroectoderm during the first and second wave of NB segregation (stage 8/9), and transiently in most NBs at stage 10 and 11. In the procephalon the expression of the Pdm1 protein is highly dynamic. Until stage 10, Pdm1 is roughly restricted to the neuroectoderm of the antennal and ocular segments. Later, it is also found in the intercalary and labral ectoderm. At stage 9, NBs derived from Pdm1-positive neuroectoderm appear to be Pdm1 negative and initiate pdm1 expression at stage 10 or stage 11. At late stage 11, approximately one half of the brain NBs (about 52 NBs) express pdm1, including most deuto- and trito-cerebral NBs, as well as central ocular NBs and part of the labral NBs (Urbach, 2003b).
Expression of the homeodomain gene unplugged (unpg) the trunk starts at stage 8 in the ventral midline and becomes detectable in NBs of the ventral nerve cord at late stage 11. Using an unpg-lacZ line, unpg expression is observed in the head at stage 9 in a large domain encompassing the intercalary, antennal and most of the ocular ectoderm. Until stage 11, the expression is gradually lost in the intercalary ectoderm, but upregulated in the dorsal part of the antennal and adjacent ocular ectoderm. In contrast to trunk NBs, which have already divided several times before expressing unpg at late stage 11, unpg-lacZ is weakly expressed already at stage 9 in all deutocerebral and almost all protocerebral NBs. At late stage 11, it is strongly expressed in almost all deutocerebral NBs (except for some ventral ones), and in some ocular NBs close to the deutocerebral/ocular border. Until the end of embryogenesis, unpg expression is observed in the putative progeny cells of the unpg-lacZ-positive deuto- and protocerebral NBs (Urbach, 2003b).
For thoracic and abdominal segments, each NB acquires a unique ideny, which corresponds to a particular position in the neuroectoderm and (upon delamination) in the subectodermal NB layer, to a certain time point of its delamination, and to the combination of genes expressed. Descriptions of gene expression has provided an important basis for the elucidation of mechanisms controlling cell fate specification during early neurogenesis in the trunk region. In contrast to the truncal CNS, in which the segmental organization is obvious and the composition of the neuromeres is almost identical, the brain neuromeres are much more diverse and complex. Accordingly, information on identified brain cells and their gene expression is hardly available so far, and thus essential tools for investigating the mechanisms underlying pattern formation and cell diversity in the brain are lacking (Urbach, 2003b).
Comparison of the combination of markers expressed in individual NBs ell as their relative position within the NB layer of each segment suggests that several NBs exist in the brain that are serially homologous to NBs in the ventral nerve cord (VNC). This mainly applies to the posterior brain (deuto- and trito-cerebrum), which is less derived than the anterior brain (protocerebrum). For example, according to these criteria, NB5-6 in all abdominal, thoracic and gnathal neuromeres would be serially homologous to Td4 in the tritocerebrum and to Dd7 in the deutocerebrum. These NBs exhibit a similar posterodorsal position within the respective neuromer immediately anterior to the En-positive NBs, and are the only NBs which specifically co-express the following molecular markers: lbe (which is generally expressed in only one NB per hemisegment), wg, gsb-d, slp1 (except Td4), msh, cas, seven-up (except Td4), pdm1, klu and asense. Furthermore, some of the daughter cells of Td4 and NB5-6 co-express ladybird and the glia-specific marker reversed polarity. The existence of serially homologous NBs is intriguing since the number of NBs in the tritocerebrum and deutocerebrum is considerably reduced, the timecourse of neurogenesis within the brain and VNC is different (especially in the tritocerebrum the development of NBs is significantly delayed), and the development of head segments (and consequently of brain neuromeres) has been assumed to be differently regulated (Urbach, 2003b).
In the VNC, serially homologous NBs that express the same combination of mocular markers give rise to almost identical cell lineages, suggesting that similar regulatory interactions take place during the development of these NBs and their cell lineages. However, some of the serially homologous VNC lineages have been shown to include a subset of progeny cells that specifically differ between thoracic and abdominal neuromeres. It is expected that such segment-specific differences are even more pronounced among serially homologous lineages within the brain and between the brain and VNC. Differences in the combination of marker genes expressed by putative serially homologous NBs may point to candidate genes conferring segment-specific characteristics to their lineages. Thus, unravelling the lineages of serially homologous NBs and the genetic network that controls their development will help to elucidate how region-specific structural and functional diversity in the CNS evolves from a basic developmental ground state (Urbach, 2003b).
Animals from flies to humans are able to distinguish subtle gradations in temperature and show strong temperature preferences. Animals move to environments of optimal temperature and some manipulate the temperature of their surroundings, as humans do using clothing and shelter. Despite the ubiquitous influence of environmental temperature on animal behaviour, the neural circuits and strategies through which animals select a preferred temperature remain largely unknown. This study identified small set of warmth-activated anterior cell (AC) neurons located in the Drosophila brain, the function of which is critical for preferred temperature selection. AC neuron activation occurs just above the fly's preferred temperature and depends on dTrpA1, an ion channel that functions as a molecular sensor of warmth. Flies that selectively express dTrpA1 in the AC neurons select normal temperatures, whereas flies in which dTrpA1 function is reduced or eliminated choose warmer temperatures. This internal warmth-sensing pathway promotes avoidance of slightly elevated temperatures and acts together with a distinct pathway for cold avoidance to set the fly's preferred temperature. Thus, flies select a preferred temperature by using a thermal sensing pathway tuned to trigger avoidance of temperatures that deviate even slightly from the preferred temperature. This provides a potentially general strategy for robustly selecting a narrow temperature range optimal for survival (Hamada, 2008).
Although the physiology of all cells is affected by temperature, the expression of temperature-activated members of the transient receptor potential (TRP) family (thermoTRPs) can make cell excitability highly temperature-responsive (Dhaka, 2006). ThermoTRPs are cation channels with highly temperature-dependent conductances that participate in thermosensation from insects to humans. The Drosophila TRP channel dTrpA1 promotes larval heat avoidance (Rosenzweig, 2005) and can be activated by warming in ooctyes (Viswanath, 2003). This study addressed whether dTrpA1 contributes to the selection of a preferred temperature in the adult fly. When allowed to distribute along a thermal gradient for 30 min, wild-type D. melanogaster adults prefer ~25°C, their optimal growth temperature. Compared to wild-type controls, dTrpA1 loss-of-function mutant animals showed increased accumulation in the warmest (28-32°C) regions of the gradient, but not in the coolest (18-22°C) regions. A dTrpA1 genomic minigene rescued the phenotype. Animals heterozygous for dTrpA1 loss-of-function mutations also preferred slightly elevated temperatures. Thus, dTrpA1 function is important for determining thermal preference and specifically contributes to avoidance of warm regions (Hamada, 2008).
If dTrpA1 was involved in thermotransduction, it should regulate the warmth responsiveness of thermosensors. As the identity of the adult Drosophila thermosensors was unknown, dTrpA1 protein expression was examined (using anti-dTrpA1 antisera). dTrpA1 expression was detected in three sets of previously uncharacterized cells in the brain: lateral cell (LC), ventral cell (VC) and AC neurons. dTrpA1 was also detected in the proboscis, but ablation studies detected no contribution of the proboscis to warmth avoidance. To focus on the neurons that contribute to thermal preference, where the rescuing dTrpA1 minigene restored dTrpA1 expression was examined; dTrpA1 expression was restored specifically within AC neurons, but not LC or VC neurons. This suggested that dTrpA1 expression in AC neurons (two pairs of neurons at the brain's anterior) sufficed to restore thermal preference and that AC neurons might act as thermosensors (Hamada, 2008).
Temperature responsiveness of AC neurons was examined using the fluorescent calcium indicator G-CaMP. When exposed to increasing temperature, AC neurons showed robust increases in G-CaMP fluorescence, reflecting warmth-responsive increases in intracellular calcium. Ten out of the 27 AC neurons imaged had fluorescence increases between 4% and 39%, with a mean increase over baseline among these cells of 15%. The average temperature at which fluorescence increases were initially observed was 24.9°C, compatible with AC activation as temperatures rise above preferred. In contrast, none of the 21 dTrpA1 mutant AC neurons imaged had fluorescence increases. As a control that mutant AC neurons remained physiologically active, it was confirmed that they showed robust responses on potassium chloride addition. Notably, AC responses did not depend on an intact periphery, since all G-CaMP studies were performed using isolated brains from which peripheral tissues had been removed. These observations identify AC neurons as warmth-activated, dTrpA1-dependent thermosensors (Hamada, 2008).
AC neurons project towards several brain regions, including the antennal lobe. The antennal lobe is implicated in cockroach thermosensation, but has been studied exclusively for olfaction in Drosophila. So far, 11 of the ~50 antennal lobe glomeruli remain unassociated with identified olfactory receptors. AC neurites elaborated within two such unassociated glomeruli, VL2a and VL2p. Thus the Drosophila antennal lobe contains both thermosensory and olfactory neuron processes. VL2a is also innervated by Fruitless-expressing neurons implicated in pheromone transduction, suggesting that even individual glomeruli receive multi-modal sensory information. AC processes also branched within the subesophageal ganglion and superior lateral protocerebrum, although these target regions are less defined than in the antennal lobe. These regions have been previously implicated in processing other types of sensory input (Hamada, 2008).
As dTrpA1 expression in AC neurons seemed sufficient to restore normal thermal preference, whether such expression was also necessary was also examined. dTrpA1 was knocked down selectively in AC neurons using tissue-specific RNA interference targeting dTrpA1 controlled by dTrpA1SH-GAL4, a promoter expressed in AC but not LC or VC neurons. Consistent with the importance of dTrpA1 expression in AC neurons in thermal preference, AC knockdown increased the fraction of animals present in the 28-32°C region compared to controls. Similar results were obtained when dTrpA1 expression was knocked down using a broad neuronal promoter (Appl-GAL4). All knockdowns were assessed by dTrpA1 immunohistochemistry. dTrpA1 knockdown with the general cholinergic neuron promoter Cha(7.4)-GAL4 eliminated detectable dTrpA1 expression in AC (and LC and VC ) neurons, decreasing warmth avoidance. In contrast, dTrpA1 RNAi expressed using Cha(1.2)-GAL4 -- which is expressed in many brain cholinergic neurons but not AC neurons -- did not disrupt warmth avoidance. Taken together, these data suggest that dTrpA1 expression in AC (but not LC or VC) neurons is both necessary and sufficient for normal thermal preference behaviour. Whether LC and VC neurons participate in other warmth-activated responses is unknown (Hamada, 2008).
The identification of an internal sensor controlling temperature preference conflicts with the established view that Drosophila sense moderate warming using thermosensors in the third antennal segment. The effects were tested of surgically removing either one third antennal segment and arista (unilateral ablation) or both (bilateral ablation). Both unilateral and bilateral ablation increased the fraction of animals in cool (18-22°C), but not warm (28-32°C), regions. Thus these tissues were dispensable for warmth avoidance, but essential for cool avoidance. When dTrpA1 mutants were subjected to bilateral ablation, they accumulated in both cool and warm regions: the fraction between 18-22°C did not differ from wild-type ablation animals; the fraction between 28-32°C did not differ from non-ablated dTrpA1 mutants. Thus dTrpA1-expressing cells and antennal cells function additively to set preferred temperature, promoting avoidance of elevated and reduced temperatures, respectively (Hamada, 2008).
These data are consistent with warmth activation of dTrpA1 serving as the molecular basis of warmth sensing by AC neurons. As thermal activation of mammalian TRPA1 proteins is controversial, whether dTrpA1 could act as a molecular sensor of warming in the fly was tested. Indeed, misexpression of dTrpA1 throughout the fly nervous system (using C155-GAL4) caused a dramatic phenotype not observed in controls: heating these flies to 35°C for 60 s caused incapacitation, an effect reversed on return to 23°C. Similar effects were observed using electrophysiology, with moderate warming (above ~25°C) triggering a barrage of excitatory junction potentials at the neuromuscular junction. These data strongly support dTrpA1 acting as a molecular sensor of warming. The ability of dTrpA1 mis-expression to confer warmth activation also suggests that dTrpA1 can be used as a genetically encoded tool for cell-specific, inducible neuronal activation. dTrpA1 might be particularly useful in tissues such as the fly brain where thermal stimulation is easier to deliver than the chemical or optical stimulation that controls other tools for modulating neuronal activity (Hamada, 2008).
To test whether warmth activation is a property of other insect TrpA1s, the malaria mosquito Anopheles gambiae TrpA1 (agTrpA1) was examined. dTrpA1 is warmth-activated when expressed in Xenopus laevis oocytes. agTrpA1 also showed robust warmth activation. These currents were specific, they were not observed in uninjected oocytes and were inhibited by ruthenium red (which antagonizes other TRPs). Similar to mammalian thermoTRPs, both dTrpA1 and agTrpA1 showed outward rectification. Closely related TrpA1s are present in the flour beetle Tribolium castaneum and in disease vectors such as Pediculus humanus corporis (body lice), Culex pipiens (common house mosquito) and Aedes aegypti (yellow and dengue fever mosquito) which use warmth-sensing for host location and habitat selection. Such insect TrpA1s constitute potential targets for disrupting thermal preference and other thermosensory behaviours in agricultural pests and disease vectors (Hamada, 2008).
Environmental temperature affects the physiology of all animals. Increasing temperatures associated with climate change are linked to poleward redistributions of hundreds of species including insects, fish, birds and mammals, AC neurons are internal. As a ~1 mg fly is readily penetrated by ambient temperature variations, such an internal sensor should monitor environmental temperature effectively. dTrpA1 activation seems to be critical for AC neuron activation, suggesting that dTrpA1 threshold and expression changes could modulate thermal preferences. More speculatively, changes in insect TrpA1 function and expression could facilitate movements into novel environments or development of novel behaviours such as host seeking (Hamada, 2008).
Although effects of environmental temperature on behaviour are ubiquitous, the mechanisms animals use to seek out optimal temperatures are largely unknown. AC neurons become active as temperatures rise above the preferred temperature, suggesting that they may function as 'discomfort' receptors that, together with putative antennal cool receptors (similar to those described in other insect antennae), repel the fly from all but the most optimal temperatures. Notably, mice lacking the cool-activated channel TRPM8 prefer abnormally cool temperatures, whereas mice lacking heat-activated TRPV4 prefer warmer temperatures, indicating that similar strategies may be used in mammals (Hamada, 2008).
Animals react to environmental changes over timescales ranging from seconds to days and weeks. An important question is how sensory stimuli are parsed into neural signals operating over such diverse temporal scales. This study uncover a specialized circuit, from sensory neurons to higher brain centers, that processes information about long-lasting, absolute cold temperature in Drosophila. Second-order thermosensory projection neurons (TPN-IIs) were identified exhibiting sustained firing that scales with absolute temperature. Strikingly, this activity only appears below the species-specific, preferred temperature for D. melanogaster (~25°C). The inputs and outputs of TPN-IIs were traced, and they were found to be embedded in a cold 'thermometer' circuit that provides powerful and persistent inhibition to brain centers involved in regulating sleep and activity. These results demonstrate that the fly nervous system selectively encodes and relays absolute temperature information and illustrate a sensory mechanism that allows animals to adapt behavior specifically to cold conditions on the timescale of hours to days (Alpert, 2020).
This work uncovered a complete circuit, from sensory neurons to circadian and sleep centers, that processes information about absolute cold temperature to exert influence on fly behavior in the timescale of minutes to hours to days (Alpert, 2020).
The circuit is composed of sensory neurons of the antenna (including newly identified thermosensory neurons only active in the cold) and of specialized second-order thermosensory projection neurons of the PAL and provides persistent inhibition to the DN1a cluster of circadian neurons to adapt sleep/activity patterns specifically to cold conditions (Alpert, 2020).
The data show that 'absolute temperature' and 'temperature change' signals can be extracted by second-order neurons from the activity of peripheral thermoreceptors and demonstrate that persistent signaling in sensory circuits mediates long-lasting changes in behavior, beyond the rapid responses that are generally well understood. Moreover, the results illustrate how the fly nervous system selectively encodes and relays absolute cold temperature information to adapt behavior specifically to cold conditions (Alpert, 2020).
What may be the significance of this sensory mechanism for the animal's natural behavior? Thermal conditions are well known to exert long-lasting changes in physiology and behavior, but due to the pervasive nature of temperature itself, such changes do not necessarily require input from a sensory circuit. For example, on the timescale of days and weeks, cold temperature promotes the alternative splicing of clock genes, directly affecting the dynamics of the molecular clock. The sensory mechanism we discover here allows the animal to respond both rapidly and persistently to cold conditions. Such a mechanism may be important to bridge the gap between behavioral responses on the timescale of minutes to hours and biochemical changes that may take days to fully set in (and may be difficult to reverse) (Alpert, 2020).
In a small poikilotherm, cold (the range of temperature below the optimal species-specific value determined by the biochemistry of the animal) profoundly impacts motility and the ability to process stimuli. Cold temperature can quickly render a fly unable to move rapidly or fly away, and it is well known that larger insects, such as bumble bees, have evolved adaptations to ensure that their internal temperature is sufficient to support flight once they leave the hive. It is speculated that, for example, it may be adaptive for a fly to 'sleep in' on a cold, dark morning until the conditions are met for it to warm up sufficiently as to rapidly avoid predation. If cold conditions indeed persist, the new sleep/wake pattern may become further reinforced by stable biochemical/molecular changes and become part of a new seasonal pattern of activity (Alpert, 2020).
Following up on the TPN-II targets, this work also identifies DN1a neurons as a key node for the integration of sensory information with internally regulated drives for rest and activity. DN1as were shown to be not only powerfully and persistently inhibited by cold temperature but also that they have clock-regulated rhythms of activity, respond to light, and receive excitatory drive from sLNvs (which are part of the endogenous pacemaker and are in turn also activated by light). Together, the results demonstrate how information about external conditions (light and temperature) is directly relayed to a circadian/sleep center in the brain and integrated with internal drives to adapt sleep and wake cycles to changing external conditions (Alpert, 2020).
These results open a window on the temporal structure of sensory signaling in the fly thermosensory system and reveal how-even within sensory modality-distinct neural circuits can operate on different temporal scales to drive appropriate behavioral responses (Alpert, 2020).
Animals exhibit innate and learned preferences for temperature and humidity-conditions critical for their survival and reproduction. Leveraging a whole-brain electron microscopy volume, this study examined the adult Drosophila melanogaster circuitry associated with antennal thermo- and hygrosensory neurons. Two new target glomeruli were identified in the antennal lobe, in addition to the five known ones, and the ventroposterior projection neurons (VP PNs) that relay thermo- and hygrosensory information to higher brain centers, including the mushroom body and lateral horn, seats of learned and innate behavior. This paper presents the first connectome of a thermo- and hygrosensory neuropil, the lateral accessory calyx (lACA), by reconstructing neurons downstream of heating- and cooling-responsive VP PNs. A few mushroom body-intrinsic neurons solely receive thermosensory input from the lACA, while most receive additional olfactory and thermo- and/or hygrosensory PN inputs. Furthermore, several classes of lACA-associated neurons form a local network with outputs to other brain neuropils, suggesting that the lACA serves as a hub for thermo- and hygrosensory circuitry. For example, DN1a neurons link thermosensory PNs in the lACA to the circadian clock via the accessory medulla. Finally, this study surveyed strongly connected downstream partners of VP PNs across the protocerebrum; these include a descending neuron targeted by dry-responsive VP PNs, meaning that just two synapses might separate hygrosensory inputs from motor circuits. These data provide a comprehensive first- and second-order layer analysis of Drosophila thermo- and hygrosensory systems and an initial survey of third-order neurons that could directly modulate behavior (Marin, 2020).
Knowing one's own behavioral state has long been theorized as critical for contextualizing dynamic sensory cues and identifying appropriate future behaviors. Ascending neurons (ANs) in the motor system that project to the brain are well positioned to provide such behavioral state signals. However, what ANs encode and where they convey these signals remains largely unknown. Through large-scale functional imaging in behaving animals and morphological quantification, this study reports the behavioral encoding and brain targeting of hundreds of genetically identifiable ANs in the adult fly, Drosophila melanogaster. It is revealed that ANs encode behavioral states, specifically conveying self-motion to the anterior ventrolateral protocerebrum, an integrative sensory hub, as well as discrete actions to the gnathal ganglia, a locus for action selection. Additionally, AN projection patterns within the motor system are predictive of their encoding. Thus, ascending populations are well poised to inform distinct brain hubs of self-motion and ongoing behaviors and may provide an important substrate for computations that are required for adaptive behavior (Chen, 2023).
The brain is the central organizer of food intake, matching the quality and quantity of the food sources with organismal needs. To ensure appropriate amino acid balance, many species reject a diet lacking one or several essential amino acids (EAAs) and seek out a better food source. This study shows that, in Drosophila larvae, this behavior relies on innate sensing of amino acids in dopaminergic (DA) neurons of the brain. The amino acid sensor GCN2 acts upstream of GABA signaling in DA neurons to promote avoidance of the EAA-deficient diet. Using real-time calcium imaging in larval brains, this study shows that amino acid imbalance induces a rapid and reversible activation of three DA neurons that are necessary and sufficient for food rejection. Taken together, these data identify a central amino-acid-sensing mechanism operating in specific DA neurons and controlling food intake (Bjordal, 2014).
All organisms need to sense and adapt to changes in nutrient
levels and nutrient demand. In vertebrates, this is achieved
through close monitoring of available nutrients by sentinel tissues
such as the gut, adipose tissue, and the pancreas, which,
in turn, signals the nutritional status to the brain, ultimately leading
to changes in metabolism and food intake. In the brain, nutrient-sensing neurons also respond directly to fuel-related stimuli like glucose, fatty acids, or amino
acids, engaging neurophysiological responses that control
energy intake. However, the neurochemical
identity of these neurons and the molecular sensors
used are in many instances unknown (Bjordal, 2014).
The complexity of the vertebrate brain presents a challenge to
understand the integration of nutrient signals and the molecular
and cellular mechanisms of neuronal nutrient sensing. A possible
alternative is to use genetically tractable organisms with simpler
brain structures like the fruit fly Drosophila.
Indeed, Drosophila recapitulates many of the hallmarks of
peripheral and central nutrient sensing seen in mammals. At
the periphery, a main nutrient sensor located in fat cells signals
to the brain and controls the release of Drosophila insulin-like
peptides (Dilps). In conditions of nutrient restriction,
the drop in general insulin signaling affects growth of peripheral
tissues as well as the function of specific neuropeptides such as
the Drosophila orthologs of neuropeptide Y, called NPF and sNPF, leading to changes in feeding behavior (Bjordal, 2014).
Recent reports also point to the presence of central nutrient
sensors regulating food intake. Experiments made on tasteless
animals have revealed that mice and flies are able to evaluate
the caloric content of carbohydrates independently of sweet
tasting. Interestingly, the Drosophila fructose receptor Gr43a is expressed
in specific neurons of the adult brain and controls
feeding according to circulating hemolymph fructose. Therefore, central fructose-sensing neurons could represent a new type of sensor for carbohydrates, although the
cellular and molecular mechanisms by which GR43a acts to
regulate food intake remain elusive (Bjordal, 2014).
Besides sugar, adult flies sense changes in dietary amino acid
levels, and a deprivation in amino acid induces a change in their
feeding preference toward amino acids. The downstream effector of the target of rapamycin (TOR) pathway, S6-kinase, and the neurotransmitter serotonin are involved in this regulation.
However, the detailed molecular mechanisms and the cellular
identity of such amino acid sensor are unknown (Bjordal, 2014).
One aspect of amino acid sensing concerns the necessity to
provide essential amino acids (EAAs) that cannot be synthesized
or stored. Earlier experiments in rodents have demonstrated that
animals rapidly evaluate the lack of one essential amino acid in
the food and initiate a series of drastic changes in behavioral
strategies, starting with food avoidance. Injection of imbalanced
amino acid mixes in defined areas of the rodent brain is sufficient
to trigger a reduction in food intake, suggesting that the sensor
for EAA deficiency (EAAD) is located in the brain. Additionally, mice with a mutation in the gene encoding the conserved GC nonderepressing 2 (GCN2) kinase do not
reject the imbalanced diet, indicating a role for this cell-based amino acid sensor in
triggering the EAAD response. The neural circuitry involved in this behavior remains
uncharacterized (Bjordal, 2014).
This study has identified a neural circuitry involved in amino acid
sensing and the control of feeding behavior in the Drosophila
larval brain. Drosophila larvae were shown to reduce food intake
when encountering an EAAD, and it was demonstrated that amino acid
sensing takes place in a limited number of dopaminergic (DA)
neurons. Calcium imaging in live brain show that DA cells are
rapidly and reversibly activated by EAAD in a GCN2-dependent
manner. Finally, using tissue-targeted genetic loss and gain-of-function
tools, it was demonstrated that EAAD-induced food avoidance
involves a GCN2-dependent inhibition of GABA signaling
in dopaminergic neurons. This demonstrates the existence of a
dopaminergic circuitry providing homeostatic control on feeding
through a central amino acid sensing mechanism (Bjordal, 2014).
Drosophila larvae reduce their food intake on EAAD diet. This behavior does not rely on smell or taste because it can be specifically mimicked or suppressed by interfering with amino acid sensing in the brain. In addition, ex vivo brain imaging demonstrates that DA
cells directly and rapidly activate in response to EAAD. The fast kinetics of
the response observed in DA cells suggests that uncharged tRNA levels are
instantly linked to variations in intracellular amino acid concentrations and translated into changes in GCN2 activity. GCN2 activation leads to several cellular
responses, including a block in translation initiation through eIF2a phosphorylation
and the consequent activation of a specific transcription program in which the ATF4 transcription
factor plays a key role. This study demonstrates that dATF4 in DA cells is required for the rejection of EAAD food. Given the very fast kinetics of neuronal
activation, it is unlikely that transcription participates in acute
EAAD avoidance. Using genetic interactions and ex vivo
calcium imaging, it was shown that EAAD-induced feeding inhibition
requires the repression of GABA signaling by dGCN2
activation in DA cells. In addition, bioluminescence resonance energy transfer (BRET) analysis demonstrates that ATF4 and GABA(B)R1 directly interact in living cells. These data are supported by observations made in
rodents indicating that suppression of GABAergic inhibition
contributes to EAAD-induced food avoidance. A model is proposed whereby, in response to EAAD, activation of dGCN2 induces dATF4-mediated
GABA signaling inhibition, dopamine release, and food rejection (Bjordal, 2014).
TOR signaling couples amino acid availability with the systemic
control of growth in fat body cells and ecdysone production
in the larval ring gland. Interestingly, TOR inhibition in DA cells does not attenuate EAAD-induced food avoidance. Similarly, rapamycin
injection in the antero-piriform complex of rodent brain does
not alter EAAD-induced feeding inhibition, supporting
the notion that GCN2, but not TOR signaling, is the
sensor for EAAD response. How independent these pathways
are is still an open question. Work in yeast suggests that TOR
acts upstream of GCN2. Such functional epistasis has not been established in
metazoan cells, and the present data suggest that the two pathways
operate independently in vivo (Bjordal, 2014).
Not all DA neurons are activated by EAAD. Using live imaging,
this study repeatedly observed that the DM1 and DL1 cluster, but not
the DL2 cluster, are activated by EAAD. Interestingly, this cluster
was recently implicated in olfactory reward-driven feeding, indicating subfunctionality among different dopamine circuits. Using subtraction analysis, it was possible to show that only three neurons in the DL1 cluster are responsible for EAAD-induced food avoidance. Nevertheless, additional DA
cells are activated by EAAD, suggesting that they could contribute
to other EAAD-induced behaviors. Indeed, EAAD induces
long-term effects such as the development of a learned aversion to a deficient
or imbalanced food and memory for the place associated with EAAD food. Hence, activation
of other DA cells by EAAD may contribute to these additional behaviors (Bjordal, 2014).
This work demonstrates a direct role of DA in nutrient sensing and food rejection
in flies; however, its role in aversive learning is well established. The activity
of the PPL1 cluster of DA neurons in the adult fly brain can produce aversive
memory when paired with an odor. Distinct DA neurons in the PPL1 cluster provide
motivational control over memory expression, suggesting that
these neurons constitute a dopaminergic circuitry that regulates
the internal motivational state of hunger and satiety. Direct lineage tracing remains to be done, but the
D0 and C1 Gal4 drivers targeting the subdomains of the larval
DL1 cluster also target the PPL1 cluster in the adult brain, suggesting that DL1 and PPL1 cells may be related.
Therefore, DA signaling in the DL1/PPL1 cluster could act as a
general satiation signal, reducing food intake and abolishing
appetitive performance (Bjordal, 2014).
The dopaminergic circuitry is known for its role in the motivational
control of feeding. This study has shown that it also plays a key
role in the homeostatic regulation of food intake. In light of recent
studies showing that metabolic hormones also exert their effect
on the dopamine reward circuit, the emerging picture is that dopamine
is a central player in the regulation of food intake through
the integration of nutrient sensing and motivational drives (Bjordal, 2014).
Animals exhibit innate behaviours to a variety of sensory stimuli including olfactory cues. In Drosophila, one higher olfactory centre, the lateral horn (LH), is implicated in innate behaviour. However, structural and functional understanding of the LH is scant, in large part due to a lack of sparse neurogenetic tools for this region. This study generated a collection of split-GAL4 driver lines providing genetic access to 82 LH cell types. These were used to create an anatomical and neurotransmitter map of the LH and link this to EM connectomics data. It was found that ~30% of LH projections converge with outputs from the mushroom body, site of olfactory learning and memory. Using optogenetic activation, this study identify LH cell types that drive changes in valence behavior or specific locomotor programs. In summary, this study has generated a resource for manipulating and mapping LH neurons, providing new insights into the circuit basis of innate and learned olfactory behavior (Dolan, 2019).
Previous work has classified LH cell types with either electrophysiology or calcium imaging. However, due to the lack of sparse driver lines, specific genetic access was not possible for most LH cell types. Two recent studies have identified more driver lines labelling LH neurons but these are broad and suitable mostly for electrophysiology and dye-filling identification of individual cells. This study has generated split-GAL4 lines that enable anatomical analyses and specific neurogenetic control of LH cell types. These reagentswere used to generate an atlas of identified LH neurons, defined their major neurotransmitter and polarity; classifying most cell types into LHONs, LHLNs and LHINs (Output, Local and Input)). Connectomic analysis of neural circuits in EM volumes is providing an unprecedented window into the structure of neural circuits, especially in Drosophila. To facilitate connecting these genetic tools with EM data this study has provided traced backbones of neurons from many identified LH cell types in a whole-brain EM volume, in addition to light-level single-neuron labelling. These resources will allow future studies to correlate functional and connectomic data to determine how the LH generates olfactory behaviour (Dolan, 2019).
This study used driver lines to identify major output zones and convergence sites of LH neurons. The diversity of inter- and intra-regional connections demonstrates the complexity of the LH and the innate olfactory circuitry, even in the relatively 'simple' brain of Drosophila. In terms of output, anatomical analysis identified the superior lateral protocerebrum (SLP) as the next major site of olfactory processing, although LHONs projected to many other brain regions, with a broader range than MBONs. Across identified cell types, no LHONs were identified that project to the ventral nerve cord, suggesting at least one more layer of processing before motor output. Within the LH, previous studies characterized one population of GABAergic local neurons. Other than these and MB associated neurons however, the neurotransmitter profiles of neurons in the protocerebrum, were unknown. This study identified populations of LH neurons that were cholinergic, GABAergic and/or glutamatergic. LHONs, LHINs and LHLNs were clustered into groups based on the sign of their neurotransmitter. Several distinct populations of both GABAergic and glutamatergic cells were found that were local to the LH, some potentially interacting with many LHON dendrites. This indicates that both lateral excitation and inhibition may exist in the LH. It was also found that the LH integrates visual, gustatory, thermosensory and mechanosensory input in a restricted, ventral region (Dolan, 2019).
The LH and MB are thought to mediate innate and learned behaviour respectively, yet their interactions remain little understood. Several MBONs project to the LH and this study has identified potential downstream targets. Indeed, one prediction (connectivity from MBON-α2sc to PD2a1/b1 neurons) has been validated and shown to be necessary for innate and learned behaviour (Dolan, 2018). In addition this study found that AD2b2 neurons extend their dendrite outside the LH and receive MB input. Previous studies have identified three instances of MB and LHON axonal convergence, however it was unclear if this was the case for all LHONs. This study systematically compared overlap for all the LHONs identified in the split-GAL4 screen. ~30% of LHONs were found to converge with MB-associated neurons, several of which appear placed to interact with more than one DAN and/or MBON. Therefore while this is a significant circuit motif, it does not explain all LHON projections. AD1b2 interact extensively with three MBONs in a unique axoaxonic integration motif, and stimulation of AD1b2 neurons drives approach behaviour when stimulated. Although the MB and LH both receive input from the AL, the coding logic of these two regions is strikingly different and the new tools described in this study will greatly facilitate studies of how these regions together orchestrate behaviour (Dolan, 2019).
To apply these reagents experimentally, cell type specific optogenetic activation were used, and several LH neurons were identified whose activation drives behaviour. Two cell types (one LHON and one LHLN) were identified that drive aversion, one LHON which drives attraction and several LHONs whose stimulation leads to changes in motor behaviours. The optogenetic screen identified only 3/50 LH cell types that could alone consistently drive changes in valence. This was a lower proportion than in the MB, where 6/20 MBON cell types tested were implicated solely in attraction and aversion rather than specific motor parameters (Dolan, 2019).
In total, the split-GAL4 lines covered 82 different cell types in the LH, although 24 of these were consistently colabeled in split-GAL4 lines. In addition, five anatomical cell types (PV4a1:5) occur in overlapping but disjoint split-GAL4 lines. Anatomical analysis of single neurons labeled with either stochastic genetic labelling or dye fills during recording identified ~165 different 'core' cell types in the LH (Frechter, 2019), that is cell types that seemed to have a majority of their dendrite within the LH and in the vicinity of olfactory uniglomerular PNs. Core LH cell types excluded LHINs and any neurons with a low fraction of their dendritic arbor overlapping the PN terminals bounded by the LH (Frechter, 2019). To calculate coverage of this study compared to these datasets, the same definition was used to categorize LH cell types identified in split-GAL4 lines as core or non-core. For LHON and LHLN cell types (68 in total), 63 were classified as core LH cell types. Based on this definition, the split-GAL4 collection covers 63/165 or ~38% of known core LH cell types (Dolan, 2019).
For the five cell types not defined as core, connectomic analysis indicates that they receive substantial input from multiglomerular PNs and/or some level of input from canonical uniglomerular PNs (Dolan, 2019).
Although this 38% is lower than estimates of coverage for the MB, which exceeds 80%, it is similar to the estimated coverage achieved to date (30-50%) for split-GAL4 lines targeting descending interneurons from the brain to VNC (Namiki, 2018). There are several possible hypotheses for these differences. The first is simply that there appear to be fewer MB cell types, for example 23 MBONs, than LH or descending interneurons. The second is that the dense, compartmental arborization of most MB neurons makes them easier to identify in lines containing many cell types, which are the starting point for split-GAL4 screens. Finally, it may well be that neurons in different brain regions have more or less distinct transcriptomic and epigenetic profiles and that this impacts the genetic isolation of their constituent cells. It is also emphasized that full and comprehensive reconstruction of all these neuronal classes from whole brain EM data will likely reveal additional cell types not discovered in these annotations of genetic data (Dolan, 2019).
This split-GAL4 driver line resource will now allow investigators to manipulate many LH neurons with cell type specificity. Driver lines can be used for imaging, functional activation or silencing experiments with genetically-encoded optogenetic or thermogenetic tools. However, the final expression pattern of any driver line also depends on the insertion site of the effector transgene; expression patterns should be verified for effectors at locations other than attP18 (the site used during this screening) (Dolan, 2019).
In the course of this work this study has produced large-scale anatomical maps of overlap between different populations of LH and MB neurons. This is a rapid approach to generate hypotheses for synaptic connectivity and circuit motifs. However, all potential synaptic interactions are subject to both registration error and biological variability across different brains. While several highly overlapping pairs with EM reconstruction were validated, even neurons that are very close in space may not be synaptically connected and this must be confirmed with connectomics or physiology (Dolan, 2019).
During behavioural studies, this study focused only on phenotypes that were consistent across multiple 'ideal' split-GAL4 lines for the same cell type. Indeed one of the four PD2a1/b1 lines tested had an aversive phenotype when activated, contrary to previous data (Dolan, 2018). This is most likely due to off-target expression in the VNC, as described (Dolan, 2018) and it is noted that stimulation of PD2a1/b1 neurons, using a different split-GAL4 line, in flying animals drove attraction. Despite this, not all differences between split-GAL4 lines ostensibly targeting the same cell type should be discarded. Some of the differences in phenotype between split-GAL4 lines targeting the same cell type could be due to differences in expression strength or simply experimental variation during a large-scale screen (that was not designed to identify small effect sizes). Even within an anatomically and genetically defined cell type, neurons can have different connectivities (Dolan, 2018). Therefore it is quite possible that this study has discounted viable phenotypes in this analysis and subsequent replication experiments. Finally, it also possible that the relatively low number of phenotypes in optogenetic activation screen is because activation of just one of >150 LH cell types is often too selective to produce strong behavioural consequences, especially in the absence of an olfactory stimulus. Future studies will be able to combine split-GAL4 lines to control multiple LH cell typesor use transsynaptic labelling to control large populations postsynaptic to specific cell types (Dolan, 2019).
This paper has directly demonstrated the existence of a diversity of genetically defined cell types with highly stereotyped dendritic arbours in the LH. Additionally, it was demonstrated that stimulation of LH neurons can drive valence or motor behaviours. While gain-of-function experiments do not alone demonstrate the role of the LH in innate behaviour, when combined with published data there is now clear anatomical, functional and behavioural evidence that the LH mediates instinctive olfactory responses. Firstly, anatomical and functional experiments demonstrate stereotyped connectivity between PNs and LH cell types between animals, implying a role in the generation of innate behaviour in contrast to the nearly random connectivity between PNs and the MB. Neuronal silencing experiments demonstrate that abolition of KC neurotransmission leads to a silencing of memory and a reset to innate olfactory responses . In addition to this work, two contemporaneous studies from the Jenilia group has used a subset of the new split-GAL4 lines to interrogate the role of this region in specific learned and innate olfactory behaviours. While Dolan (2018) shows that PD2a1/b1 LHONs are required for both innate attraction and aversive memory recall, the memory recall phenotype seems best understood as the modulation by a mushroom body output pathway of a hardwired pathway required for naive behaviour (Dolan, 2019).
Given the extensive interactions between LH and MB identified in this study it is proposed that these 'horizontal' pathways (e.g., MBONs projecting to LH, LHON to DAN) orchestrate additional functions such as memory retrieval, provide categorical information and MB modulation. It is suggested that LH-to-MB information flow (e.g, via LHON-to-DAN synapses) may also explain why some MBONs exhibit valence-specific responses, a hypothesis which can now be tested using the split-GAL4 lines developed in this study (Dolan, 2019).
The tools described in this study provide a critical resource for cell type specific dissection of the LH. Stimulation of AV1a1 and PV4a1:5 drive avoidance while silencing AV1a1 abolishes the response to geosmin (an ecologically relevant aversive odour) in an oviposition assay (Chen, 2018). This suggests AV1a1 LHONs may be a major pathway for ethologically-relevant aversion in the fly brain and that PV4a1:5 interneurons may pool other inputs to drive aversion via this pathway (Dolan, 2019).
This study also found that AD1b2 LHONs drive approach behaviour when activated. These cholinergic neurons project axons to the Superior Medial Protocerebrum (SMP) and have a distinctive additional dendritic projection extending out to the SMP/SIP. AD1b2 integrates MBON input in both an axodendritic and axoaxonic manner, including two MBONs which bidirectionally drive valence behaviour. Interestingly, previous work has identified another LHON (PD2a1/b1) involved in innate attraction. PD2a1/b1 also received MBON input onto both its dendrite and axon, and both AD1b2 and PD2a1/b1 receive axonal input from MBON-α'2 a memory-relevant MBON. These parallels suggest both a shared circuit motif and a relation between (innate) approach behaviour and memory (Dolan, 2019).
In general however, naive olfactory responses are more diverse than attraction or aversion, and olfactory stimulation can drive changes in locomotion/flight speed, stopping and turning or exploratory behaviour. This study identified one LH cell type which drives forward locomotion when stimulated, but does not impact valence behaviour. Much olfactory sensation occurs during flight and by recording the wingbeat responses of flying Drosophila before and after optogenetic stimulation it was possible to identify LH neurons that modulate different parameters (turning, wing thrust and wingbeat frequency) in flying animals. Interestingly, the effects of stimulating these neurons had different impacts on behaviour depending on visual stimulus (see multimodal integration below) while the persistence of these effects post-stimulation also varied. Therefore, different LH neurons likely drive different motor programs at various timescales. This diversity in downstream functions may explain why the LH has a large number of cell types which show distinct but frequently overlapping odour responses or pool different PN channels. Behavioural responses to odours (e.g. exploratory behaviour) may be a composite of different motor programs (e.g. locomotion increase, decrease turning) where each program is driven by a small assembly of different LH cell types. However, this study found only one LH neuron projecting to the sensorimotor integration circuitry of the central complex. Taken together these results imply that LH output is not immediately sent to motor neurons, likely it is integrated downstream with other information such as memory or internal state. This model is supported by the observation of specific impacts on motor behaviour by manipulations of different olfactory sensory neurons, although such manipulations also impact downstream MB circuits (Dolan, 2019).
Extensive overlap was found between a minority of LHONs and MB neurons, and an additional function of the LH may be to modulate and monitor the distributed coding in the MB. LHONs may modulate or even implant memories via DANs and the axoaxonic integration of MBON inputs by LHONs may occur more generally (Dolan, 2018). Higher olfactory neurons downstream of MB and LH likely read out olfactory information from these two different coding regimes to integrate decorrelated and hardwired olfactory representations (Dolan, 2019).
In addition to olfactory responses, the data also demonstrates the LH integrates multimodal input. This non-olfactory sensory information could provide context for the innate olfactory processing system which may be relevant for naive responses to odours or for pheremonally-driven behaviours such as courtship or aggression. Multimodal sensory input may also provide a route for direct integration of context and olfactory stimulation. Indeed, it was found the effects of optogenetic stimulation of LH neurons differed depending on the visual stimulation displayed to flying animals (Dolan, 2019).
In future experiments, the split-GAL4 tools described in this study will help investigators to determine the exact functions of the LH during olfaction at single cell type resolution. While the functions of individual LH neurons can range from aversion to memory retrieval, many LH neurons do not appear to be solely involved in avoidance or attraction. Higher-resolution behavioural assays will be needed. These will identify the subtle olfactory responses beyond avoidance or attraction and can be combined with connectomic information to design precise loss-of-function experiments (Dolan, 2019).
Both the mammalian and insect olfactory systems split into two parallel processing tracks, one hardwired or chemotropic and one distributed or random with functional roles in innate and learned behaviour respectively. This striking similarity suggests it may be able to obtain important and general insights into how olfactory perception is transformed into action using the simpler nervous system of Drosophila. Using the split-GAL4 lines generated in this study it was possible to provide a map of the inputs and outputs of this region, identifying many previously unknown features, such as multimodal input and extensive interactions with the MB. In addition this study has identified LH neurons that are sufficient to drive valence behaviour or specific motor movements, providing the first functional evidence that this hardwired region can direct diverse olfactory responses. This work represents a step towards a full model of how olfactory stimuli are processed in the Drosophila brain, from sensory input to motor output (Dolan, 2019).
The Drosophila brain contains about 50 distinct morphological types of dopamine neurons. Physiological studies of Drosophila dopamine neurons have been largely limited to one brain region, the mushroom body, where they are implicated in learning. By comparison, little is known about the physiology of other Drosophila dopamine neurons. Interestingly, a recent whole-brain imaging study found that dopamine neuron activity in several fly brain regions is correlated with locomotion. This is notable because many dopamine neurons in the rodent brain are also correlated with locomotion or other movements; however, most rodent studies have focused on learned and rewarded behaviors, and few have investigated dopamine neuron activity during spontaneous (self-timed) movements. This study monitored dopamine neurons in the Drosophila brain during self-timed locomotor movements, focusing on several previously uncharacterized cell types that arborize in the superior-lateral brain, specifically the lateral horn and superior-lateral protocerebrum. It was found that activity of all of these dopamine neurons correlated with spontaneous fluctuations in walking speed, with different cell types showing different speed correlations. Some dopamine neurons also responded to odors, but these responses were suppressed by repeated odor encounters. Finally, the same identifiable dopamine neuron encode different combinations of locomotion and odor in different individuals. If these dopamine neurons promote synaptic plasticity-like the dopamine neurons of the mushroom body-then, their tuning profiles would imply that plasticity depends on a flexible integration of sensory signals, motor signals, and recent experience (Marquis, 2022).
Most sensory systems are organized into parallel neuronal pathways that process distinct aspects of incoming stimuli. In the insect olfactory system, second order projection neurons target both the mushroom body, required for learning, and the lateral horn (LH), proposed to mediate innate olfactory behavior. Mushroom body neurons form a sparse olfactory population code, which is not stereotyped across animals. In contrast, odor coding in the LH remains poorly understood. This study combine genetic driver lines, anatomical and functional criteria to show that the Drosophila LH has ~1400 neurons and >165 cell types. Genetically labeled LHNs have stereotyped odor responses across animals and on average respond to three times more odors than single projection neurons. LHNs are better odor categorizers than projection neurons, likely due to stereotyped pooling of related inputs. These results reveal some of the principles by which a higher processing area can extract innate behavioral significance from sensory stimuli (Frechter, 2019).
The principal finding of this study is that LH neurons (LHNs) as a population are genetically and anatomically defined cell types with stereo- typed odor responses. Starting from recordings of genetically defined populations, this study cross-validated fine scale anatomical differences and odor tuning for 37 LHN cell types; this confirms that stereotypy is a general feature of the LH and not particular to specialist odor pathways such as those that process pheromone information, which may retain a labeled line logic all the way from the periphery. Although evidence is seen of narrowly tuned LHNs dedicated to the processing of specific odors, the population as a whole shows 3x more odor responses than their PN inputs. The increased tuning breadth may reflect a transition to a more behaviorally relevant coding scheme. This is consistent with the findings that LHNs show significantly improved odor categorization compared with PNs, apparently due to stereotyped pooling of related odor channels. The chemical categories that were analyzed are probably not of direct ethological relevance to the fly, but serve as proxies: further explorations of olfactory neuroecology are clearly necessary. For example limited evidence was seen for simple representations of olfactory valence in LHN responses. It is instructive to compare the odor tuning properties found in this study across the lateral horn with those reported for the Drosophila mushroom body. Major differences in the MB include the lack of response stereotypy and sparser odor tuning; the distribution of odor tuning in the LH also appears to be wider - i.e. LHNs appear more functionally heterogeneous. However, there are also similarities - there is divergence of PNs onto a larger population of third order neurons in both cases. Furthermore baseline firing rates are very low in both LHNs and Kenyon cells and the evoked firing rates are also lower than in their PN input. This could reflect energetic, spike economy considerations or a need to binarize neural responsesprior to memory formation or organizing behaviors (Frechter, 2019).
It is also interesting to compare response properties with recent recordings from the mammalian posterolateral cortical amygdala, which has been compared to the LH, since it receives spatially stereotyped input from the olfactory bulb and is required for innate olfactory behaviors. It has been found that odor tuning properties were very similar to the mammalian piriform cortex (which has been compared to the mushroom body). Both regions showed decorrelated odor representations (whereas LHN odor responses were found to show significant correlations suggestive of a focus on particular combinations of olfactory channels) and odor tuning in the cortical amygdala was actually somewhat sparser. In further contrast to the current observations in the LH no evidence was found for categorization of odors by chemical class and crucially no evidence for response stereotypy in a way suggestive of stereotyped integration of defined odor channels. Caution should be applied with respect to the last point; had this study recorded from a small fraction of randomly selected neurons of the Drosophila LH, response stereotypy might easily have been missed. It is only because it was possible to use genetics to bias the sampling, and also to record from
a significant fraction of the whole LH population, that it was possible to obtain clear evidence for odor response stereotypy. Nevertheless, these differences seem marked and it will be very interesting to compare the logic of these systems across organisms. One point to note is that the circuits in the fly may be more compact: LHNs can in a few cases connect directly to fourth order neurons with descending projections to the nerve cord likely to have a direct impact on motor behavior (Frechter, 2019).
There are some similarities between the increase in tuning breadth that was observed at the PN-LHN transition and what has previously been reported at the first synaptic layer of the olfactory system (the olfactory receptor neuron to PN synapse). In the antennal lobe broadening appears to depend on a compressive non-linearity, which boosts weaker inputs and possible excitatory local interactions. Although a direct comparison between the extent of broadening in the antennal lobe and LH is not possible without measuring odor responses from many receptor neurons under the same stimulus conditions (as waa done for PNs and LHNs) it seems likely that the effect is larger in the LH. Importantly the mechanism here appears quite different, with direct pooling of feed-forward inputs (Frechter, 2019).
The initial EM connectomics observations suggest that a typical LHON receives strong inputs from 3-7 excitatory PNs albeit with a long tail of weaker connections, some of which are likely to have an impact. Intriguingly this number (referred to as the synaptic degree, K) is not that different
from the 7 inputs reported for Kenyon cells in the mushroom body. How is it that LHONs and KCs listen to rather similar information but produce very different responses? It is true that the inputs received by LHNs will in general be more highly correlated; this is both because LHNs appear to receive input from all the PNs originating from a given glomerulus (when there are >1) and because those PNs coming from different glomeruli often have related odor tuning. Nevertheless, it is argued strongly that the rules of integration that result in broadening in LHONs and a sharp reduction in tuning breadth in KCs are likely to differ significantly. It has been shown that LHON firing rates scale linearly with their PN inputs, while it has been shown that KC membrane potential linearly integrates dendritic inputs. Differences in the integrative properties could result from both intrinsic and circuit mechanisms (i.e., local interneuron interactions), but two factors likely to have a major impact are the spatial distribution of synapses and the spike threshold. PN inputs are broadly distributed across LHON dendrites, whereas PN inputs onto KCs are highly clustered at individual dendritic claws. The many individual connections at each KC claw may be integrated to produce a reliable response that is nevertheless usually below the spike threshold. Therefore multiple input PNs must be co-active and KCs act as coincidence detectors. In contrast the inputs on LHON dendrites may be integrated in a more graded fashion with a lower spike threshold. Of course the biggest difference is that LHNs receive stereotyped inputs according to their anatomical/genetic identity and this provides a mechanism for the odor response stereotypy that was observe (Frechter, 2019).
Some additional differences in circuit architecture between the MB and LH are highlighted that may be of functional significance. First the MB calyx receives only excitatory PN input, whereas, there is a population of almost 100 inhibitory PNs that project to the LH . Second this study found that the LH contains an estimated 580 local neurons (most of which are inhibitory, whereas the
mushroom body contains just one local inhibitory neuron, the APL. It is suspect that a major reason for this difference is again related to the stereotyped vs non-stereotyped design of these two centers (Frechter, 2019).
The APL is not selective but appears to pool all KC inputs to implement a winner take all gain control mechanism, suppressing more weakly activated KCs. The preliminary EM results show that at least some LHLNs integrate small numbers of input channels (2-3 strong inputs). It is suggested that they then make stereotyped connections either reciprocally onto their input PNs or onto other specific neurons in the LH (Frechter, 2019).
There is renewed interest in the identification of cell types in the
brain as an important step in the process of characterizing circuits and behavior. Historically, cell types have been best classified by morphology and the most detailed work has been in the sensory periphery (e.g., 55 cell types in the mouse retina). Recently single cell transcriptomics has begun to match this morphological classification and also to enable more detailed exploration of diversity in deeper brain regions (e.g., 133 cell types in mammalian cortex: Tasic et al., 2018). However, relating cell types to functional and network properties especially
in higher brain areas remains challenging (Frechter, 2019).
One of the major surprises from this work is the identification
of 165 anatomically distinct LHN cell types; cross-validation of anatomical and odor response properties for 37 leads to belief that most of these will turn out to be functionally stereotyped as well. Furthermore the light level survey is incomplete; it is predicted that complete EM data could reveal more than 250 LHN cell types. In short there are more cell types in the lateral horn than have yet been identified in the whole of the mammalian neocortex. This disparity raises a number of issues (Frechter, 2019).
One interesting observation is that it was easier to identify cell types anatomically than by odor response profile alone. It has recently proven possible to characterize 30 retinal ganglion cell types in the mouse based solely on their visual response properties. It may be that this highlights a difference between the richness of achievable visual stimulation protocols with odor delivery; although the core 36 odor set was large by the standards of the field, this is still a small fraction of the world of possible odors for the fly. Nevertheless there appear to be many more LHNs than retinal ganglion cell types and examples of were found of neurons that appear to be solely distinguished by their projection patterns (presumably defining different downstream partners) which are only revealed through anatomical characterization. For these reasons it is believed that response properties alone are insufficient to define cell type and this seems likely to be the case in other higher brain areas (Frechter, 2019).
Initial evidence from EM connectomics has shown that two specific LHN cell types integrate stereotyped sets of olfactory channels with similar odor response profiles. This is paralleled by the recent work of that showed that morphologically similar neurons sampled from the same or different GAL4 lines showed similar functional connectivity; furthermore they showed that the patterns of co-integration were not random, but that certain pairings of PN inputs were overrepresented in the PN population. These observations are likely to be at the heart of the category selectivity that was observe in LHON responses. It will be exciting to integrate functional and anatomical properties more deeply with circuit properties. Furthermore genetic screening identifies at least 69 molecular profiles based on expression of driver lines. This molecular diversity underlies the ability to generate cell type specific split-GAL4 lines. The existence of such a rich and coupled genetic and anatomical diversity raises interesting questions about how connection specificity can beachieved during development in this integrative brain area (Frechter, 2019).
The lateral horn is one of two major olfactory centers in the fly. The hypothesis that it might play a specific role in unlearned olfactory behaviors dates back almost 20 years. This has been strengthened by observations about the relative anatomical stereotypy of input projections to the mushroom body and lateral horn. Nevertheless in spite of this general model of a division of labour between LH and MB, functional evidence has been hard to come by. Some arguments about LH function have been based on experiments that manipulate mushroom body neurons; here it is worth noting that there are olfactory projections neurons that target areas outside of these two principal centers so the lateral horn cannot rigorously be concluded to mediate behaviors for which the mushroom body appears dispensable (Frechter, 2019).
In this experimental vacuum a large number of hypotheses
have been proposed for LH function. One obvious suggestion based on anatomy was that LHNs should integrate across olfactory channels. Of course integration can have opposing effects on tuning. For example it has been proposed that LHNs might have highly selective odor responses and early recordings from narrowly tuned pheromone responsive neurons are consistent with this idea. However another study also observed more broadly tuned neurons that clearly integrated across olfactory channels, and quite linear integration of two identified olfactory channels has been shown. Electrophysiological recordings together with first EM connectomics results suggest that integration across multiple odor channels and broadening of odor responses are the norm (Frechter, 2019).
Turning to the biological significance of LHNs for the fly, one
suggestion, based on anatomically discrete domains for food and pheromone odors, is that the LH might organize odors by behavioral significance. Other studies have suggested that the LH might mediate innate responses to repulsive odors only or that the LH might organize odor information by hedonic valence. Although the current survey of LHN odor responses is not yet conclusive on any of these points, clear evidence was found for an improved ability to categorize chemical groups of odorants. Further work integrating more information about the behavioral significance of different odors should be instructive (Frechter, 2019).
One synthesis of these different ideas is that the mushroom bodies perform odor identification, whereas the lateral horn/protocerebrum performs odor evaluation, both learned and innate. Although there is no evidence to support a direct role for the LH in evaluation of learned olfactory signals, new work from has identified a class of lateral horn neurons that integrates both innate (directly from the antennal lobe) and learned olfactory information (from MB output neurons) of specific valence; these LHNs are required for innate appetitive behavior as well as learned aversive recall. This study also identified multiple LHN axon terminals as targets of mushroom body output neurons, suggesting that mushroom body modulation of innate olfactory pathways may be a general strategy of learned behavioral recall. These results emphasize the extensive interconnection between these brain areas and should caution against oversimplifying their distinct roles in olfactory behavior. Nevertheless synthesizing the results in this study with other new work does support the hypothesis that stereotyped integration in the LH could support genetically determined categorical odor representations, while the MB may enable identification of specific learned odors (Frechter, 2019).
A key question posed at the start of the manuscript is why does the LH need so many cells and cell types? At this stage it is suggested that LHNs are likely to show both stereotyped selectivity for odor categories and specificity for different aspects of odor-guided behavior. Specific combinations of the same odor information could be used to regulate distinct behaviors by targeting different premotor circuits. Indeed a requirement of a specific LHN cell type (AV1a1) has been found in egg-laying aversion to the toxic mold odorant geosmin even though this is one of more than 70 cell types that receive geosmin information from olfactory PNs within the LH. The picture that this paints is of a complex switchboard for olfactory information with many more outputs than can yet be understood (Frechter, 2019).
It seems likely that different paths for information flow through the LH may be modulated by external signals such as the internal state of the animal. The next few years should see very rapid progress in understanding the logic of circuits within the LH and their downstream targets through the impact of connectomics approaches combined with the anatomical and functional characterization and tool development. In conclusion, the Drosophila lateral horn now offers a very tractable model to understand the transition between sensory coding and behavior (Frechter, 2019).
The Insulin signaling pathway couples growth, development and lifespan to nutritional conditions. This study demonstrates a function for the Drosophila lipoprotein LTP (FlyBase term: Apolipoprotein lipid transfer particle) in conveying information about dietary lipid composition to the brain to regulate Insulin signaling. When yeast lipids are present in the diet, free calcium levels rise in blood brain barrier (BBB) glial cells. This induces transport of LTP across the Blood Brain Barrier by two LDL receptor-related proteins: LRP1 and Megalin. LTP accumulates on specific neurons that connect to cells that produce Insulin-like peptides, and induces their release into the circulation. This increases systemic Insulin signaling and the rate of larval development on yeast-containing food compared with a plant-based food of similar nutritional content (Brankatschk, 2014).
Nutrient sensing by the central nervous system is emerging as an important regulator of systemic metabolism in both vertebrates and invertebrates. Little is known about how nutrition-dependent signals pass the blood brain barrier to convey this information. Like the vertebrate BBB, the BBB of Drosophila forms a tight barrier to passive transport, and is formed by highly conserved molecular components. Its simple structure and genetic accessibility make it an ideal model to study how nutritional signals are communicated to the CNS. Insulin and Insulin-like growth factors are conserved systemic signals that regulate growth and metabolism in response to nutrition. Although fruit flies do not have a single pancreas-like organ, they do produce eight distinct Drosophila Insulin/IGF-like peptides (Dilps) that are expressed in different tissues. A set of three Dilps (dILP2,3,5), released into circulation by Dilp-producing cells (IPCs) in the brain, have particularly important functions in regulating nutrition-dependent growth and sugar metabolism; ablation of IPCs in the CNS causes diabetes-like phenotypes, slows growth and development, and produces small, long-lived adult flies. Systemic Insulin/IGF signaling (IIS) increases in response to dietary sugars, proteins and lipids. Sugars act on IPCs directly to promote Dilp release, but other nutrients are sensed indirectly through signals from the fat body (an organ analogous to vertebrate liver/adipose tissue) (Brankatschk, 2014).
The Drosophila fat body produces two major types of lipoprotein particles: Lipophorin (LPP; Retinoid- and fatty acid-binding glycoprotein), the major hemolymph lipid carrier, and Lipid Transfer Particle (LTP). LTP transfers lipids from the intestine to LPP. These lipids include fatty acids from food, as well as from endogenous synthesis in the intestine. LTP also unloads LPP lipids to other cells (Van Heusden, 1989; Canavoso, 204; Parra-Peralbo, 2011). LPP crosses the BBB and accumulates throughout the brain. It is required for nutrition-dependent exit of neural stem cells from quiescence (Brankatschk, 2010). This study investigated possible functions of LTP in the brain (Brankatschk, 2014).
This work demonstrates a key requirement for lipoproteins in conveying nutritional information across the BBB to specific neurons in the brain. As particles that carry both endogenously synthesized and diet-derived lipids, lipoproteins are well-positioned to perform this function. The data suggest that transport of LTP across the BBB to Dilp2-recruiting neurons (DRNs) influences communication between DRNs and the Dilp-producing IPCs, increasing the release of Dilp2 into circulation. Since the IPCs also deliver Dilp2 to the DRNs, this indicates that these two neuronal populations may communicate bidirectionally. How might LTP affect the function of DRNs? One possibility is that it acts to deliver a signaling lipid to the DRNs. It could do so either directly, or indirectly by promoting lipid transfer from LPP, which is present throughout the brain. LTP enrichment on specific neurons may increase lipid transfer to these cells (Brankatschk, 2014).
This work highlights a key function for BBB cells in transmitting nutritional information to neurons within the brain. Feeding with yeast food increases free calcium in BBB glia, which then increases transport of LTP to DRNs. How might BBB cells detect the difference between yeast and plant food? The data suggest differences in the lipid composition of yeast and plant-derived foods are responsible. Previous work has shown that the lipids in these foods differ in their fatty acid composition. Yeast food has shorter and more saturated fatty acids than plant food (24). How could these nutritional lipids affect the activity of BBB glia? Interestingly, differences in food fatty acid composition are directly reflected in the fatty acids present in membrane lipids of all larval tissues including the brain. Thus, it is possible that the bulk membrane properties of BBB glia are different on these two diets. Membrane lipid composition is known to affect a variety of signaling events. Alternatively, yeast food may influence the specific fatty acids present in signaling lipids that activate BBB glia (Brankatschk, 2014).
This study demonstrates an unexpected functional specialization of the BBB glial network, which permits specific and regulated LTP transport to particular neurons. How this specificity arises is an important question for the future. It is noted that a subset of glial cells within the brain also accumulates LTP derived from the fat body. Could these represent specific transport routes from the BBB (Brankatschk, 2014)?
An alternative possibility is that transport depends on neuronal activity. Mammalian LRP1 promotes localized transfer of IGF in response to neuronal activity. Could LTP delivery by LRP1 and LRP2 (Megalin) in the Drosophila brain depend on similar mechanisms? The remarkable specificity of LTP trafficking in the Drosophila CNS provides a novel framework for understanding information flow between the circulation and the brain (Brankatschk, 2014).
To what extent might this be relevant to vertebrate systems? While it is clear that the vertebrate brain (unlike that of Drosophila) does not depend on lipoproteins to supply it with bulk sterols, this does not rule out possible functions for these particles in nutrient sensing. The vertebrate cerebrospinal fluid is rich in many types of HDL particles, including those containing ApoA-1, which is not expressed in the brain - this suggests that at least some lipoprotein particles in the brain may derive from the circulation. Consistent with this idea, ApoA-I can target albumin-containing nanoparticles across the BBB in rodents. Recent work suggests that lipoproteins may be the source of specific Long Chain Fatty Acids that signal to the hypothalamus to regulate glucose homeostasis, since neuronal lipoprotein lipase is required for this process. Thus, it would be interesting to investigate whether circulating mammalian lipoproteins might reach a subset of neurons in the hypothalamus (Brankatschk, 2014).
It has been known for some time that increasing the amount of yeast in the diet of lab grown Drosophila melanogaster increases the rate of development and adult fertility, but reduces lifespan. This study shows that flies have evolved specific mechanisms to increase systemic IIS in response to yeast, independently of the number of calories in the diet or its proportions of sugars proteins and fats. What pressures could have driven the evolution of such mechanisms? In the wild, Drosophila melanogaster feed on rotting plant material and their diets comprise both fungal and plant components. Drosophila disperse yeasts and transfer them to breeding sites during oviposition improving the nutritional resources available to developing larvae. Yeast that are able to induce more rapid development of the agents that disperse them may propagate more efficiently. On the other hand, it has been noted that Drosophila species that feed on ephemeral nutrient sources like yeasts or flowers have more rapid rates of development than other species. It may be that, even within a single species, the ability to adjust developmental rate to the presence of a short-lived resource is advantageous. Humans subsist on diets of both plant and animal materials that during most of evolution have differed in their availability. It would be interesting to investigate whether Insulin/IGF signaling in humans might respond to qualitative differences in the lipid composition of these nutritional components (Brankatschk, 2014).
Body size constancy and symmetry are signs of
developmental stability. Yet, it is unclear exactly how developing
animals buffer size variation. Drosophila insulin-like peptide
Dilp8 is responsive to growth perturbations and controls homeostatic
mechanisms that co-ordinately adjust growth and maturation to maintain
size within the normal range. This study shows that Lgr3 is a Dilp8 receptor. By functional and cAMP assays, a pair of Lgr3 neurons were found to mediate the homeostatic regulation. These neurons have
extensive axonal arborizations, and genetic and GFP reconstitution
across synaptic partners (GRASP) show these neurons connect with the insulin-producing cells and PTTH-producing neurons to attenuate
growth and maturation. This previously unrecognized circuit suggests how
growth and maturation rate are matched and co-regulated according to
Dilp8 signals to stabilize organismal size (Vallejo, 2015).
The impressive consistency and fidelity in size of developing organisms reflects both the robustness of genetic programs and the developmental plasticity necessary to counteract the variations in size arising from genetic noise, erroneous morphogenesis, disease, or injury. To counterbalance growth abnormalities, systemic homeostatic mechanisms are implemented that delay the onset of the reproductive stage of adulthood until a correct size of the individual and its body parts has been reached. Indeed, most animals initiate a pubertal transition only once a critical size and body mass has been achieved and generally, in the absence of tissue damage or growth abnormalities. However, the mechanisms underlying such homeostatic regulation have yet to be fully defined (Vallejo, 2015).
Recently, the secreted peptide Dilp8, a member of the insulin/relaxin-like family has been identified as a factor mediating homeostatic control in Drosophila melanogaster. During the larval (growth) stage, the expression of dilp8 declines as maturation proceeds, whereas its expression is activated when growth is disturbed. Hence, fluctuating Dilp8 levels provides a reliable read-out of overall growth status (e.g., deficit) and of the time needed to complete growth and Dilp8 also orchestrates hormonal responses that stabilize body size. This includes inhibiting the production of the steroid hormone ecdysone by the prothoracic gland (PG) until the elements or organs affected are recomposed and also slowing down growth rates of undamaged tissues to ensure affected organs catch up with normal tissues in order to the adult flies reach a normal body size, maintain body proportions and symmetry. Accordingly, in the absence of dilp8, mutant flies are incapable of maintaining such strict control over their size, as reflected by the exaggerated variation in terms of overall proportionality and imperfect bilateral symmetry. However, the receptor that transduces Dilp8 signals and its site of action remained unknown (Vallejo, 2015).
Two models can be envisioned to establish such homeostatic regulation: a 'central' mechanism that dictates coordinated adjustments in both the duration and rate of growth, and an 'endocrine' mechanism that involves sensing and processing Dilp8 signals directly by hormone-producing cells. In Drosophila, several anatomically separate neural populations regulate growth and maturation time by impinging directly on the ring gland (comprising the PG and the juvenile hormone-producing corpus allatum, CA). Thus, the receptors that transduce the Dilp8 signals of growth status may act directly or communicate with neurons that produce the prothoracicotropic hormone (PTTH) and/or the neurons of the pars intercerebralis, including the insulin-producing cells (IPCs), that synthesize and release insulin-like peptides Dilp2, Dilp3 and Dilp5. Insect PTTH neurons, which are analogous to the gonadotropin-releasing hormone (GnRH) neurons in mammals, signal the commitment to sexual reproduction by stimulating the production of ecdysone in the PG in order to terminate growth. The IPCs in the pars intercerebralis, a functional equivalent of the mammalian hypothalamus, integrate nutritional signals and modulate tissue growth accordingly. Manipulation of IPCs by genetic ablation, starvation, or mutations in the single insulin receptor leads to the generation of animals with smaller size. Similarly manipulations of the PTTH neuropeptide and neurons result in variations in size of the adult flies, leading to larger or smaller than normal flies due to an extension or acceleration of the larval period and delayed pupariation. The insulin receptor also directly activates synthesis of the juvenile hormone (JH) in the CA, a hormone that promotes growth and the juvenile programs, and of ecdysone production in the PG, again augmenting the variation in normal adult size. These observations may explain how environmental and internal influences by operating through individual IPCs or PTTH neurons enable body size variation and plasticity in developmental timing that can be vital for survival in changing environments. However, the origin of developmental stability and invariant body size may require different or more complex neural mechanisms from those involved in adaptive size regulation (Vallejo, 2015).
By employing a candidate approach and biochemical assays, this study demonstrates that the orphan relaxin receptor Lgr3 acts as a Dilp8 receptor. This study identifies the neuronal population molecularly defined by the lgr3 enhancer fragment R19B09 (Jenett, 2012) and shows it is necessary and sufficient to mediate such homeostatic regulation. Using a cyclic AMP sensor as an indicator of Lgr3 receptor activation in vivo and tools for circuit mapping, it was determined that a pair of these Lgr3 neurons is highly sensitive to Dilp8. These neurons display extensive axonal arborizations and appear to connect with IPCs and PTTH neurons to form a brain circuit for homeostatic body size regulation. These data identify the insulin genes, dilp3 and dilp5, the JH, and ecdysone hormone as central in developmental size stability. Collectively, these findings unveil a homeostatic circuit that forms a framework for studying how the brain stabilize body size without constraining the adaptability of the system to reset body size in response to changing needs (Vallejo, 2015).
The data presented provide strong evidence that Dilp8 signals for organismal and organ homeostatic regulation of size are transduced via the orphan relaxin receptor Lgr3 and that activation of Lgr3 in molecularly defined neurons mediates the necessary hormonal adjustments for such homeostasis. Human insulin/relaxin-like peptides are transduced through four GPCRs, RXFP1 to 4. RXFP1 and 2 are characterized by large extracellular domains containing leucine-rich repeats similar to fly Lgr3 and Lgr4 receptors, and like Lgr3 (this study), their activation by their cognate ligand binding results in an increase in cAMP production. RXFP3 is distinctly different in structure from fly Lgr3 and its biochemical properties are also distinct, but RXPF3 is analogous to fly Lgr3 in the sense that it is found in highest abundance in the brain, suggesting important central functions for relaxin 3/RXFP3. However, a function in pubertal development and/or growth control for vertebrate relaxin receptors is presently unknown (Vallejo, 2015).
The neuronal populations that regulate body size and, in particularly, how their regulation generate variations in body size (plasticity) in response to internal and environmental cues such as nutrition have been intensely investigated. Less is known about how the brain stabilizes body size to ensure developing organisms reach the correct, genetically determined size. In particular, it remains unknown how limbs, and other bilaterally symmetric traits, grow to match precisely the size of the contralateral limb and maintain proportion with other parts even when they are faced with perturbations. Paired organs are controlled by an identical genetic program and grow in the same hormonal environment, and yet, small deviations in size can happen as result of developmental stress, genetic noise, or injury. Imperfections in symmetry thus reflect the inability of an individual to counterbalance variations and growth abnormalities (Vallejo, 2015).
This study shows that without lgr3, the brain is unable to detect growth disturbances and more importantly, it is not able to adjust the internal hormonal environment to allocate additional time during development to restore affected parts or catch-up on growth. Without lgr3, the brain also cannot slow down the growth rate to compensate for the extra time for growth so that unaffected and affected tissues can grow in a harmonious manner so as to sustain normal size, proportionality and symmetry. Using a cAMP sensor, this study has been able to define a pair of neurons that are highly sensitive to Dilp8 (Vallejo, 2015).
Communication in neuronal networks is essential to synchronize and perform efficiently. Notably, although most neurons have only one axon, Lgr3 responding neurons display extensive axonal arborizations reminiscent of hub neurons (Bonifazi, 2009). GRASP analyses show that Lgr3 neurons are broadly connected with the IPCs, and to a lesser extent with PTTH neurons, linking (Dilp8) inputs to the neuronal populations that regulate the key hormonal outputs that modulate larval and imaginal disc growth. Furthermore, the information flow from Lgr3 neurons to IPCs and to PTTH may explain how the brain matches growth with maturation in response to Dilp8. This brain circuit provides the basis for studying how the brain copes with genetic and environmental perturbations to stabilize body size, proportions and symmetry that is vital for the animal's survival (Vallejo, 2015).
The Drosophila genome encodes eight insulin/IGF-like peptide (dilp) paralogs, including tandem-encoded dilp1 and dilp2. This study finds that dilp1 is highly expressed in adult dilp2 mutants under nondiapause conditions. The inverse expression of dilp1 and dilp2 suggests these genes interact to regulate aging. Dilp1 and dilp2 single and double mutants were used to describe interactions affecting longevity, metabolism, and adipokinetic hormone (AKH), the functional homolog of glucagon. Mutants of dilp2 extend lifespan and increase Akh mRNA and protein in a dilp1-dependent manner. Loss of dilp1 alone has no impact on these traits, whereas transgene expression of dilp1 increases lifespan in dilp1 - dilp2 double mutants. dilp1 and dilp2 interact to control circulating sugar, starvation resistance, and compensatory dilp5 expression. Repression or loss of dilp2 slows aging because its depletion induces dilp1, which acts as a pro-longevity factor. Likewise, dilp2 regulates Akh through epistatic interaction with dilp1. Akh and glycogen affect aging in Caenorhabditis elegans and Drosophila. The data suggest that dilp2 modulates lifespan in part by regulating Akh, and by repressing dilp1, which acts as a pro-longevity insulin-like peptide (Post, 2018).
Based on mutational analyses of the insulin receptor (daf-2, InR) and its associated adaptor proteins and signaling elements, numerous studies in C. elegans and Drosophila established that decreased insulin/IGF signaling (IIS) extends lifespan. Studies on how reduced IIS in Drosophila systemically slows aging also reveal systems of feedback where repressed IIS in peripheral tissue decreases DILP2 production in brain insulin-producing cells (IPC), which may then reinforce a stable state of longevity assurance. This study finds that expression of dilp1 is required for loss of dilp2 to extend longevity. This novel observation contrasts with conventional interpretations where reduced insulin ligand is required to slow aging: Elevated dilp1 is associated with longevity in dilp2 mutants, and transgene expression of dilp1 increases longevity (Post, 2018).
dilp1 and dilp2 are encoded in tandem, likely having arisen from a duplication event. Perhaps as a result, some aspects of dilp1 and dilp2 are regulated in common: Both are expressed in IPCs, are regulated by sNPF, and have strongly correlated responses to dietary composition. Nonetheless, the paralogs are differentially expressed throughout development. While dilp2 is expressed in larvae, dilp1 expression is elevated in the pupal stage when dilp2 expression is minimal. In reproductive adults, dilp1 expression decreases substantially after eclosion and dilp2 expression increases (Post, 2018).
Furthermore, DILP1 production is associated with adult reproductive diapause. IIS regulates adult reproductive diapause in Drosophila, a somatic state that prolongs survival during inclement seasons. DILP1 may stimulate these diapause pro-longevity pathways, while expression in nondiapause adults is sufficient to extend survival even in optimal environments (Post, 2018).
The current data suggest a hypothesis whereby dilp1 extends longevity in part through induction of adipokinetic hormone (AKH), which is also increased during reproductive diapause and acts as a functional homolog of mammalian glucagon. Critically, AKH secretion has been shown to increase Drosophila lifespan and to induce triacylglycerides and free fatty acid catabolism. Here, it is noted that dilp1 mutants were more sensitive to starvation than wild-type and dilp2 mutants, as might occur if DILP1 and AKH help mobilize nutrients during fasting and diapause. Mammalian insulin and glucagon inversely regulate glucose storage and glycogen breakdown, while insulin decreases glucagon mRNA expression. It is propose that DILP2 in Drosophila indirectly regulates AKH by repressing dilp1 expression, while DILP1 otherwise induces AKH (Post, 2018).
A further connection between dilp1 and diapause involves juvenile hormone (JH). In many insects, adult reproductive diapause and its accompanied longevity are maintained by the absence of JH. Furthermore, ablation of JH-producing cells in adult Drosophila is sufficient to extend lifespan, and JH is greatly reduced in long-lived Drosophila insulin receptor mutants. In each case, exogenous treatment of long-lived flies with a JH analog (methoprene) restores survival to the level of wild-type or nondiapause controls. JH is a terpenoid hormone that interacts with a transcriptional complex consisting of Met (methoprene tolerant), Taimen, and Kruppel homolog 1 (Kr-h1). As well, JH induces expression of kr-h1 mRNA, and this serves as a reliable proxy for functionally active JH. This study finds that dilp2 mutants have reduced kr-h1 mRNA, while the titer of this message is similar to that of wild-type in dilp1 - dilp2 double mutants. DILP1 may normally repress JH activity, as would occur in diapause when DILP1 is highly expressed. Such JH repression may contribute to longevity assurance during diapause as well as in dilp2 mutant flies maintained in laboratory conditions (Post, 2018).
Does DILP1 act as an insulin receptor agonist or inhibitor? Inhibitory DILP1 could directly interact with the insulin receptor to suppress IIS, potentially even in the presence of other insulin peptides. Such action could induce programs for longevity assurance that are associated with activated FOXO. Alternatively, DILP1 may act as a typical insulin receptor agonist that induces autophosphorylation and represses FOXO. In this case, to extend lifespan, DILP1 should stimulate cellular responses distinct from those produced by other insulin peptides such as DILP2 or DILP5. Through a third potential mechanism, DILP1 may interact with binding proteins such as IMPL2 or dALS to indirectly inhibit IIS output. These distinctions may be resolvednin a future study using synthetic DILP1 applied to cells in culture (Post, 2018).
A precedent exists from C. elegans where some insulin-like peptides are thought to function as antagonists. In genetic analyses, ins-23 and ins-18 stimulate larval diapause and longevity, while ins-1 promotes Dauer formation during development and longevity in adulthood. Moreover, C. elegans ins-6 acts through DAF-2 to suppress ins-7 expression in neuronal circuits to affect olfactory learning, where ins-7 expression inhibits DAF-2 signaling. These studies propose that additional amino acid residues of specific insulin peptides contribute to their distinct functions, and notably, the B-chain of DILP1 has an extended N-terminus relative to other DILP sequences (Post, 2018).
While dFOXO and DAF-16 are intimately associated with how reduced IIS regulates aging in Drosophila and C. elegans, in the current work, the behavior of FOXO does not correspond with how longevity is controlled epistatically by dilp1 and dilp2. Mutation of dilp2 did not impact FOXO activity, as measured by expression of target genes InR and 4eBP, and interactions with dilp1 did not modify this result. Some precedence suggests only a limited role for dfoxo as the mediator of reduced IIS in aging, as dfoxo only partially rescues longevity benefits of chico mutants, revealing that IIS extends lifespan through some FOXO-independent pathways. On the other hand, dilp1 expression from a transgene in the dilp1-2 double mutant background did induce FOXO targets. Differences among these results might arise if whole animal analysis of dFOXO targets obscures its role when IIS regulates aging through actions in specific tissues. In this vein, this study found that dilp2 controls thorax ERK signaling but not AKT, suggesting that dilp2 mutants may activate muscle-specific ERK/MAPK anti-aging programs (Post, 2018).
Dilp1 and dilp2 redundantly regulate glycogen levels and blood sugar, while these dilp loci interact synergistically to modulate dilp5 expression and starvation sensitivity. In contrast, dilp1 and dilp2 interact in a classic epistatic fashion to modulate longevity and AKH. Such distinct types of genetic interactions may reflect unique ways DILP1 and DILP2 stimulate different outcomes from their common tyrosine kinase insulin-like receptor, along with outcomes based on cell-specific responses. Understanding how and what is stimulated by DILP1 in the absence of dilp2 will likely reveal critical outputs that specify longevity assurance (Post, 2018).
Recent studies in mammals have documented the neural expression and mobility of retrotransposons and have suggested that neural genomes are diverse mosaics. This study found that transposition occurs among memory-relevant neurons in the Drosophila brain. Cell type-specific gene expression profiling revealed that transposon expression is more abundant in mushroom body (MB) αβ neurons than in neighboring MB neurons. The Piwi-interacting RNA (piRNA) proteins Aubergine and Argonaute 3, known to suppress transposons in the fly germline, are expressed in the brain and appear less abundant in αβ MB neurons. Loss of piRNA proteins correlates with elevated transposon expression in the brain. Paired-end deep sequencing identified more than 200 de novo transposon insertions in αβ neurons, including insertions into memory-relevant loci. These observations indicate that genomic heterogeneity is a conserved feature of the brain (Perrat, 2013).
Transposons constitute nearly 45% of the human genome and 15% to 20% of the fly genome. Mobilized transposons can act as insertional mutagens and create lesions where they once resided. Recombination between homologous transposons can also delete intervening loci. Specific regions of the mammalian brain, such as the hippocampus, might be particularly predisposed to transposition. LINE-1 (L1) retrotransposons mobilized during differentiation appear to insert in the open chromatin of neurally expressed genes. One such insertion in neural progenitor cells altered the expression of the receiving gene and the subsequent maturation of these cells into neurons. The mosaic nature of transposition could therefore provide additional neural diversity that might contribute to behavioral individuality and/or neurological disorders (Perrat, 2013).
The Drosophila melanogaster mushroom bodies (MBs) are brain structures critical for olfactory memory. The approximately 2000 intrinsic MB neurons are divisible into α'β', γ, and αβ according to their morphology and roles in memory processing. This study used cell type-specific gene expression profiling to gain insight into cellular properties of MB neurons. Intersectional genetics allowed the exclusive labeling of MB α'β', γ, and αβ neurons in the brain with green fluorescent protein (GFP). For comparison, a 'no MB' genotype, in which GFP labels other neurons in the brain but not MB neurons, was also examined. Sixty brains per genotype were dissected from the head capsule and dissociated by proteolysis and agitation; GFP-expressing single cell bodies were then collected by fluorescence-activated cell sorting (FACS). Total RNA was isolated from 10,000 cells per genotype, and polyadenylated RNA was amplified and hybridized to Affymetrix Drosophila 2.0 genome expression arrays. Each genotype was processed in four independent replicates (Perrat, 2013).
Routine statistical analysis for differentially expressed genes, including a multiple-testing correction across all 16 data sets, did not reveal significant differences at a false discovery rate (FDR) of <0.05. Therefore CARMAweb was used to identify 146 mRNAs whose average signal was >7 in αβ neurons and that were also higher than in α'β' neurons by a factor of >2. Of the top 60 transcripts from this list, 29 were significantly different from α'β' signals and represent transposons. Alignment of the corresponding values from the γ and no-MB profiles showed a similarly significant bias in transposon expression over these samples. Retrotransposons were identified that transpose via a replicative mechanism involving an RNA intermediate and DNA elements that use nonreplicative excision and repair. Retrotransposons can be subdivided into long-terminal repeat (LTR) elements and long interspersed nuclear elements (LINEs). Rleven LTR elements (Tabor, mdg1, roo, qbert, gypsy, invader3, gypsy2, microcopia, 412, accord, and blood), 11 LINE-like elements (G6, RT1b, HeT-A, Ivk, Cr1a, F element, Doc2, baggins, R2, Doc3, and Doc), and four DNA elements (Bari1, pogo, Tc3, and transib3) were identified (Perrat, 2013).
Fourteen transposons, representing the most abundant in each class, were further analyzed. Quantitative reverse transcription polymerase chain reaction (qRT-PCR) of RNA from independently purified cell samples confirmed that transposon expression was significantly higher in αβ neurons than in other MB neurons. All transposons, other than R2, were also significantly higher in αβ neurons than in the rest of the brain. R2 is unique, because it exclusively inserts in the highly repeated 28S rRNA locus and heterochromatin (Perrat, 2013).
Transposition is ordinarily regulated by chromatin structure and posttranscriptional degradation of transposon mRNA guided by complementary RNAs. The small interfering RNA (siRNA) pathway has been implicated in somatic cell. In contrast, the Piwi-interacting RNA (piRNA) pathway has a more established role in the germline. The microarray analysis skewed attention toward piRNA because the expression level of the translocated Stellate locus, Stellate12D orphon (Ste12DOR) mRNA, was higher in αβ than in other MB neurons and the rest of the brain by a factor of >20. Stellate repeat transcripts are usually curtailed by piRNA, not siRNA. Stellate repeats encode a casein kinase II regulatory subunit, and piRNA mutant flies form Stellate protein crystals in testis. Immunostaining Stellate in the brain labeled puncta within αβ dendrites in the MB calyx, consistent with high Ste12DOR expression in wild-type αβ neurons (Perrat, 2013).
piRNAs are loaded into the Piwi clade argonaute proteins Piwi, Aubergine (Aub), and Argonaute 3 (Ago3). Piwi and Aub can amplify piRNA pools with Ago3. To investigate piRNA involvement in differential transposon expression, Piwi proteins and colocalized GFP were immunolocalized to assign signals to MB neuron type. Aub and Ago3 differentially labeled MB subdivisions in addition to structures throughout the brain, but no Piwi was detected. The ellipsoid body of the central complex stained strongly for Aub but not at all for Ago3, which suggests possible functional exclusivity of Piwi proteins in the brain (Perrat, 2013).
Differential Aub and Ago3 labeling was most evident within axon bundles in the peduncle and lobes, where MB neuron types are anatomically discrete. Aub protein colocalized with γ and α'β' neurons in the peduncle and lobes but was reduced in αβ neurons in both locations. Ago3 did not label MB lobes but colocalized with γ neurons in the peduncle. Ago3 labeled core αβ (αβc) neurons but did not label outer αβ neurons. Therefore, outer αβ neurons do not abundantly express Aub or Ago3, which implies that transposon suppression is relaxed. In contrast, γ neurons express Aub and Ago3, providing potential for piRNA amplification, and α'β' neurons express Aub. These patterns of Aub and Ago3 in the MB peduncle appear conserved in brains from D. erecta, D. sechellia, and the more distantly related D. pseudoobscura species (Perrat, 2013).
Loss of siRNA function elevates transposon expression in the head. These findings were replicated with ago2414 and dcr-2L811fsX mutant flies. In parallel, trans-heterozygous aub (aubHN2/aubQC42) and ago3 heads (ago3t2/ago3t3) and trans-heterozygous armitage heads (armi1/armi72.1) were used to test whether piRNA suppressed transposon expression. Levels of the 14 LTR, LINE-like, and TIR group transposons verified to be expressed in αβ neurons were assayed by qRT-PCR; of these 14 transposons, 13 were significantly elevated in siRNA-defective ago2 and dcr-2 mutants. The piRNA-defective aub, ago3, and armi mutants also exhibited significantly elevated levels of 9 of the 14 elements. Levels of the LTR elements gypsy, Tabor, and qbert; the LINE-like elements HeT-A, RT1b, and R2; and the TIR element pogo were higher in ago3 mutants. In addition, blood, Tabor, and R2 were elevated in aub mutants, and blood, gypsy, Tabor, invader3, qbert, HeT-A, and R2 were elevated in armi mutants. Therefore, the piRNA pathway contributes to transposon silencing in the brain, and low levels of Aub and Ago3 may permit expression in αβ neurons (Perrat, 2013).
To determine whether transposons are mobile, new insertions were mapped by deep sequencing of αβ DNA. αβ neurons were purified by FACS, as for transcriptome analysis, but isolated genomic DNA. Insertions were defined by paired-end reads in which one end mapped to the annotated genome and the other to the transposon sequence. To identify de novo transposition events in αβ neurons, the genomic position of transposons within the αβ sequence were compared to those located by sequencing DNA from genetically identical embryos. In addition, DNA was sequenced from the remainder of the brain tissue from the FACS separation of αβ neurons (Perrat, 2013).
These studies identified 3890 transposon insertions in embryo DNA that differed from the published Drosophila genome sequence. In comparison, αβ neuron DNA revealed 215 additional sites. The remaining brain tissue uncovered 200 new insertions, including 19 that were identical to those in αβ neurons. The sequencing depth for embryos was an order of magnitude greater than for neurons because embryo material could be collected more easily; hence, the αβ and other brain insertions are likely de novo. By randomly sampling reads to yield 1x genome coverage, 129 new transposon insertions were calculated per αβ neuron genome. Sequencing single neurons would reveal the exact cellular frequency and heterogeneity of transposition events (Perrat, 2013).
New αβ insertions occurred across all chromosomes, without obvious regional bias. In addition, insertions resulted from 49 different transposons representing LTR, LINE-like, TIR, and Foldback (FB) classes. They included 11 of the 29 transposons in the αβ transcriptome, and the number of insertions per class was consistent with their prevalence in the genome. Therefore, many transposons mobilize in αβ neurons (Perrat, 2013).
Of the 215 de novo αβ insertions, 108 mapped close to identified genes. Of these, 35 disrupted exons, 68 disrupted introns, and 5 fell in promoter regions (<1 kb from transcription start site). The remaining 107 insertions mapped to piRNA clusters or intergenic regions and were not assigned to a particular gene. A similar distribution was observed for the 200 new insertions in the rest of the brain. The reference fly genome has 258 transposon insertions in exonic regions, 11,110 insertions in intronic regions, 502 insertions in promoter regions, and 33,008 insertions in intergenic regions. Therefore, both groups of brain cells had a significantly larger fraction of insertions within exons, and fewer in intergenic regions, than the transposons that are annotated in the genome. To test whether such a distribution was unique to neurons, de novo insertions were analyzed in ovary DNA, again using embryo sequence as the comparison. New insertions in ovary DNA revealed a similar skew toward exons (Perrat, 2013).
In mammals, active L1 elements appear to disrupt neurally expressed genes. New αβ neuron insertions, but not those in other tissue, were significantly enriched in 12 Gene Ontology (GO) terms, all of which are related to neural functions. Moreover, promoter regions from 18 of 20 of the targeted genes drive expression in αβ neurons. Exonic insertions were found in gilgamesh, derailed, and mushroom body defect and intronic insertions in dunce and rutabaga, all of which have established roles in MB development and function. In addition, MB neurons are principally driven by cholinergic olfactory projection neurons and receive broad GABA-ergic inhibition and dopaminergic modulation through G protein–coupled receptors. Intronic insertions were in nicotinic Acetylcholine Receptor α 80B, G protein-coupled receptor kinase 1, and cyclic nucleotide gated channel-like and an exonic insertion in GABA-B-receptor subtype 1. Transposon-induced mosaicism could therefore alter integrative and plastic properties of individual MB αβ neurons (Perrat, 2013).
These data establish that transposon-mediated genomic heterogeneity is a feature of the fly brain and possibly other tissues. Together with prior work in rodents and humans, thede results suggest that genetic mosaicism may be a conserved characteristic of certain neurons. Work in mammals indicates that L1 expression occurs because the L1 promoter is released during neurogenesis. The data are consistent with such a model and also support the idea that transposons avoid posttranscriptional piRNA silencing in adult αβ neurons (Perrat, 2013).
A recent study described a role for piRNA in epigenetic control of memory-related gene expression in Aplysia neurons. It is therefore possible that MB neurons differentially use piRNA to control memory-relevant gene expression and that transposon mobilization is an associated cost. Because tansposon expression was found in αβ neurons of adult flies, it is conceivable that disruptive insertions accumulate throughout life, leading to neural decline and cognitive dysfunction. Alternatively, permitting transposition may confer unique properties across the 1000 neurons in the αβ ensemble and potentially produce behavioral variability between individual flies in the population (Perrat, 2013).
Controlling the orientation of cell division is important in the context of cell fate choices and tissue morphogenesis. However, the mechanisms providing the required positional information remain incompletely understood. This study used stem cells of the Drosophila larval brain that stably maintain their axis of polarity and division between cell cycles to identify cues that orient cell division. Using live cell imaging of cultured brains, laser ablation and genetics, this study reveals that division axis maintenance relies on their last-born daughter cell. It is proposed that, in addition to known intrinsic cues, stem cells in the developing fly brain are polarized by an extrinsic signal. It was further found that division axis maintenance allows neuroblasts to maximize their contact area with glial cells known to provide protective and proliferative signals to neuroblasts (Loyer, 2018).
Deciphering the signals that provide positional information is a central issue in understanding how cell divisions are oriented. This study addressed this question in the highly proliferative NBs in the Drosophila larval brain, which maintain their division axis from one cell cycle to the next in part by using an apical microtubule network as a spatial cue to specify their apico-basal polarity axis and consequently the orientation of mitosis. Attempts were made to understand why NBs only partially fail to maintain their division axis upon loss of this intrinsic polarizing cues, and it was found out that the last-born daughter cell of NBs participates to their division axis maintenance. These results also shed light on some aspects of the physiological importance of division axis maintenance in larval NBs, which has remained elusive. Control of NB division orientation may provide a means to maximize NB/cortex glia surface area to allow optimum protection against environmental stresses by the cortex glia (Loyer, 2018).
Under normal conditions, about 80% of the surface of NBs is in direct contact with a cortex glia and NBs with partially defective division axis maintenance display reduced contact with cortex glia. This most likely directly results from NBs producing progeny between themselves and the cortex glia when the last-born daughter cell derived cue that positions normally the apico-basal polarity axis is damaged. This seems to be important for the protective function of these glial cells on NB proliferation under stress conditions. Indeed, NBs with reduced surface contact to cortex glia appear to be less well protected by glial cells, as was observed a significant increase of sensitivity to oxidative stress using an established assay. However, despite this reduction being statistically significant, only a 9% reduction was measured in NB/cortex glia contact area. On a normal diet, addition of the oxidant tert-butyl hydroperoxide (tbh) results in a 14% drop in NB proliferation when the formation of lipid droplets mediating this protection is prevented. It is therefore surprising that in these experiments reducing the NB/cortex glia contact area by only ~9% in Cindr-depleted NB is already accompanied by a similar drop in proliferation upon tbh treatment. Therefore, although this decrease may directly result from interfering with the protection provided by cortex glia, other unrelated functions of Cindr in protecting NBs against the effect of tbh cannot be ruled out (Loyer, 2018).
It was initially hypothesized that the last-born GMC could act as an additional, extrinsic cue maintaining NB division orientation. A number of observations are consistent with this possibility: it was observed that, upon (perhaps artefactual) last-born GMC movements, NBs realign their division axis toward this GMC; ablation of the last-born GMC and depletion of proteins specifically observed at the last-born GMC/NB interface affect division axis maintenance by misorienting the apico-basal polarity of NBs. It cannot be excluded that the entire NB and any intrinsic spatial cue that it carries simply rotate upon migration or ablation of the last-born GMC or depletion of proteins specifically observed at the last-born GMC/NB interface. Thus the last-born GMC may participate in division axis maintenance by preventing NB rotation. This function could be mediated by specific adhesive contacts at the interface with the NB, plausible given the numerous specific characteristics that were observed at that interface. In particular, the midbody carried by this interface, although not likely to act itself as a stable physical link given its possible ability to migrate within the fluid mosaic of the plasma membrane and the fact that its internalization does not affect division orientation maintenance, may be able to organize specific adhesive contacts at the NB/last-born GMC interface (Loyer, 2018).
An alternative hypothesis is that the last-born GMC provides a cue that more directly functions in specifying the orientation of the apico-basal polarity axis by polarizing Baz, which functions upstream of NB division orientation control. Consistently, despite affecting division orientation maintenance, neither GMC ablation nor RNAi of Cindr disrupt alignment of the mitotic spindle with the polarity axis. In this case, the molecular mechanism through which a positional information provided by the last-born GMC is transduced to the NB polarization machinery remains to be determined. Although bearing similarities with division axis maintenance in budding yeasts, relying on a Septin-rich cytokinesis remnant, the midbody of NBs is unlikely to directly control polarization as midbody internalization does not affect division axis maintenance. Instead, it is proposed that the midbody may organize various other specific components of the last-born GMC/NB interface that in turn may directly control NB polarization. This could be the case of cell-cell contacts organized by the midbody, consistent with the involvement of an adhesion molecule such as Roughest, whose mammalian orthologue physically interacts with Septins, and the fact that GMC ablation, although not directly targeting the interface, affects division axis maintenance. Another promising candidate potentially controlling NB polarity are the plasma membrane tubules probably organized by the midbody, given their physical origin (the midbody) and the timing (immediately after cytokinesis) of their appearance. Interestingly, a physical interaction was observed between Septins and the mammalian orthologue of Cindr, found enriched at the tubules and involved in division axis maintenance. Tubules function might be linked to the integrity of the last-born GMC/NB interface, which itself probably depends on the integrity of the last-born GMC. While these tubules do not disappear upon GMC ablation, it would be of particular interest to monitor whether tubules morphology, dynamics or the enrichment of Flare and Cindr are affected by ablation of the last-born daughter cell (Loyer, 2018).
Interestingly, proteins that were found to be involved in division axis maintenance were described to interact with polarity complexes in other contexts: Septins genetically interact with Baz during Drosophila embryogenesis, and the mammalian orthologues of Roughest regulate podocyte polarity by physically interacting with Par-3. However, both Septins and Roughest localize to the basal pole of NBs, whereas Baz polarizes apically. Therefore, how could a cue received at the basal pole direct polarization of Baz, at the opposite apical pole of the NB? In the C. elegans zygote, the sperm entry point acts as a cue inducing an actomyosin flow establishing Par complex polarity at the opposite end of the cell. Septins, Cindr, Roughest and Flare can be linked in one way or another to the regulation of actomyosin, and at least the maintenance of Baz localization in mitotic NBs is also actin-dependent. Intracellular long-range control of polarization has been further observed in eight-cell stage mouse blastomeres, where cell-cell contacts induce apical polarization at the opposite end of the cell. A promising lead for future work is the possible involvement of actomyosin-dependent mechanical forces in such long-range control of polarity in NBs. Indeed, tensions participate in polarization in the C. elegans zygote, were proposed to mediate polarization of eight-cell stage mouse blastomeres and maintain polarity in migrating neutrophils (Loyer, 2018).
During CNS development, there is prominent expansion of the anterior region of the brain. In Drosophila, anterior CNS expansion emerges from three rostral features: (1) increased progenitor cell generation, (2) extended progenitor cell proliferation, (3) more proliferative daughters. This study finds that tailless (mouse Nr2E1/Tlx), otp/Rx/hbn (Otp/Arx/Rax) and Doc1/2/3 (Tbx2/3/6) are important for brain progenitor generation. These genes, and earmuff (FezF1/2), are also important for subsequent progenitor and/or daughter cell proliferation in the brain. Brain TF co-misexpression can drive brain-profile proliferation in the nerve cord, and can reprogram developing wing discs into brain neural progenitors. Brain TF expression is promoted by the PRC2 complex, acting to keep the brain free of anti-proliferative and repressive action of Hox homeotic genes. Hence, anterior expansion of the Drosophila CNS is mediated by brain TF driven 'super-generation' of progenitors, as well as 'hyper-proliferation' of progenitor and daughter cells, promoted by PRC2-mediated repression of Hox activity (Curt, 2019).
Detailed analysis of Drosophila CNS development has revealed that there is 'super-generation' of NBs in the B1 segment; ~160 NBs in B1 compared to 28-70 NBs/segment for each of the 18 posterior segments (B2-A10). In the ventral neurogenic regions (generating the nerve cord) a single NB delaminates from each proneural cluster. In contrast, the NB super-generation in B1 stems, at least in part, from group delamination of NBs. The specification of NB cell fate depends upon low, or no, Notch activity. In line with this notion, evidence points to reduced Notch signalling in the procephalic neuroectoderm (Curt, 2019).
Head gap genes, such as tll, were previously shown to be important for B1 NB generation, and in line with this strikingly reduced NB generation was observed in tll. Does tll intersect with Notch signalling? tll mutants show loss of expression of the proneural gene l'sc, which is negatively regulated by Notch. Recent studies furthermore reveal an intimate interplay between tll and Notch signalling in the developing Drosophila embryonic optic placodes. In addition, the C. elegans tll orthologue nhr-67 regulates both lin-12 (Notch) and lag-2 (Delta) during uterus development. Strikingly, in the mouse brain, the tll orthologue Nr2E1 (aka Tlx) was recently shown to negatively regulate the canonical Notch target gene Hes1. Against this backdrop, it is tempting to speculate that the group NB delamination normally observed in the procephalic region results, at least in part from tll repression of the Notch pathway. Indeed, tll was the only one of the four TFs that could act alone to trigger ectopic NBs in the wing disc (Curt, 2019).
Other previously identified head gap genes are oc (also known as orthodenticle: otd), buttonhead (btd) and ems. However, it was not observed that misexpression of oc or ems from elav-Gal4 efficiently drove ectopic proliferation in the nerve cord. Moreover, oc acts both in B1-B2, ems in B2-B3, being repressed from B1 by tll, and btd acts in B2-B3. Because B2 and B3 segments do not display super-generation of NBs these findings point to tll as the key head gap gene driving the super-generation of NBs specifically observed in the B1 segment (Curt, 2019).
Reduced NB generation was observed in the triple otp/Rx/hbn and Doc1/2/3 mutants. This would tentatively place them in the category of head gap genes, at least as far as being important for NB generation. However, their effects on NB generation is weaker than that observed in tll mutants. In addition, otp/Rx/hbn and Doc1/2/3 show genetic redundancy. The combination of genetic redundancy and their weaker effects on NB generation, likely explain why they were not previously categorised as head gap genes (Curt, 2019).
The connection between the brain TFs examined in this study herein and NB super-generation is not only evident from the mutant phenotypes, but also from their potent gain-of-function effects. Strikingly, it was found that brain TF co-misexpression was sufficient to generate ectopic NBs in the embryonic ectoderm and developing wing discs. A number of markers indicate that these ectopic NBs undergo normal CNS NB lineage progression, generating neurons and glia. Moreover, the ectopic expression of the brain-specific factors Rx and Hbn, the apparently higher neuron/glia ratio, the reduced GsbN expression, the generation of Dpn+/Ase- NBs (Type II-like) in both the embryonic ectoderm and wing discs, in combination suggest that brain TF co-misexpression specifically triggered reprogramming towards a B1 brain-like phenotype (Curt, 2019).
One surprising finding pertains to the clear difference between the potency of the tll,erm double and the Tetra (tll, erm, Doc2 and otp) in the embryonic ectoderm versus the wing disc, with the double being more potent in the wing disc and the Tetra more potent in the embryo. Indeed, in the wing disc the strong effect of tll,erm is suppressed by the addition of any combination of otp and Doc2. There is no obvious explanation for the different responsiveness to brain TF misexpression in the two tissues, but it may reflect the fact the embryonic neuroectoderm is already primed for the generation of NBs (Curt, 2019).
Another surprising finding pertains to the role of erm in embryonic versus larva Type II NBs. Previous studies of erm function in the larvae found that erm mutants displayed more Type II NBs. Larval MARCM clone induction and marker analysis demonstrate that this is due to de-differentiation of INPs back to type II NBs, rather than excess generation of Type II NBs in the embryo. No extra Type II or Type I NBs were found in erm mutants but rather reduced number of cells generated in the embryonic Type II lineages, showing that erm is important for lineage progression. Hence, the role of erm appears to be different in the embryonic versus larval Type II lineages (Curt, 2019).
In addition to the NB super-generation in B1, recent studies reveal that three different lineage topology mechanisms underlie the hyper-proliferation of the brain. First, the majority of NBs (136 out 160 NB) display a protracted phase of NB proliferation, and do not show evidence of switching from Type I to Type 0 daughter proliferation (Yaghmaeian Salmani, 2018). Second, the eight MBNBs, which appear to divide in the Type I mode and never enter quiescence, also generate large lineages. Third, the 16 Type II NBs progress by budding off INP daughter cells, which divide multiple times to generate daughter cells that in turn divide once, hence resulting in lineage expansion. In contrast, in the nerve cord many NBs switch from Type I to Type 0, and all halt neurogenesis by mid-embryogenesis. The Hox anti-proliferation gradient further results in a gradient of the Type I-->0 switch and NB exit along the nerve cord. The combined effects of these alternate lineage topology behaviours translate into striking differences in the average lineage size in the brain when compared to the nerve cord (Yaghmaeian Salmani, 2018) (see Mechanisms underlying the anterior expansion of the Drosophila CNS). Moreover, the three different modes of more extensive NB and daughter cell proliferation combine with the super-generation of NBs in B1 to generate many more cells in the B1 brain segment, when compared to all posterior segments (Curt, 2019).
The brain TFs examined in this study are expressed in several or all (Tll) of the three brain NB types, and are important for both NB and daughter cell proliferation. In line with this, brain TF ectopic expression, with the late neural driver elav-Gal4, drives aberrant nerve cord proliferation and blocks both the Type I-->0 daughter cell proliferation switch and NB cell cycle exit. This results in the generation of supernumerary cells, evident both by the expansion of specific lineages and an increase in overall nerve cord cell numbers. This study found that both the double and Tetra misexpression can trigger the ectopic generation of what appears to be a mix of Type I and Type II-like NBs. The mix of these two NB types may reflect that the misexpression scenario does not accurately and reproducibly recreate the temporal order of the brain TFs, with for example tll expressed prior to erm in the wild type (Curt, 2019).
The ectopic appearance of symmetrically dividing NBs in the brain TF co-misexpression nerve cords is more difficult to explain. However, since there normally are divisions of cells in the neuroectodermal layer prior to NB delamination, and given the early expression of the brain TFs (prior to NB delamination), it is tempting to speculate that brain TF co-misexpression to some extent can trigger an early neuroectodermal cell fate (Curt, 2019).
It was recently found that NB and daughter proliferation is also promoted by a set of early TFs expressed by most, if not all NBs. Strikingly, these TFs are expressed at higher levels in the brain, due to the lack of Hox expression therein, thereby contributing to the extended NB proliferation and more proliferative daughter cells observed in the brain. It will be interesting to address the possible regulatory interplay between these broadly expressed early NB factors and the brain TFs described in this study (Curt, 2019).
Gene expression studies have revealed the mutually exclusive territory of brain TF and Hox gene expression in the Drosophila CNS. In line with this notion, it was found that co-misexpression of brain TFs in the nerve cord repressed expression of the posterior Hox genes of the BX-C, and conversely that BX-C co-misexpression repressed several brain TFs; Bsh, Rx, Hbn, Tll and Doc2 (Curt, 2019).
A key 'gate-keeper' of the brain versus nerve cord territories appears to be the PRC2 epigenetic complex. Removing PRC2 function results in complete loss of the H3K27me3 repressive epigenetic mark and anterior expansion of the expression of all Hox genes. This furthermore results in repression of brain TF expression, that is Tll and Doc2, as well as Rx. Surprisingly, in spite of the many roles that PRC2 may play, this study found that transgenic brain TF co-expression could rescue the PRC2 mutant proliferation defects. Given the repressive action of BX-C Hox genes on brain TFs, this suggests that the principle role of PRC2 during early CNS development, at least regarding proliferation, is to ensure that Hox genes are prevented from being expressed in the brain, thus ensuring brain TF expression. Indeed, it was recently demonstrated that the reduced brain proliferation observed in esc mutants could also be fully rescued by the simultaneous removal of the posterior-most and most anti-proliferative Hox gene, Abd-B (Curt, 2019).
In mammals, the precise number of neural progenitors present at different axial levels during embryonic development has not yet been mapped. However, the wider expanse of the anterior embryonic neuroectoderm would suggest the generation of more progenitors anteriorly. There is also an extended phase of neurogenesis in the forebrain, when compared to the spinal cord. Dividing daughter cells (most often referred to as basal progenitors; bP) have been identified along the entire A-P axis of the mouse CNS. Intriguingly, the ratio of dividing bPs to apical progenitors (radial glial cells) was found to be higher in the telencephalon than in the hindbrain. Similarly, recent studies revealed a higher ratio of dividing cells in the outer layers than in the lumen, when comparing the developing telencephalon to the lumbo-sacral spinal cord. Albeit still limited in their scope, these studies suggest that a similar scenario is playing out along the A-P axis of mouse CNS as that observed in Drosophila, with an anteriorly extended phase of progenitor proliferation and a higher prevalence of proliferating daughter cells (Curt, 2019).
In addition to the similarities between Drosophila and mouse regarding progenitor generation, as well as progenitor and daughter cell proliferation, the genetic mechanisms controlling these events may also be conserved. Mouse orthologues of the Drosophila brain TFs examined in this study, that is Nr2E1/Tlx (Tll); Otp, Rax and Arx (Otp); Tbx2/3/6 (Doc1/2/3); and FezF1/2 (Erm), are restricted to the brain and are known to be critical for normal mouse brain development, and in several cases for promoting proliferation. Furthermore, Hox genes are not expressed in the mouse forebrain and there is a generally conserved feature of brain TFs expressed anteriorly and Hox genes posteriorly. Mutation and misexpression has revealed that Hox genes are anti-proliferative also in the vertebrate CNS. Moreover, PRC2 (EED) mouse mutants show extensive expression of Hox genes into the forebrain and reduced gene expression of for example Nr2E1, Fezf2 and Arx. This is accompanied by reduced proliferation in the telencephalon and a microcephalic brain, while the spinal cord does not appear effected (Curt, 2019).
Gene expression and phylogenetic consideration recently led to the proposal that the CNS may have evolved by 'fusion' of two separate nervous systems, the apical and basal nervous systems, present in the common ancestor. Interestingly, in arthropods for example Drosophila, the brain and nerve cord initially form in separate regions only to merge during subsequent development. Recent studies of the role of the PRC2 complex and Hox genes in controlling A-P differences in CNS proliferation, in both Drosophila and mouse, lend support for the notion of a 'fused' CNS (Yaghmaeian Salmani, 2018). This idea is further supported by recent studies of the epigenomic signature and early embryonic cell origins of the anterior versus posterior developing CNS. The findings outlined in this study, showing that brain hyperproliferation is driven not only by the lack of Hox homeotic gene expression, but also by the specific expression of highly conserved brain TFs, lend further support to the notion of a separate evolutionary origin of brain and nerve cord (Curt, 2019).
It is tempting to speculate that the possibly separate evolutionary origins of the brain and nerve cord may manifest not only as distinct modes of neurogenesis, but also be reflected by separate regulatory mechanisms. These would involve brain TFs acting anteriorly, generating an abundance of progenitors, as well as driving progenitor and daughter cell proliferation. Conversely, Hox genes would act posteriorly, counteracting progenitor generation, as well as tempering progenitor and daughter cell proliferation. In this model, PRC2 would act as a 'gate keeper', ensuring that Hox genes are restricted from the brain and thereby promoting brain TF expression. This model clearly represents an over-simplification, but may serve as a useful launching point for future comparative studies in many model systems (Curt, 2019).
Neural progenitors undergo temporal patterning to generate diverse neurons in a chronological order. This process is well-studied in the developing Drosophila brain and conserved in mammals. During larval stages, intermediate neural progenitors (INPs) serially express Dichaete (D), grainyhead (Grh) and eyeless (Ey/Pax6), but how the transitions are regulated is not precisely understood. In this study a method was developed to isolate transcriptomes of INPs in their distinct temporal states to identify a complete set of temporal patterning factors. This analysis identifies odd-paired (opa), as a key regulator of temporal patterning. Temporal patterning is initiated when the SWI/SNF complex component Osa induces D and its repressor opa at the same time but with distinct kinetics. Then, high opa levels repress D to allow Grh transcription and progress to the next temporal state. It is proposed that Osa and its target genes opa and D form an incoherent feedforward loop (FFL) and a new mechanism allowing the successive expression of temporal identities (Abdusselamoglu, 2019).
Drosophila has been crucial to understanding stem cell biological mechanisms and in particular distinct temporal patterning processes. During embryonic neurogenesis, Drosophila NSCs, called Neuroblasts (NBs), undergo temporal patterning through a cascade of transcription factors. During larval neurogenesis, NB temporal patterning relies on opposing gradients of two RNA-binding proteins. Temporal patterning is also seen in intermediate neural progenitors (INPs), the transit-amplifying progeny of a discrete subset of larval NBs called type II NBs (Bayraktar, 2013). Once they arise from an asymmetric division of a type II NB, newborn INPs undergo several maturation steps before they resume proliferation: they first turn on earmuff (erm), and asense (ase), and finally deadpan (dpn) expression to become mature INPs (mINP). Then mINPs divide 3-6 times asymmetrically to generate ganglion mother cells (GMCs), which in turn divide to generate a pair of neurons or glia. Analogous to embryonic NBs, recent reports suggest that a transcription factor cascade regulates temporal patterning of INPs (Bayraktar, 2013). Indeed, the sequential expression of Dichaete (D), Grainyhead (Grh) and Eyeless (Ey) is required to generate different neurons: D+ INPs produce Brain-specific homeobox (Bsh)+ neurons, while Ey+ INPs produce Toy+ neurons (Bayraktar, 2013; Abdusselamoglu, 2019).
The three temporal identity factors are regulated through various regulatory interactions (Bayraktar, 2013): D is necessary, but not sufficient, for activating Grh. Grh instead is required for repression of D and activation of Ey (Bayraktar, 2013). Therefore, INP temporal patterning is thought to be regulated by a 'feedforward activation and feedback repression' mechanism. Intriguingly however, INP temporal patterning also critically requires the SWI/SNF chromatin remodeling complex subunit Osa (Eroglu, 2014). Although Osa is not considered a specific temporal identity factor, it is required to initiate temporal patterning by activating the initial factor D. While the Osa target gene hamlet is required for the Grh-to-Ey transition (Eroglu, 2014), regulation of the first transition is less well understood. This result suggests that in addition to feedforward activation and feedback repression, temporal switch genes are required to ensure correct INP temporal patterning. Nevertheless, D and ham double knock down (k.d.) phenotypes do not recapitulate the complete loss of temporal patterning initiation observed in Osa-depleted type II NB lineages, suggesting the contribution of additional unidentified factors (Abdusselamoglu, 2019).
This study describes a FACS-based method to isolate INPs from three different temporal identities. By comparing the transcriptomic profiles of each set of INPs, odd-paired (opa), a transcription factor whose expression is enabled by direct binding of Osa to its TSS, was identified as a regulator of temporal patterning and repressor of D. Though Osa enables both D and opa expression, Opa's slower activation kinetics allow D to function in a short time window before being repressed by Opa. This mode of action resembles an incoherent feedforward-loop (FFL) motif, where an upstream gene directly activates the target gene, meanwhile indirectly repressing it by activating its repressor. Thus, this study uncovered a novel mechanism controlling temporal patterning during neurogenesis (Abdusselamoglu, 2019).
This study establish two different roles of the SWI/SNF complex subunit, Osa, in regulating INP temporal patterning. Initially, Osa initiates temporal patterning by activating the transcription factor D. Subsequently, Osa regulates the progression of temporal patterning by activating opa and ham, which in turn downregulate D and Grh, respectively. The concerted, but complementary action of opa and ham ensures temporal identity progression by promoting the transition between temporal stages. For instance, opa regulates the transition from D to Grh, while ham regulates the transition from Grh to Ey. It is proposed that opa achieves this by repressing D and activating grh, as indicated by the lack of temporal patterning in D and opa-depleted INPs. Loss of opa or ham causes INPs to lose their temporal identity and overproliferate. Moreover, it is proposed that D and Opa activate Grh expression against the presence of ham, which represses Grh expression. As D and Opa levels decrease as INPs age and become Grh positive, ham is capable of repressing Grh later on in temporal patterning. This explains how opa and ham act only during specific stages even though they are expressed throughout the entire lineage (Abdusselamoglu, 2019).
An open question pertains to the fact that the double knock-down of opa and ham, as well as that of D and opa, failed to recapitulate the Osa phenotype. Even though opa and ham RNAi caused massive overproliferation in type II lineages, no Dpn+ Ase- ectopic NB-like cells (as occurs in Osa mutant clones, Eroglu, 2014) were detected. It is proposed that this is caused by D expression which is still induced even upon opa/ham double knockdown, but not upon Osa knock-down where D expression fails to be initiated. Thus, the initiation of the first temporal identity state may block the reversion of INPs to a NB-state. In the future, it will be important to understand the exact mechanisms of how opa regulates temporal patterning (Abdusselamoglu, 2019).
This study further demonstrated that Osa initiates D expression earlier than opa expression. Osa is a subunit of SWI/SNF chromatin remodeling complex, and it guides the complex to specific loci throughout the genome, such as the TSS of both D and opa. The differences in timing of D and opa expression may be explained by separate factors involved in their activation. Previous work suggests that the transcription factor earmuff may activate. However, it remains unknown which factor activates opa expression. One possibility is that the cell cycle activates opa, since its expression begins in mINPs, a dividing cell unlike imINPs, which are in cell cycle arrest (Abdusselamoglu, 2019).
It is proposed that balanced expression levels of D and opa regulate the timing of transitions between temporal identity states. Indeed, Osa initiates D and opa, the repressor of D, at slightly different times, which could allow a time window for D to be expressed, perform its function, then become repressed again by Opa. Deregulating this pattern, for example by overexpressing opa in the earliest INP stage, results in a false start of temporal patterning and premature differentiation. This elegant set of genetic interactions resembles that of an incoherent feedforward loop (FFL). In such a network, pathways have opposing roles. For instance, Osa promotes both the expression and repression of D. Similar examples can be observed in other organisms, such as in the galactose network of E. coli, where the transcriptional activator CRP activates galS and galE, while galS also represses galE. In Drosophila SOP determination, miR-7, together with Atonal also forms an incoherent FFL. Furthermore, mammals apply a similar mechanism in the c-Myc/E2F1 regulatory system (Abdusselamoglu, 2019).
The vertebrate homologues of opa consist of the Zinc-finger protein of the cerebellum (ZIC) family, which are suggested to regulate the transcriptional activity of target genes, and to have a role in CNS development. In mice, during embryonic cortical development, ZIC family proteins regulate the proliferation of meningeal cells, which are required for normal cortical development. In addition, another member of the ZIC family, Zic1, is a Brn2 target, which itself controls the transition from early-to-mid neurogenesis in the mouse cortex. Along with these lines, it has been shown that ZIC family is important in brain development in zebrafish. Furthermore, the role of ZIC has been implicated in variety of brain malformations and/or diseases. These data provide mere glimpses into the roles of ZIC family proteins in neuronal fate decisions in mammals, and this study offers an important entry point to start understanding these remarkable proteins (Abdusselamoglu, 2019).
The current findings provide a novel regulatory network model controlling temporal patterning, which may occur in all metazoans, including humans. In contrast to existing cascade models, this study instead shows that temporal patterning is a highly coordinated ensemble that allows regulation on additional levels than was previously appreciated to ensure a perfectly balanced generation of different neuron/glial cell types. Together, these results demonstrate that Drosophila is a powerful system to dissect the genetic mechanisms underlying the temporal patterning of neural stem cells and how the disruption of such mechanisms impacts brain development and behavior (Abdusselamoglu, 2019).
Molecular interactions at the cellular interface mediate organized assembly of single cells into tissues and, thus, govern the development and physiology of multicellular organisms. This study has developed a cell-type-specific, spatiotemporally resolved approach to profile cell-surface proteomes in intact tissues. Quantitative profiling of cell-surface proteomes of Drosophila olfactory projection neurons (PNs) in pupae and adults revealed global downregulation of wiring molecules and upregulation of synaptic molecules in the transition from developing to mature PNs. A proteome-instructed in vivo screen identified 20 cell-surface molecules regulating neural circuit assembly, many of which belong to evolutionarily conserved protein families not previously linked to neural development. Genetic analysis further revealed that the lipoprotein receptor LRP1 cell-autonomously controls PN dendrite targeting, contributing to the formation of a precise olfactory map. These findings highlight the power of temporally resolved in situ cell-surface proteomic profiling in discovering regulators of brain wiring (Li, 2020).
Despite its functional importance, the cell-surface proteome in intact tissues is difficult to characterize by traditional methods. Biochemical fractionation of membrane proteins not only includes mitochondrial, endoplasmic reticulum (ER), and Golgi contaminants but also omits secreted and extracellular matrix proteins that form an integral part of the cell-surface proteome. Moreover, fractionation does not provide cell-type specificity, especially in the nervous system, where many cell types intermingle within a compact region. To capture the cell-type-specific cell-surface proteome in intact tissues, this study modified a peroxidase-mediated proximity biotinylation procedure for cultured neurons and developed a method for cell-surface biotinylation in intact tissues (including optimized chemical delivery, the in situ biotinylation reaction, and tissue sample processing). Horseradish peroxidase (HRP) fused to the N terminus of a generic transmembrane protein, rat CD2, targets HRP to the extracellular side of the plasma membrane. Transgenic expression of HRP-CD2 in genetically defined cell populations thus confers cell-type specificity. Although HRP is cell-surface targeted, it is also enzymatically active in the secretory pathway intracellularly. To avoid biotinylating proteins there, the biotin-xx-phenol (BxxP) substrate, which is unable to cross the plasma membrane, was used. Combining surface-targeted HRP and membrane-impermeable BxxP, this dual-gate approach should ensure cell-surface specificity. In the presence of hydrogen peroxide (H2O2), HRP converts BxxP into phenoxyl radicals that promiscuously biotinylate endogenous proteins in proximity. Because of its rapid kinetics, this HRP-mediated reaction requires only a few minutes to complete, providing a cell-surface snapshot with high temporal resolution and minimizing the potential toxicity of H2O2 (Li, 2020).
This cell-surface biotinylation strategy was applied to Drosophila olfactory PNs. In the antennal lobe, the first-order olfactory processing center of Drosophila, PNs of the same type target their dendrites to a specific glomerulus to form synaptic connections with their partner olfactory receptor neurons (ORNs). Precise one-to-one pairings of 50 types of PNs and ORNs at 50 discrete glomeruli build an anatomically stereotyped olfactory map in the antennal lobe, providing an excellent system for studying neural development and computation. To examine HRP-mediated cell-surface biotinylation in intact brains, adult brains were first stained with fluorophore-conjugated neutravidin, which specifically recognizes biotin. Extensive HRP-dependent biotinylation of PN dendrites and axons was observed. Confocal optical sectioning revealed that only the surface of PN cell bodies was biotinylated, in contrast to the broad distribution of membrane-targeted GFP (Li, 2020).
Biochemical characterization by streptavidin blot and silver stain of the brain lysate or its post-enrichment eluate showed that the PN surface-targeted HRP biotinylated a wide range of proteins compared with the control lacking HRP-CD2 expression, although fly brains also expressed endogenously biotinylated proteins that require filtering in proteomic analysis. Immunoblotting revealed that the neuronal surface marker N-cadherin, but not the cytosolic proteins actin or bruchpilot (a pan-neural synaptic marker), was detected in streptavidin bead eluates in an HRP-dependent manner. Thus, surface biotinylation procedure provides a way to label and enrich cell-surface proteins of a chosen cell type in intact tissues, enabling mass-spectrometry-based proteomic profiling (Li, 2020).
It is generally assumed that cell-surface proteins change as neurons mature, but this has not been systematically examined because of technical limitations. This study profiled the PN surface proteome at two time points: 36 h after puparium formation (APF), when developing PNs elaborate their dendrites and axons to build synaptic connections, and 5 days after eclosion into adults, when mature PNs actively process olfactory information. To better quantify protein changes and filter out contaminants captured in negative controls, an 8-plex tandem mass tag (TMT)-based quantitative strategy was used and each time point was profiled with two biological replicates. Each time point also contained two negative controls that captured endogenously biotinylated proteins and non-specific binders to streptavidin beads. Freshly dissected intact brains (~1,000 per TMT plex and ~8,000 for this 8-plex experiment) were pre-incubated with the BxxP substrate for 1 h before a 5-min H2O2 reaction. Biotinylated proteins were then enriched from brain lysates using streptavidin beads, followed by on-bead trypsin digestion, TMT labeling, and liquid chromatography-tandem mass spectrometry analysis (LC-MS/MS) (Li, 2020).
From this 8-plex experiment, a total of 2,020 proteins were detected with 2 or more unique peptides. To remove potential contaminants, a ratiometric strategy was used in which the TMT ratio of each protein reflects its differential enrichment in the experimental group versus the negative control group. A bona fide PN surface protein should exhibit a high TMT ratio because it should be extensively biotinylated in the experimental group but not in the control group. A false positive, such as an endogenously biotinylated protein or a non-specific bead binder, would have a low ratio because it should be captured similarly in both groups. Each experimental group was paired with a control group, for calculating TMT ratios. Each biological replicate was then ranked in descending order by the TMT ratio and its receiver operating characteristic (ROC) curve, which depicts the true positive rate against the false positive rate of detected proteins, was plotted. Notably, the top 20% of proteins exhibited almost vertical ROC curves, demonstrating high specificity. To maximize the signal-to-noise ratio of the proteome, each biological replicate was cut off at the position where the value of true positive rate - false positive rate was maximal and only the overlapping proteins of the two replicates were collected at each time point to further minimize potential contaminants (Li, 2020).
Of the 712 proteins in the PN surface proteomes, only 252 proteins were shared by the developing and mature proteomes, whereas 460 were specific to either developing or mature PNs. These data suggest a profound difference between the PN surface proteomes at these time points. The developing and mature PN surface was systematically examined and compared (Li, 2020).
In contrast to their nearly identical cell-surface annotations, the developing and mature proteomes exhibited distinct and non-overlapping signatures regarding their biological functions. In accord with the primary task of PNs at each time point, the developing surface highlighted the processes for neural development, whereas the mature surface featured categories covering ion channels, receptors, and transporters. The most enriched proteins from the developing PN surface with known functions were predominantly neural development molecules, in particular those involved in axon guidance and neural circuit wiring. This is surprising because guidance molecules are generally thought to act at growth cones-only the tip of arborizing dendrites and axons. It is possible that these proteins are either expressed at exceptionally high levels at growth cones or broadly distributed on the neuronal surface, not limited to their sites of action. Strikingly, the top 100 cell-surface-enriched proteins of developing PNs contained almost all cell-surface regulators of olfactory circuit wiring identified in the past 15 years as well as many wiring molecules discovered in other systems. In contrast, the most enriched proteins of the mature PN surface with known functions were not synaptic molecules, consistent with the notion that synapses are highly restricted at select loci of a mature neuron. The enrichment of many previously uncharacterized proteins on the mature PN surface suggests that future investigations of these proteins may reveal new facets of PN physiology (Li, 2020).
MS-based proteomics provides a systematic way for understanding proteomes and their dynamics in biological systems, including the nervous system. Despite its central roles in neural development and function, the cell-surface proteome of a specific neuronal type in intact brains has been difficult to characterize because of a lack of appropriate methodology. Chemical labeling enables enrichment of cell-surface proteins but does not provide cell type specificity. Alternative technologies for cell-type-specific proteomics, such as bio-orthogonal metabolic labeling or organelle tagging, are not amenable to analysis of cell-surface proteomes specifically (Li, 2020).
This study reports a spatiotemporally resolved approach to profile the cell-surface proteome of a genetically defined cell population in intact tissues. Compared with previous methods for profiling the cell-surface molecular composition, this approach simultaneously enables cell type, subcellular, and temporal specificities in an in situ tissue context. The quantitative profiling of PNs not only provides a detailed view of neuronal cell-surface proteomes at the wiring and functional stages but also systematically uncovers the expression change of individual proteins across these two stages. It was found that the cell surface of developing PNs is highly enriched for wiring molecules and that there is a proteome-wide coordinated change in PN surface molecules during the transition from developing to mature stages. Notably, changes in cell-surface molecule expression at the protein and RNA levels showed considerable discrepancies, suggesting a prominent role of post-transcriptional regulation in shaping the neuronal cell-surface proteome (Li, 2020).
This approach should be readily applicable to profiling proteins on the surface of other cell types, tissues, and organisms by expressing cell-surface-delivered HRP from a transgene in the desired cell type(s). In addition to studying cell-cell communications under physiological conditions, in situ quantitative cell-surface proteomic profiling can be used to decipher proteomic changes in mutants or under pathological conditions. Cell-surface proteins are also major targets for drug development. Probing cell-surface proteome changes under pathogenic conditions and in response to drug application can help identify therapeutic targets and monitor treatment efficacy. From a technological perspective, recent innovations in MS instrumentation, such as parallel accumulation-serial fragmentation, have remarkably enhanced the sensitivity and coverage of low-abundance samples, which could empower cell-surface proteomic analysis of exceedingly small populations of cells. It is anticipated that the combination of genetic strategies for capturing cell-surface proteomes and innovations in MS will further expand the ability to interrogate cell-cell communication mechanisms (Li, 2020).
In the past decades, identification and mechanistic studies of classic wiring molecules have revealed fundamental principles governing neural circuit assembly. Despite these advances, current knowledge is far from explaining the striking precision of connectivity observed in the nervous system. Remarkably, cell-surface proteomic profiling enriched almost all olfactory wiring regulators identified in the past 15 years. This inspired the performance of a proteome-instructed, unbiased screen for previously uncharacterized molecules. Indeed, this screen was exceptionally efficient in uncovering cell-surface molecules controlling circuit assembly. The discovery of LRP1 as a cell-autonomous regulator of dendrite targeting was unexpected because it has been extensively studied in the nervous system for its involvement in Alzheimer's disease, but no in vivo neurodevelopmental function has been reported previously. Given its widely observed role as an endocytic receptor, it is likely that LRP1 controls, via endocytosis, the dynamics of specific ligands or receptors in distinct PN types, accounting for cell-type-specific and stereotyped loss-of-function wiring defects (Li, 2020).
In summary, the precision of neural circuit assembly requires multiple previously unexpected classes of proteins that are conserved from flies to humans. Because many of these proteins would not be identified by a molecular family-based screen, whereas a genome-wide unbiased screen in vivo is technically challenging, this study highlights the power of using temporally resolved cell-surface proteomic profiling to discover regulators of brain wiring. Future investigations of these proteins should expand understanding of the mechanisms by which neural circuits are assembled with exquisite specificity (Li, 2020).
Wiring a complex brain requires many neurons with intricate cell specificity, generated by a limited number of neural stem cells. Drosophila central brain lineages are a predetermined series of neurons, born in a specific order. To understand how lineage identity translates to neuron morphology, this study mapped 18 Drosophila central brain lineages. While large aggregate differences between lineages were found, shared patterns of morphological diversification were discovered. Lineage identity plus Notch-mediated sister fate govern primary neuron trajectories, whereas temporal fate diversifies terminal elaborations. Further, morphological neuron types may arise repeatedly, interspersed with other types. Despite the complexity, related lineages produce similar neuron types in comparable temporal patterns. Different stem cells even yield two identical series of dopaminergic neuron types, but with unrelated sister neurons. Together, these phenomena suggest that straightforward rules drive incredible neuronal complexity, and that large changes in morphology can result from relatively simple fating mechanisms (Lee, 2020).
In order to discern how NBs guide neuronal diversification, It is necessary to appreciate neuronal development at the single-cell level. In other words, individual neurons need to be mapped back to their developmental origins. Achieving this with stochastic clone induction (i.e. labeling GMC offspring as isolated, single-neuron clones and assigning the neurons to specific lineages based on lineage-characteristic morphology) is possible but laborious and can be extremely challenging for lineages that contain similar neurons. Targeted cell-lineage analysis using lineage-restricted genetic drivers is therefore preferred for mapping specific neuronal lineages of interest with single-cell resolution. To date, only three of the about 100 distinct neuronal lineages have been fully mapped at the single-cell level in adult fly brains: the mushroom body (MB), anterodorsal antennal lobe (ALad1) and lateral antennal lobe (ALl1) lineages. These mapped lineages consist of 1) MB Kenyon cells (KC), 2) AL projection neurons (PN), and 3) AL/AMMC PNs and AL local interneurons (LN), respectively. All three lineages produce morphologically distinct neuron types in sequential order, indicating a common temporal cell-fating mechanism. However, the progeny's morphological diversity varies greatly from one lineage to the next. The four identical MB lineages are composed of only three major KC types; moreover, paired KCs from common GMCs show no evidence for binary sister fate determination. By contrast, the two AL NBs produce progeny that rapidly change type (producing upwards of 40 neuron types) and the GMCs generate discrete A/B sister fates. In the ALl1 lineage, differential Notch signaling specifies PNs versus LNs. Notably, the paired PN and LN hemilineages show independent temporal-fate changes, as evidenced by windows with unilateral switches in production of distinct PNs or LNs. Moreover, the ALl1 PN hemilineage alternately yields Notch-dispensable AL and Notch-dependent AMMC PNs. Together, these phenomena demonstrate a great versatility in lineage-guided neuronal diversification (Lee, 2020).
Assembling complex region-specific intricate neural networks for an entire brain requires exquisite cell specificity. In fact, cellular diversity-as characterized by gene expression-is higher during development than in mature brains, signifying that the underpinnings of the connectome can be understood by studying development. Such developmental diversity is reflected by characteristic neurite projection and elaboration patterns. This study therefore aimed to elucidate the roles of NB lineage specification, temporal patterning and binary sister-fate decisions upon neuronal morphology. By doing this, it is hoped to gain insight about how a limited number of NBs can specify such enormous brain complexity. This study mapped a large subset of NB lineages, enough to make generalizations but not so many to confound analysis. To this end, NBs expressing the conserved spatial patterning gene vnd were selected. With single-neuron resolution, 25 hemilineages derived from 18 vnd-expressing NBs were mapped. Hemilineage-dependent morphological diversity were observed at two levels. First, neurons of the same hemilineage may uniformly innervate a common neuropil or differentially target distinct neuropils. Second, neurons show additional structural diversity in terminal elaboration, which depends on neuropil targets rather than lineage origins. Once you factor in the differences of the neuropil targets, hemilineages which seem grossly distinct actually show comparable temporal patterns in the diversification of neuron morphology. Many hemilineages exhibit recurrent production of analogous neuron types and/or cyclic appearance of characteristic morphological features, implicating dynamic fating mechanisms. Moreover, non-sister hemilineages were discovered that make similar or even identical neuron types with common temporal patterns. These observations suggest involvement of conserved lineage-intrinsic cell-fating mechanisms in the derivation of diverse neuronal lineages (Lee, 2020).
Despite having predetermined fates, it is hard to imagine how complex neuronal morphology is controlled. Mapping neuron morphology for 25 hemilineages in this study reveals that primary trajectories and thus neuropil targets are mainly dependent upon both lineage identity and Notch signaling-that is hemilineage identity. Notably, sister hemilineages can vary greatly in the extent of innervation. Hemilineages with excessive coverage areas are consistently associated with the Notchoff (Noff) state. However, the larger coverage area of the B hemilineage could be a result of only a subset of neurons with lengthy projections. At the single-cell level, the average length of the main trajectory (defined as the total length of the segmented neuron after pruning branches shorter than 50 microns) is significantly greater on the Noff than Non side in only two (CREa1 and SMPp&v1) of the seven Vnd lineages composed of dual hemilineages. Instead, evidence was found in support of presence of more diverse morphological groups and/or topological classes of B neurons (as opposed to a dominant group/class of A neurons). First, the degree (coefficient) of variation in the length of the main trajectory is significantly higher on the Noff than Non side. Second, there indeed exist significantly higher numbers of topological neuronal classes in the B than A hemilineages. However, it is unclear if the hemilineage Notch state also affects diversity of neuronal topology in the Drosophila thoracic ganglion with well-defined hemilineages (Lee, 2020).
Notch signaling as a binary switch delivers context-dependent outcomes, including grossly opposite phenotypes. For instance, Notch can promote or suppress neuronal cell death depending on lineage identity. This can lead to unpaired hemilineages, in which only one viable neuron is produced after each GMC division. From the 11 Vnd unpaired hemilineages, seven are Non and four are Noff. Given this random association, it is curious that correlation is seen between Notch state and hemilineage complexity. Higher gross diversity is frequently seen on the Noff side. The same applies to the previously mapped ALl1 lineage where the Non hemilineage consists exclusively of AL local interneurons, whereas its Noff sister hemilineage consists of projection neurons innervating diverse neuropils, including AL, AMMC, LH, PLP, and VLP. However, the striking Notch-dependent LN/PN fate separation appears to be a characteristic of only the ALl1 lineage. This study found instead that neurons of the same hemilineage can adopt various topologies. In fact, both general topology and terminal arborization seem primarily tailored by the targets innervated. Further, unrelated hemilineages with striking similarities (e.g., CREa1A/CREa2A and LALv1A/AOTUv4A) consistently have the same Notch state. Such resemblance across non-sister hemilineages could simply reflect their evolutionary relatedness at the lineage level. Nonetheless, Notch may directly regulate neuropil targeting, as implicated by the complete segregation of the Non and Noff neuronal processes observed in six of the seven (not LALv1) dual lineages. Further, Notch can promote cell adhesion, either by acting as a cell adhesion molecule or by upregulating canonical cell adhesion molecules such as integrins. It is speculated that Notch may strengthen neurite-neurite affinity, as higher affinity in A hemilineages could suppress neurite defasciculation resulting in more uniform trajectory, and facilitate extension of long neurite fascicles. By contrast, reduced affinity in B hemilineages could promote gross diversity through serial defasciculation of primary projections. Further, reduced affinity across sister branches could enhance neurite elaboration within targeted neuropils. Nonetheless, additional factors (e.g., neuropil-characteristic topographic maps) might modulate the gross manifestation of Notch's morphogenetic effects (Lee, 2020).
The orderly derivation of morphologically distinct neuron types within a given hemilineage is indicative of temporal fating. However, the final neuron morphology depends not only on temporal fate, but also on lineage identity and Notch binary sister fate, as well as the anatomy of target neuropils. Despite the complexities, similar temporal features are observable across diverse hemilineages. First, it is common to see beginning neurons with uniquely elaborate projections and ending neurons with reduced morphology (see Common themes of neuronal lineage temporal patterning). Second, there are temporally ordered neuropil targets characteristic of each hemilineage. Although rarely restricted to a single window, most morphological groups show select windows of production. These phenomena indicate long-range temporal patterning. However, recurrent neuropil targeting is also common. Moreover, a comparable series of related neuron types or progressive morphological changes can appear multiple times in a hemilineage. These recurrences suggest repetition of dynamic factors. Taken together, the temporal changes in neuron morphology and targeting indicate the combination of both long-range temporal patterning and reiteration of temporal windows (Lee, 2020).
As to underlying molecular mechanisms, the observed birth order-dependent neuronal morphogenesis is unlikely due to the environmental differences over the course of larval neurogenesis, since the final neuronal targeting and innervation occur in a rather synchronized manner at the early pupal stage. Further, some of the temporal patterning phenomena may be explained by the intrinsic temporal factors that have been previously described in the literature. The Cas and Svp embryonic temporal transcription factors are expressed in NBs during early postembryonic neurogenesis and are thus candidates to promote the uniquely exuberant neurite projections in first-born neurons. Opposing temporal gradients of Imp and Syp RNA binding proteins in cycling NBs have been shown to control neuronal temporal fate in MB, AL, and complex type II lineages as well as global NB termination. Imp/Syp are likely to govern long-range temporal patterning of most, if not all, neuronal lineages. Imp and Syp gradients shape the descending protein gradient of Chinmo. The hierarchical temporal gradients of Imp/Syp and Chinmo could define serial temporal windows with expression of various terminal selector genes. For instance, Mamo (a temporally patterned terminal selector gene) is selectively expressed in the window defined by both weak Chinmo and abundant Syp in MB and AL lineages (Liu, 2019; Lee, 2020 and references therein).
However, it is not known exactly how the opposite Imp/Syp protein gradients can define ~30 serial temporal fates in a protracted neuronal lineage. Interestingly, recurrent production of related neuron types has emerged as a dominant theme in the temporal patterning of Vnd lineages. Dynamic Notch signaling may underlie some alternating temporal fates, as Notch has been shown to control alternate production of AL and AMMC projection neurons in the lateral AL lineage. However, it is unlikely that Notch alone can mediate multiple recurring features as seen in most Vnd hemilineages. Therefore involvement of parallel recurring factors is proposed to elicit unsynchronized repetition of distinct features. In sum, there likely exist multiple temporal fating mechanisms that act in concert to expand neuron diversity, thus resulting in complex temporal patterns (Lee, 2020).
Given the need for large neuronal diversity, it was surprising to see the production of two identical, long series of dopaminergic neuron types by CREa1A and CREa2A. Strikingly, the NB homology extends throughout postembryonic neurogenesis. The FB neurons born prior to dopaminergic neurons in CREa1A are also morphologically indistinguishable from those in CREa2A. The only difference is the selective loss of the first larval-born CREa2A neuron. Contrasting the almost identical CREa1A and CREa2A hemilineages, their paired sister hemilineages (CREa1B and CREa2B) are easily distinguishable, as only CREa1B neurons cross the midline. These phenomena implicate that neighboring CREa1 and CREa2 lineages may have arisen from NB duplication followed by a change in midline crossing. Thus, it is believed that one way for brain complexity to increase is through lineage duplication and subsequent divergence (Lee, 2020).
In conclusion, this high-resolution, comprehensive analysis of Vnd lineages reveals how a complex brain can be reliably built from differentially fated neural stem cells. This seminal groundwork lays an essential foundation for unraveling brain development from genome to connectome (Lee, 2020).
Object-based attention describes the brain's capacity to prioritize one set of stimuli while ignoring others. Human research suggests that the binding of diverse stimuli into one attended percept requires phase-locked oscillatory activity in the brain. Even insects display oscillatory brain activity during visual attention tasks, but it is unclear if neural oscillations in insects are selectively correlated to different features of attended objects. This question was addressed by recording local field potentials in the Drosophila central complex, a brain structure involved in visual navigation and decision making. Attention was found to selectively increase the neural gain of visual features associated with attended objects; attention could be redirected to unattended objects by activation of a reward circuit. Attention was associated with increased beta (20- to 30-Hz) oscillations that selectively locked onto temporal features of the attended visual objects. These results suggest a conserved function for the beta frequency range in regulating selective attention to salient visual features (Grabowska, 2020).
Selective attention refers to the brain's capacity to focus on a subset of stimuli while ignoring others. While subjectively intuitive in humans, selective attention has also been documented in a wide variety of animals, such as other primates, birds, and even insects. What is attended to depends on stimulus salience (e.g., loudness or brightness), as well as on the perceived value of a stimulus and the motivational state of the animal. What is attended to also depends on what is perceived as a singular object. Object-based attentionrefers to the capacity to direct attention to a conjunction of different features linked as part of the same object. Attending to one feature of a given object would thus enhance not only the neural representation of that particular feature, but also other features that are associated with the object. How this form of generalization works is not entirely understood but seems to require some form of feature binding to first determine which stimuli belong together as a unified object and then to link the object to some inherent value, or valence. Hence, feature binding appears to be essential for object-based attention, as neural gain has to be allocated to specific features first in order to perceive an object as a whole. At the same time, distinct stimulus features can become unbound from an attended object if they are selectively ignored (Grabowska, 2020).
In the mammalian brain, feature binding and object-based attention have been proposed to be associated mechanisms, both of which seem to be facilitated by synchronized activity of neuronal assemblies, which can be detected as phase-locked neural oscillations. In particular, oscillations in the range of 13 to 30 Hz (beta) and 30 to 80 Hz (gamma) seem to reflect this form of binding based on their strong synchronization at various time points following visual or auditory stimulation, with distinct oscillatory processes potentially reflecting different levels of perception. For example, early (<100-ms) stimulus-evoked synchronization in the gamma range has been suggested to represent rapid integration of unconscious sensory processes, whereas later (200- to 400-ms) synchronization in both the beta and gamma range is hypothesized to reflect feature binding and conscious perception. Stimulus-evoked beta and gamma oscillations would thus represent a phase reset of ongoing neuronal activity associated with enhancing attentional gain for specific features, by facilitating information transfer or binding among different brain regions (Grabowska, 2020).
While there is neural evidence for object-based attention in nonhuman primates, it is unknown if the smallest animal brains, such as those of insects, combine diverse sensory stimuli into unified percepts, or if they even have a subjective awareness. Behavioral studies in honeybees suggest that some insects can detect illusory contours as single objects and can group distinct stimuli into abstract concepts such as 'sameness' or 'difference', which could indicate a form of categorization through object-based attention. Similarly, visual learning paradigms for Drosophila melanogaster have uncovered a capacity for context generalization, where flies perceived visual objects as the same despite changes in color, suggesting they were attending to the object shape feature and ignoring color cues. There is growing evidence for attention-like processes in insects, such as during visual fixation, decision making, and novelty detection in Drosophila flies, as well as multiple object tracking in dragonflies. The latter electrophysiological study uncovered motion-detecting neurons in dragonflies that selectively lock onto the timing or phase of salient objects, which was shown by 'tagging' competing objects with distinct flicker frequencies. However, it is unknown how such selective neural processes are controlled in the insect brain or whether these neural measures are relevant to behavioral decision making (Grabowska, 2020).
It is possible that the insect brain, like the mammalian brain, employs oscillatory activity and stimulus-evoked phase locking to prioritize and bind stimulus features, and to enhance attentional gain. Indeed, earlier electrophysiological studies revealed oscillatory activity in the 20- to 30-Hz range in the Drosophila brain that was associated with detecting salience effects, such as visual novelty, suggesting that these endogenously generated oscillations might be more broadly involved in regulating attention-like processes in the fly brain. However, it remained unclear which neurons might be generating these oscillations. One likely neuropil is the central complex (CX), a heterogeneous structure in the central brain that has been associated with visual pattern learning. Recent studies in behaving Drosophila also identified the CX as a key brain region for visual navigation. This suggests a broader role for the CX in directing attention-like processes, which could also reflect ring attractor dynamics within CX circuits. While its role in visual perception is increasingly evident, whether the CX produces neural oscillations relevant to visual attention and feature binding is unknown. To address this question requires not only measuring electrical activity in the CX of behaving flies, but also correlating any endogenous brain activity to distinct neural signatures associated with competing visual stimuli or stimulus features (Grabowska, 2020).
In tethered virtual reality experiments, flies tend to fixate on large objects and avoid small objects, whether they are flying or walking. This innate visual dichotomy was exploited to examine mechanisms underlying visual selective attention in Drosophila. To disambiguate between the attractive and aversive stimuli in the fly brain, and to relate neural activity to ongoing behavioral choices, local field potentials (LFPs) were recorded from the CX and the competing visual stimuli flicker at distinct frequencies, thereby evoking steady-state visually evoked potentials (SSVEPs) in the fly brain. It was first shown that the SSVEPs varied in amplitude depending on the visual objects being fixated upon, allowing investigation of how attention guided the binding of different visual features, such as object size, brightness, and flicker frequency. By calculating phase-locking strength between the distinct SSVEPs and endogenous brain activity, this study then examined how oscillations in the CX interacted with one another. Frequency-specific phase locking was found between endogenous oscillations in the 20- to 30-Hz frequency range and the object features that the fly paid attention to, suggesting that beta-like oscillations could be employed for object-based attention in the insect brain (Grabowska, 2020).
The brain's ability to link complex patterns of sensory input into coherent objects has been termed the binding problem, with the 'problem' being that it remains unclear how diverse sensory streams are unified into a single conscious percept. Subjective experience of the world in humans is of discretely bound units rather than segregated sensory streams, and this capacity of the human brain is probably adaptive as sensory cues are often correlated, such as voices with faces or fruits with colors. The adaptive advantage of perceiving the world in such a unitary fashion raises the question of whether other brains do this and if so, whether the selective attention mechanisms observed in simpler animals such as insects facilitate a form of feature binding (Grabowska, 2020).
In humans, the modulation of endogenous beta (15- to 30-Hz) oscillations is associated with the perception and integration of visual stimuli, as well as decision making, among other cognitive functions. Intriguingly, beta oscillations have also been associated with task-related engagement and reward processing as well as stimulus-locked attentional load effects. However, a full understanding of how beta oscillations are deployed to achieve these functions is lacking, and there remains debate regarding their functional role. The finding of beta-like oscillations involved in object-based attention in the insect brain lends support to the view that these oscillations perform a conserved function relevant to perception, as it seems unlikely that a completely different neuroanatomy (an insect brain) would have preserved a neural epiphenomenon. Consistent with a causal role for oscillations in the insect brain, this study found that dNPF circuit activation in Drosophila increased 20- to 30-Hz activity in the CX, which promoted phase locking to attended visual stimuli. Interestingly, in open-loop conditions dNPF activation seemed to produce a valence reversal, suggesting that attention was redirected covertly to the smaller, aversive object. Why the smaller object should have higher value in this specific open-loop context remains unclear. An alternative interpretation of this result is that salience for the smaller object was increased, rather than it having been rendered more attractive. Thus, the dNPF circuit might be more involved with regulating salience rather than valence, and the salience of the aversive small bar could thus have been magnified by dNPF activation in open-loop conditions, when the fly is not in control. Interestingly, the increased salience assigned to the competing small bar was associated with increased walking speed, suggesting a motivation to respond behaviorally. Electroencephalography (EEG) studies have found that when humans have no control over an array of emotionally laden visual images, these images evoke a higher SSVEP response compared with emotionally neutral images; however, strongly aversive images evoked the greatest SSVEP responses of all. Perhaps similarly in the fly brain, an uncontrollable aversive object becomes much more salient upon dNPF activation. This highlights the importance of accounting for behavioral control in any understanding of brain functions underlying perception, including flies in open- vs. closed-loop experiments (Grabowska, 2020).
Beta-like oscillations have been observed previously in the insect brain. For example, recordings in the locust have identified 20- to 30-Hz oscillations associated with processing of olfactory stimuli, and comparable oscillations have also been associated with visual attention in flies. Additionally, there is increasing evidence that insect brains employ a variety of oscillations, comparable in range with the mammalian brain. These include 7 to 12 Hz (alpha) 20 to 50 Hz, and even 1 Hz (delta). These oscillations have been shown to be involved in processes such as olfaction, vision, and sleep, suggesting conserved functions that might transcend the differences in brain architecture between insects and mammals. Whether any of these oscillations are functionally comparable remains to be seen. Nevertheless, the current findings suggest that beta-like oscillations might be employed by the insect brain to bind different stimulus features into unified percepts that guide the animal's attention. Although this study did not investigate nonvisual stimulus modalities in this study, previous work has demonstrated that odors modulate the amplitude of visually evoked 20- to 30-Hz activity (van Swinderen, 2003), suggesting these oscillations might govern cross-modal binding as well. Whether endogenous 20- to 30-Hz activity in the fly brain is performing a similar function to beta oscillations in the human brain remains an open question. It is, however, possible that oscillatory processes are supported by different brain architectures that have conserved circuit timing relationships through evolution. Such conservation might be expected if these oscillations were performing a key function for a variety of adaptive behaviors, such as navigation, finding food, or avoiding predators. The current study suggests that oscillations in the beta range (20 to 30 Hz) are indeed performing an important phase-locking function to choreograph meaningful information and thereby, guide selective attention. Although mammalian and fly brains are obviously different, they share some organizational principles that could support the preservation of such oscillatory functions (Grabowska, 2020).
To determine whether the significant phase-amplitude correlations observed were due to a physiologically relevant shift between SSVEP phases and endogenous 20- to 30-Hz oscillations, rather than just due to increases in stimulation frequency amplitudes, a simulation was performed where the amplitudes of only the SSVEPs were artificially increased while keeping other frequencies constant. In the simulation, no effect was found of local field potentials (LFPs) amplitudes on envelope to signal correlation (ESC). The simulation was then repeated with a specific increase in 20- to 30-Hz and 30- to 40-Hz amplitudes and no correlation to the SSVEPs was seen. This indicates that the phase correlations observed in real fly brain activity are functionally relevant and not a by-product of multiple superimposed oscillations of varying amplitudes (Grabowska, 2020).
Although 20- to 30-Hz activity stood out as relevant for phase locking to attended objects, other endogenous frequencies showed significant changes upon visual stimulation. In open-loop conditions, visual stimulation alone (without NPF activation) led to an increase in phase-amplitude coupling between SSVEPs and endogenous frequencies in the gamma range (30 to 50 Hz). In the mammalian brain, gamma oscillations have been proposed to provide different functions in sensory processing, depending on the frequency range and timing poststimulus induction. For example, EEG activity in the lower gamma range (30 to 40 Hz) can be elicited by brief and steady visual stimuli, and an increase in oscillatory power for this frequency range can be observed up to 100 ms after stimulation. One idea is that these stimulus-locked gamma oscillations might be relevant for rapid (i.e., unconscious) integration processes that might not necessarily be stimulus relevant. Nevertheless, gamma oscillations in humans can also be significantly modulated by attention and stimulus saliency. In contrast, a nonstimulus-locked component in the gamma range, occurring around 250 to 350 ms after stimulus presentation, has been proposed to be more relevant for object representation. Intriguingly, a similar frequency shift is seen in the fly brain. An increase of 30- to 50-Hz phase locking was seen when visual stimuli were presented, while 20- to 30-Hz phase locking predominated upon NPF circuit activation. In humans, it has been shown that synchronized oscillations in the gamma and beta ranges have a high degree of interdependence, showing a so-called 'gamma-to-beta' transition in response to novel auditory stimuli, for example. Whether a gamma-to-beta transition is also occurring in the insect brain, associated with visual perception, remains difficult to address because any evidence for perception in flies must ultimately depend on behavior, which occurs on a slower timescale than stimulus-evoked neural oscillations (Grabowska, 2020).
By grounding this study on innate visual preferences, it could be, however, inferred what the flies were most likely paying attention to. It was found that an innately attractive visual object evokes a greater response in the fly brain than an aversive object and that this effect is preserved even under open-loop conditions, when flies are not in control. This suggests a neural correlate of object-based attention or in other words, a brain signal that correctly identifies what a fly is paying attention to-even in the absence of correlated behavior. Although this remains speculative, future experiments tapping directly from this brain signal in closed-loop paradigms should be able to test if it indeed provides a level of cognitive control (Grabowska, 2020).
Various insect species serve as valuable model systems for investigating the cellular and molecular mechanisms by which a brain controls sophisticated behaviors. In particular, the nervous system of Drosophila melanogaster has been extensively studied, yet experiments aimed at determining the number of neurons in the Drosophila brain are surprisingly lacking. Using isotropic fractionator coupled with immunohistochemistry, this study counted the total number of neuronal and non-neuronal cells in the whole brain, central brain, and optic lobe of Drosophila melanogaster. For comparison, neuronal populations were counted in three divergent mosquito species: Aedes aegypti, Anopheles coluzzii and Culex quinquefasciatus. The average number of neurons in a whole adult brain was determined to be 199,380 ±3,400 cells in D. melanogaster, 217,910 ±6,180 cells in Ae. aegypti, 223,020 ± 4,650 cells in An. coluzzii and 225,911±7,220 cells in C. quinquefasciatus. The mean neuronal cell count in the central brain vs. optic lobes for D. melanogaster (101,140 ±3,650 vs. 107,270 ± 2,720), Ae. aegypti (109,140 ± 3,550 vs. 112,000 ± 4,280), An. coluzzii (105,130 ± 3,670 vs. 107,140 ± 3,090), and C. quinquefasciatus (108,530 ±7,990 vs. 110,670 ± 3,950) was also estimated. Each insect brain was comprised of 89% ± 2% neurons out of its total cell population. Isotropic fractionation analyses did not identify obvious sexual dimorphism in the neuronal and non-neuronal cell population of these insects. This study provides experimental evidence for the total number of neurons in Drosophila and mosquito brains (Raji, 2021).
Anatomical connectivity can constrain both a neural circuit's function and its underlying computation. This principle has been demonstrated for many small, defined neural circuits. For example, connectome reconstructions have informed models for direction selectivity in the vertebrate retina as well as the Drosophila visual system. In these cases, the circuit in question is relatively compact, well-defined, and has known functions. However, how the connectome constrains global properties of large-scale networks, across multiple brain regions or the entire brain, is incompletely understood. As the availability of partial or complete connectomes expands to more systems and species it becomes critical to understand how this detailed anatomical information can inform understanding of large-scale circuit function. This study used data from the Drosophila connectome in conjunction with whole-brain in vivo imaging to relate structural and functional connectivity in the central brain. A strong relationship was found between resting-state functional correlations and direct region-to-region structural connectivity. The relationship between structure and function varies across the brain, with some regions displaying a tight correspondence between structural and functional connectivity whereas others, including the mushroom body, are more strongly dependent on indirect connections. Throughout this work, features were observed of structural and functional networks in Drosophila that are strikingly similar to those seen in mammalian cortex, including in the human brain. Given the vast anatomical and functional differences between Drosophila and mammalian nervous systems, these observations suggest general principles that govern brain structure, function, and the relationship between the two (Turner, 2021).
Previous work looking at specific circuits, for example, in the mushroom body and central complex, described relationships between connectome structure and circuit function. The release of a nearly complete central brain connectome combined with the ability to measure whole-brain functional activity in defined brain regions has allowed relating synapse-level structural connectivity to mesoscale functional connectivity both within and across brain regions. It was found that cell-level anatomical connectivity provides a strong constraint on resting-state functional connectivity, with over half of the variance in functional correlations being accounted for by the variance in the number of cells that connect two regions (Turner, 2021).
Although the direct, inter-region structural connectivity was broadly predictive of functional correlations, indirect pathways could also contribute significantly to functional connectivity. Indirect pathways are thought to be important in many circuits, including in human cortex. Indirect connections are especially important for some brain regions in particular, including the mushroom body, which is associated with multisensory integration and learning, suggesting that dense, indirect connectedness might be important for these computations. Conversely, brain regions in the SNP (superior neuropils), VMNP (ventromedial neuropils), and VLNP (ventrolateral neuropils) showed strong structural connectedness to other regions compared to their functional connectivity. The structural connectivity among these regions may support activity patterns not explored in the fly's resting state, and other behavioral states or stimulation conditions may reveal different functional connectivity for regions like these (Turner, 2021).
One metric of indirect connections between two regions was used, namely the shortest path distance, but one could imagine other indirect paths also contributing to functional connectivity in a meaningful way. Indeed, recent work developed a graph-embedding procedure to predict functional connectivity from structural connectivity in the human brain, which allows for the influence of higher order interactions in general. It is suspected that, rather than the specific shortest path being of particular importance in shaping functional connectivity between a pair of regions, the shortest path distance is a proxy for more-general indirect path connectedness. Disentangling which higher order interactions most strongly shape functional connectivity will be an important task in future work (Turner, 2021).
An illuminating approach to understanding large-scale structure-function relationships in the brain has emerged in human neuroscience, where macro-scale structural connectivity patterns can be related to functional connectivity using noninvasive imaging or electrical recording techniques. Several of these studies have shown that functional connectivity can be predicted by structural connectivity to some degree. The structure-function correlation found in the Drosophila central brain is higher than typically reported in macro-scale networks in the human brain, where MRI-based structural and functional connectivity is similarly positively correlated but generally with lower correlation coefficients. There are many factors that could contribute to this difference, including different levels of temporal and spatial precision in the functional measurements, vastly different amounts of biological detail in the structural connectivity measurements, and/or genuine biological differences between the two systems. Interestingly, the inclusion of detailed synaptic information (i.e., the number of synapses associated with a given inter-region connection) did not increase the correlation between structural and functional connectivity. This suggests that synapse-level connectivity does not much constrain functional connectivity at the mesoscale level, despite synapse count being very important for smaller, functionally defined circuits. Mesoscale functional correlations reflect the influence of many neurons that individually belong to different functional circuits. Although these individual neurons can vary substantially in their synapse number, when averaging over large brain regions, the overall synapse count connectivity is highly correlated with the cell count connectivity (Turner, 2021).
In addition to the general correspondence between structural and functional connectivity, this study found a number of other intriguing similarities between Drosophila central brain networks and networks in mammalian cerebral cortex. For example, the structural network in the Drosophila central brain showed signs characteristic of a small-world network, which has been shown before in human structural and functional networks. This study also found that inter-region connectivity strengths are log-normally distributed, which is also the case in mouse and monkey cortex. The observation that brain networks with such different biological and anatomical features share a key network topology suggests a deep correspondence in either developmental rules and/or functional constraints. It is not unusual for biological variables to be log-normally distributed. Indeed, multiplicative interactions among many variables will tend to produce log-normally distributed variables. In the context of nervous system development, interactions between pre- and post-synaptic cells that are shaped by cooperative combinations of adhesion and signaling molecules will be multiplicative. It is speculated that the shared features between fly and vertebrate brains that can be observed at the mesoscale are the result of these universal phenomena at the microscale (Turner, 2021).
Brains contain networks of interconnected neurons and so knowing the network architecture is essential for understanding brain function. This study therefore mapped the synaptic-resolution connectome of an entire insect brain (Drosophila larva) with rich behavior, including learning, value computation, and action selection, comprising 3016 neurons and 548,000 synapses. Neuron types, hubs, feedforward and feedback pathways, as well as cross-hemisphere and brain-nerve cord interactions were characterized. Pervasive multisensory and interhemispheric integration, highly recurrent architecture, abundant feedback from descending neurons (DNs), and multiple novel circuit motifs were found. The brain's most recurrent circuits comprised the input and output neurons of the learning center. Some structural features, including multilayer shortcuts and nested recurrent loops, resembled state-of-the-art deep learning architectures. The identified brain architecture provides a basis for future experimental and theoretical studies of neural circuits (Winding, 2023).
This study presents a synaptic-resolution connectivity map of an entire Drosophila larva brain and a detailed analysis of the associated brain circuit architecture. Each neuron was split into two compartments, axon and dendrite, resulting in a rich multiplexed network with four connection types, facilitating analysis. To characterize long-range brainwide anatomical pathways, an algorithm was developed that utilizes synapse numbers between neurons to track signal propagation across polysynaptic pathways (Winding, 2023).
Neuron types have been classified based on their functional role, morphology, gene expression, or combinations of features. Although these features are likely correlated, it is still unclear which is ideal for defining neuron types and how neuron types based on different features correspond to each other. An unbiased hierarchical clustering of all neurons was performed using synaptic connectivity alone and 93 types were identified. The morphology of neurons within clusters was notably similar. Furthermore, neurons that had similar known functions were usually found in the same or related clusters. Thus, clustering neurons based on synaptic connectivity resulted in clusters that were internally consistent for other features, when those features were known. However, many clusters contained uncharacterized neurons with unknown gene expression and function (Winding, 2023).
Noncanonical connection types are pronounced in learning and action-selection circuits
Although most connections in the brain were a-d (66.4%), a significant number of a-a (26.4%), d-d (5.4%), and d-a (1.8%) connections were found. Most neurons that received prominent axonic input were in the learning center: DANs that provide the teaching signals for learning and KCs that encode stimuli. Modulatory a-a DAN-to-KC input drives heterosynaptic plasticity of the KC-to-MBON synapse. DANs also receive excitatory a-a input from KCs, which provides positive feedback that facilitates memory formation. KCs also receive a-a input from other KCs. In the adult Drosophila, a-a connections between otherwise excitatory (cholinergic) KCs were found to be inhibitory due to expression of inhibitory mAChR-B in axon terminals. Lateral inhibition between KCs could improve stimulus discrimination and reduce memory generalization. A subset of pre-DNsVNC and a few somatosensory projection neurons (PNs), lateral horn neurons (LHNs), and MBONs, and fan-shaped body neurons (FBNs) also had a high axonic input/output ratio. If a-a connections in these neurons are inhibitory they could enhance contrast between representations of distinct stimuli and actions (Winding, 2023).
Edges were observed with multiple connection types between neurons, including up to all four types simultaneously. The most common combination, axo-dendritic with axo-axonic, may grant the presynaptic neuron post- and presynaptic control of the downstream neuron, as has been observed in triad motifs in mammals (Winding, 2023).
Pathways from sensory to output neurons form a multilayered distributed network
Multiple parallel pathways of varying depths were observed downstream of each modality, albeit with extensive interconnectivity between different pathways. This architecture suggests that distinct features may not be processed independently but rather that each feature may potentially influence the computation of many other features in a distributed network. Such architecture has the potential to generate a diversity of neural responses with mixed selectivity for specific combinations of features thereby expanding the dimensionality of neural representations and increasing output flexibility (Winding, 2023).
The shortest paths from sensory neurons to output neurons were found to be surprisingly shallow. All output neurons receive input from sensory neurons within a maximum of 3 hops. However, most output neurons also received input from the same modality through multiple longer pathways. Such an architecture, with connections that skip layers, is characteristic of prominent machine learning networks, including deep residual learning and U-Net architectures. Although predictive accuracy improves with depth, features can become too abstract at deep layers leading to performance degradation. Shortcuts between layers can solve this problem by combining lower-level features as an additional teaching signal. Shallower networks with shortcuts can therefore exceed the performance of deeper networks lacking shortcuts. The layer skipping observed may therefore increase the brain's computational capacity, overcoming physiological constraints on the number of neurons that limit network depth (Winding, 2023).
Recurrence has been observed in many brain circuits and implicated in a range of computations. However, the architecture of long-range recurrent pathways and the nature of the feedback that each neuron receives is still poorly understood. Signal cascades were used to systematically identify all connected pairs of brain neurons (with up to 5 hops) that had a reciprocal connection (of up to 5 hops). It was found that 41% of brain neurons received long-range recurrent input (up to 5 hops) from at least one of their downstream partners with recurrent pathways of varying lengths forming multiple nested loops. Recurrent nested structure can compensate for a lack of network depth in artificial neural networks and supports arbitrary, task-dependent computation depth (Winding, 2023).
Learning center dopaminergic neurons are amongst the most recurrent in the brain
DANs were amongst the most recurrent neurons in the brain. Dopaminergic neurons, referred to as DANs in insects, are central for learning, motivation, and action across the animal kingdom and are implicated in a range of human mental disorders. The highly recurrent connectivity of DANs might deliver high-dimensional feedback, enabling them to encode a range of features and flexibly engage in parallel computations. Recurrent excitatory loops could also play roles in working memory (Winding, 2023).
Previous studies have reported that DANs receive extensive feedback from neurons that integrate learned and innate values. This study found that DANs also receive long-range feedback (up to 5 hops) from descending neurons, which likely encode motor commands. Furthermore, this study found that DANs receive polysynaptic feedforward inputs from all sensory modalities. DAN activity correlates with movement in both vertebrates and flies, which could be explained by the observed input from DNsVNC or from proprioceptive neurons (Winding, 2023).
Hub neurons have been shown to play essential roles in behavior. Most (73%) of the larval brain's in-out hubs were postsynaptic to the learning center output neurons (MBONs) and/or presynaptic to the learning center modulatory neurons (mostly DANs). Many were also postsynaptic to the LH that mediates innate behaviors, thus integrating learned and innate values. One of these hubs, MBON-m1, has been shown to compute overall predicted value by comparing input from neurons encoding positive and negative values. MBON-m1 bidirectionally promotes approach or avoidance when its activity is increased or decreased, respectively. Several additional hubs identified in this study have similar patterns of input to MBON-m1, suggesting that they may play similar roles in computing predicted values. These hubs provide direct feedback to the MB DANs and could therefore play roles in regulating learning (Winding, 2023).
This study identified all contralaterally projecting neurons and their connections, providing a basis for understanding how information from both hemispheres is used by the brain. Notably, neurons with contralateral axons were disproportionately represented amongst in-out hubs, suggesting that they have important roles in behavior. Contralateral neurons tended to synapse onto each other. Thus, after integration of contra- and ipsilateral information in one hemisphere, the integrated information is often passed back to the other hemisphere. Multiple consecutive hemisphere crossings could potentially enable better discrimination between ipsilateral, contralateral, or bilateral events and better coordination between the two hemispheres. This study also discovered multiple reciprocal pair loops between contralateral left-right homologs. If inhibitory, pair loops could mediate interhemispheric comparisons, and if excitatory, they could be involved in signal perpetuation or short-term memory. Consistent with this idea, many pair loops occurred between neurons presynaptic to the MB DANs (Winding, 2023).
This study sheds light on brain-nerve cord interactions. DNs targeted only a small fraction of premotor elements that could play important roles in switching between locomotor states. A subset of DNs targeted low-order post-sensory interneurons likely modulating sensory processing. DNs and ascending neurons (ANs) also synapsed onto each other, often forming zigzag motifs (DN1→AN→DN2). A recent study has demonstrated that an AN can activate the downstream DN and drive the same action as the DN. Thus, ANs may facilitate DN activation and transitions between actions based on proprioceptive feedback or somatosensory context. Somatosensory neurons have been shown to activate descending neurons in vertebrates, raising the possibility that ascending-descending connectivity may be a general feature of brain-nerve cord interactions (Winding, 2023).
Recent connectome analyses of the entire synaptic circuit in the nervous system have provided tremendous insights into how neural processing occurs through the synaptic relay of neural information. Conversely, the extent to which ephaptic transmission which does not depend on the synapses contributes to the relay of neural information, especially beyond a distance between adjacent neurons and to neural processing remains unclear. This study shows that ephaptic transmission mediated by extracellular potential changes in female Drosophila melanogaster can reach >200 μm, equivalent to the depth of its brain. Furthermore, ephaptic transmission driven by retinal photoreceptor cells mediates light-evoked firing rate increases in olfactory sensory neurons. These results indicate that ephaptic transmission contributes to sensory responses that can change momentarily in a context-dependent manner (Ikeda, 2022).
This study found that ephaptic transmission of light information from photoreceptor cells in the retina mediates the increase in firing rate in the olfactory sensory neurons (OSNs) during odor stimulations. This study has not revealed whether the ephaptic transmission directly changes the firing rate of the OSNs. Amputation of the antennal nerve abolished the firing rate increases during sustained light, suggesting that once the light information might be received by neurons in the brain, the information would be relayed by the neurons through the antennal nerve to the antenna, resulting in the firing rate increases in the OSNs (Ikeda, 2022).
While ephaptic coupling has been reported earlier, such as between neighboring neurons within the same sensillum, or between Purkinje cells, which is at a distance of <100 μm, this study shows that ephaptic transmission reaches >200 μm in vivo, equivalent to the depth of the entire fly brain, beyond the distance between neighboring neurons. Light stimulations cause -10 mV field potential deflections in a retina. If endogenous fields in the brain are neglected, light stimulations may induce ~33.3 mV/mm electric field between the retina and center of the brain (0 mV), since the distance between them is ~300 μm. This electric field is strong enough to modulate neural activities, as even weaker electric fields (<0.5 mV/mm) changed the firing patterns of neurons in vitro (Weiss and Faber, 2010) (Ikeda, 2022).
In rodents, the firing rate of cerebellar Purkinje cells either decreased or increased when a current was injected into the extracellular field around their axons, causing field potential changes of 0.2 mV. In insects, odor-evoked field potential oscillations whose amplitude is comparable with that caused by the current injection in the rodents, are induced by synchronous firing of olfactory neurons in the antennal lobe which are mediated by GABAergic neurons forming reciprocal synapses with excitatory projection neurons. Changes in the extracellular field potential are commonly observed in many nervous systems. While such extracellular field potential activities have been considered as a side effect of synchronized spiking of neurons, this study suggests that such field potential changes evoked by a sensory stimulus can control the excitability of distant neurons, in addition to adjacent neurons. As ephaptic transmission is more effective at a short distance, the ephaptic transmission from the retinae may contribute significantly to firing rate changes in downstream neurons of the photoreceptor cells in the optic lobe (Ikeda, 2022).
This study also revealed that odor responses of OSNs were clearly modulated when light conditions changed transiently. This mechanism may help flies switch attention to newly presented sensory cues or maintain attention toward those remaining after the change. Turning the light on, for example, reduces the firing rates of the OSNs, which may enable the flies to pay more attention to visual information, whereas turning the light off increases the firing rates of the OSNs, which may help them attend to olfactory sensory cues (Ikeda, 2022).
Recent connectome analyses have revealed the entire synaptic network in the CNS in Drosophila and provides insight into how neural information is subject to synaptic relays to determine the behavioral output. This study has shown that ephaptic relays also contribute to modulating the firing rate of distant neurons and modify the sensory responses that can change momentarily in a context-dependent manner should also be considered. To build an integrated model of the fly brain, ephaptic relay of neural information should be considered. The compound eye-antenna model would be a suitable model to determine the role of ephaptic transmission in neural processing (Ikeda, 2022).
Nociception is a neural process that animals have developed to avoid potentially tissue-damaging stimuli. While nociception is triggered in the peripheral nervous system, its modulation by the central nervous system is a critical process in mammals, whose dysfunction has been extensively implicated in chronic pain pathogenesis. The peripheral mechanisms of nociception are largely conserved across the animal kingdom. However, it is unclear whether the brain-mediated modulation is also conserved in non-mammalian species. This study shows that Drosophila has a descending inhibitory mechanism of nociception from the brain, mediated by the neuropeptide Drosulfakinin (DSK), a homolog of cholecystokinin (CCK) that plays an important role in the descending control of nociception in mammals. Mutants lacking dsk or its receptors are hypersensitive to noxious heat. Through a combination of genetic, behavioral, histological, and Ca(2+) imaging analyses, neurons involved in DSK-mediated nociceptive regulation were subsequently revealed at a single-cell resolution, and a DSKergic descending neuronal pathway was identified that inhibits nociception. This study provides the first evidence for a descending modulatory mechanism of nociception from the brain in a non-mammalian species that is mediated by the evolutionarily conserved CCK system, raising the possibility that the descending inhibition is an ancient mechanism to regulate nociception (Oikawa, 2023).
This study has demonstrated that (1) DSK and its receptors CCKLR-17D1 and CCKLR-17D3 are involved in negatively regulating thermal nociception. (2) Two sets of brain neurons, MP1 and Sv, express DSK in the larval nervous system. (3) One of the DSK receptors (CCKLR-17D1) functions in Goro neurons in the VNC to negatively regulate thermal nociception, and (4) Thermogenetic activation of DSK-2A-GAL4 neurons attenuates both larval nociception and the activity of Goro neurons. Based on these results, it is proposed that the DSKergic neurons regulating the activity of Goro neurons constitute a descending inhibitory pathway of nociception from the brain to the VNC in larval Drosophila. These findings represent the first evidence of a descending mechanism modulating nociception from the brain in a non-mammalian species (Oikawa, 2023).
DSK has been implicated in multiple physiological and developmental processes in Drosophila. Previous studies have shown that dsk and CCKLR-17D1 mutants exhibit significant reductions of synaptic growth and excitability in larval neuromuscular junctions (NMJ) and larval locomotion under bright light, suggesting the importance of the DSK/CCKLR-17D1 signaling pathway in the developmental processes of motoneurons. However, in this study, no major developmental defects were observed in Goro neurons with CCKLR-17D1 RNAi. Furthermore, the simultaneous thermogenetic activations of DSK-2A-GAL4 neurons and C4da nociceptors inhibited larval nociception and the activity of Goro neurons. Given the pronociceptive role of Goro neurons, their reduced synaptic or neuronal activity should cause nociceptive insensitivity or reduced Ca2+ responses, which is contradicting to the observation that CCKLR-17D1 knockdown in Goro produced the thermal hypersensitivity and exaggerated Ca2+ responses. Thus, these data consistently support a physiological role for DSKergic descending signals in the modulation of the activity of Goro neurons rather than a developmental role, highlighting the functional differences of the DSK/CCKLR-17D1 pathway between the NMJ and the nociceptive system in the CNS (Oikawa, 2023).
A previous study has reported the expression of DSK in a population of larval insulin-producing cells (IPCs) and its functions in responding to starving conditions. However, this study observed that anti-FLRFa signals in IPCs persisted in dsk null mutants and that anti-FLRFa signals in IPCs, detected with an antibody against a crustacean neurohormone, hardly overlap with either DSK-GAL4 or DSK-2A-GAL4 expressions. Although a different antibody from that used in the study mentioned above, the number of IPCs visualized by anti-FLRFa was comparable to that of cells visualized by anti-DSK. Regarding the functions of IPCs in nociception it has been shown that silencing dilp2-GAL4 positive IPCs has no significant effect on the baseline nociceptive responses. In CaMPARI2 imaging, IPCs marked by DSK-2A-GAL4 showed very low baseline activity and no responsiveness to noxious heat. Thus, the current and previous studies strongly suggest that IPCs are likely irrelevant for DSK-mediated nociceptive regulation, at least under normal conditions, although it is still possible for DSK to be expressed stochastically in a limited population of IPCs (Oikawa, 2023).
Although the data suggest the existence of a DSKergic descending inhibitory system from the brain to the Goro neurons in the VNC, there are multiple remaining questions for future studies on the mechanisms by which DSK-expressing neurons in the brain regulate Goro neurons: First, whether both MP1 and Sv neurons are involved in nociceptive inhibition remains unclear. The findings consistently point to MP1 neurons over Sv neurons as a major source of DSK regulating Goro neurons. However, given the potential of DSK for distant actions through diffusion, the possibility of Sv neurons communicating with Goro neurons cannot be eliminated. Further analysis using finer genetic/neuronal tools will be necessary to tease the functions of MP1 and Sv neurons apart and understand the functional mechanism of the DSKergic system in the regulation of larval nociception (Oikawa, 2023).
Second, how this DSKergic system is activated to regulate nociception in vivo needs further investigation. Very high CaMPARI2 photoconversion was observed in MP1 neurons without noxious heat stimuli, which did not further increase in response to noxious heat application. These data raise the hypothesis that MP1 neurons may function as a tonically active brake for nociceptive circuits, rather than a simple negative feedback system that is activated by nociceptive stimuli as input. The tonically active model of MP1 function is also consistent with data showing that dsk and CCKLR-17D1 mutations, as well as CCKLR-17D1 RNAi in Goro neurons, all caused thermal hypersensitivity under normal conditions. Up to this point, no external or internal signals have been directly shown to be the input into larval DSKergic neurons. However, since DSK has been implicated in the regulation of several physiological processes such as hunger and stress, sensory or molecular signals that are involved in these physiological processes have the potential to serve as upstream cues that activate DSKergic neurons. Alternatively, it is also possible that MP1 neurons may have spontaneous activity. Further physiological studies on the activity and responsiveness of DSKergic neurons would be required to elucidate how the DSKergic system functions to regulate nociception in larvae under normal conditions (Oikawa, 2023).
Third, the transmission mechanisms of DSK from the brain to Goro neurons need further clarification.nSyb-GRASP and trans-Tango experiments consistently showed negative results, indicating no synaptic connectivity between MP1 descending axons and Goro neurons. Thus, the interaction between DSK from DSKergic neurons in the brain and Goro neurons in the VNC may be mediated non-synaptically through volume transmission, as described in many neuropeptidergic systems. Many of the CCKLR-expressing neurons in the VNC are distantly located from the descending MP1 axons. Since it is unlikely that all these CCKLR-expressing neurons are synaptically connected to MP1 axons, neuronal communications through volume transmission can be fairly assumed for DSKergic systems in the larval VNC. More detailed analyses at an electron microscopic level of the circuitry connectivity between MP1 and Goro neurons and of the localization of DSK-containing vesicles in DSKergic neurons as well as DSK receptors in Goro neurons would be necessary to further clarify the transmission mechanisms of DSK from the brain to Goro neurons (Oikawa, 2023).
The data presented in this study also suggest that DSK signaling could regulate larval nociception through multiple pathways other than the DSK/CCKLR-17D1 system. For example, while CCKLR-17D3 mutants exhibited as severe thermal hypersensitivity as CCKLR-17D1 mutants, RNAi and rescue experiments failed to locate the function of CCKLR-17D3 to Goro neurons. GCaMP Ca2+ imaging also revealed that Goro neurons lacking CCKLR-17D3 showed modestly sensitized responses to noxious heat. These results suggest the major functioning of CCKLR-17D3 in Goro-independent nociceptive pathways. It was also observed that DSK receptor mutants exhibited more severe thermal phenotypes than dsk mutants , which may indicate the presence of unidentified DSK- or CCKLR-dependent signaling pathways in nociception-related cells. Further research is apparently necessary to reveal the whole picture of nociceptive regulations mediated by DSK and its receptors in larval Drosophila (Oikawa, 2023).
The CCK system is thought to be one of the most ancient neuropeptide systems, suggested to have multiple common physiological functions across taxa. In this study, it was demonstrated that CCKergic signaling in Drosophila participates in nociceptive modulation through a descending inhibitory pathway similarly to the mammalian CCK system, adding new evidence about the conserved physiological roles of CCK (Oikawa, 2023).
Unlike the peripheral nociceptive systems, it is still challenging to align the Drosophila and mammalian CCKergic descending pathways due to low homologies of the CNS structures between Drosophila and mammals, wide-spread CCK expression in the mammalian CNS, and the multiple roles of the CCKergic systems in mammalian nociceptive controls. However, the common usage of an orthologous molecular pathway in descending controls of nociception between the two evolutionarily distant clades raises a fascinating new hypothesis that the descending control from the brain may also be an ancient, conserved mechanism of nociception, which has emerged in the common ancestor of protostomes and deuterostomes. It will be of interest for future research to investigate whether the role of CCK signaling in descending nociceptive controls is also present in other species (Oikawa, 2023).
Non-mammalian model systems have been increasingly recognized as powerful tools to identify novel pain-related molecular pathways. However, the utilization of these models has so far been mostly limited to the research on peripheral pain pathophysiology, and few studies have used them to investigate central pain pathophysiology. Descending nociceptive control mechanisms are crucial for central pain modulation and have been implicated in the development of chronic pain states in humans. Thus, the current study opens the door to a new approach to using powerful neurogenetic tools and the simpler nervous system of Drosophila for elucidating the functional principles of descending nociceptive systems, which may potentially contribute to our understanding of the mechanisms underlying central pain modulation and pain pathology due to dysfunctions of descending modulatory pathways (Oikawa, 2023).
Sleep is vital for most animals, yet its mechanism and funwction remain unclear. This study found that permeability of the BBB (blood-brain barrier)-the organ required for the maintenance of homeostatic levels of nutrients, ions, and other molecules in the brain-is modulated by sleep deprivation (SD) and can cell-autonomously effect sleep changes. Increased BBB permeability was observed in known sleep mutants as well as in acutely sleep-deprived animals. In addition to molecular tracers, SD-induced BBB changes also increased the penetration of drugs used in the treatment of brain pathologies. After chronic/genetic or acute SD, rebound sleep or administration of the sleeping aid gaboxadol normalized BBB permeability, showing that SD effects on the BBB are reversible. Along with BBB permeability, RNA levels of the BBB master regulator moody are modulated by sleep. Conversely, altering BBB permeability alone through glia-specific modulation of moody, g&slpha;o, loco, lachesin, or i>neuroglian -each a well-studied regulator of BBB function-was sufficient to induce robust sleep phenotypes. These studies demonstrate a tight link between BBB permeability and sleep and indicate a unique role for the BBB in the regulation of sleep (Axelrod, 2023).
This study establish the BBB as a sleep regulatory center in Drosophila. It was observed that the BBB dynamically and rapidly—both functionally and molecularly—responds to acute as well as chronic-genetic SD and that altering BBB function alone, through mutations of the moody signaling pathway, can affect sleep, suggesting that the BBB senses sleep need and responds to it by altering its permeability. The correct BBB permeability state seems to be required for normal sleep, as disrupting BBB function causes sleep loss (Axelrod, 2023).
In normal physiology, the BBB controls influx and efflux of molecules in and out of the brain, including ions, amino acids, neurotransmitters, vitamins, proteins, carbohydrates, lipids, and hormones. Transport is facilitated via various mechanisms which include active transporters, transcytosis, and passive diffusion. The function of the BBB is not static but is affected by circadian rhythms, neuronal activity, and the metabolism. the data add sleep as an additional parameter modulating BBB function, as has been suggested by studies in mammals, and this study found that the relationship between the BBB and sleep is reciprocal. The results show that the BBB rapidly and dynamically responds to SD by increasing its permeability and that permeability increases conversely cause sleep loss. It is conceivable that BBB permeability changes allow the passage of one or multiple molecules that affect sleep into or out of the brain. Defective xenobiotic exclusion of steroid hormones from the brain via the BBB has been shown to affect sleep in Drosophila, demonstrating that molecules outside the CNS can affect sleep. Future studies may determine whether the BBB mediates a rapid global switch from a sleeping to an awake brain, for example, by controlling ionic diffusion and potassium concentrations, which could broadly raise neuronal activation thresholds and lead to the observed extracellular ionic changes in the sleeping mammalian brain (Axelrod, 2023).
The BBB creates a highly controlled microenvironment in the brain, perturbations of which affect neuronal function, as evidenced by neuropathologies at least in part caused by BBB disruption including stroke, Alzheimer's, Multiple Sclerosis, epilepsy and brain tumor metastasis. BBB disruption also occurs during normal aging, under chronic stress and during systemic inflammation and infection; however, whether the relationship between BBB disruption and these pathologies is causal or merely correlative is less clear. The magnitude of BBB changes in disease states can be an order of magnitude larger than the observed changes after sleep loss described in this study, illustrating the physiological relevance of these experiments. All of the pathologies listed above are also marked by sleep disruption, which in some instances increases the risk for developing the disease in the first place and the severity of which correlates with disease severity. The fly ortholog of a human autism gene has been reported to act in the BBB and elicit sleep fragmentation via developmental hyperserotonemia. However, a direct involvement of the BBB in sleep regulation had not been shown until this present study (Axelrod, 2023).
BBB and sleep disruptions have been, so far independently, associated with causing illness, and this study provides additional evidence for the notion that the two processes are linked. If BBB permeability and sleep regulation are fundamentally connected as part of the same system, it will be interesting to study how they relate to disease risk, development, and prognosis, and whether interventions that affect one also augment the other (Axelrod, 2023).
Adequate drug delivery to the brain is a problem plaguing the treatment of many neuropathologies. This puts strong importance on finding possible regulatory sites for controlling the BBB as a means of finding better ways to deliver drugs to the brain. The BBB has distinct mechanisms to exclude molecules from the brain including xenobiotic transporters in the cell membrane that efflux most small hydrophobic molecules taken up transcellularly through the membrane, and the paracellular barrier formed by junctions between the BBB forming cells. The latter creates a diffusion barrier for hydrophilic molecules in a size-dependent manner with a maximum passage size of about 0.5 kDa. Due to these mechanisms, the BBB poses a major obstacle to treatment of brain pathologies, because 100% of large molecules (>0.5 kDa), like the brain cancer drug vinblastine (molecular weight: 1 kDa), and 98% of small molecule drugs (<0.5 kDa), like the non-steroidal anti-inflammatory drug paracetamol, cannot penetrate it in therapeutic amounts. Current solutions to this problem include pharmacologically increasing the diffusion barrier permeability, implantation of ultrasonic devices to temporarily permeabilize the diffusion barrier, fusing molecules to receptor ligands to engage the transcellular route, or intracranial administration via injection. Provided the observed effects are conserved in humans, a patient's sleep and nighttime administration of drugs could have a significant effect on drug penetration into the CNS and possibly treatment outcomes (Axelrod, 2023).
In Drosophila, the BBB's subperineurial glia form a layer of polarized epithelial cells, with septate junctions between them creating a seal for water-soluble molecules over 500 Da. The observation that SD rapidly induces increases in BBB permeability raises the question: What are the molecular and cellular mechanisms that underlie the remodeling of septate junctions on the scale of hours? The Moody GPCR pathway has been shown to play a key role in the development of the epithelial barrier, as well as its maintenance in adulthood. Moody activates a signaling cascade which, via PKA activation, controls the developmental assembly of the septate junction belt in a highly coordinated spatiotemporal manner that involves rearrangements of the actin cytoskeleton and vesicular trafficking. This study has shown that RNAi-based suppression of moody pathway genes causes increased BBB permeability. Strong downregulation of moody was observed in every genetic sleep mutant tested. These findings suggest that altered moody pathway function may be primarily responsible for the leaky barriers of these mutants (Axelrod, 2023).
During SD and recovery, it is conceivable that similar processes occur. Indeed, it has been shown that perturbation of endocytosis at the Drosophila BBB alters sleep, possibly interfering with Moody-activated remodeling during sleep/wake cycles. Scute SD has been shown to rapidly depress moody expression in wild-type Drosophila, which is accompanied by increased BBB permeability. Moody is expressed on the neuronal-facing apical cell surface of the subperineurial glia. While its ligand is not known, this topology could permit neuronal states related to sleep to be communicated to glia via this GPCR pathway. As many known sleep genes in Drosophila have been shown to be expressed in neurons and this study observed BBB changes upon sleep-suppressing neuronal activation, neuron-glia signaling seems likely to be part of the process, and it will be interesting to elucidate the involved signaling molecules including Moody's ligand (Axelrod, 2023).
The two-step model of sleep regulation posits that circadian rhythm and sleep homeostasis control the timing of sleep. Sleep homeostasis is a process that controls sleep duration to meet an organism's daily sleep need. Rebound sleep after SD is interpreted as proof of the existence of sleep homeostasis. The physiological and cellular nature of the sleep homeostat is-like the nature of sleep itself-the subject of current research. Sleep homeostasis has been shown to be affected by oxidative stress, neuronal, glial, and microglial activity, neuropeptide signaling, infection, the circadian clock, temperature, and starvation. Synaptic downscaling, glymphatic clearance of toxins, peripheral glucose homeostasis, reactive oxygen species clearance, and inhibition of motor function have been proposed as processes occurring during sleep homeostasis, constituting functions of sleep. As at least some of these processes must involve yet to be identified body-brain signals, future studies will reveal how the BBB changes shown here affect known or novel functions of sleep (Axelrod, 2023).
Together the data indicate a system where an increase of sleep need is measured by the BBB, resulting in an adjustment of Moody levels, an opening of the BBB, and the initiation of rebound sleep. This process is likely to involve neuronal to BBB signaling (dopaminergic SD also alters the BBB, and many of the sleep mutants are expressed in neurons) as well as yet-to-be-characterized molecular exchange across the BBB (Axelrod, 2023).
This study presents a functionally relevant, quantitative characterization of the neural circuitry of Drosophila melanogaster at the mesoscopic level of neuron types as classified exclusively based on potential network connectivity. Starting from a large neuron-to-neuron brain-wide connectome of the fruit fly, this study use stochastic block modeling and spectral graph clustering to group neurons together into a common "cell class" if they connect to neurons of other classes according to the same probability distributions. This study then characterize the connectivity-based cell classes with standard neuronal biomarkers, including neurotransmitters, developmental birthtimes, morphological features, spatial embedding, and functional anatomy. Mutual information indicates that connectivity-based classification reveals aspects of neurons that are not adequately captured by traditional classification schemes. Next, using graph theoretic and random walk analyses to identify neuron classes as hubs, sources, or destinations, pathways and patterns of directional connectivity were detected that potentially underpin specific functional interactions in the Drosophila brain. This study uncovered a core of highly interconnected dopaminergic cell classes functioning as the backbone communication pathway for multisensory integration. Additional predicted pathways pertain to the facilitation of circadian rhythmic activity, spatial orientation, fight-or-flight response, and olfactory learning. This analysis provides experimentally testable hypotheses critically deconstructing complex brain function from organized connectomic architecture (Mehta, 2023).
Over the last two decades, Drosophila has been a popular model organism for studying how structural connections in the brain give rise to functional interactions. Numerous studies have previously focused on the functional dissection of regional neural circuitry in Drosophila, including that of vision, olfaction, mushroom body, and central complex, to name a few. With advancements in data acquisition and processing, recent large-scale projects have been successful in generating detailed connectomes of the brain-wide neural circuitry, that go well beyond a single region. While a complete brain-wide wiring diagram is a necessary prerequisite, unraveling the underlying mechanistic pathways to interpolate function and behavior is a problem that cannot be solved by gathering more data alone. One of the fundamental obstacles is the absence of a quantitative specification of neuron types and how they relate to neural circuitry (Mehta, 2023).
The prevalent approach to analyzing brain-wide microscopic connectomes is to group neurons into spatially compact regions, for example, local processing units or compartments, and then attempt to characterize the connectivity between these groups. While there is undoubtedly a strong spatial component to synaptic connectivity, anatomical division is by itself insufficient to deconstruct brain computation. Modeling connectomes as a network comprising solely of assortative, anatomically segregated regions interacting with each other may crucially miss many functionally relevant characteristics pertaining to the complex mesoscale organization of the brain (Mehta, 2023).
In contrast, the stochastic block model (SBM) framework groups neurons by their patterns of potential synaptic connectivity to capture assortative and nonassortative features in the underlying neuron-to-neuron brain-wide network. The information captured by connectivity-based classification is complementary, yet synergistic, to that provided by traditional neuronal classification-together enabling functional interpretation not previously possible. By inferring the directed connectivity patterns on the derived circuit and simultaneously characterizing the circuit with respect to multiple biomarkers, this study has identified functional pathways supporting multiple sensory modalities and cross-region integration (Mehta, 2023).
A key pathway identified by the circuit is the backbone communication pathway comprising of six interconnected dopaminergic-centric hubs working together to facilitate integration of the sensory and motor pathways. The neural circuitry for vision, olfaction, and hearing are well characterized, but the pathways for how these modalities combine to guide behavior, for example, during navigation, remain elusive. Interestingly, two out of these six hubs are the only classes overall that consist of all three neuromodulators-DA, OA, and 5-HT. Further, since the hubs innervated multiple neuropils, this pathway would not have been revealed by a classification purely based on anatomy, neurotransmitter, or morphology. Dopamine is a known neuromodulator central to reinforcement signaling, and responding to salient stimuli related to olfaction and vision. Also, while dopamine affects visual tracking, its role in place learning or spatial memory remains to be elucidated (Mehta, 2023).
Another prominent pathway revealed in this analysis is the glutamate-dominant pathway for circadian activity. The clock neuron subsets in the Drosophila brain are well described including their spatial location, morphology, neurotransmitter, and role in behavioral rhythmicity. Nevertheless, the inputs from other neurons onto clock neurons, and the exact neuronal pathways underlying the entrainment of the clock for synchronization with the environment are poorly known. The identified pathways link the motor and vision regions interlaterally, suggesting involvement in coordination of rhythmic activity. Unfortunately, however, it is unknown which, if any, of the neurons in the 19,902-neuron FlyCircuit dataset are actually clock neurons (Mehta, 2023).
Over the last several years, optogenetics have been invaluable in studying how dynamical interactions on the neural circuit give rise to specific Drosophila behavior. These include circuits relating to rhythmic motor function, olfaction, central complex function, and learning and memory. The derived circuit is amenable to optogenetic manipulation, thus allowing for experimental verification of the predicted functional behavior of the identified pathways. Specifically, for each of the characterized circuit pathways, this study identified GAL4-UAS driver lines that could potentially be used for selective neuronal targeting. This study also predicted the associated loss (or gain) of function that would accompany the selective deactivation (or activation) of the corresponding neurons in these pathways (Mehta, 2023).
The connectomic dataset used in this analysis consists of 19,902 neurons reconstructed from multiple specimens using light microscopy, corresponding to a sample size of approximately 12% of the total number of neurons in a Drosophila brain. As with all research based on sampled data, the possibility that this analysis may miss some essential features of the brain-wide circuitry cannot be exclused, due to important neural classes being absent or not represented with enough neurons in the considered dataset. This derived mesoscale circuit does, however, demonstrate robustness to the proportion of neurons sampled, suggesting that small variations in the size of the dataset are unlikely to change conclusions significantly. The constructed circuit additionally shows excellent left-right symmetry, indicating bilaterally uniform sampling of the neurons (Mehta, 2023).
A practical limitation of applying the spectral graph clustering framework is that it requires the size of the dataset (number of vertices in the network) to be much larger than the number of blocks/classes κ. In particular, κ << n may not hold for connectomic datasets with a limited number of sampled neurons, or with a large number of classes in relation to total neurons. On the other hand, the SBM inference model scales extremely well as n increases; thus, this analysis can be progressively refined as more neurons continue to be morphologically reconstructed, registered to a common atlas, and shared with the community (Mehta, 2023).
This model could also be applied to dense connectomic reconstructions from electron microscopy, which is considered the gold standard in the field. However, the knowledge derived from sparse light microscopy collected from multiple brains valuable is viewed in its own merit, as it provides complementary information on the shared circuit elements across individuals. In this regard, it is useful to note that the probabilistic model, currently restricted to Bernoulli adjacency matrices, can be generalized to cluster nonbinary connectomes using a weighted SBM, which would allow accounting for the number of synapses between connected neuron pairs in the identification of connectivity-based classes (Mehta, 2023).
This study applied community detection methods to analyze the synapse-level reconstruction of an adult female Drosophila brain containing over 20 thousand neurons and 10 million synapses. Using a machine-learning algorithm, the most densely connected communities of neurons were found by maximizing a generalized modularity density measure. The community structure was resolved at a range of scales, from large (on the order of thousands of neurons) to small (on the order of tens of neurons). The network was found to be is organized hierarchically and larger-scale communities are composed of smaller-scale structures. These methods identify well-known features of the fly brain, including its sensory pathways. Moreover, focusing on specific brain regions, it was possible to identify subnetworks with distinct connectivity types, including the fan-shaped body and the superior neuropil, with distinct clusters of upstream and downstream brain regions dividing the neuropil into several pathways. These methods show that the fine-scale, local network reconstruction made possible by modern experimental methods are sufficiently detailed to identify the organization of the brain across scales, and enable novel predictions about the structure and function of its parts (Kunin, 2023).
Coordinated activity across networks of neurons is a hallmark of both resting and active behavioural states in many species. These global patterns alter energy metabolism over seconds to hours, which underpins the widespread use of oxygen consumption and glucose uptake as proxies of neural activity. However, whether changes in neural activity are causally related to metabolic flux in intact circuits on the timescales associated with behaviour is unclear. This study combine two-photon microscopy of the fly brain with sensors that enable the simultaneous measurement of neural activity and metabolic flux, across both resting and active behavioural states. Neural activity was demonstrated to drive changes in metabolic flux, creating a tight coupling between these signals that can be measured across brain networks. Using local optogenetic perturbation, it was demonstrated that even transient increases in neural activity result in rapid and persistent increases in cytosolic ATP, which suggests that neuronal metabolism predictively allocates resources to anticipate the energy demands of future activity. Finally, these studies reveal that the initiation of even minimal behavioural movements causes large-scale changes in the pattern of neural activity and energy metabolism, which reveals a widespread engagement of the brain. As the relationship between neural activity and energy metabolism is probably evolutionarily ancient and highly conserved, these studies provide a critical foundation for using metabolic proxies to capture changes in neural activity (Mann, 2021).
The Integrator complex is a multi-subunit protein complex that regulates the processing of nascent RNAs transcribed by RNA polymerase II (RNAPII), including small nuclear RNAs, enhancer RNAs, telomeric RNAs, viral RNAs, and protein-coding mRNAs. Integrator subunit 11 (INTS11) is the catalytic subunit that cleaves nascent RNAs, but, to date, mutations in this subunit have not been linked to human disease. This study describes 15 individuals from 10 unrelated families with bi-allelic variants in INTS11 who present with global developmental and language delay, intellectual disability, impaired motor development, and brain atrophy. Consistent with human observations, this study has found that the fly ortholog of INTS11, dIntS11, is essential and expressed in the central nervous systems in a subset of neurons and most glia in larval and adult stages. Using Drosophila as a model, the effect of seven variants was investigated. Two (p.Arg17Leu and p.His414Tyr) fail to rescue the lethality of null mutants, indicating that they are strong loss-of-function variants. Furthermore, it was found that five variants (p.Gly55Ser, p.Leu138Phe, p.Lys396Glu, p.Val517Met, and p.Ile553Glu) rescue lethality but cause a shortened lifespan and bang sensitivity and affect locomotor activity, indicating that they are partial loss-of-function variants. Altogether, these results provide compelling evidence that integrity of the Integrator RNA endonuclease is critical for brain development (Tepe, 2023).
Left-right (LR) asymmetry of the brain is fundamental to its higher-order functions. The Drosophila brain's asymmetrical body (AB) consists of a structural pair arborized from AB neurons and is larger on the right side than the left. This study found that the AB initially forms LR symmetrically and then develops LR asymmetrically by neurite remodeling that is specific to the left AB and is dynamin dependent. Additionally, neuronal ecdysone signaling inhibition randomizes AB laterality, suggesting that ecdysone signaling determines AB's LR polarity. Given that AB's LR asymmetry relates to memory formation, this research establishes AB as a valuable model for studying LR asymmetry and higher-order brain function relationships (Sakamura, 2023).
The level of each RNA species depends on the balance between its rates of production and decay. Although previous studies have measured RNA decay across the genome in tissue culture and single-celled organisms, few experiments have been performed in intact complex tissues and organs. It is therefore unclear whether the determinants of RNA decay found in cultured cells are preserved in an intact tissue, and whether they differ between neighboring cell types and are regulated during development. To address these questions, this study measured RNA synthesis and decay rates genome wide via metabolic labeling of whole cultured Drosophila larval brains using 4-thiouridine. This analysis revealed that decay rates span a range of more than 100-fold, and that RNA stability is linked to gene function, with mRNAs encoding transcription factors being much less stable than mRNAs involved in core metabolic functions. Surprisingly, among transcription factor mRNAs there was a clear demarcation between more widely used transcription factors and those that are expressed only transiently during development. mRNAs encoding transient transcription factors are among the least stable in the brain. These mRNAs are characterized by epigenetic silencing in most cell types, as shown by their enrichment with the histone modification H3K27me3. This data suggests the presence of an mRNA destabilizing mechanism targeted to these transiently expressed transcription factors to allow their levels to be regulated rapidly with high precision. This study also demonstrates a general method for measuring mRNA transcription and decay rates in intact organs or tissues, offering insights into the role of mRNA stability in the regulation of complex developmental programs (Thompson, 2022).
Left-Right (LR) asymmetry of the nervous system is widespread across animals and is thought to be important for cognition and behaviour. But in contrast to visceral organ asymmetry, the genetic basis and function of brain laterality remain only poorly characterized. In this study, RNAi screening was performed to identify genes controlling brain asymmetry in Drosophila. The conserved NetrinB (NetB) pathway was found to be required for a small group of bilateral neurons to project asymmetrically into a pair of neuropils (Asymmetrical Bodies, AB) in the central brain in both sexes. While neurons project unilaterally into the right AB in wild-type flies, netB mutants show a bilateral projection phenotype and hence lose asymmetry. Developmental time course analysis reveals an initially bilateral connectivity, eventually resolving into a right asymmetrical circuit during metamorphosis, with the NetB pathway being required just prior symmetry breaking. This study shows using unilateral clonal analysis that netB activity is required specifically on the right side for neurons to innervate the right AB. It was shown that loss of NetB pathway activity leads to specific alteration of long-term memory, providing a functional link between asymmetrical circuitry determined by NetB and animal cognitive functions (Lapraz, 2023).
Recruitment of stem cells is crucial for tissue repair. Although stem cell niches can provide important signals, little is known about mechanisms that coordinate the engagement of disseminated stem cells across an injured tissue. In Drosophila, adult brain lesions trigger local recruitment of scattered dormant neural stem cells suggesting a mechanism for creating a transient stem cell activation zone. This study found that injury triggers a coordinated response in neuro-glial clusters that promotes the spread of a neuron-derived stem cell factor via glial secretion of the lipocalin-like transporter Swim. Strikingly, swim is induced in a Hif1-α-dependent manner in response to brain hypoxia. Mammalian Swim (Lcn7) is also upregulated in glia of the mouse hippocampus upon brain injury. These results identify a central role of neuro-glial clusters in promoting neural stem cell activation at a distance, suggesting a conserved function of the HIF1-α/Swim/Wnt module in connecting injury-sensing and regenerative outcomes (Simoes, 2022).
Injury is known to stimulate diverse forms of plasticity, which serve to restore organ function. Many tissues harbor a small number of undifferentiated adult stem cells that are engaged in tissue turnover or become activated following injury to replace damaged cells. Some tissues, such as muscle or brain, contain mainly dormant stem cells that are not dividing and reside in a reversible state of quiescence. Niche cells in intimate contact with quiescent stem cells have been found to provide activating cues upon tissue damage. However, little is known how the activation of multiple dispersed stem cell units is coordinated to establish an adequate stem cell response zone across an injured tissue (Simoes, 2022).
Quiescent progenitor cells in muscle and the brain respond to injury in mammals, but also in fruit flies (Drosophila). This allows to harness the extensive genetic tools available in Drosophila to dissect injury-dependent stem cell activation. Although still unclear, the presence of dormant stem cells in short-lived insects indicates that these cells may play a beneficial role for tissue plasticity or repair upon predator attacks or inter-species aggressions (Simoes, 2022).
In the adult fly brain, experimental stab lesions to the optic lobes (OLs) or the central brain trigger a proliferative response resulting in local neurogenesis several days after injury (AI), which has been linked to activation of normally quiescent neural progenitor cells (qNPs). qNPs have also been found to promote adult brain plasticity in contexts unrelated to injury. On the other hand, stab lesions can also trigger glial divisions shortly after injury (Simoes, 2022).
Despite extensive knowledge on neural stem cell proliferation during fly development, the signals governing qNP activation in response to injury are unknown. A ubiquitous pulse of Drosophila Myc (dMyc) overexpression has been previously shown to promote qNP division, but the signals detected by qNPs remained enigmatic (Simoes, 2022).
In mammals, a wide variety of signals are known to regulate quiescent neural stem cells (qNSCs) in homeostatic conditions, whereas their response to tissue damage is less well understood. qNSCs are located in two main niches, the subventricular zone and the dentate gyrus of the hippocampus, buried within the brain. Upon brain injury, qNSCs only partially enter an activated state, and neuroblast recruitment to infarcted brain regions and local neurogenesis is limited (Simoes, 2022).
Strikingly, the initial consequences triggered by brain injury, which include neural cell death, upregulation of antioxidant defense, and c-Jun N-terminal kinase (JNK) stress signaling, are very conserved in flies and mice suggesting that injury sensing of qNSCs/qNPs may rely on common principles.
In this work, injury-induced changes were studied in the adult fly brain leading to recruitment of isolated qNPs near the injury site. A crucial role was identified of damage-responsive neuro-glial clusters (DNGCs), which enable proliferation of distant qNPs by promoting an enlarged stem cell activation zone. Evidence is provided that these multicellular units orchestrate the spatial and temporal availability of an essential, but localized stem cell factor for qNPs via injury-stimulated secretion of the transport protein Swim. As Swim production is dependent on the injury-sensitive transcription factor HIF1-α, the identified mechanism may serve to spatially and temporary adjust the stem cell activation zone to the extent of damage suffered in a given tissue area, resulting in locally calibrated pulses of stem cell activity (Simoes, 2022).
How tissue damage is sensed and how the recruitment of multiple stem cell units is coordinated in response to local, heterogeneous tissue damage represents a fundamental question. By investigating how dispersed qNPs are locally recruited to injury, we have identified a mechanism that creates a defined zone of stem cell activation in the adult fly brain. The process is dependent on DNGCs, which depending on their size and possibly composition may regulate the extent by which a localized stem cell factor such as Wg/Wnt can travel to rare qNPs in the vicinity. Whereas the neuronal cells provide Wg/Wnt, the glial component supplies the carrier protein Swim, thereby promoting the dispersion of the signal. This cooperative interaction of two different cell types to gain long range function of Wg/Wnt is rather unique (Simoes, 2022).
At the cellular level, a model is proposed whereby injury-sensitive HIF1-α directs Swim synthesis in glial cells. Swim transporters diffuse and facilitate the spread of localized neural-derived Wg ligands, probably by binding to and shielding the lipid-residues of Wg/Wnt in the aqueous extracellular space. Mobile Wg-Swim complexes are consequently able to reach and activate qNPs in the injured brain domain. Wg/Wnt signal transduction and downstream upregulation of dmyc is shown to be crucial for the proliferation of this novel cell type. Overall, it is proposed that the described mechanism provides a means to match recruited stem cell activity to the spatial and temporal persistence of damage in the injured brain.
Activation of dormant neural progenitors by high levels of Wg/Wnt
Wg/Wnt signaling is probably one of the most universal pathways driving stem cell proliferation. Nevertheless, an understanding of Wg/Wnt signals for dormant stem cells has only recently emerged. Dormant muscle stem cells, for example, maintain quiescence by raising their threshold for Wnt transduction via cytoplasmic sequestration of &betal-catenin, and qNSC in the hippocampus do not rely on Wnt signaling under homeostasis but display a high capability to respond to Wnt in a graded manner when exposed. Similarly, the results demonstrate that qNPs start proliferating when high Wg/Wnt levels are provided in an autocrine fashion (Simoes, 2022).
Overall, the results suggest that activation of qNPs in the adult fly brain is mainly prevented by the low availability of Wg/Wnt ligands under homeostatic conditions. Although Wnt signaling normally occurs between adjacent cells, this study provides evidence that Wg functions at a tissue scale in the injured fly brain (Simoes, 2022).
This study describes the property of Swim to extend the signaling range of Wg/Wnt. Further research will be required to determine whether other stem cell-relevant factors can be transported by Swim.
In zebrafish, reduced levels of Swim/Lcn7 produce craniofacial defects due to compromised Wnt3 signaling, highlighting a different context of Wnt/Swim interaction. A Wg/Swim interaction has previously been proposed in developing epithelia in flies, although the effect was not observed in a more recent study (Simoes, 2022).
Swim::mCherry is strongly expressed in the adult ovary germline of flies, in agreement with data from the recently published Fly Cell Atlas. Remarkably, swim KO flies showed reduced fertility, a phenotype which has also been reported for lcn7/tinagl1 KO mice (Takahashi, 2016). Interestingly, Swim expression in the germarium strongly overlapped with Wg::GFP, in line with previous findings describing a requirement of extensive Wg travel from the niche to distant follicular stem cells (Simoes, 2022).
Finally, this study elucidated how the Swim/Wg stem cell-activating signal is connected to damage sensing in the injured brain. Both in flies and mice, swim/lcn7 induction occurs in glial cells in response to brain injury. Remarkably, stroke-induced lcn7 induction is not observed in mouse brains, in which Hif1-α has been deleted from mature neurons and glial cells. This suggests that HIF1-α-dependent swim regulation is conserved in mammals (Simoes, 2022).
According to the current model, the damage responsiveness of stem cells is strongly gated by the availability of stable HIF1-α during acute hypoxia. Such a limited activation pulse would effectively restrict the mitotic effect of Swim/Wg complexes to the acute phase of repair, acting as a safeguard mechanism against overgrowth. Moreover, the hypoxia-dependent secretion of Swim would allow to temporally and locally fine-tune the realm of the stem cell activation zone to injury.
Local oxygen concentrations modulate the activity of adult stem cells in different niches. In the fly larval OL, Dpn-expressing neural progenitors proliferate in a pronounced hypoxic environment, which bears parallels to the situation following brain injury (Simoes, 2022).
In the mammalian brain, injury-induced Wnt ligands may not efficiently reach qNSCs in distant neurogenic niches, resulting in poor stem cell activation. As such, Wnt pathway stimulating approaches hold promise as possible treatment for brain injury as they are known to support regeneration at several levels including qNSC activation, neurogenesis, and axon outgrowth. Increasing the mobility or stability of Wg/Wnts by Swim-like transporters may therefore represent a successful strategy to engage endogenous progenitors into regeneration. Given the fact that Wg/Wnts can support tissue renewal and regeneration in numerous tissues, the properties of Swim to transform a restricted tissue area into a temporary stem cell-activating zone, uncovered in this study may have important applications in regenerative medicine (Simoes, 2022).
Although the current experiments have revealed an impaired distribution of Wg in the injured brain in the absence of Swim transporters, it cannot be completely rule out that Swim may alter Wg function by other means than physical binding and direct transport. Ideally, the injury-induced formation of Wg-Swim complexes should be observable in the extracellular space. Although colocalization of Swim and Wg signals was detected, it was not possible to image Wg-Swim complexes at high resolution due to elevated background of Wg and mCherry antibodies when performing extracellular stainings. Overcoming these current limitations with overexpression systems or optimized immunodetection should allow to capture the dynamics of Wg-Swim interactions in injured brain tissue in the future (Simoes, 2022).
The brain as a central regulator of stress integration determines what is threatening, stores memories, and regulates physiological adaptations across the aging trajectory. While sleep homeostasis seems to be linked to brain resilience, how age-associated changes intersect to adapt brain resilience to life history remains enigmatic. This study provides evidence that a brain-wide form of presynaptic active zone plasticity ("PreScale"), characterized by increases of active zone scaffold proteins and synaptic vesicle release factors, integrates resilience by coupling sleep, longevity, and memory during early aging of Drosophila. PreScale increased over the brain until mid-age, to then decreased again, and promoted the age-typical adaption of sleep patterns as well as extended longevity, while at the same time it reduced the ability of forming new memories. Genetic induction of PreScale also mimicked early aging-associated adaption of sleep patterns and the neuronal activity/excitability of sleep control neurons. Spermidine supplementation, previously shown to suppress early aging-associated PreScale, also attenuated the age-typical sleep pattern changes. Pharmacological induction of sleep for 2 days in mid-age flies also reset PreScale, restored memory formation, and rejuvenated sleep patterns. The data suggest that early along the aging trajectory, PreScale acts as an acute, brain-wide form of presynaptic plasticity to steer trade-offs between longevity, sleep, and memory formation in a still plastic phase of early brain aging (Huang, 2022).
Neurogenesis in the Drosophila central brain progresses dynamically in order to generate appropriate numbers of neurons during different stages of development. Thus, a central challenge in neurobiology is to reveal the molecular and genetic mechanisms of neurogenesis timing. This study found that neurogenesis is significantly impaired when a novel mutation, Nuwa, is induced at early but not late larval stages. Intriguingly, when the Nuwa mutation is induced in neuroblasts of olfactory projection neurons (PNs) at the embryonic stage, embryonic-born PNs are generated, but larval-born PNs of the same origin fail to be produced. Through molecular characterization and transgenic rescue experiments, it was determined that Nuwa is a loss-of-function mutation in Drosophila septin interacting protein 1 (sip1). Furthermore, it was found that SIP1 expression is enriched in neuroblasts, and RNAi knockdown of sip1 using a neuroblast driver results in formation of small and aberrant brains. Finally, full-length SIP1 protein and truncated SIP1 proteins lacking either the N- or C-terminus display different subcellular localization patterns, and only full-length SIP1 can rescue the Nuwa-associated neurogenesis defect. Taken together, these results suggest that SIP1 acts as a crucial factor for specific neurogenesis programs in the early developing larval brain (Wei, 2022).
The actin cytoskeleton is a key determinant of cell and tissue homeostasis. However, tissue-specific roles for actin dynamics in aging, notably brain aging, are not understood. This study shows that there is an age-related increase in filamentous actin (F-actin) in Drosophila brains, which is counteracted by prolongevity interventions. Critically, modulating F-actin levels in aging neurons prevents age-onset cognitive decline and extends organismal healthspan. Mechanistically, autophagy, a recycling process required for neuronal homeostasis, was shown to be disabled upon actin dysregulation in the aged brain. Remarkably, disrupting actin polymerization in aged animals with cytoskeletal drugs restores brain autophagy to youthful levels and reverses cellular hallmarks of brain aging. Finally, reducing F-actin levels in aging neurons slows brain aging and promotes healthspan in an autophagy-dependent manner. These data identify excess actin polymerization as a hallmark of brain aging, which can be targeted to reverse brain aging phenotypes and prolong healthspan (Walker, 2023).
The Drosophila brain is a frequently used model in neuroscience. Single-cell transcriptome analysis, three-dimensional morphological classification and electron microscopy mapping of the connectome have revealed an immense diversity of neuronal and glial cell types that underlie an array of functional and behavioural traits in the fly. The identities of these cell types are controlled by gene regulatory networks (GRNs), involving combinations of transcription factors that bind to genomic enhancers to regulate their target genes. This study characterize GRNs at the cell-type level in the fly brain, the chromatin accessibility of 240,919 single cells was characterized spanning 9 developmental time points and integrated these data with single-cell transcriptomes. More than 95,000 regulatory regions were identified that are used in different neuronal cell types, of which 70,000 are linked to developmental trajectories involving neurogenesis, reprogramming and maturation. For 40 cell types, uniquely accessible regions were associated with their expressed transcription factors and downstream target genes through a combination of motif discovery, network inference and deep learning, creating enhancer GRNs. The enhancer architectures revealed by DeepFlyBrain lead to a better understanding of neuronal regulatory diversity and can be used to design genetic driver lines for cell types at specific timepoints, facilitating their characterization and manipulation (Janssens, 2022).
The Drosophila brain contains around 220,000 cells and is an excellent model for investigating the diversity of cell types. Advances in electron microscopy have yielded connectome maps across the fly brain, while driver lines provide genetic access to many cell types. This diversity of cell types has been bolstered by single-cell transcriptomics analyses of the brain and the ventral nerve cord (Janssens, 2022).
The role that transcription factors (TFs) have in the determination of neuronal fate was previously highlighted by imputing unique TF combinations for each cell type that activate or repress target genes. These combinations arise in neural progenitors through TF-guided temporal and spatial axes of differentiation15. Furthermore, TFs govern key neuronal features such as dendritic targeting and neurotransmitter determination, and altering a single TF can change neuronal fate. Inferring TFs and their putative target genes is crucial, but transcriptomics analysis leads to high false-positive rates, as TF activity often cannot be predicted from TF expression levels because it depends on many variables, such as protein activity and localization as well as the presence of co-binding TFs and co-factors (Janssens, 2022).
The recent development of the single-cell assay for transposase accessible chromatin by sequencing (scATAC-seq) provides additional understanding of the mechanisms that underlie neuronal identity, by enabling the analysis of which genomic regions encode regulatory information for creating and maintaining each cell type. Integrating genomic enhancers with gene expression would yield precise regulatory programs (Janssens, 2022).
A single-cell multi-omics atlas across fly brain development was built for this study, covering neurogenesis, maturation and maintenance. Key regulators of neuronal and glial cell identity were identified, the enhancer code for specific neuronal subtypes was deciphered, informed enhancer driver lines and map enhancer GRNs (eGRNs, all of which are available for exploration online (Janssens, 2022).
To study regulatory programs of neuronal diversity, chromatin accessibility was profiled of 240,919 cells from whole brains at 9 timepoints from larvae to adult, covering crucial stages of development. This atlas is accompanied by a single-cell RNA sequencing (scRNA-seq) atlas of the adult brain, containing 118,687 high-quality cells divided into 204 clusters, of which 66 are annotated (Janssens, 2022).
First, the chromatin landscape of 60,624 cells from adult flies and late-stage pupa (72 h after puparium formation (APF)) was analyzed, and 79 stable cell states using were identified cisTopic. Next the accessibility of regions upstream and within the gene body was analyzed, and scATAC clusters were linked to cell types in the scRNA-seq atlas using co-clustering, marker gene enrichment and non-negative least squares (NNLS) regression. After manual curation, 43 of the 79 ATAC clusters were one-to-one linked to RNA clusters (Janssens, 2022).
The annotated clusters include six glial subtypes (approximately 10-15%), non-brain cells (1%, plasmatocytes and photoreceptors) and neurons (85-90%). Notably, optic lobe (OL) neurons form distinct clusters, whereas central brain (CB) neurons form a continuum. In the CB, three Kenyon cell (KC) subtypes and two smaller cell types of the central complex (ellipsoid body ring neurons, protocerebral bridge neurons) were identified. CB clusters can be split into Imp+ or pros+ cells on the basis of scATAC-seq, recapitulating the differences that were found in scRNA-seq analyses of the brain and ventral nerve cord. To validate cell type annotation, driver lines were used that label the three KC subtypes and two OL cell types (Tm1 and T4/T5), and bulk ATAC-seq was performed after fluorescence-activated cell sorting (FACS), matching the scATAC aggregates with high concordance (Janssens, 2022).
Every cluster has a unique chromatin accessibility profile, with a range of 105 to 4,732 differentially accessible regions (DARs) out of a total of 24,543 median accessible regions per cluster, many of which are located close to validated marker genes. The transcription start site (TSS) is often ubiquitously accessible, meaning specificity is more distally controlled. Although T4/T5 neurons are inseparable in adult scRNA-seq, 110 DARs were identified between them, and subclustering also separated the a/b and c/d subtypes. Given this high resolution in scATAC-seq, the missing CB cell types were investigated by examining olfactory projection neurons (OPNs). OPNs are identified in a 57,000-entry scRNA-seq dataset, but not in the 60,000-entry scATAC-seq dataset nor in an expanded dataset with additional timepoints nor using different clustering approaches. However, when OPNs were FACS-sorted for scATAC-seq, 876 peaks were revealed near OPN marker genes, showing that, despite unique chromatin profiles, OPNs, and probably other CB cell types, are more difficult to identify by scATAC-seq compared with scRNA-seq (Janssens, 2022).
To investigate how neuronal diversity is generated, chromatin accessibility changes during development were studied by analysing 135,275 cells from third instar larvae to 12 h APF using cisTopic, obtaining 54 clusters. A support vector machine (SVM) classifier was trained on the adult cell types to transfer labels to earlier stages, enabling the detection of core sets of specific regions per cell type that remain continuously accessible, as well as DARs that vary over time. Similar to RNA-seq analyses in which a maximum number of differentially expressed genes was detected at 48 h APF and a minimum in adults, this study found a decrease in DARs over time, with a relative spike at 48 h APF during synaptogenesis (Janssens, 2022).
Progenitor cell types, which are characterized by accessible regions near the neuroblast markers dpn and ase, form the roots of two main branches in the uniform manifold approximation and projection (UMAP) analysis-a continuous CB branch and a tree-like OL branch, indicating different neurogenesis modes. A spike was detected in TF motifs from the neuronal remodelling factors EcR and Sox14 in Imp+ neurons, but not in pros+ neurons, consistent with their respective pruning roles and with a potential embryonic origin of Imp+ neurons. In the OL, six branches emerge, each enriched for the motif of one major class of TFs (such as POU, bHLH and ETS), linked to synaptic partner recognition and neurotransmitter determination (Janssens, 2022).
To identify cell-type-specific key regulators, a 'cistrome' as the combination of a TF with its target enhancers. A dual approach of conventional motif discovery and deep learning (DL) was developed, integrating information from the adult scRNA-seq and scATAC-seq data, to identify TFs per cell type that are both expressed and of which the motif is enriched in the accessible regions (Janssens, 2022).
First, using the conventional approach, the correlation between TF expression and motif enrichment was calculated in DARs per cell type. In total, 116 TFs show a strong positive correlation, suggesting that they open chromatin as activators. These cover pan-neuronal, pan-glial and cell-type-specific TFs, including known regulators for glia: Repo and Kay; for KCs: Ey and Mef2; for ellipsoid body neurons: Grn and D; for T1 neurons: Ets65A and Oc; and combinations of Acj6, Fkh, TfAP-2 and SoxN/Sox102F in the other T neurons. Moreover, 131 TFs display a negative correlation between expression and motif enrichment (such as Mamo and Lola-N), suggesting a repressive role (Janssens, 2022).
Second, a convolutional neural network, called DeepFlyBrain, was trained using sequences of co-accessible regions (topics) from KCs, T neurons and glia as input, and topic accessibility as output. DeepExplainer was used to calculate the contribution of each nucleotide in the prediction of region accessibility, and TF-MoDISco to identify motifs from recurring patterns in the contribution scores. This revealed that KC enhancers are characterized by Ey, Onecut, Mef2, Mamo and Dati motifs, matching their expression. Mamo and Dati have negative nucleotide importance, suggesting that they correlate with closing chromatin. For T neurons, the most important motifs include Fkh, TfAP-2 and Acj6; and, for glia, Ct, Repo, Zfh2 and Klu were found. Scanning accessible regions with TF-MoDISco patterns, DeepFlyBrain provides high-confidence, cell-type-specific genome-wide binding-site predictions that show increased sequence conservation compared with the flanking sequences, supporting their functionality (Janssens, 2022).
To validate the predicted binding sites, Eyeless and Repo CUT&Tag analysis was performed on whole-brain samples, and found a significant overlap. Targeted DamID (TaD5) analysis was performed, finding 8,543 Mef2 peaks in γ-KCs and 10,900 Acj6 peaks in T4/T5 neurons that overlap significantly with predicted cell-type-specific DL and cistrome regions. Interestingly, TaDa analysis of Mef2 in Tm1 neurons detects different sites from those in γ-KCs, whereas the Acj6 sites from T4/T5 neurons contain all those found in Acj6 TaDa analysis of all Acj6-expressing neurons, suggesting a stronger pioneering role for Acj6. As a third validation experiment, cell-type-specific TF knockdowns were performed followed by FACS and ATAC-seq analysis. Mef2, Acj6, Onecut and TfAP-2 knockdowns all resulted in a decreased accessibility of regions with their respective motifs, whereas knocking down the predicted repressor Mamo in γ-KCs increased the accessibility of Mamo sites, and led to a partial switch from the γ-KC-type chromatin landscape to the Mamo-negative α/β-KC-type chromatin landscape (Janssens, 2022).
Finally, bulk ATAC-seq analysis of the adult brain was performed across 44 homozygous fly lines, identifying 4,063 single-nucleotide polymorphisms (SNPs) that correlate with chromatin accessibility. DeepFlyBrain identified affected TF motifs for 20% of the SNPs, consistent with previous studies, with SNPs destroying Mamo or Lola-N repressor binding sites, leading to increased accessibility. When Lola-N was expressed in glia, the Lola-N GATC sites decreased in accessibility, confirming its repressive role in neurons (Janssens, 2022).
The atlas of enhancers, linked to cell types and regulators, enables the design of reporter lines to target cell populations throughout development. Previous efforts to create driver lines for the fly brain (FlyLight and the Vienna Drosophila Resource Center) used random regions of 2-3 kb around neuronal genes, causing many lines to be non-specific: 2,551 out of 3,456 FlyLight lines contain more than one ATAC peak, of which 1,796 are DARs. These lines can be made specific by subcloning the individual ATAC-seq peaks. Furthermore, split-GAL4 lines that combine two enhancers through AND logic are recapitulated as the intersection of their ATAC-seq signal (Janssens, 2022).
Using a more systematic approach, this study selected an additional 60 regions for a total of 63 enhancers and tested their enhancer activity in vivo using transgenic fly lines. The selected regions are accessible in either KCs, OL neurons, glia or mixed with a size range of 300-1,732 bp, and three negative controls were added that are either ubiquitously accessible or inaccessible. Overall, all of the enhancers show reporter activity in the brain, with 73% showing high activity at any developmental stage; whereas, for 65%, the adult reporter activity is specific to the predicted cell type (Janssens, 2022).
Next, the relationship between motif architecture and reporter activity was examined using DeepFlyBrain. An enhancer near the sNPF gene is predicted to be accessible in γ-KCs and α/β-KCs; contains candidate Ey-, Mef2-, Onecut- and Sr-binding sites; and has specific GFP reporter activity in KCs. In silico mutagenesis identified nucleotides with a high impact on accessibility in the Mef2- or Ey-binding site, and mutating these nucleotides abolished activity. A second enhancer near Eip93F is active in T4 neurons, with binding sites for Fkh, Acj6 and Tfap-2. Mutating either Fkh- or Tfap-2-binding sites abolish enhancer activity, confirming their predicted activator roles. Similar analyses were performed on enhancers near Bx, gish, Pkc53E, Appl and CG15117, highlighting TF activator binding sites; GFP reporter activity is lost after mutation of these sites. In the enhancers near sNPF, Bx and Appl, the model predicted that changes in Mamo sites would increase enhancer activity. Indeed, when these sites were mutated, the enhancer activity is not lost, but expanded to additional KC (Janssens, 2022).
Every enhancer was scored based on their KC, OL, glia or CB activity. Calculating receiver operating characteristic (ROC) curves, shows that accessibility alone is able to distinguish positive from negative enhancers with an accuracy of 0.89 for KC, 0.87 for OL and 0.79 for glia. Taking KC activator motif content into account, the accuracy increases to an AUC of 0.935 and 90.5% of enhancers that contain at least two activator motifs are active in KCs (Janssens, 2022).
Current descriptions of GRNs have mostly focused on co-expression, but the availability of transcriptome and chromatin accessibility profiles of cell types enables their regulatory code to be scrutinized. In particular, this study aimed to map cell-type specific eGRNs, including key TFs, as well as their enhancers and target genes (Janssens, 2022).
To link cistrome regions to target genes, a co-variability score of gene expression and region accessibility was calculated for a window of 100 kb around each gene (50 kb upstream and downstream, plus introns), leading to an average of 6 positively linked regions per gene. Enhancer-gene links within BEAF-32 domains (average size, 57.7 kb) have higher correlation scores and a lower proportion of negative links, so links crossing these domains (45%) were pruned. The strength of the region-gene links correlates with enhancer activity, and intronic and distal intergenic regions correlated better with gene expression compared with promoter/TSS accessibility, confirming previous observations. The target genes were pruned using gene set enrichment analysis, retaining those of which the expression co-varies with the cistrome TF (Janssens, 2022).
This procedure resulted in 171 cistromes forming eGRNs for 45 cell types, including 87 activator TFs, with 4,995 enhancers linked to 2,025 genes, covering 17% of the adult DARs (13% promoters, 43% intronic, 44% distal), and 39% of the variable genes in the brain. In particular, cell-type eGRNs have on average 5 activator TFs (range, 1-15) that regulate 67 target genes through 81 enhancers. Indeed, 62% of the genes are regulated by multiple regions within the same cell type and 93% of enhancers have multiple TF inputs. The overlap of predicted binding sites for a TF is only high between similar cell types, suggesting a dependency on the presence of co-factors. The network for γ-KCs reveals that 2/3 of genes are co-regulated by at least two TFs, with the auto-regulatory factors Mef2 and Ey/Toy regulating 97 to 122 genes, alongside Onecut and Sr regulating an average of 38 genes (Janssens, 2022).
Next, different modes of repression were detected. The first type of repressors represses target genes by reducing chromatin accessibility. Mamo is involved in regulating the typical two-out-of-three pattern in KCs by repressing α/β-KC marker genes in α'/β'- and γ-KCs. Similarly, Lola-N represses glial genes in neurons. In the second type, TFs cause nucleosome displacement and chromatin opening, similar to activator TFs, but would recruit co-repressors to repress target genes. This would be manifested by a negative correlation between accessibility and target gene expression. However, these relationships are less common compared with chromatin repression, and were detected for only TFs that also have positive correlation targets, such as Acj6, suggesting more complex mechanisms (Janssens, 2022).
Finally, the cistrome regions were investigated throughout development and found that 45% (14,051) become more accessible at late time-points, with 28.8% increasing after the ecdysone pulse. 458 regions were found that undergo an enhancer switch, as their accessibility increases in one cell type while decreasing in another. One of these regions is a T1 enhancer that drives CG15117 expression that is accessible in early glia and switches to T1 at 24 h APF. Using a scRNA-seq atlas of OL development, it was confirmed that an analogous switch in CG15117 expression occurs, from a developmental glial marker to an adult T1 marker, with a small delay between accessibility and expression changes, as previously observed. When co-staining the enhancer GFP with the glial marker Repo, the overlapping signal in development disappears in the adult, coinciding with the closing of the enhancer. To study this phenomenon at the sequence level, a DL model was trained using developmental topics as input. Inspecting the CG15117 enhancer, the model detects the same TAATTA motif in glia and T1 neurons. Given that only a few TFs overlap, this suggests the binding of different factors to the same motif in different cell types. The reuse of the same enhancer in different cell types at different timepoints can also be noticed by differences in expression between larval and adult brains for 47 out of 54 tested enhancers (Janssens, 2022).
This study generated the first single-cell chromatin accessibility atlas of the whole fly brain throughout development, tracing neuronal and glial cell types from birth to maturity. Using an integrated multi-omics approach, the concept of eGRNs was introduced in which TFs are linked to high-confidence binding sites that are linked to target genes. eGRNs can soon be derived for other datasets, given the pioneering work in scATAC-seq in mouse and human and developments in single-cell multi-omics (Janssens, 2022).
The atlas showed that all cell types in the brain have unique chromatin profiles, often combinatorial, with tens of thousands of accessible regions and hundreds to thousands specific, identifying over 95,000 candidate enhancers, covering 34.4% of the genome. To accurately predict enhancer activity on the basis of the DNA sequence, DL models were integrated with omics data. This 'smarter' motif discovery has been shown to reveal motifs that are missed by conventional algorithms, and leads to prediction of TF binding at the base-pair resolution. These annotations and TF roles were validated using CUT&Tag, TaDa and knockdown experiments, confirming the high quality of the cistromes and DL-based annotations. The library of annotated enhancers was then used as a starting point to clean up or design new driver lines for cell-type-specific genetic access. Developmental dynamics open up the possibility of creating spatiotemporal driver lines with enhancers corresponding to different maturation modules (Janssens, 2022).
By linking TF cistromes, enhancer accessibility and target gene expression, 45 eGRNs covering 87 activator TFs were generated of which 90% have lethal mutations, 62% are linked to known brain phenotypes and 64% are linked to human diseases, providing a foundation for follow-up studies. Many enhancers were regulated by multiple TFs, as highlighted by Ey and Mef2 in KCs. Furthermore, mutating either binding site led to the abolishment of GFP activity in in vivo assays, suggesting cooperativity. The switching enhancers described here present a considerable fraction of enhancers (~500) encoding multilayered motif architectures, reminiscent of the phenotypic convergence phenomenon of different TFs in different cell types regulating the same enhancer (Janssens, 2022).
This regulatory atlas of the brain covers cell types, TFs and enhancers, together with their joint representation as eGRNs and transitions through development. To be of further value to the community, all data has been made publicly available online, enabling users to explore eGRNs with links to SCOPE and UCSC. Finally, the adult DL model DeepFlyBrain is available at Kipoi (Janssens, 2022).
Temporal patterning of neural progenitors is an evolutionarily conserved strategy for generating neuronal diversity. Type II neural stem cells in the Drosophila central brain produce transit-amplifying intermediate neural progenitors (INPs) that exhibit temporal patterning. However, the known temporal factors cannot account for the neuronal diversity in the adult brain. To search for new temporal factors, NanoDam, which enables rapid genome-wide profiling of endogenously-tagged proteins in vivo with a single genetic cross, was developed. Mapping the targets of known temporal transcription factors with NanoDam identified Homeobrain and Scarecrow (ARX and NKX2.1 orthologues) as novel temporal factors. Homeobrain and Scarecrow define middle-aged and late INP temporal windows and play a role in cellular longevity. Strikingly, Homeobrain and Scarecrow have conserved functions as temporal factors in the developing visual system. NanoDam enables rapid cell type-specific genome-wide profiling with temporal resolution and can be easily adapted for use in higher organisms (Tang, 2021).
The nervous system is generated by a relatively small number of neural stem cells (NSCs) and progenitors that are patterned both spatially and temporally. Spatial patterning confers differences between populations of NSCs, while changes in gene expression over time direct the birth order and subtype identity of neuronal progeny. Temporal transcription factor cascades determine neuronal birth order in the Drosophila embryonic central nervous system (CNS) and the larval central brain and optic lobe. In the central brain, Type II NSCs generate transit-amplifying intermediate neural progenitors (INPs), which divide asymmetrically to self-renew and generate daughter cells (ganglion mother cells or GMCs) in a manner analogous to human outer radial glial cells. GMCs in turn undergo a terminal cell division, generating neurons or glial cells that contribute to the adult central complex. The sequential divisions of INPs increase the quantity of neurons, which in turn creates a platform for generating wider neuronal diversity: eight Type II NSCs in each brain lobe give rise to at least 60 different neuronal subtypes. The tight control of progenitor temporal identity is crucial for the production of neuronal subtypes at the appropriate time and in the correct numbers. The INPs produced by the six dorsal-medial Type II lineages (DM1-6) sequentially express the temporal transcription factors Dichaete (D, a member of the Sox family), Grainy head (Grh, a Grh/CP2 family transcription factor) and Eyeless (Ey, a homologue of Pax6). These temporal factors were discovered initially by screening Type II lineages for restricted expression of neural transcription factors, using 60 different antisera. This non-exhaustive approach was able to find a fraction of the theoretically necessary temporal factors, leaving the true extent of temporal regulation and the identity of missing temporal factors open. Furthermore, the cross-regulatory interactions predicted in a temporal cascade, in which each temporal transcription factor activates expression of the next temporal factor and represses expression of the temporal factor preceding it, are not fulfilled solely by D, Grh and Ey (Tang, 2021).
Three further factors contribute to INP temporal progression, but they are expressed broadly rather than in discrete temporal windows: Osa, a SWI/SNF chromatin remodelling complex subunit, and two further transcription factors, Odd-paired (Opa) and Hamlet (Ham). Therefore, there must exist other transcription factors that are expressed in defined temporal windows and exhibit the regulatory interactions expected in a temporal cascade. It was postulated that other temporal factors, that contribute to generating the diversity of neuronal subtypes arising within each INP lineage, remain to be identified (Tang, 2021).
Given the feed-forward and feed-back transcriptional regulation previously observed in temporal transcription cascades, it was surmised that novel temporal factors would be amongst the transcriptional targets of D, Grh or Ey. Therefore, a novel approach, NanoDam, was devised to identify the genome-wide targets of transcription factors within their normal expression windows in vivo without cell isolation, cross linking or immunoprecipitation. Temporal factors are expressed transiently in a small pool of rapidly dividing progenitor cells. NanoDam provides a simple, streamlined approach to obtain genome-wide binding profiles in a cell-type-specific and temporally restricted manner (Tang, 2021).
Using NanoDam, the transcriptional targets of D, Grh and Ey were determined in INPs and, by performing single cell RNA sequencing, determined which of the directly bound loci were activated or repressed. Next, which of the target loci encoded transcription factors was assessed and whether these were expressed in restricted temporal windows within INPs. Where in the INP transcriptional cascade these factors acted was surveyed and whether they cross regulate the expression of other temporal transcription factor genes was ascertained, as expected for temporal factors. Finally, it was shown that the newly discovered temporal factors play the same roles, and exhibit the same cross regulatory interactions, in the temporal cascade in the developing visual system. This is particularly striking as theINPs and the NSCs of the developing optic lobe have different cells of origin and yet the mechanism they use to generate neural diversity is conserved (Tang, 2021).
Temporal patterning leads to the generation of neuronal diversity from a relatively small pool of neural stem or progenitor cells. Temporal regulation is achieved by the restricted expression of temporal transcription factors within precise developmental windows. The onset and duration of each temporal window in neural stem or progenitor cells must be regulated tightly in order for the appropriate subtypes of neurons to be generated at the correct time to establish functional neuronal circuits (Tang, 2021).
This study focused on the INPs of the Type II NSC lineages that generate the central complex of the Drosophila brain. Previously, INPs were shown to express sequentially the temporal factors D, Grh and Ey. Given the expectation that other temporal factors remained to be discovered, and that these were likely to be the transcriptional targets of the known temporal factors, a new technique called was used NanoDam to profile the binding targets of D, Grh and Ey with cell-type specificity and within their individual temporal windows (Tang, 2021).
NanoDam enables genome-wide profiling of any endogenously tagged chromatin-binding protein with a simple genetic cross, bypassing the need to generate Dam-fusion proteins, or the need for specific antisera or cell isolation. Furthermore, NanoDam profiles binding only in cells where the tagged protein is normally expressed. Binding within a subset of the protein's expression pattern can be achieved by controlling NanoDam with specific GAL4 drivers. To date, collaborative efforts have produced more than 3900 Drosophila lines expressing GFP-tagged proteins in their endogenous patterns. Approximately 93% of all transcription factors have been GFP-tagged in lines that are publicly available at stock centres. Lines that are not yet available can be rapidly generated by CRISPR/Cas9-mediated tagging (Tang, 2021).
NanoDam is thus a versatile tool that can be used as a higher throughput method to profile genome-wide binding sites of any chromatin associated protein. NanoDam can be readily adapted for use in other organisms to facilitate simpler and easier in vivoprofiling experiments, as hads been demonstrated previously for TaDa (Tang, 2021).
By combining the power of NanoDam with scRNA-seq, it was possible to identify scro and hbn as novel temporal factors in the INPs of Type II NSC lineages. It was shown that hbn and scro regulate the maintenance and transition of the middle-aged and late temporal windows. The mammalian homologues of ey (Pax6), hbn (Arx) and scro (Nkx2.1) are restricted to distinct progenitor populations in the developing mouse forebrain. This study found that scro regulates the late INP identity by repression of Ey. Interestingly, the loss of Nkx2.1 in the mouse forebrain leads to aberrant expression in ventral regions of the dorsal factor Pax6, suggesting that the repressive relationship between scro and ey may be conserved between Nkx2.1 and Pax6. Not all relationships appear to be conserved, however. It esd found that Hbn promotes progression through the middle-aged temporal stage and that maintenance of the middle-aged temporal window is regulated in part by interactions between Hbn and Grh. Arx mutant mice exhibit loss of upper layer (later-born) neurons but no change in the number of lower layer (early-born) neurons (Tang, 2021).
Intriguingly, the novel temporal factors identified in the INPs were also temporally expressed in optic lobe NSCs and the regulatory relationships between scro and Ey appeared to be conserved. This suggests that similar regulatory strategies may be shared between neural stem cells or progenitor cells inorder to regulate longevity and neuronal subtype production. The remarkable conservation of the regulatory interactions of scroin two different progenitor cell types with different origins in the Drosophila brain may also be translated to the context of mammalian neurogenesis, highlighting the possibility of a more generalised regulatory network used by stem and progenitor cells to regulate cell fate, progeny fate and proliferation (Tang, 2021).
The Type II lineages in Drosophila divide in a very similar manner to the outer radial glia (oRGs) that have been attributed to the rapid evolutionary expansion of the neocortex seen in humans and other mammals. Interestingly, oRGs show a shortened cell cycle length in primates in comparison to rodent progenitors, which increase cell cycle duration as development progresses . Investigating whether oRGs use temporally expressed factors to control longevity and cell cycle dynamics at different developmental stages in order to regulate neuronal subtype generation would be important for understanding neocortex development (Tang, 2021).
There is significant heterogeneity between the Type II lineages and this study has identified differences in the regulatory relationships of hbn and scro. For example, misexpression of Ey leads to an increase in scro in all lineages except DM 2 and 3, where scro expression is reduced. To date, Hbn is the only factor identified that activates Grh in DM1, the lineage that does not normally express Grh. The heterogeneity between lineages may be a consequence of variations in combinatorial binding of temporal factors, as the NanoDam data indicate. Although INPs share temporal factors, different DM lineages display subtle to striking differences when the temporal cascade is manipulated, demonstrating the likelihood that each DM employs unique temporal cascades. Combinatorial binding would enable more complex regulatory interactions that could refine or sub-divide temporal windows in the INPs (Tang, 2021).
Kolb, D., Kaspar, P., Kloppel, C. and Walldorf, U. (2021). The Drosophila homeodomain transcription factor Homeobrain is involved in the formation of the embryonic protocerebrum and the supraesophageal brain commissure. Cells Dev 165: 203657. PubMed ID: 33993980
During the embryonic development of Drosophila melanogaster many transcriptional activators are involved in the formation of the embryonic brain. This study shows that the transcription factor Homeobrain (Hbn), a member of the 57B homeobox gene cluster, is an additional factor involved in the formation of the embryonic Drosophila brain. Using a Hbn antibody and specific cell type markers a detailed expression analysis during embryonic brain development was conducted. Hbn is expressed in several regions in the protocerebrum, including fibre tract founder cells closely associated with the supraesophageal brain commissure and also in the mushroom bodies. During the formation of the supraesophageal commissure, Hbn and FasII-positive founder cells build an interhemispheric bridge priming the commissure and thereby linking both brain hemispheres. The Hbn expression is restricted to neural but not glial cells in the embryonic brain. In a mutagenesis screen two mutant hbn alleles were generated that both show embryonic lethality. The phenotype of the hbn mutant alleles is characterized by a reduction of the protocerebrum, a loss of the supraesophageal commissure and mushroom body progenitors and also by a dislocation of the optic lobes. Extensive apoptosis correlates with the impaired formation of the embryonic protocerebrum and the supraesophageal commissure. These results show that Hbn is another important factor for embryonic brain development in Drosophila melanogaster (Kolb, 2021).
The homeobox gene homeobrain (hbn) Honey bees (Apis mellifera) express remarkable social interactions and cognitive capabilities that have been studied extensively. In many cases, behavioral studies were accompanied by neurophysiological and neuroanatomical investigations. While most studies have focused on primary sensory neuropils, such as the optic lobes or antennal lobes, and major integration centers, such as the mushroom bodies or the central complex, many regions of the cerebrum (the central brain without the optic lobes) of the honey bee are only poorly explored so far, both anatomically and physiologically. To promote studies of these brain regions, this study used anti-synapsin immunolabeling and neuronal tract tracings followed by confocal imaging and 3D reconstructions to demarcate all neuropils in the honey bee cerebrum and close this gap at the anatomical level. 35 neuropils and 25 fiber tracts were demarcated in the honey bee cerebrum, most of which have counterparts in the fly (Drosophila melanogaster) and other insect species that have been investigated so far at this level of detail. The role of cerebral neuropils in multisensory integration in the insect brain is discussed, the importance of this brain atlas for comparative studies is emphasized, and specific architectural features of the honey bee cerebrum are discussed (Habenstein, 2023).
The level of each RNA species depends on the balance between its rates of production and decay. Although previous studies have measured RNA decay across the genome in tissue culture and single-celled organisms, few experiments have been performed in intact complex tissues and organs. It is therefore unclear whether the determinants of RNA decay found in cultured cells are preserved in an intact tissue, and whether they differ between neighboring cell types and are regulated during development. To address these questions, this study measured RNA synthesis and decay rates genome wide via metabolic labeling of whole cultured Drosophila larval brains using 4-thiouridine. This analysis revealed that decay rates span a range of more than 100-fold, and that RNA stability is linked to gene function, with mRNAs encoding transcription factors being much less stable than mRNAs involved in core metabolic functions. Surprisingly, among transcription factor mRNAs there was a clear demarcation between more widely used transcription factors and those that are expressed only transiently during development. mRNAs encoding transient transcription factors are among the least stable in the brain. These mRNAs are characterized by epigenetic silencing in most cell types, as shown by their enrichment with the histone modification H3K27me3. This data suggests the presence of an mRNA destabilizing mechanism targeted to these transiently expressed transcription factors to allow their levels to be regulated rapidly with high precision. This study also demonstrates a general method for measuring mRNA transcription and decay rates in intact organs or tissues, offering insights into the role of mRNA stability in the regulation of complex developmental programs (Thompson, 2022).
Abdusselamoglu, M. D., Eroglu, E., Burkard, T. R. and Knoblich, J. A. (2019). The transcription factor odd-paired regulates temporal identity in transit-amplifying neural progenitors via an incoherent feed-forward loop. Elife 8. PubMed ID: 31329099
Alpert, M. H., Frank, D. D., Kaspi, E., Flourakis, M., Zaharieva, E. E., Allada, R., Para, A. and Gallio, M. (2020). A circuit encoding absolute cold temperature in Drosophila. Curr Biol. PubMed ID: 32442464
Axelrod, S., Li, X., Sun, Y., Lincoln, S., Terceros, A., O'Neil, J., Wang, Z., Nguyen, A., Vora, A., Spicer, C., Shapiro, B., Young, M. W. (2023). The Drosophila blood-brain barrier regulates sleep via Moody G protein-coupled receptor signaling. Proc Natl Acad Sci U S A, 120(42):e2309331120 PubMed ID: 37831742
Bayraktar, O. A. and Doe, C. Q. (2013). Combinatorial temporal patterning in progenitors expands neural diversity. Nature 498: 449-455. PubMed ID: 23783519
Bjordal, M., Arquier, N., Kniazeff, J., Pin, J. P. and Leopold, P. (2014). Sensing of amino acids in a dopaminergic circuitry promotes rejection of an incomplete diet in Drosophila. Cell 156: 510-521. PubMed ID: 24485457
Bonifazi, P., Goldin, M., Picardo, M. A., Jorquera, I., Cattani, A., Bianconi, G., Represa, A., Ben-Ari, Y. and Cossart, R. (2009). GABAergic hub neurons orchestrate synchrony in developing hippocampal networks. Science 326: 1419-1424. PubMed ID: 19965761
Brankatschk, M. and Eaton, S. (2010). Lipoprotein particles cross the blood-brain barrier in Drosophila. J Neurosci 30: 10441-10447. PubMed ID: 20685986
Brankatschk, M., Dunst, S., Nemetschke, L. and Eaton, S. (2014). Delivery of circulating lipoproteins to specific neurons in the brain regulates systemic insulin signaling. Elife 3 [Epub ahead of print]. PubMed ID: 25275323
Canavoso, L. E., Yun, H. K., Jouni, Z. E. and Wells, M. A. (2004). Lipid transfer particle mediates the delivery of diacylglycerol from lipophorin to fat body in larval Manduca sexta. J Lipid Res 45: 456-465. PubMed ID: 14679163
Chen, C. L., Aymanns, F., Minegishi, R., Matsuda, V. D. V., Talabot, N., Gunel, S., Dickson, B. J. and Ramdya, P. (2023). Ascending neurons convey behavioral state to integrative sensory and action selection brain regions. Nat Neurosci 26(4): 682-695. PubMed ID: 36959417
Curt, J. R., Yaghmaeian Salmani, B. and Thor, S. (2019). Anterior CNS expansion driven by brain transcription factors. Elife 8. PubMed ID: 31271353
Dhaka, A., Viswanath, V. and Patapoutian, A. (2008). TRP ion channels and temperature sensation. Annu. Rev. Neurosci. 29: 135-161. PubMed Abstract: 16776582
Dolan, M. J., Belliart-Guerin, G., Bates, A. S., Frechter, S., Lampin-Saint-Amaux, A., Aso, Y., Roberts, R. J. V., Schlegel, P., Wong, A., Hammad, A., Bock, D., Rubin, G. M., Preat, T., Placais, P. Y. and Jefferis, G. (2018). Communication from learned to innate olfactory processing centers is required for memory retrieval in Drosophila. Neuron 100(3): 651-668 e658. PubMed ID: 30244885
Dolan, M. J., Frechter, S., Bates, A. S., Dan, C., Huoviala, P., Roberts, R. J., Schlegel, P., Dhawan, S., Tabano, R., Dionne, H., Christoforou, C., Close, K., Sutcliffe, B., Giuliani, B., Li, F., Costa, M., Ihrke, G., Meissner, G. W., Bock, D. D., Aso, Y., Rubin, G. M. and Jefferis, G. S. (2019). Neurogenetic dissection of the Drosophila lateral horn reveals major outputs, diverse behavioural functions, and interactions with the mushroom body. Elife 8: e43079. PubMed ID: 31112130
Eroglu, E., Burkard, T. R., Jiang, Y., Saini, N., Homem, C. C., Reichert, H. and Knoblich, J. A. (2014). SWI/SNF complex prevents lineage reversion and induces temporal patterning in neural stem cells. Cell 156: 1259-1273. PubMed ID: 24630726
Frechter, S., Bates, A. S., Tootoonian, S., Dolan, M. J., Manton, J. D., Jamasb, A. R., Kohl, J., Bock, D. and Jefferis, G. S. (2019). Functional and anatomical specificity in a higher olfactory centre. Elife 8: e44590. PubMed ID: 31112127
Grabowska, M. J., Jeans, R., Steeves, J. and van Swinderen, B. (2020). Oscillations in the central brain of Drosophila are phase locked to attended visual features. Proc Natl Acad Sci U S A. PubMed ID: 33177231
Habenstein, J., Grubel, K., Pfeiffer, K. and Rossler, W. (2023). 3D atlas of cerebral neuropils with previously unknown demarcations in the honey bee brain. J Comp Neurol. PubMed ID: 37070301
Hamada, F. N., et al. (2008). An internal thermal sensor determining temperature preference in Drosophila. Nature 454: 217-220. PubMed ID: 18548007
Hildebrandt, K., Kolb, D., Kloppel, C., Kaspar, P., Wittling, F., Hartwig, O., Federspiel, J., Findji, I. and Walldorf, U. (2022). Regulatory modules mediating the complex neural expression patterns of the homeobrain gene during Drosophila brain development. Hereditas 159(1): 2. PubMed ID: 34983686
Huang, S., Piao, C., Beuschel, C. B., Zhao, Z. and Sigrist, S. J. (2022). A brain-wide form of presynaptic active zone plasticity orchestrates resilience to brain aging in Drosophila. PLoS Biol 20(12): e3001730. PubMed ID: 36469518
Ikeda, K., Kataoka, M. and Tanaka, N. K. (2022). Nonsynaptic
Mediates Light Context-Dependent Odor Responses in Drosophila melanogaster. J Neurosci 42(46): 8621-8628. PubMed ID: 36180227
Janssens, J., Aibar, S., Taskiran, II, Ismail, J. N., Gomez, A. E., Aughey, G., Spanier, K. I., De Rop, F. V., Gonzalez-Blas, C. B., Dionne, M., Grimes, K., Quan, X. J., Papasokrati, D., Hulselmans, G., Makhzami, S., De Waegeneer, M., Christiaens, V., Southall, T. and Aerts, S. (2022). Decoding gene regulation in the fly brain. Nature 601(7894): 630-636. PubMed ID: 34987221
Jenett, A., et al. (2012). A GAL4-driver line resource for Drosophila neurobiology. Cell Rep 2: 991-1001. PubMed ID: 23063364
Kolb, D., Kaspar, P., Kloppel, C. and Walldorf, U. (2021). The Drosophila homeodomain transcription factor Homeobrain is involved in the formation of the embryonic protocerebrum and the supraesophageal brain commissure. Cells Dev 165: 203657. PubMed ID: 33993980
Kunin, A. B., Guo, J., Bassler, K. E., Pitkow, X. and Josic, K. (2023). Hierarchical Modular Structure of the Drosophila Connectome. J Neurosci. PubMed ID: 37591738
Lapraz, F., Boutres, C., Fixary-Schuster, C., De Queiroz, B. R., Placais, P. Y., Cerezo, D., Besse, F., Preat, T. and Noselli, S. (2023). Asymmetric activity of NetrinB controls laterality of the Drosophila brain. Nat Commun 14(1): 1052. PubMed ID: 36828820
Lee, Y. J., Yang, C. P., Miyares, R. L., Huang, Y. F., He, Y., Ren, Q., Chen, H. M., Kawase, T., Ito, M., Otsuna, H., Sugino, K., Aso, Y., Ito, K. and Lee, T. (2020). Conservation and divergence of related neuronal lineages in the Drosophila central brain. Elife 9. PubMed ID: 32255422
Li, J., Han, S., Li, H., Udeshi, N. D., Svinkina, T., Mani, D. R., Xu, C., Guajardo, R., Xie, Q., Li, T., Luginbuhl, D. J., Wu, B., McLaughlin, C. N., Xie, A., Kaewsapsak, P., Quake, S. R., Carr, S. A., Ting, A. Y. and Luo, L. (2020). Cell-surface proteomic profiling in the fly brain uncovers wiring regulators. Cell 180(2): 373-386 e315. PubMed ID: 31955847
Li, Y., Chang, P., Sankaran, S., Jang, H., Nie, Y., Zeng, A., Hussain, S., Wu, J. Y., Chen, X. and Shi, L. (2023). Bioorthogonal Stimulated Raman Scattering Imaging Uncovers Lipid Metabolic Dynamics in Drosophila Brain During Aging. GEN Biotechnol 2(3): 247-261. PubMed ID: 37363411
Liu, L. Y., Long, X., Yang, C. P., Miyares, R. L., Sugino, K., Singer, R. H. and Lee, T. (2019). Mamo decodes hierarchical temporal gradients into terminal neuronal fate. Elife 8. PubMed ID: 31545163
Loyer, N. and Januschke, J. (2018). The last-born daughter cell contributes to division orientation of Drosophila larval neuroblasts. Nat Commun 9(1): 3745. PubMed ID: 30218051
Mann, K., Deny, S., Ganguli, S. and Clandinin, T. R. (2021). Coupling of activity, metabolism and behaviour across the Drosophila brain. Nature 593(7858): 244-248. PubMed ID: 33911283
Marquis, M. and Wilson, R. I. (2022). Locomotor and olfactory responses in dopamine neurons of the Drosophila superior-lateral brain. Curr Biol 32(24): 5406-5414. PubMed ID: 36450284
Marin, E. C., Buld, L., Theiss, M., Sarkissian, T., Roberts, R. J. V., Turnbull, R., Tamimi, I. F. M., Pleijzier, M. W., Laursen, W. J., Drummond, N., Schlegel, P., Bates, A. S., Li, F., Landgraf, M., Costa, M., Bock, D. D., Garrity, P. A. and Jefferis, G. (2020). Connectomics Analysis Reveals First-, Second-, and Third-Order Thermosensory and Hygrosensory Neurons in the Adult Drosophila Brain. Curr Biol. PubMed ID: 32619476
Mehta, K., Goldin, R. F. and Ascoli, G. A. (2023). Circuit analysis of the Drosophila brain using connectivity-based neuronal classification reveals organization of key communication pathways. Netw Neurosci 7(1): 269-298. PubMed ID: 37339321
Namiki, S., Dickinson, M. H., Wong, A. M., Korff, W. and Card, G. M. (2018). The functional organization of descending sensory-motor pathways in Drosophila. Elife 7: e34272. PubMed ID: 29943730
Oikawa, I., Kondo, S., Hashimoto, K., Yoshida, A., Hamajima, M., Tanimoto, H., Furukubo-Tokunaga, K. and Honjo, K. (2023). A descending inhibitory mechanism of nociception mediated by an evolutionarily conserved neuropeptide system in Drosophila. Elife 12. PubMed ID: 37310871
Parra-Peralbo, E. and Culi, J. (2011). Drosophila lipophorin receptors mediate the uptake of neutral lipids in oocytes and imaginal disc cells by an endocytosis-independent mechanism. PLoS Genet 7: e1001297. PubMed ID: 21347279
Perrat, P. N., DasGupta, S., Wang, J., Theurkauf, W., Weng, Z., Rosbash, M. and Waddell, S. (2013). Transposition-driven genomic heterogeneity in the Drosophila brain. Science 340: 91-95. PubMed ID: 23559253
Post, S., Liao, S., Yamamoto, R., Veenstra, J. A., Nassel, D. R. and Tatar, M. (2018). Drosophila insulin-like peptide dilp1 increases lifespan and glucagon-like Akh expression epistatic to dilp2. Aging Cell: e12863. PubMed ID: 30511458
Raji, J. I. and Potter, C. J. (2021). The number of neurons in Drosophila and mosquito brains. PLoS One 16(5): e0250381. PubMed ID: 33989293
Rosenzweig, M., et al. (2005). The Drosophila ortholog of vertebrate TRPA1 regulates thermotaxis. Genes Dev. 19: 419-424. PubMed ID: 15681611
Sakamura, S., Hsu, F. Y., Tsujita, A., Abubaker, M. B., Chiang, A. S. and Matsuno, K. (2023). Ecdysone signaling determines lateral polarity and remodels neurites to form Drosophila's left-right brain asymmetry. Cell Rep: 112337. PubMed ID: 37044096
Simoes, A. R., Neto, M., Alves, C. S., Santos, M. B., Fernandez-Hernandez, I., Veiga-Fernandes, H., Brea, D., Dur, I., Encinas, J. M. and Rhiner, C. (2022). Damage-responsive neuro-glial clusters coordinate the recruitment of dormant neural stem cells in Drosophila. Dev Cell. PubMed ID: 35716661
Tang, J. L. Y., Hakes, A. E., Krautz, R., Suzuki, T. Contreras, E. G., Fox, P. M. and Brand, A. H. (2021). NanoDam identifies novel temporal transcription factors conserved between the Drosophila central brain and visual system Biorxiv
Tepe, B., Macke, E. L., Niceta, M., Weisz Hubshman, M., Kanca, O., Schultz-Rogers, L., Zarate, Y. A., Schaefer, G. B., Granadillo De Luque, J. L., Wegner, D. J., Cogne, B., Gilbert-Dussardier, B., Le Guillou, X., Wagner, E. J., Pais, L. S., Neil, J. E., Mochida, G. H., Walsh, C. A., Magal, N., Drasinover, V., Shohat, M., Schwab, T., Schmitz, C., Clark, K., Fine, A., Lanpher, B., Gavrilova, R., Blanc, P., Burglen, L., Afenjar, A., Steel, D., Kurian, M. A., Prabhakar, P., Goswein, S., Di Donato, N., Bertini, E. S., Wangler, M. F., Yamamoto, S., Tartaglia, M., Klee, E. W. and Bellen, H. J. (2023). Bi-allelic variants in INTS11 are associated with a complex neurological disorder. Am J Hum Genet. PubMed ID: 37054711
Thompson, M. K., Ceccarelli, A., Ish-Horowicz, D. and Davis, I. (2023). Dynamically regulated transcription factors are encoded by highly unstable mRNAs in the Drosophila larval brain RNA. PubMed ID: 37041032
Turner, M. H., Mann, K. and Clandinin, T. R. (2021). The connectome predicts resting-state functional connectivity across the Drosophila brain. Curr Biol. PubMed ID: 33770490
Urbach, R. and Technau, G. M. (2003a). Segment polarity and DV patterning gene expression reveals segmental organization of the Drosophila brain. Development 130: 3607-3620. PubMed ID: 12835379
Urbach, R. and Technau, G. M. (2003b). Molecular markers for identified neuroblasts in the developing brain of Drosophila. Development 130: 3621-3637. PubMed ID: 12835380
P>Vallejo, D. M., Juarez-Carreño, S., Bolivar, J., Morante, J. and Dominguez, M.
(2015). A brain circuit that synchronizes growth and maturation revealed through Dilp8 binding to
Lgr3. Science [Epub ahead of print]. PubMed ID: 26429885 Van Heusden, M. C. and Law, J. H. (1989). An insect lipid transfer particle promotes lipid loading from fat body to lipoprotein. J Biol Chem 264: 17287-17292. PubMed ID: 2793856
van Swinderen, B. and Greenspan, R. J. (2003). Salience modulates 20-30 Hz brain activity in Drosophila. Nat Neurosci 6(6): 579-586. PubMed ID: 12717438
Viswanath, V., et al. (2003). Opposite thermosensor in fruitfly and mouse. Nature 423: 822-823. PubMed Abstract: 12815418
Walker, D., Schmid, E., Schinaman, J. and Williams, K. (2023). Accumulation of F-actin drives brain aging and limits healthspan in Drosophila. Res Sq. PubMed ID: 37577708
Wei, J. Y., Chu, S. Y., Huang, Y. C., Chung, P. C. and Yu, H. H. (2022). Drosophila septin interacting protein 1 regulates neurogenesis in the early developing larval brain. Sci Rep 12(1): 292. PubMed ID: 34997175
Winding, M., Pedigo, B. D., Barnes, C. L., Patsolic, H. G., Park, Y., Kazimiers, T., Fushiki, A., Andrade, I. V., Khandelwal, A., Valdes-Aleman, J., Li, F., Randel, N., Barsotti, E., Correia, A., Fetter, R. D., Hartenstein, V., Priebe, C. E., Vogelstein, J. T., Cardona, A. and Zlatic, M. (2023). The connectome of an insect brain. Science 379(6636): eadd9330. PubMed ID: 36893230
Yaghmaeian Salmani, B., Monedero Cobeta, I., Rakar, J., Bauer, S., Curt, J. R., Starkenberg, A. and Thor, S. (2018). Evolutionarily conserved anterior expansion of the central nervous system promoted by a common PcG-Hox program. Development 145(7). PubMed ID: 29530878
genes expressed in brain morphogenesis
Genes involved in organ development
Home page: The Interactive Fly © 1995, 1996 Thomas B. Brody, Ph.D.
The Interactive Fly resides on the
Society for Developmental Biology's Web server.