InteractiveFly: GeneBrief

traffic jam: Biological Overview | References

BIOLOGICAL OVERVIEW

Interactions between somatic and germline cells are critical for the normal development of egg and sperm. This study shows that the gene traffic jam (tj) produces a soma-specific factor that controls gonad morphogenesis and is required for female and male fertility. tj encodes the only large Maf (musculoaponeurotic fibrosarcoma) factor in Drosophila, an orthologue of the atypical basic Leu zipper transcription factors c-Maf and MafB/Kreisler in vertebrates. Expression of tj occurs in somatic gonadal cells that are in direct contact with germline cells throughout development. In tj mutant gonads, somatic cells fail to inter-mingle and properly envelop germline cells, causing an early block in germ cell differentiation. In addition, tj mutant somatic cells show an increase in the level of expression for several adhesion molecules. It is proposed that tj is a critical modulator of the adhesive properties of somatic cells, facilitating germline–soma interactions that are essential for germ cell differentiation (Li, 2003).

During the process of becoming a gamete, a Drosophila germline cell encounters different types of somatic gonadal cells (SGCs) that influence its development. Although these SGC types differ in function, morphology and the use of signalling molecules, they all share the ability to interact with the germline. Whether this ability is promoted by a common molecular mechanism has remained unknown. This study shows that all SGCs contacting the germline during gonadogenesis and gametogenesis express a transcription factor called Traffic Jam (Li, 2003).

tj mutant female and male flies are sterile (Schüpbach, 1991). tj mutants (tj^-/-) have severe gonadal defects, but are viable and display no other gross morphological abnormalities, suggesting that tj may have a specific function in gonad development. tj mutant ovaries were small, disorganized, and lacked mature germ cells (Schüpbach, 1991). Germline cells were completely absent in 26–27% of the ovaries from 1–2-day-old tj^-/- females, and within 1–2 weeks the number of such ovaries doubled, suggesting that germ cells were lost over time. Persisting female germ cells form irregular clusters of variable size that seem randomly distributed in the ovary and reach various degrees of early differentiation. Some germ cells developed to the follicle stage but were not enveloped by somatic follicle cells. Follicle cells seemed to be missing in tj^-/- ovaries. The observed germ-cell defects were not germline-dependent, as mosaic follicles with a tj mutant germline and wild-type follicle cells developed normally. In addition, ovaries that lack a germline, as in oskar (osk) mutants, still developed follicle cells, whereas tj osk double mutants displayed the same somatic defects as tj mutants. These findings suggest that the lack of follicle cells is not caused by a defective germline but by the loss of tj function in the soma (Li, 2003).

tj^-/- testes are irregularly shaped balls instead of long coiled tubes. Wild-type testes are filled with a developmental series of differentiating germ cells, showing mitotically active cells at the anterior tip, followed by spermatocytes and germ cells at later stages of differentiation. tj^-/- testes only contained a few groups of germ cells that remained in a pre-spermatocyte state, as suggested by their small size and the lack of fully branched fusomes. The remainder of tj mutant testes were filled with somatic cells that are segregated from the germ cells, whereas normally each differentiating germline cyst is ensheathed by a pair of somatic cyst cells. Together, this analysis indicates that tj has a soma-dependent function that contributes to normal development of somatic tissues and gametogenesis in both sexes (Li, 2003).

Deletion mapping confined tj to a 50–60 kb region in chromosomal interval 37E1. As tj mutations cause gonad-specific defects, this region was screened for transcripts expressed in larval and adult gonads by tissue in situ-hybridization. A 7.1-kb genomic BglII–SmaI fragment detected a transcription unit that was expressed almost exclusively in the gonads of both sexes, and was therefore an attractive candidate for tj. Corresponding cDNAs identified an open reading frame (ORF) that is contained within a single exon and represents the predicted gene CG10034, translating into a predicted protein of 509 amino acids. Two transcripts were detected by northern blot analysis, with the smaller, more abundant, transcript being similar in length (3.5 kbs) to the combined length of two overlapping cDNAs (3.2 kbs). To determine whether this transcription unit corresponds to the tj gene, four mutant tj alleles were sequenced. Each tj allele contains a premature stop codon in the identified ORF, indicating that the tj gene had been cloned (Li, 2003).

TJ is a homologue of the retroviral oncoprotein v-Maf (Kataoka, 1993; Blank, 1997) and the large Maf transcription factors of vertebrates (Blank, 1997; Kerppola, 1994; Dlakic, 2001). Similar to all Maf factors, TJ has a Leu zipper domain, an atypical DNA-binding basic domain and a Maf-specific extended homology region, which functions as an ancillary DNA-binding domain (Blank, 1997). The basic domains of TJ and the mammalian large Mafs show 85% identity. The conserved Gly and Tyr residues of the basic domain are a signature of all Maf factors, and distinguish them in their DNA-binding properties from classic basic Leu zipper transcription factors. Similar to other large Mafs, and in contrast to small Mafs, TJ has an extended amino-terminal portion that contains sequences rich in acidic amino acids implicated in transactivation of transcription (Kurschner, 1995), followed by sequences of low complexity. Among mammalian large Mafs, TJ is most similar to c-Maf and MafB, as indicated by the high similarity in the three conserved domains, the common motif SSPEF at the carboxyl terminus and overall protein architecture. TJ is the only large Maf in Drosophila. In addition, Drosophila has one small Maf, MafS, which is involved in head development (Veraksa, 2000). A comparison of the conserved region of Maf factors across phyla indicates that the split into a small and a large Maf gene has occurred before the separation of arthropods and chordates, making them two distinct Maf subfamilies that subsequently underwent duplications in the chordate lineage (Li, 2003).

TJ is a nuclear protein that is expressed in the somatic cells of gonads. No TJ protein was found in gonads of embryos lacking the tj locus, whereas four tj mutants carrying point mutations still expressed protein, as detected with antibodies directed against the N-terminal region of TJ. The distribution and level of TJ expression, however, were abnormal. In tj^eo2, tj^z3434, and tj^z1949 mutants, the protein was distributed throughout the cell, and the amount seemed to be reduced. These tj alleles encode truncated proteins that lack the two DNA-binding domains, which harbour a putative bipartite nuclear localization signal, and the Leu zipper domain. As expected, these alleles behaved as genetic null mutations, with the exception of tj^z1949, which for unknown reasons only caused a partial loss-of-function phenotype. The protein encoded by the tj^z4735 allele (a genetic null), which has an incomplete Leu zipper localizes normally to the nucleus, but seemed to be more abundant, suggesting that the Leu zipper may modulate the nuclear concentration of TJ (Li, 2003).

To understand the developmental basis of the tj mutant phenotype, embryonic and larval tj^-/- gonads were analysed. Expression of tj in SGCs began at stage 12 of embryogenesis, when primordial germ cells (PGCs) first make contact with SGCs, and expression was maintained in germline-associated SGCs throughout fly development. The germline was not required to induce or maintain tj expression in SGCs, as tj transcript and protein were still expressed in embryonic and larval gonads of osk mutants lacking a germline. After gonad formation and coalescence, PGCs are inter-mingled and surrounded by SGCs in wild-type embryos. Gonads also formed in tj^-/- embryos and coalesced into a round organ composed of PGCs and SGCs, but they displayed defects in PGC number and the arrangement of SGCs and PGCs. Whereas in wild-type gonads the average number of PGCs increases from 11 cells at stage 15 to 14 cells at early stage 17, the average number of PGCs in tj mutant gonads remains at 11. The difference between wild-type and tj^-/- gonads at stage 17 was highly significant, suggesting that PGCs are normally incorporated into tj^-/- gonads, but that their number does not increase properly. As tj seems to be neither expressed nor required in the germline, the effect of tj on the proliferation or maintenance of PGCs is indirect and mediated through contact with tj-expressing SGCs. Gonad formation in tj mutants demonstrated that PGCs and SGCs recognize each other and are able to make contact. However, mutant SGCs did not mix with PGCs, instead their main cell body remained in the periphery of the gonad. tj mutant SGCs expressed several SGC-specific markers, including the TJ protein, suggesting that SGCs are properly specified. The fact that gonads formed, coalesced and incorporated PGCs in tj mutants corroborates this conclusion, as defects in SGC specification disrupt these processes. These findings suggest that TJ is required for the intermingling of somatic cells and germ cells in the embryo (Li, 2003).

The segregation of SGCs and germ cells is particularly evident in larval gonads. At the end of the larval period and before oogenesis begins, several distinct somatic cell populations develop in a wild-type ovary. PGCs, located in the centre, were densely mixed with so-called 'interstitial cells' that expressed TJ. In contrast, tj mutant interstitial cells were not mixed with PGCs and instead formed a separate layer of cells surrounding the germ cells. tj^-/- larval ovaries were also considerably smaller than wild-type ovaries and contained a reduced number of PGCs. In late third instar larval testes where spermatogenesis is already underway, TJ was expressed at high levels in the hub, in cyst progenitor cells (that together with the germline stem cells surround the hub) and in the early cyst cells that wrap around mitotically active germ cells. TJ was detected at low levels in cyst cells that ensheathed early primary spermatocytes and was not detectable afterwards. In tj^-/- larval testes, the association between cyst cells and germ cells was lost. The interior of the testis was filled with germ cells and the majority of cyst cells were located in the periphery. The hub was always present. Cyst cells did not seem to differentiate properly, as they maintained expression of TJ. All germ cells in tj^-/- testes were homogeneously small, expressed the vas^BC69 marker, and had dot- or dumbbell-shaped fusomes. All these traits are characteristic of early mitotically active germ cells, indicating a block in germ cell differentiation. These findings show that tj is essential for the normal association between SGCs and germ cells in developing ovaries and testes. Although tj^-/- SGCs seem to be able to make contact with germ cells they do not penetrate between germ cells to wrap them individually, as in wild-type gonads (Li, 2003).

Cell sorting is driven by differential adhesion between cells, which can originate from differences in the expression level of adhesion molecules. Analysis of tj mutant gonads revealed striking changes in the level of expression for several adhesion molecules. It was first noticed that in contrast to wild type, Fasciclin-III (Fas3) is not only expressed in the hub of tj^-/- testes, but is also seen in cyst cells. The adult ovary was used to further analyse the function of tj in the regulation of adhesion molecules, since te ovary facilitates studies at a single cell level. During oogenesis, tj was expressed in all somatic cell types that contact the germline. This comprises cap, inner sheath and follicle precursor cells in the germarium, and all follicle cells of developing follicles until stage 12 of oogenesis, including polar, border and centripetal cells. At late stage 10, tj expression was down-regulated in most follicle cells, and higher levels of expression were restricted to the centripetal and the most posterior follicle cells. SGCs that are not in contact with the germline (terminal filaments, interfollicular stalks and epithelial sheaths) did not express tj (Li, 2003).

tj^-/- follicle cell clones, induced by mitotic recombination, expressed unusually high levels of the adhesion molecules Fas3, E-cadherin (Ecad) and Neurotactin (Nrt). In wild-type follicles, Fas3 was detected in all follicle cells until stage 7, and from stage 8–10 only in polar cells. In contrast, tj^-/- follicle cells maintained Fas3 expression until at least stage 10. Ecad is normally seen in follicle and germline cells throughout oogenesis and mediates interactions between these two cell types. The abnormally high concentration of Ecad in response to the loss of tj function was particularly prominent in oocyte-associated follicle cells at stage 9–10, when wild-type follicle cells reduced their level of Ecad expression. An increase in Ecad was also detected at earlier stages, suggesting that Ecad expression was not only maintained at high levels, but also upregulated in tj^-/- follicle cells. The transcripts that encode Fas3 and Ecad were also overexpressed in tj^-/- follicle cells, suggesting that TJ function is needed for the transcriptional repression of these genes. Overexpression of TJ did not cause a detectable decrease in the amount of Ecad in follicle cells, which suggests either that TJ does not affect the expression of Ecad directly, or that the normal amount of TJ in follicle cells is already sufficient to down-regulate Ecad expression to a basic level independent of tj function. In addition, Nrt, a heterophilic adhesion molecule that is normally only expressed in somatic cells of the germarium, was heavily expressed in tj^-/- follicle cells of developing follicles. Ectopic expression of Nrt was detected only when tj^-/- clones were already induced in the germarium, but not when they were generated in developing follicles, suggesting that lack of tj function affects the expression of nrt only when this gene is in an active state. In contrast to Fas3, Nrt and Ecad, the adhesion molecule N-cadherin that was expressed in follicle cells during stages 1–10 revealed no change in expression in tj^-/- clones, indicating that TJ influences the expression of a specific subset of adhesion molecules in follicle cells (Li, 2003).

Polar follicle cells express higher levels of Ecad and Fas3 than other follicle cells. This, together with morphological similarities between polar cells and tj^-/- follicle cells raises the question of whether tj^-/- follicle cells are transformed into polar cells. The finding that the polar-cell-specific marker neur^A101 is not expressed in tj^-/- follicle cells suggests that there is no shift in cell fate. tj^-/- follicle cells demonstrated abnormalities in their morphology and behaviour. Instead of having a cuboidal or columnar shape, tj^-/- cells had a more rounded and irregular shape. At stage 9–10, they seemed smaller in size than expected, suggesting that they may not undergo a normal degree of endoreplication. Many tj^-/- cells moved out of the follicular epithelium, away from the germline. Occasionally, a mutant clone was sandwiched between follicular epithelium and oocyte at stage 10. If mutant follicle cells remained in the epithelium, they usually formed a tight, round cluster and their apical surfaces (which face the germline) were strongly constricted. These observations suggest that tj mutant follicle cells try to maximize their contact with each other, and minimize those with surrounding wild-type cells (Li, 2003).

Collectively, these data demonstrates that TJ is a key regulator of gonad morphogenesis. TJ may regulate different aspects of SGC differentiation, but its most apparent function is to allow somatic cells to invade and properly envelop germ cells. The failure of cell mixing in tj mutant gonads is reminiscent of classic cell sorting experiments, in which two different cell types that differ in their adhesive properties sort out when mixed, resulting in two-layered tissue aggregates. Interestingly, the loss of tj function resulted in increased expression of several adhesion molecules in SGCs that is controlled at the transcriptional level, and does not seem to result from a change in cell fate. This suggests that TJ functions either directly or indirectly as a modulator of adhesion molecules, such as Ecad and Fas3, that have been shown to mediate cell sorting in vitro and in vivo. Increased adhesion between somatic cells may interfere with their motility, their ability to acquire the appropriate shape, or their ability to form normal contacts with germ cells. Any of these effects could disturb the interaction between somatic cells and germ cells. In recent years, Ecad has emerged as having an important function in mediating germline–soma contact in Drosophila. Interestingly, Ecad is needed for the wrapping of primordial germ cells by SGCs in the embryonic gonad, and overexpression of Ecad in germ cells prevents SGCs from intermingling ^{(Jenkins, 2003). Whether an increased level of Ecad is solely responsible for the abnormal behaviour of tj mutant follicle cells was tested, but it was found that overexpression of Ecad in wild-type follicles caused a different type of abnormal cell behaviour. This suggests that the misregulation of several adhesion molecules is involved in the abnormal behaviour of tj mutant cells. Vertebrate large Maf factors have multiple functions in cellular differentiation (Black, 1997), including the regulation of cell behaviour and cell interactions (Cooke, 2001; Sadl, 2002). Since MafB and c-Maf are expressed in the soma of mouse gonads, the function of large Maf factors in gonad development and gametogenesis is possibly conserved between flies and mammals (Li, 2003).}

Specification and spatial arrangement of cells in the germline stem cell niche of the Drosophila ovary depend on the Maf transcription factor Traffic jam

Germline stem cells in the Drosophila ovary are maintained by a somatic niche. The niche is structurally and functionally complex and contains four cell types, the escort, cap, and terminal filament cells and the newly identified transition cell. The large Maf transcription factor Traffic jam (Tj) is essential for determining niche cell fates and architecture, enabling each niche in the ovary to support a normal complement of 2-3 germline stem cells. In particular, this study focused on the question of how cap cells form. Cap cells express Tj and are considered the key component of a mature germline stem cell niche. It is concluded that Tj controls the specification of cap cells, as the complete loss of Tj function caused the development of additional terminal filament cells at the expense of cap cells, and terminal filament cells developed cap cell characteristics when induced to express Tj. Further, it is proposed that Tj controls the morphogenetic behavior of cap cells as they adopted the shape and spatial organization of terminal filament cells but otherwise appeared to retain their fate when Tj expression was only partially reduced. The data indicate that Tj contributes to the establishment of germline stem cells by promoting the cap cell fate, and controls the stem cell-carrying capacity of the niche by regulating niche architecture. Analysis of the interactions between Tj and the Notch (N) pathway indicates that Tj and N have distinct functions in the cap cell specification program. It is proposed that formation of cap cells depends on the combined activities of Tj and the N pathway, with Tj promoting the cap cell fate by blocking the terminal filament cell fate, and N supporting cap cells by preventing the escort cell fate and/or controlling the number of cap cell precursors (Panchal, 2017).

Stem cells retain the capacity for development in differentiated organisms, which is important for tissue growth, homeostasis and regeneration, and for long-term reproductive capability. Stem cells are often associated with a specialized microenvironment, a niche that is essential for the formation, maintenance, and self-renewal of stem cells by preventing cell differentiation and controlling rate and mode of cell division. The niche for the germline stem cells (GSCs) in Drosophila serves as an important model for the analysis of interactions between niche and stem cells. The astounding fecundity of Drosophila females that can lay dozens of eggs per day over several weeks depends on approximately 100 GSCs that are sustained by 40 stem cell niches. To understand the formation and maintenance of these GSCs, it is important to understand how stem cell niches form and how they function (Panchal, 2017).

The GSC niche of the Drosophila ovary consists of three somatic cell types: cap cells, escort cells, and terminal filament (TF) cells. GSCs are anchored to cap cells by DE-cadherin-mediated adhesion and require close proximity to cap cells to retain stem cell character. Cap cells secrete the BMP homolog Decapentaplegic (Dpp), activating the TGFβ signaling pathway in adjacent GSCs, which leads to the repression of the germline differentiation factor Bag-of-Marbles (Bam). Through Hedgehog (Hh) signaling, cap cells also appear to stimulate escort cells to secrete Dpp. The combined pool of Dpp from cap and escort cells, together with mechanisms that concentrate Dpp in the extracellular space around GSCs, promotes the maintenance of 2-3 GSCs, whereas the adjacent GSC daughter cells that have lost the contact to cap cells will enter differentiation as cystoblasts. In contrast, TFs are not in direct contact with GSCs but serve important functions in the development and probably also in the maintenance and function of GSC niches (Panchal, 2017).

Formation of GSC niches begins with the progressive assembly of TFs by cell intercalation during the 3rd larval instar. The process of TF cell specification is not understood but might start in 2nd instar when the first TF precursor cells appear to leave the cell cycle. TF morphogenesis depends on the Bric à brac transcriptional regulators that control the differentiation of TF cells and their ability to form cell stacks, and involves the Ecdysone Receptor (EcR), Engrailed, Cofilin, and Ran-binding protein M (RanBPM). The number of TFs that form at the larval stage determine the number of GSC niches at the adult stage, and are regulated by several signaling pathways that control cell division and timing of cell differentiation in the larval ovary, including the EcR, Hippo and Jak/Stat, Insulin and Activin pathways. Despite the recent advance in elucidating mechanisms that control the number of GSC niches and the temporal window in which they form, relatively little is known about the origin and specification of the somatic cell types of the GSC niche (Panchal, 2017).

Notably, the origin and specification of cap cells, the main component of an active GSC niche is little understood. Cap cells (also called germarial tip cells) are first seen at the base of completed TFs at the transition from the 3rd larval instar to prepupal stage. They appear to derive from the interstitial cells (also called intermingled cells) of the larval ovary that are maintained by Hh signaling from TFs. The formation of cap cells is accompanied by the establishment of GSCs. The N pathway contributes to the development of cap cells. A strongly increased number of functionally active cap cells per niche form in response to overexpression of the N ligand Delta (Dl) in germline or somatic cells, or the constitutive activation of N in somatic gonadal cells. The ability of N to induce additional cap cells seems to depend on EcR signaling. Loss of Dl or N in the germline had no effect on cap cells. However, loss of N in cap cell progenitors or Dl in TF cells caused a decrease in the number of cap cells. A current model suggests that Dl signaling from basal-most TF cells to adjacent somatic cells together with Dl signaling between cap cells allows for a full complement of cap cells to form. Furthermore, N protects cap cells from age-dependent loss as long as its activity is maintained by the Insulin receptor. The Jak/Stat pathway, which operates downstream or in parallel to the N pathway in the niche, is not required for cap cell formation. As cap cells were reduced in number but never completely missing when the N pathway components were compromised, the question remains whether N signaling is the only factor that is important for cap cell formation. Furthermore, no factor that operates downstream of N has been identified that is crucial for cap cell formation (Panchal, 2017).

This study finds that Traffic jam (Tj) is both required for cap cell specification and for the morphogenetic behavior of cap cells, enabling them to form a properly organized niche that can accommodate 2-3 GSCs. Tj is a large Maf transcription factor that belongs to the bZip protein family. Its four mammalian homologs control differentiation of several cell types and are associated with various forms of cancer. Tj is essential for normal ovary and testis development, and is only expressed in somatic cells of the gonad. Interestingly, Tj is present in cap cells and escort cells but not in TFs. This study shows that Tj is essential for the formation of the GSC niche. First, Tj regulates the behavior of cap cells, enabling them to form a cell cluster instead of a cell stack, which appears to be important for the formation of a normal-sized GSC niche with the capacity to support more than one GSC. Second, cap cells adopt the fate of TF cells in the absence of Tj function, and TF cells develop cap cell-like features when forced to express Tj, indicating that Tj specifies the cap cell fate. Genetic interactions suggest that Tj and N are required together for cap cell formation, but have different functions in this process. For somatic gonadal cells to adopt the cap cell fate, it is proposed that Tj has to be present to inhibit the TF cell fate and N has to be present to prevent the escort cell fate and/or produce the correct number of cap cell precursors (Panchal, 2017).

Loss of Tj has a profound negative effect on the establishment, number, and maintenance of GSCs. Effects of Tj on the germline were previously shown to be indirect as Tj is neither expressed nor cell-autonomously required in the germline. Therefore, it is proposed that the dramatic change in the structure of the somatic niche affects GSCs when Tj function is compromised. An inverse causal relationship, where a reduced number of GSCs would trigger the somatic niche defects was ruled out by showing that cap cells can still look and behave normally in the absence of any germ cells. It is concluded that Tj controls GSCs indirectly by controlling somatic cell fate and cell arrangement in the stem cell niche (Panchal, 2017).

By controlling the morphology and behavior of the cap cells, Tj regulates the GSC-carrying capacity of the niche. When Tj expression was moderately reduced, the number of GSCs per niche was reduced, with the remaining GSC properly maintained over several weeks. The decrease of GSCs per niche correlated with a decrease of cap cells in the germarium. Two cap cells were on average required to sustain one GSC, similar to what has been proposed for a wild-type ovary. The data indicate that the reduced niche capacity is due to a reduction in the available contact surface between cap cells and GSCs. Tj-depleted cap cells that convert from forming a cluster inside the germarium to forming a stalk outside the germarium minimize their availability for GSC attachment. A connection between the GSC-cap cell contact area and niche capacity is similarly reflected in the increased number of GSCs that accompanies an increase in cap cell size due to loss of RanBPM. This study shows that the spatial arrangement of the cap cells has a crucial impact on the number of stem cells per niche (Panchal, 2017).

When Tj function was completely abolished, the number of GSCs was drastically reduced, as expected in the absence of cap cells. The very few pMad-positive GSC-like cells in tj mutant prepupal ovaries were always associated with a TF, suggesting that TFs might temporarily provide enough Dpp to activate Mad in a few germline cells, consistent with the finding that Dpp is expressed in TFs at the late larval stage\. This is not sufficient, however, to maintain GSCs and adult ovaries rarely contain pMad-positive germline cells. This is in agreement with the finding that Dpp is not detected in adult TFs, and corroborates that cap cells are required for GSC maintenance. In addition, the rapid loss of the entire germ cell pool in Tj-depleted ovaries during the pupal stage might be precipitated by loss or defects in escort cells. Escort cell precursors are not properly intermingled with germ cells at the larval stage and differentiated escort cells appear to be missing in adult ovaries that lack Tj. As escort cells are crucial for germ cell differentiation, the defect in escort cell differentiation could be responsible for the demise of the germline in tj mutants (Panchal, 2017).

GSCs have broad cellular protrusions, which they use to reach and tightly ensheath the accessible surface of cap cells. In wild type, relatively short protrusions are sufficient to make extensive contact with more than one cap cell. However, when cap cells formed a stalk, GSCs were often observed to produce unusually long extensions that allowed them not only to contact the immediate cap cell neighbor but also a more distantly located cap cell. This suggests that GSCs respond to a chemotactic signal from cap cells and send protrusions toward this signal. It remains to be investigated whether this is a response to Dpp signaling or signaling through another pathway. The importance of cellular protrusions in signaling events in the stem cell niche has recently come to light with the discovery of nanotubes that mediate Dpp signaling between GSCs and hub cells in the Drosophila testes, and cytonemes that contribute to Hh signaling from cap to escort cells in the ovary (Panchal, 2017).

This analysis shows that Tj is required for the specification of cap cells. In the absence of Tj function, additional TF cells form at the expense of cap cells, resulting in unusually long TFs while the cap cell fate is not established. Whereas the formation of cap cell precursors appears not to require Tj, this transcription factor is essential for the ability of these precursors to take on the cap cell fate and to prevent the TF cell fate that is otherwise adopted as a default state. The following findings support this conclusion: (1) In the absence of Tj function, cap cells were missing while additional cells that displayed TF cell-characteristic morphology, behavior and marker expression were integrated into the TF. The number of additional TF cells was comparable to the normal number of cap cells. (2) Prospective cap cells cell-autonomously adopted a TF-specific morphology and behavior in the absence of functional Tj. (3) A hypomorphic tj mutant provided direct evidence for the incorporation of cap cells into TFs, forming the basal portion of these stalks. (4) Ectopic expression of Tj in TF cells caused a change toward cap cell-typical marker expression and morphology. Together, these data demonstrate that Tj promotes cap cell specification (Panchal, 2017).

The expression pattern of Tj supports the notion that Tj has a function in cap cells but not in TF cells. Tj is continuously expressed in cap cells. Tj is also present in the anterior interstitial cells of the larval ovary, which are thought to develop into cap cells. In contrast, Tj is neither detected in the cell population that gives rise to TFs during 3rd larval instar, nor in differentiated TFs. Interestingly, even in the absence of Tj function, the tj gene remains differentially expressed in the anterior niche, being inactive in regular TF cells but active in the additional TF cells, which form the apical and basal portion of a TF, respectively. This differential expression of Tj indicates that a regionally or temporally regulated mechanism operates upstream of Tj that initiates differences in anterior niche cells. Although it is conspicuous that Tj expression from 3rd instar onwards is restricted to cells that are in direct contact with germline cells, which includes cap cells but excludes TF cells, it has previously been shown that Tj expression is not dependent on the germline. This suggests that a soma-specific mechanism is responsible for the differential expression of Tj in anterior niche cells. Interestingly, a recent study uncovered the importance of Hh signaling from TFs to neighboring interstitial cells in the larval ovary and proposes that tj is a direct target of the Hh signaling pathway (Panchal, 2017).

The current findings suggest the presence of a new cell type in the GSC niche that has been named 'transition cell' as it is located between the cap cell cluster and the TF, connecting these two structures of the niche. Notably, the one or occasionally two transition cells have the morphology of TF cells and align with neighboring TF cells despite displaying a cap cell-like marker profile that includes the expression of Tj—although Tj expression is substantially lower than in cap cells. Interestingly, cap cells from ovaries with reduced Tj expression (tj^hypo) similarly displayed a TF cell-like morphology and behavior while their expression profile remained cap cell-like. A similar, although weaker effect was noted in a tj hemizygous condition, suggesting that Tj function is haplo-insufficient in cap cells. Thus, when Tj levels are reduced, cap cells adopt very similar molecular and morphogenetic properties as the transition cell in a wild-type niche, and might have adopted this cell fate (Panchal, 2017).

Together, the current findings indicate that Tj has an important role in the establishment of three cell types in the GSC niche: TF cells, transition cells, and cap cells. As lack of Tj function seems to cause a transformation of cap and transition cells into TF cells, and a mild reduction of Tj a cap to transition cell transformation, it is proposed that different Tj expression levels establish different cell fates and morphogenetic traits. It is proposed that a high concentration of Tj leads to the formation of cap cells and a lower concentration to the formation of the transition cell, whereas absence of Tj is required for the formation of TF cells. This model implies that different levels of Tj have different effects on target genes. It is predicted that Tj has at least one target gene that only responds to high levels of Tj and that specifically controls the morphogenetic behavior of cap cells, allowing them to adopt a round morphology and organize into a cell cluster. Whether this relates to an effect of Tj on the expression of adhesion molecules as observed in other gonadal tissues awaits further analysis (Panchal, 2017).

This study identifies Tj as essential for cap cell formation. In addition, this process depends on the N pathway. Therefore, it was asked how the functions of Tj and N in cap cell formation relate to each other. A comparison between the loss and gain-of-function phenotypes suggests that Tj and N have different functions in the establishment of cap cells. In the absence of Tj function, cap cell precursor cells are present but take on the fate of TF cells, whereas depletion of N leads to a loss of cap cells but does not cause the formation of additional TF cells. Ectopic activation of N can induce a strong increase in the number of cap cells, whereas overexpression of Tj did not appear to affect the number of cap cells. Therefore, both factors are important for cap cell formation but contribute differently to this process. The questions then are: What is the respective contribution of Tj and N to cap cell formation, and how are their functions related (Panchal, 2017)?

The function of N in cap cell formation is still not fully understood. The observation that depletion of N reduces the number of cap cells confirms previous findings. However, neither in this nor any previously published experiments were cap cells lost completely when the N pathway was compromised, and it remains therefore unclear whether N is de facto essential for cap cell formation or primarily functions in regulating the size of the cap cell pool. Interestingly, evidence amounts to a function of the N pathway in a decision between the cap cell and escort cell fate: First, Dl signal from TF cells activates the N pathway in adjacent interstitial cells, inducing them as cap cells, whereas the remaining interstitial cells are thought to develop into escort cells. Second, escort cells expressing activated N can develop into cap cells. Third, when tj-Gal4 was used to express active N in interstitial cells, the number of cap cells dramatically increased while the escort cell region became smaller, and some germaria seemed to lack escort cells all together. These germaria also lacked germline cells, although a larger pool of cap cells was expected to increase the number of GSCs. However, the absence of germline cells is consistent with an absence of escort cells, as escort cells have been shown to be important for maintaining the germline. Together, these observations support the hypothesis that N is involved in a cap cell versus escort cell fate decision, and suggest that the N pathway might promote the formation of cap cells by inhibiting the escort cell fate (Panchal, 2017).

To determine how the functions of Tj and N depend on each other, genetic interactions were examined. The N pathway seems to be still functional in tj mutants. First, the expression of N and Dl appeared unaffected and E(spl) was activated in the additional TF cells (= transformed cap cells) similarly to normal cap cells. Second, the formation of additional TF cells in the absence of Tj depended on the presence of N, as only very few additional TF cells formed in a N compromised background. These findings indicate that the N pathway is still active in cap cell precursors when Tj function is abolished. This together with the observation that constitutively active N cannot suppress the tj mutant phenotype suggests that Tj does not act upstream of N in regulating cap cell fate (Panchal, 2017).

Therefore, it was asked whether Tj might operate downstream of N. Loss of Tj was not detected upon N depletion, and this together with the finding that Tj is expressed in all interstitial cells, and not only in those that receive Dl signaling argues against a requirement of N signaling for tj expression. If at all, one would expect tj to be negatively regulated by N as cap cells express a lower level of Tj than escort cells. The maintenance of somatic cell types in N mutant ovaries that are lost in tj mutant ovaries, including the escort cells is also not consistent with a linear relationship. Nevertheless, the ability of Tj to promote the formation of cap cells appears to depend on the activity of the N pathway in cap cell precursors. Again, this is suggested by the finding that when N and Tj were both compromised, the number of additional TF cells were much smaller than when N was fully active. Therefore, it is proposed that N activity sets aside a pool of percursor cells that in the presence of Tj take on the cap cell fate, and in its absence the TF fate (Panchal, 2017).

Similar to the ovary, N is important for the formation of the GSC niche (= hub) in the Drosophila testis. Interestingly, N contributes to hub cell specification by downregulating the expression level of Tj. Not only is the hub still present in tj mutant testes but additionally, ectopic hub cells form in the absence of Tj. Thus, Tj seems to have opposing functions in testes and ovaries, suppressing the niche cell fate in the testis, while promoting it in the ovary (Panchal, 2017).

The interplay between Tj and N seems not restricted to the cap cell fate in the ovary. Whereas neither factor alone is required for TF cell formation, as TF cells formed normally in the absence of either Tj or N, the combined loss of Tj and N led to a strong reduction in the number of TFs and number of TF cells within stalks. This suggests that their combined action is already required at an earlier stage of ovary development, when Tj is still expressed in all somatic cells of the ovary. Moreover, Tj knockdown combined with expression of activated N caused TF cells to be the only cell type remaining of the ovary, indicating that several cell types in the ovary require proper input from both factors. Taken together, the findings support a model, in which both Tj and N operate together to promote the cap cell fate but have separate functions. It is proposed that Tj and N promote the cap cell fate by blocking the TF cell fate and escort cell fate, respectively, and that the combined actions of Tj and the N pathway are required to establish the cap cell fate (Panchal, 2017).

Dmaf, a novel member of Maf transcription factor family is expressed in somatic gonadal cells during embryonic development and gametogenesis in Drosophila

Members of the maf gene family encode basic/leucine zipper transcription factors and play important roles during cell differentiation and organogenesis in vertebrate development. The maf family is evolutionarily conserved and the Drosophila maf (Dmaf) gene is expressed in somatic gonadal cells. During embryonic development, Dmaf mRNA is detected in somatic gonadal precursor cells emerging from dorsolateral mesoderm. Relatively weak expression is also observed in subset of neuronal cells in the central nervous system. In adult flies, Dmaf is expressed in somatic gonadal cells surrounding developing oocytes and spermatocytes. These results suggest a specific function for Dmaf in gonadal development, including migration and differentiation of primordial germ cells (Kawashima, 2003).

A regulatory circuit for piwi by the large Maf gene traffic jam in Drosophila

PIWI-interacting RNAs (piRNAs) silence retrotransposons in Drosophila germ lines by associating with the PIWI proteins Argonaute 3 (AGO3), Aubergine (Aub) and Piwi. piRNAs in Drosophila are produced from intergenic repetitive genes and piRNA clusters by two systems: the primary processing pathway and the amplification loop. The amplification loop occurs in a Dicer-independent, PIWI-Slicer-dependent manner. However, primary piRNA processing remains elusive. This study analysed piRNA processing in a Drosophila ovarian somatic cell line where Piwi, but not Aub or AGO3, is expressed; thus, only the primary piRNAs exist. In addition to flamenco, a Piwi-specific piRNA cluster (Brennecke, 2007), traffic jam (tj), a large Maf gene, was determined as a new piRNA cluster. piRNAs arising from tj correspond to the untranslated regions of tj messenger RNA and are sense-oriented. piRNA loading on to Piwi may occur in the cytoplasm. zucchini, a gene encoding a putative cytoplasmic nuclease, is required for tj-derived piRNA production. In tj and piwi mutant ovaries, somatic cells fail to intermingle with germ cells and Fasciclin III is overexpressed. Loss of tj abolishes Piwi expression in gonadal somatic cells. Thus, in gonadal somatic cells, tj gives rise simultaneously to two different molecules: the TJ protein, which transcriptionally activates Piwi expression, and piRNAs, which define the Piwi targets for silencing (Saito, 2009).

Genetic studies have shown that piwi and aub are essential in germline stem-cell self-renewal and pole-cell formation, respectively. Mutations introduced into piwi and aub cause de-repression of retrotransposons and a loss of piRNA accumulation in ovaries B8">8. A recent study has revealed that strong loss-of-function mutations in AGO3 also increase expression of selfish genetic elements in germ lines B5">5. Thus, the PIWI proteins, with their associated piRNAs, function in retrotransposon silencing. piRNA production in Drosophila ovaries occurs in a Dicer-independent manner B8">8. A model for piRNA biogenesis—the piRNA amplification loop B3">3, B4">4—was proposed as a result of deep-sequencing and bioinformatic analyses of Drosophila piRNAs. In this model, Aub/Piwi and AGO3 reciprocally guide the 5' end formation of piRNAs (Saito, 2009).

Classification of piRNAs according to their origins indicated that piRNAs derived from a particular piRNA cluster locus—flamenco (flam)—on the X chromosome are exclusively loaded on to Piwi (Aravin, 2007; Siomi, 2009), indicating that those piRNAs are produced by a pathway independent of the amplification loop. This pathway is called the primary processing pathway. Recently, two independent groups deduced the existence of the primary piRNA processing pathway from extensive bioinformatic analyses of piRNAs in a broad range of piRNA mutants; these studies reconfirmed that primary piRNAs derived from flam are most likely loaded directly on to Piwi and not further amplified (Li, 2009; Malone, 2009). However, a molecular mechanism of the primary processing pathway remains elusive (Saito, 2009).

A stable cell line of ovarian somatic cells (OSCs) was establised from the parental cell line fGS/OSS, comprising germline stem cells and sheets of somatic cells (OSS). fGS/OSS culture was shown to be Vasa-positive, whereas the OSCs are Vasa-negative. OSCs express Fasciclin III (FasIII) and undergo rounds of passage in culture for several months. All these data support the idea that OSCs contain only mitotically active early follicle (somatic) cells (Saito, 2009).

Piwi is expressed in somatic gonadal cells, whereas Aub and AGO3 are not expressed in this cell type. Western blot analysis revealed that Piwi, but not Aub and AGO3, was detectable in OSCs. As in ovaries, Piwi in OSCs was localized in the nucleus. The absence of AGO3 expression in OSCs was further confirmed by polymerase chain reaction with RT-PCR (Saito, 2009).

Piwi in OSCs was detected in a form bound to small RNAs of 24-30 nucleotides. The size distribution of these small RNAs was very similar to that of gonadal Piwi-associating piRNAs (Piwi piRNAs). In addition, they showed resistance to periodate oxidation andβ-elimination treatments, which are hallmarks of 2'-O-methyl modification at the 3' end, as in ovarian piRNAs. Thus, the small RNAs associated with Piwi in OSCs could be categorized as piRNAs and, importantly, these piRNAs are produced in an Aub/AGO3-independent manner (Saito, 2009).

Whether the nuclear localization of Piwi is required for piRNA production and for Piwi loading in OSCs was examined. A mutant of Piwi (Piwi-δ>N) in which its putative nuclear localization signals were deleted, resulting in cytoplasmic localization of the mutant, was loaded with piRNAs in a similar manner to wild-type Piwi. It was also observed that a Slicer-deficient Piwi mutant [Piwi-DDAA, where two aspartic acids (D614 and D685) in the PIWI domain, which are required for Slicer activity, are altered to alanines] was loaded with piRNAs, similarly to wild-type Piwi. Depletion of endogenous Piwi from OSCs did not affect piRNA loading on to a double mutant of Piwi (Piwi-δN-DDAA). Piwi does not seem to homodimerize in vivo. These results support a model in which the primary piRNA processing and the piRNA loading on Piwi may occur in the cytoplasm in a Piwi-Slicer-independent manner (Saito, 2009).

Piwi-associating piRNAs in OSCs (OSC piRNAs) are mainly derived from the antisense strand of retrotransposons, similar to the derivation of ovarian Piwi piRNAs. Examination of their nucleotide bias indicated that OSC piRNAs mostly have a bias for U as the first nucleotide in the sequence (1st-U), but no other prominent bias was observed throughout their entire sequence. Exclusion of piRNAs with 1st-U from the piRNA pool did not uncover any obvious bias, including 10th-A. piRNA pairings through the ten nucleotides from the 5' ends were negligible. Thus, Piwi does not self-amplify piRNAs. All these observations correlate well with the data obtained from a recent deep-sequencing study that was performed using a Drosophila ovarian somatic sheet (OSS) (Saito, 2009).

The origins of OSC piRNAs were examined. Unique mapping of OSC piRNAs on the Drosophila genome revealed that flam is the main source (1,365 perfectly matched and 186 one-base mismatched piRNAs), as it is for ovarian Piwi piRNAs and OSS piRNAs. In addition to flam, another locus was found on chromosome 2L that also produces piRNAs uniquely mapped to the particular region (322 perfectly matched and 29 one-base mismatched piRNAs). This locus corresponds to the 3' UTR of a protein-coding, single-exon gene, traffic jam (Saito, 2009).

TJ is a soma-specific large Maf factor necessary for controlling gonad morphogenesis in Drosophila. TJ is the only Drosophila orthologue of the transcriptional factors c-Maf and MafB/Kleisler in vertebrates. In tj mutant gonads, somatic cells fail to intermingle and properly develop germ cells. This eventually causes an early blockage in germ-cell differentiation and no follicle cells are detected in adult ovaries of tj mutants. All of the OSC piRNAs derived from tj were sense-oriented, indicating that the tj transcript may serve not only as the template for TJ synthesis but also as the precursor of the piRNAs. The tj transcript in OSCs was not dedicated to piRNA production, because the TJ protein was strongly expressed in OSCs. The existence of piRNAs derived from the 3' UTR of tj was further confirmed by northern blot analysis (Saito, 2009).

An assessment was made as to whether the transcriptional unit of tj is first divided into two parts, each with an individual function - one for TJ synthesis and the other for piRNA production - or if one full-length tj transcriptional unit contains both functions. Northern blot analysis, using two probes corresponding to either the open reading frame (ORF) or the 3' UTR of tj, visualized a single discrete band of the same length, indicating that the latter scenario might be the case (Saito, 2009).

piRNAs derived from 3' UTRs (and from slightly extended regions) of protein-coding genes other than tj were also found. For example, the 3' UTRs of brat B21">21 and Klp10A B22">22 also produce piRNAs. Interestingly, they are all derived from the sense strand; thus the parental genes are apparently not the targets for gene silencing by Piwi. The parental genes are also not repetitive. By contrast, piRNAs derived from retrotransposons or flam in OSCs are mainly antisense-oriented and are thought to arise from much longer, repetitive precursors (Saito, 2009).

Mutations in zucchini (zuc), a gene encoding a putative nuclease, cause female sterility and a reduction of roo- and flam-derived piRNAs in ovaries. However, whether or not zuc is expressed in gonadal somatic cells and, if it is expressed, where Zuc accumulates was unknown. It was observed that OSCs express zuc and that Zuc is predominantly localized in the cytoplasm of OSCs, particularly in the perinuclear region. Attempts were made to determine whether zuc is necessary for tj piRNA production in OSCs. Depletion of zuc by RNA interference (RNAi) significantly reduced the expression level of tj piRNAs, but not of microRNAs, suggesting that zuc is involved in the tj piRNA production pathway. Other data also indicate that the read number of tj piRNAs in zuc mutants was much lower compared with those in other mutants (Saito, 2009).

It was previously reported that tj mutant somatic cells show a failure to intermingle with germ cells in third instar larval ovaries. Notably, it was noticed in this study that the piwi mutants piwi² and piwi³ showed a similar phenotype: somatic cells in the ovaries of third instar larvae of piwi mutants adhere to each other and exclude Vasa-positive germ cells. aub mutants did not phenocopy the piwi mutants. piwi mutants express the tj transcript and the TJ protein at approximately wild-type levels. Thus it is unlikely that tj piRNAs target the parental tj gene (Saito, 2009).

Next the expression of Piwi was examined in larval ovaries of tj mutants. Vasa and TJ are known to be expressed in germline stem cells (and in their developing cells) and somatic cells, respectively, whereas Piwi is known to be expressed in both cell types. In wild-type larval ovaries, Vasa- and TJ-positive cells are mutually exclusive, but Piwi is expressed in both cell populations. By contrast, in tj mutants, Piwi expression was restricted to Vasa-positive cells. These results indicate that TJ is the activator of piwi expression in gonadal somatic cells (Saito, 2009).

Expression of Piwi in tj mutant testes was also examined. As in ovaries, tj mutations in testes caused a lack of Piwi expression in gonadal somatic cells, including hub, cyst progenitor cells and early cyst cells. In contrast, Piwi signals were clearly seen in these cell types of wild-type testes. Interestingly, without functional TJ, germline stem cells and their developing cells in testes, which normally show a faint signal for Piwi, highly express Piwi. It seems that TJ in testes negatively controls piwi expression and indirectly controls expression in germline stem cells and their developing cells (Saito, 2009).

This study has uncovered two functions of tj in the regulation of piwi's functions. In gonadal somatic cells, TJ supposedly controls transcription of various genes. This study indicates that piwi is highly likely to be one of the genes under strong TJ control because loss of tj in gonadal somatic cells abolished Piwi expression. Further support for the hypothesis that TJ controls piwi expression was provided by DNA sequences near the putative transcriptional start site of the piwi gene, which show a weak but significant similarity to the Maf binding consensus sequence, and which were bound with TJ in OSCs. Thus, the first function of tj is to activate the expression of Piwi in gonadal somatic cells. The second function is to supply piRNAs for Piwi. Without the supplement of tj piRNAs, Piwi would lose the activity to target genes that should be silenced by Piwi and the tj-piRNA complex. A likely target of such silencing is FasIII because FasIII, a cell adhesion molecule concentrated at the hub cell junction, is ectopically overexpressed in other somatic cells in tj larval testes. Indeed, the FasIII expression level was higher in piwi mutant testes than in control testes. Some of the tj piRNAs identified in this study showed strong complementarity to the FasIII primary transcript. Although all the tj-derived piRNAs are sense-oriented and thus unlikely to target the tj mRNA, given the nuclear localization of Piwi, it is conceivable that the Piwi-piRNA complex could associate with the tj gene (Saito, 2009).

These findings suggest a novel regulatory circuit where tj mRNA simultaneously produces two types of functional molecules: TJ protein, which activates expression of Piwi, and piRNAs, which are loaded on to Piwi to silence specific target genes, such as FasIII and other, as yet undiscovered, genes (Saito, 2009).

A broadly conserved pathway generates 3'UTR-directed primary piRNAs

Piwi-interacting RNAs (piRNAs) are ~24-30 nucleotide regulatory RNAs that are abundant in animal gonads and early embryos. The best-characterized piRNAs mediate a conserved pathway that restricts transposable elements, and these frequently engage a 'ping-pong' amplification loop. Certain stages of mammalian testis also accumulate abundant piRNAs of unknown function, which derive from noncoding RNAs that are depleted in transposable element (TE) content and do not engage in ping-pong. The 3' untranslated regions (3'UTRs) of an extensive set of messenger RNAs (mRNAs) are processed into piRNAs in Drosophila ovaries, murine testes, and Xenopus eggs. Analysis of different mutants and Piwi-class immunoprecipitates indicates that their biogenesis depends on primary piRNA components, but not most ping-pong components. Several observations suggest that mRNAs are actively selected for piRNA production for regulatory purposes. First, genic piRNAs do not accumulate in proportion to the level of their host transcripts, and many highly expressed transcripts lack piRNAs. Second, piRNA-producing mRNAs in Drosophila and mouse are enriched for specific gene ontology categories distinct from those of simply abundant transcripts. Third, the protein output of traffic jam, whose 3'UTR generates abundant piRNAs, is increased in piwi mutant follicle clones. This study reveals a conserved primary piRNA pathway that selects and metabolizes the 3'UTRs of a broad set of cellular transcripts, probably for regulatory purposes. These findings strongly increase the breadth of Argonaute-mediated small RNA systems in metazoans (Robine, 2009).

Our studies highlight the prevalence and conservation of mRNA substrates for the primary piRNA pathway and provide insight into primary piRNA biogenesis. Although clear orthologs between mouse and Drosophila Piwi proteins are not established, germline-associated cells in both systems engage a primary piRNA biogenesis pathway that preferentially generates abundant piRNAs from 3'UTRs. This class of piRNAs is especially abundant in Drosophila follicle cells (via Piwi) and in prepachytene-stage (via Mili) and pachytene-stage (via both Miwi and Mili) mammalian testes; a similar Aub-dependent pathway was identified that probably operates in the Drosophila germline. While this work was under review, Siomi (2009) reported that a piwi- and zucchini-dependent primary piRNA pathway operates on the 3'UTRs of some Drosophila mRNAs. The conclusions on biogenesis are generally consistent. However, because the current sequence data were ~1000 fold deeper, it was demonstrated that the primary piRNA pathway is not selective for a few mRNAs but instead acts broadly across > 1000 cellular transcripts and operates in both flies and vertebrates (Robine, 2009).

The mechanism of primary piRNA biogenesis remains largely mysterious. It is believed that mRNA/3'UTR-directed piRNAs derive from a primary processing pathway, based on their 5' U bias, their extraordinary abundance in Drosophila OSS cells (which lack ping-pong), and their genetic independence from ping-pong components in the animal. Previous studies noted that some 3'UTRs harbored TE sequences, suggesting a potential relationship to TE-piRNA production. However, no overall enrichment for TEs was observed in the 3'UTRs of piRNA-generating mRNAs. Indeed, many Drosophila and mammalian 3'UTRs with highest piRNA production were strongly depleted for repeat elements relative to their introns, and most mRNA-derived piRNAs mapped uniquely. Moreover, the clearly distinct patterns of siRNA and piRNA biogenesis within cis-NAT-3'UTRs (endo-siRNAs from convergently transcribed protein-coding genes), along with the fact that most abundant piRNA-generating transcripts are not arranged in cis-NATs, suggest that the endo-siRNA and primary piRNA pathways act independently on cellular transcripts (Robine, 2009).

The production of piRNAs from spliced transcripts and preferentially from 3'UTRs suggests that they may be processed from cytoplasmic transcripts engaged with ribosomes. This scenario is consistent with the cosedimentation of Miwi and Mili with polysome fractions. This does not exclude the possibility of 3'UTR-specific isoforms that generate piRNAs, but no such transcripts were detected from loci with highly abundant 3'UTR piRNAs. Instead, it is speculated that the abrupt reduction of piRNA production from adjacent coding exons may reflect competition between the translation machinery and primary piRNA biogenesis machinery. Although the status of most putative noncoding loci has not generally been examined experimentally, it was observed that many of them exhibit distributive patterns across their exons. In contrast, some protein-coding genes also generate substantial CDS-piRNAs (e.g., piwi). A careful comparison of mRNA partitioning between free and translating pools with their capacity for piRNA generation would be an informative test of the competition model (Robine, 2009).

The Tudor-domain containing gene TDRD-1 interacts with Miwi and Mili and was suggested to regulate the formation of intergenic piRNAs versus genic piRNAs. This study confirmed that genic piRNAs derived predominantly from 3'UTR piRNAs were elevated in published 15 days postpartum (dpp) TDRD-1 KO library comparable to 10 dpp wild-type libraries. However, no similarly high proportion of 3'UTR piRNAs was observed in other published data from TDRD-1 null 18.5 day testis. It is noted that TDRD-1 mutants exhibit apoptotic postpachytene spermatocytes at 15 dpp. This probably reduces the contribution of postpachytene intergenic piRNAs, which may concomitantly increase the proportion of 3'UTR piRNAs in total small RNA libraries. This developmental defect provides an alternate basis besides a direct role for TDRD-1 in piRNA substrate selection (Robine, 2009).

The existence of abundant 3'UTR-derived piRNAs raises their potential relationship with pyknons (variable length sequences with a statistically significant number of intact copies in the intergenic and intronic regions of the genome and additional copies in the untranslated or amino acid coding regions of known transcripts). Genic instances of pyknons are concentrated in 3'UTRs, and these transcripts are enriched for some GO terms that are shared by genes with abundant 3'UTR piRNAs (such as transcription and nucleic acid metabolism). However, 3'UTRs broadly produce piRNAs from across their lengths with little specificity for individual sequences, and pyknons do not exhibit preference for 5' uridine as do 3'UTR piRNAs. In addition, although pyknons are defined as having multiple (>40) intergenic and genic instances, the vast majority of genic piRNAs map uniquely. These properties suggest that the primary piRNA pathway metabolizes 3'UTRs independently of pyknons (Robine, 2009).

A challenge for the future is to understand the biological consequences of the 3'UTR-directed primary piRNA pathway. Given that piRNA-generating transcripts were not consistently altered in mili-KO mutants, it is prudent to consider that these piRNAs might be incidental, perhaps by-products of poised TE defense operation. In contrast, it seems difficult to imagine that they entirely lack regulatory consequences given their sheer abundance: 3'UTR piRNAs comprise nearly 10% of small RNAs in Drosophila OSS cells and 35% of small RNAs in 10 dpp mouse testis (Robine, 2009).

Given the breadth of mRNA substrates, it is also conceivable that the primary piRNA pathway selects its substrates indiscriminately, perhaps in an attempt to identify selfish genetic material. However, this scenario is not reconciled with the observation that very substantial populations of abundant transcripts in Drosophila and mouse gonads do not generate 3'UTR piRNAs. Therefore, the piRNA pathway can no longer be seen simply as having to choose between self (endogenous transcripts) and nonself (e.g., TEs). Instead, it must further route specific populations of host transcripts for 3'UTR-directed piRNA production, which is hypothesized to have regulatory consequences (Robine, 2009).

Such regulation may operate in cis or in trans. In support of the former scenario, the tj 3'UTR is one of the most abundant sources of mRNA-derived piRNAs, and increased levels of the TJ transcription factor was observed in piwi mutant clones of appropriate age. Perhaps more compelling was the finding that mRNAs selected for abundant 3'UTR piRNA production were enriched in a variety of GO categories that were not representative of abundant transcripts per se. This is taken as an indication that these transcripts were actively selected for piRNA production. Moreover, the fact that many GO categories were shared by abundant piRNA-generating transcripts between Drosophila and mouse, such as kinase-related factors, DNA binding factors, and RNA silencing factors, suggests that they represent functionally conserved targets of the primary piRNA pathway (Robine, 2009).

It is also conceivable that mRNA/3'UTR-derived piRNAs serve as guides to regulate the expression of other transcripts. While this work was under review, Siomi (2009) proposed that tj-derived 3'UTR piRNAs regulate cell adhesion genes such as fas III, potentially via partially complementary sequences within its ~60 kb intron. The current data do not address this model; however, the candidate fas III-targeting tj piRNAs account for only 2.7% of tj piRNAs and only 0.15% of all 3'UTR piRNAs in the OSS data. Genetic modification of the tj 3'UTR in its endogenous context may clarify whether these particular tj piRNAs indeed serve critical trans-regulatory roles. Nevertheless, in light of the great abundance of mRNA/3'UTR piRNA complexes in Drosophila and mouse gonads, it is certainly plausible that they collectively have substantial trans-regulatory impact. In any case, the revelation of broad and abundant generation of piRNAs from cellular transcripts raises a new direction to understand how the piRNA pathway influences gonad and germ cell development (Robine, 2009).

Somatic primary piRNA biogenesis driven by cis-acting RNA elements and trans-acting Yb

Primary piRNAs in Drosophila ovarian somatic cells arise from piRNA cluster transcripts and the 3' UTRs of a subset of mRNAs, including traffic jam (Tj) mRNA. However, it is unclear how these RNAs are determined as primary piRNA sources. This study identified a cis-acting 100-nt fragment in the Tj 3' UTR that is sufficient for producing artificial piRNAs from unintegrated DNA. These artificial piRNAs were effective in endogenous gene transcriptional silencing. The Tudor domain RNA helicase Yb, a core component of primary piRNA biogenesis center Yb bodies, directly binds the Tj-cis element. Disruption of this interaction markedly reduces piRNA production. Thus, Yb is the trans-acting partner of the Tj-cis element. Yb-CLIP revealed that Yb binding correlates with somatic piRNA production but Tj-cis element downstream sequences produced few artificial piRNAs. It is thus proposed that Yb determines primary piRNA sources through two modes of action: primary binding to cis elements to specify substrates and secondary binding to downstream regions to increase diversity in piRNA populations (Ishizu, 2015).

PIWI-interacting RNAs (piRNAs) interact with PIWI proteins to form piRNA-induced silencing complexes (piRISCs), which repress target genes, mostly transposons, either transcriptionally or at the post-transcriptional level by cleaving transcripts in the cytoplasm. Interestingly, not all cells in the gonads use both mechanisms. Follicle cells in Drosophila ovaries use transcriptional silencing but lack piRISC-mediated post-transcriptional silencing, while germ cells possess both transcriptional and post-transcriptional piRISC machineries. In Bombyx ovaries, only posttranscriptional silencing occurs. This variation largely depends on which PIWI proteins are expressed in a given cell type; transcriptional silencing requires nuclear PIWI proteins while post-transcriptional silencing requires cytoplasmic PIWI proteins (Ishizu, 2015).

Primary piRNAs are produced from single-stranded long noncoding RNAs transcribed from piRNA clusters in a Dicer-independent manner. The Drosophila genome contains 142 piRNA clusters, whose expression is regulated differently in different cell types. flamenco (flam), a representative of unidirectional piRNA clusters, is expressed only in follicle cells, whereas the bidirectional cluster 42AB is expressed specifically in nurse cells. The types of transposon fragments inserted in individual piRNA clusters also vary; therefore, piRNA populations differ among cell types. piRNAs in nurse cells are rather complex because primary piRNAs are amplified through the amplification loop, yielding secondary piRNAs. Recent studies showed that secondary piRNAs further produce phased trailer piRNAs. Follicle cells do not use this amplification system and thus only contain primary piRNAs (Ishizu, 2015).

The biogenesis of somatic primary piRNAs has been studied using ovaries and an ovarian somatic cell (OSC). A current model suggests that upon transcription flam-piRNA precursors are localized to perinuclear Flam bodies and processed at adjacent Yb bodies. Yb bodies contain many piRNA factors besides Yb. Zucchini (Zuc), an endonuclease required for processing piRNA intermediates into mature piRNAs, is localized on the surface of mitochondria. Yb bodies tend to be observed in inter-mitochondrial regions. This arrangement of organelles appears crucial for accelerating piRNA processing because it centralizes all the necessary factors in the cytoplasm. Upon maturation, piRNAs associate with Piwi, a Drosophila PIWI protein, to form piRISCs, which are then translocated to the nucleus to implement nuclear transposon silencing through chromatin modifications on target transposon loci with support from co-factors such as GTSF1/ Asterix and Maelstrom (Ishizu, 2015).

flam is the major source of primary piRNAs in OSCs and follicle cells in the ovaries. flam is largely occupied by transposon remnants, whose orientation predominantly opposes that of the parental transposons; thus, most primary piRNAs arising from the piRNA cluster act as antisense oligos to repress parental transposons. Some protein-coding genes such as Traffic jam (Tj) also act as primary piRNA sources, and genic piRNA sources express proteins in OSCs and follicle cells. The TJ protein, encoded by Tj, is a large Maf transcriptional factor necessary for controlling gonad morphogenesis. Loss of Tj function abolishes Piwi expression in follicle cells. However, Piwi expression in nurse cells is not influenced by TJ loss. Thus, the dependence of Piwi expression on TJ differs between follicle cells and germ cells (Ishizu, 2015).

Only a limited number of transcripts serve as somatic primary piRNA precursors. However, the mechanism underlying the recognition and selection of these transcripts as piRNA precursors is poorly understood. To better understand the mechanism, we used the Tj 3' UTR as representative of somatic primary piRNA sources to identify a cis element and its trans-acting partner necessary for producing primary piRNAs in OSCs (Ishizu, 2015).

Yb bodies and Flam bodies in OSCs are considered to be the centers for primary piRNA maturation/piRISC formation and piRNA intermediate storage, respectively, and exist in close proximity. The formation of both bodies depends on the Yb protein, particularly its RNA-binding activity. In the absence of this, piRNA processing fails, resulting in piRNA loss, although piRNA intermediates and processing factors are present in the cytosol. Thus, Yb binding to piRNA sources centralizes all necessary ingredients for piRNA biogenesis, which is crucial for primary piRNA production This study discovered that the direct association of Yb with a specific ~100-nt element (i.e., cis element) within the piRNA precursors provokes somatic primary piRNA biogenesis from downstream regions. Insertion of the Yb-binding element within RNA molecules that do not otherwise serve as piRNA precursors converts the RNA transcripts into piRNA sources. Artificial primary piRNAs were mapped only downstream, but not upstream, of regions of the Yb-binding element. Previous studies demonstrated that natural genic piRNAs mostly arise from 3' UTRs rather than mRNA CDS or 5' UTRs. The present study also showed that few Tj-piRNAs mapped to the Tj CDS, and that few Yb-CLIP tags were also found in the Tj CDS. Thus, Yb determines not only substrate specificity but also processing directionality in the somatic primary piRNA biogenesis pathway. This may occur through the Yb- controlled recruitment of other piRNA factors, such as another putative RNA helicase Armi and endonuclease Zuc, only to downstream sequences (Ishizu, 2015).

Yb-CLIP tags greatly overlap with primary piRNA-producing loci in the genome. This strongly supports the idea that Yb is the central player in determining substrates in the piRNA pathway. An unexpected but intriguing observation in this study is that Tj-R1 and Tj-R2 in the Tj 3' UTR show strong Yb-binding marks, as does the Tj-cis element, but provoked very little artificial piRNA production in contrast to the Tj-cis element. Yb-CLIP experiments showed that Yb binding to Tj-R1 and Tj-R2 within the Drosophila genome largely depends on Yb binding to its upstream Tj-cis element. Therefore, a model is proposed in which Yb determines primary piRNA sources by two sequential modes of action: primary binding to cis elements that represents selection of piRNA precursors among cellular RNAs, then secondary binding to downstream regions, representing the defining domains to be processed by precursors. This complexity in determining piRNA precursors could ensure the high diversity in piRNA populations, which is a unique feature of piRNAs (Ishizu, 2015).

The RNA-binding activity of Yb is required for primary piRNA production in OSC. Yb mutants carrying a point mutation within the DEAD box showed little RNA binding activity. When these Yb mutants were expressed individually in OSC lacking endogenous Yb, piRNA precursors were not accumulated in Flam bodies, and few piRNAs were produced. As a consequence, transposons were de-silenced. Therefore, there is little doubt that the RNA-binding activity of Yb through the DEAD-box is indispensable for primary piRNA production. HITS-CLIP experiments clarified direct interaction of Yb with piRNA sources, including Tj mRNA. Insertion of a particular Yb-bound RNA element within Tj mRNA, i.e., the Tj-cis element, upstream of any given RNA molecule enables the arbitrary sequences to produce artificial piRNAs. Deletion of the Tj-cis element from the Drosophila genome significantly abolished piRNA production from its downstream region spanning at least ~200 nt. These observations strongly support the proposed model, in which Yb is the trans-acting factor that recognizes and binds cis elements within piRNA precursors to provoke primary piRNA biogenesis in ovarian somatic cells. However, it does not exclude the possibility that Yb collaboratively achieves this task with unknown factors. Moreover, it is not certain if Yb is the uppermost factor in the cytoplasmic phase of the biogenesis pathway upon nuclear transport of piRNA precursors (Ishizu, 2015).

Divergent transcriptional regulatory logic at the intersection of tissue growth and developmental patterning

The Yorkie/Yap transcriptional coactivator is a well-known regulator of cellular proliferation in both invertebrates and mammals. As a coactivator, Yorkie (Yki) lacks a DNA binding domain and must partner with sequence-specific DNA binding proteins in the nucleus to regulate gene expression; in Drosophila, the developmental regulators Scalloped (Sd) and Homothorax (Hth) are two such partners. To determine the range of target genes regulated by these three transcription factors, genome-wide chromatin immunoprecipitation experiments for each factor was performed in both the wing and eye-antenna imaginal discs. Strong, tissue-specific binding patterns are observed for Sd and Hth, while Yki binding is remarkably similar across both tissues. Binding events common to the eye and wing are also present for Sd and Hth; these are associated with genes regulating cell proliferation and 'housekeeping' functions. In contrast, tissue-specific binding events for Sd and Hth significantly overlap enhancers that are active in the given tissue, are enriched in Sd and Hth DNA binding sites, respectively, and are associated with genes that are consistent with each factor's tissue-specific functions. Overall these results suggest that both Sd and Hth use distinct strategies to regulate distinct gene sets during development: one strategy is shared between tissues and associated with Yki, while the other is tissue-specific, generally Yki-independent and associated with developmental patterning (Slattery, 2013).

The control of gene expression in multicellular eukaryotes depends on a limited set of transcription factors that are reused in different contexts and combinations to execute a diverse array of cellular functions. To gain insight into this process this study used tissue-specific, genome-wide ChIP to explore the global DNA targeting properties of three transcriptional regulators – Yki, Sd, and Hth. Yki is a transcriptional coactivator that regulates tissue growth in all tissues, and it does so in part through interactions with the DNA binding TFs Sd and Hth. However, in addition to their Yki-dependent roles in promoting tissue growth, Sd and Hth also have highly tissue-specific developmental roles. Thus, this group of regulators provides an ideal starting point for addressing the logic by which TFs execute both tissue-specific and -nonspecific gene regulatory functions in vivo. The implications of the differences this study uncovered between these modes of binding for Hth and Sd are discussed, as well as the unexpectedly large number of shared binding sites for Yki (Slattery, 2013)

Drosophila Yki was initially identified as an essential transcriptional coactivator in the Hippo tumor suppressor pathway. Loss of function clones of yki grow very poorly, while gain of function Yki clones result in tissue overgrowths that are similar to those generated when the upstream kinases (Hippo and Warts) are compromised. These observations suggested that Yki, with the help of DNA binding proteins, would target genes required for cell proliferation and survival, including the known Hippo pathway targets cycE and diap1. Consistent with this expectation, this study observed Yki binding to these and other genes that are regulated by the Hippo pathway. Unexpectedly, however, in addition to known Hippo pathway genes Yki binding to several thousands of genes was observed in both the eye-antenna and wing imaginal discs, implying that Yki targets many more genes than those regulated by the Hippo pathway, or that the Hippo pathway targets many more genes than previously thought. Consistent with the latter possibility, over 1000 of the genes identified as tissue-shared Yki targets in this study are upregulated >2-fold in wts^- wing discs relative to wild-type based on recently published RNA-seq data. In addition, Yki was recently shown to bind and activate several genes required for mitochondrial fusion. Moreover, the mammalian homologs of Yki, Yes-associated protein (YAP) and TAZ (transcriptional coactivator with PDZ-binding motif) are thought to regulate many genes in a wide variety of contexts, including human embryonic stem cells and several adult human tissues. Taken together, these results suggest that Yki may be a widely used transcriptional coactivator in Drosophila and vertebrates. The severe cell proliferation defects associated with yki mutant clones may have obscured its other functions in other pathways. These results are consistent with the idea that Yki and its vertebrate orthologs interact with a wide variety of transcription factor. Together, the data imply that DNA binding proteins in addition to Sd and Hth may recruit Yki to a large number of broadly active CRMs (Slattery, 2013)

The view that Yki is recruited to DNA by factors other than Sd was recently questioned by experiments suggesting that, in the eye imaginal disc, sd yki double mutant clones proliferate better than yki single mutant clones. These observations were interpreted to suggest that Sd is a default repressor of proliferation and survival-promoting genes. However, this conclusion is complicated by the observation that both Sd and Yki are also important for specifying non-retinal (peripodial epithelium) fates in the eye imaginal disc: thus, the partially rescued growth of sd yki clones could in part be due to a fate transformation. Further, this study found that the activity of the ban-eye enhancer is not affected in sd clones, but is lost in hth clones, arguing that at least for this direct Hippo pathway target Hth, not Sd, is the primary activator. It is noteworthy that although their activities can be separated, the ban wing and eye enhancers identified in this study are adjacent to each other in the native ban locus. It is plausible that Sd+Yki input provides a basal level of activity in both tissues and that Hth and Sd boost this level in the eye and wing, respectively. Regardless, the improved growth of sd yki clones does not argue against the idea that Yki is recruited to survival genes by Hth in wild type eye discs. Taken together with genome-wide binding and ban enhancer studies, it is suggested that the absence of Sd results in both a fate change and some derepression of survival genes, but that wild type proliferation and gene regulation in the eye disc requires the recruitment of Yki to the DNA by Hth (Slattery, 2013)

In contrast to the widespread and largely tissue-nonspecific binding observe for Yki, Sd and Hth exhibit both tissue-specific and tissue-shared binding events. Multiple characteristics distinguish these types of binding. First, tissue-shared binding by both Sd and Hth is frequently associated with Yki binding and often close to cell cycle and housekeeping genes, while tissue-specific binding is not. These observations are consistent with previous studies showing that Yki controls cell survival and proliferation in all imaginal discs, an activity that is regulated by the Hippo pathway. Second, compared to tissue-shared binding, DNA sequences bound by Sd and Hth in a tissue-specific manner are more conserved, more likely to contain the TF's consensus binding site, less likely to be promoter proximal, and more likely to be associated with key developmental regulatory loci. Third, tissue-specific Sd and Hth binding events are more likely to overlap with enhancers active in the corresponding tissue. To illustrate this point, the newly identified tissue-specific TF-CRM interactions at wg match the known roles for Sd and Hth. Taken together, these results suggest that regulation at the level of TF-DNA binding is a significant mechanism by which Sd and Hth regulate tissue-specific gene expression. Tissue-specific binding could be regulated through direct or indirect interactions with additional transcription factors, through tissue-specific differences in DNA accessibility, or through a combination of these factors (Slattery, 2013)

This study also found that distinct chromatin types are differentially correlated with tissue-specific and -nonspecific binding, even though these chromatin categories were defined in Kc cells. All tissue-shared binding events have a strong tendency to occur in actively transcribed chromatin states. Tissue-specific (W>EA) Sd and Hth binding is also enriched in RED chromatin but is uniquely enriched in BLUE chromatin. BLUE chromatin is associated with Polycomb-mediated repression. The W>EA Sd and Hth binding in Polycomb-associated chromatin indicate that these factors target tissue-specific enhancers that are also regulated by PcG proteins during development (Slattery, 2013)

Despite the importance of tissue-specific binding as a regulatory mechanism for Sd and Hth activity, both factors also displayed a significant amount of tissue-shared binding. These tissue-shared binding events can be broken down into distinct groups based on the local chromatin environment. The majority of tissue-shared binding occurs in YELLOW chromatin and is associated with ubiquitously expressed housekeeping genes. However, binding that occurs in BLUE chromatin, and to a lesser extent in RED chromatin, is more conserved and more likely to be associated with a TF's motif, both characteristics of tissue-specific binding. In the case of the bantam eye and wing enhancers, Sd and Hth binding in BLUE chromatin is direct and apparently able to drive tissue-specific, rather than ubiquitous, expression patterns. Other examples of enhancers in RED or BLUE chromatin that drive patterned expression and have tissue shared binding are shown in this study. These observations suggest that gene regulation by Sd and Hth may also be controlled at a step beyond DNA binding, perhaps via interactions with additional transcription factors at a given enhancer. Alternatively, some of the binding events called as tissue-shared may turn out to be specific binding events in distinct cell types within each imaginal disc (e.g. hinge, notum, and pouch in the wing disc and antenna, eye progenitor domain, and photoreceptors in the eye-antenna disc). Regardless, the hundreds of Sd- and Hth-CRM interactions identified in this study provide a tremendous resource for further dissecting the mechanisms by which Sd and Hth regulate patterned gene expression (Slattery, 2013)

Notably, few of the above conclusions would have been clear had genome-wide binding been measured in only one of the two tissues. Tissue-specific binding is not the most highly enriched (that is, the signal is generally weaker compared to tissue-shared events) and might have been overlooked had just one tissue been characterized, where the strongest peaks are generally the focal point. The tissue-specific binding events detected in this study may also occur in subsets of cells in the wing or eye-antennal discs, which are also heterogeneous in cell type. This would explain why tissue-specific binding signals may be weaker, because the ChIP data represent an average of all cell types in a single imaginal disc type. If correct, it would be an error to focus on only the strongest peaks when analyzing in vivo TF binding, particularly in heterogeneous tissues. It is possible that ChIP signal is more biologically meaningful in highly homogenous tissues like the blastoderm Drosophila embryo, or in cell culture. Still, distinct TF-DNA binding mechanisms (long residence time versus rapid binding turnover) with different functional outcomes can lead to indistinguishable, strong ChIP peaks, making it difficult to interpret ChIP data on strength of signal alone. Despite their lower intensity, many biologically relevant binding events, such as those identified in this study, may only stand out when looking at the influence of tissue context on binding (Slattery, 2013).

Opposite Feedbacks in the Hippo Pathway for Growth Control and Neural Fate

Signaling pathways are reused for multiple purposes in plant and animal development. The Hippo pathway in mammals and Drosophila coordinates proliferation and apoptosis via the coactivator and oncogene, YAP/Yorkie (Yki), which is homeostatically regulated through negative feedback. In the Drosophila eye, cross-repression between the Hippo pathway kinase, LATS/Warts (Wts), and growth regulator, Melted, generates mutually exclusive photoreceptor subtypes. This study shows that this all-or-nothing neuronal differentiation results from Hippo pathway positive feedback: Yki both represses its negative regulator, warts, and promotes its positive regulator, melted. This postmitotic Hippo network behavior relies on a tissue-restricted transcription factor network - including a conserved Otx/Orthodenticle-Nrl/Traffic Jam feedforward module - that allows Warts-Yki-Melted to operate as a bistable switch. Altering feedback architecture provides an efficient mechanism to co-opt conserved signaling networks for diverse purposes in development and evolution (Jukam, 2013).

A fundamental strategy in animal development is to re-purpose the same signaling pathways for a diversity of functions. This study identified a tissue-specific transcription factor network that enables the otherwise homeostatic Hippo growth control pathway to act as a bistable switch for terminal cell fate. This alteration in network level properties—such as positive versus negative feedback—within biochemically conserved pathways is an efficient means to re-use a signaling network in contexts as distinct as proliferation and terminal differentiation (Jukam, 2013).

How is the R8-specific Hippo regulatory circuit achieved? The two interlinked positive feedback loops (one with wts, one with melt) provide the R8 Hippo pathway with multiple points of potential regulation. Context-specific expression of wts and melt is defined by overlapping expression of four transcription factors: Otd, Tj, Pph13, and Sens. Otd and Pph13 are expressed in all photoreceptors and generate a permissive context that endows the initially equipotent R8s with the competence to become either subtype: Otd promotes melt/Rh5 whereas Pph13 promotes wts/Rh6 expression. This competence is further restricted by expression of Tj in R7 and R8, and Sens in R8s, which ensures that melt and wts cross-regulation is restricted to R8s. Importantly, it is the status of Yki activity and resulting feedback that assures the outcome of pR8/Rh5 vs. yR8/Rh6 (p vs. y) fate: in pR8s, Yki functions with Otd and Tj to promote melt and Rh5; in yR8s, wts inhibits Yki, preventing melt and Rh5 expression and allowing 'default' wts and Rh6 expression by Pph13 and Sens. Each of these four transcription factors regulates a partially overlapping subset of R8 subtype fate genes, and together, the network cooperates at multiple regulatory nodes to provide the specific context for repurposing the Hippo pathway (Jukam, 2013).

While other instances of pathways with both positive and negative feedback exist, these are conceptually different from R8 Hippo regulation. For example, in Sprouty (hSpry) regulation of Ras/MAPK-mediated EGFR signaling, EGFR induces hSpry2 expression but hSpry2 inhibits EGFR function (negative feedback); however, hSpry2 also promotes EGFR activity by preventing Cbl-dependent EGFR inhibition (positive feedback). hSpry2 positive feedback is likely coupled to its negative feedback to fine-tune the length and amplitude of receptor activation. In contrast, the opposite Hippo pathway feedbacks occur in vastly different cell types (mitotic epithelial cells versus post-mitotic neurons), and both forms of feedback cannot co-exist in R8 since Yki’s repression of wts expression (positive feedback) would make Yki up-regulation of Hippo regulators (negative feedback) inconsequential (Jukam, 2013).

Gaining positive feedback or losing negative feedback within Hippo signaling could permit oncogenesis. Indeed, the Yki ortholog, YAP, is an oncogene and is amplified in multiple tumors, and LATS1/2 (Wts) down-regulation is associated with non-small cell lung carcinomas, soft tissue sarcoma, metastatic prostate cancers, retinoblastoma, and acute lymphoblastic leukemia. Otx and MAF factors are also oncogenic in a number of tissues. Thus, understanding the regulatory networks identified here in other contexts will be crucial for deciphering how normal signaling pathways can go awry (Jukam, 2013).

The current findings also reveal that a Crx/Otd-Nrl/Tj feedforward module plays a conserved role in post-mitotic photoreceptor fate specification in both flies and mammals. Both induce one photoreceptor fate at the expense of another, and both regulate opsins with a feedforward loop wherein Crx/Otd activates Nrl/Tj expression and Crx-Nrl or Otd-Tj synergistically activate downstream targets (Hao, 2012). Given such deep evolutionary conservation, this module may be critical for generating photoreceptor diversity in other complex visual systems (Jukam, 2013).

This work has two main implications. First, although positive feedback is well documented in other switch-like, irreversible cell fate decisions such as in Xenopus oocyte maturation or cell cycle entry, this work suggests that positive feedback could have a broad role in terminal neuronal differentiation, which often requires permanent fate decisions to maintain a neuron’s functional identity. Second, the changes in network topology in R8 photoreceptors allows a finely tuned growth control pathway to be used as a switch in a permanent binary cell fate decision. Context-specific regulation allows the feedback architecture to change in an otherwise conserved signaling module. This may be a general mechanism to endow signaling networks with new systems properties and diversify cell fates in development and evolution (Jukam, 2013).

The Maf factor Traffic jam both enables and inhibits collective cell migration in Drosophila oogenesis

Border cell cluster (BCC) migration in the Drosophila ovary is an excellent system to study the gene regulatory network that enables collective cell migration. This study identified the large Maf transcription factor Traffic jam (Tj) as an important regulator of BCC migration. Tj has a multifaceted impact on the known core cascade that enables BCC motility, consisting of the Jak/Stat signaling pathway, the C/EBP factor Slow border cells (Slbo), and the downstream effector DE-cadherin (DEcad). The initiation of BCC migration coincides with a Slbo-dependent decrease in Tj expression. This reduction of Tj is required for normal BCC motility, as high Tj expression strongly impedes migration. At high concentration, Tj has a tripartite negative effect on the core pathway: a decrease in Slbo, an increase in the Jak/Stat inhibitor Socs36E, and a Slbo-independent reduction of DEcad. However, maintenance of a low expression level of Tj in the BCC during migration is equally important, as loss of tj function also results in a significant delay in migration concomitant with a reduction of Slbo and consequently of DEcad. Taken together, it is concluded that the regulatory feedback loop between Tj and Slbo is necessary for achieving the correct activity levels of migration-regulating factors to ensure proper BCC motility (Gunawan, 2013).

Traffic jam functions in a branched pathway from Notch activation to niche cell fate

The niche directs key behaviors of its resident stem cells, and is thus crucial for tissue maintenance, repair and longevity. However, little is known about the genetic pathways that guide niche specification and development. The male germline stem cell niche in Drosophila houses two stem cell populations and is specified within the embryonic gonad, thus making it an excellent model for studying niche development. The hub cells that form the niche are specified early by Notch activation. Over the next few hours, these individual cells then cluster together and take up a defined position before expressing markers of hub cell differentiation. This timing suggests that there are other factors for niche development yet to be defined. This study has identified a role for the large Maf transcription factor Traffic jam (Tj) in hub cell specification downstream of Notch. Tj downregulation is the first detectable effect of Notch activation in hub cells. Furthermore, Tj depletion is sufficient to generate ectopic hub cells that can recruit stem cells. Surprisingly, ectopic niche cells in tj mutants remain dispersed in the absence of Notch activation. This led to the uncovering of a branched pathway downstream of Notch in which Bowl functions to direct hub cell assembly in parallel to Tj downregulation (Wingert, 2015).

Adult stem cells are a crucial component of most organs. Having the unique ability to either self-renew or replace differentiated cells, they function in steady-state homeostasis and can also be activated in response to injury. The instructive cues that guide stem cells to self-renew or differentiate often come extrinsically from the local microenvironment called the stem cell niche. Whereas previous work focused primarily on the identification and steady-state function of the stem cell niche, the mechanisms by which a niche is established during development are only more recently being addressed. Several pathways required for niche cell specification have now been identified, including Sonic Hedgehog in the neurogenic niche, hormonal and Notch signaling in the Drosophila ovarian niche and Wnt signaling in the C. elegans germline niche. Comprehensive analysis of the pathway targets and how they individually direct niche cell fate is limited in complex systems. Fortunately, the Drosophila testis germline stem cell (GSC) niche is well characterized and is established early within the male embryonic gonad. The niche cells can be unambiguously identified with multiple markers as early as the embryo-to-larval transition, and stem cells are recruited shortly thereafter. For this reason, the male gonad provides an ideal system to identify and dissect the pathways required for niche specification (Wingert, 2015).

The male GSC niche is located at the apical tip of the testis and is composed of a small aggregate of post-mitotic somatic cells called hub cells. The hub cells secrete bone morphogenetic proteins (BMPs) and the cytokine Unpaired to activate self-renewal and adhesion in the GSCs. An additional component of the niche, the cyst stem cell (CySC) lineage, is maintained by Unpaired and Hedgehog secreted from the hub. Steady-state function of the niche is established within the larval gonad following hub cell specification and stem cell recruitment (Wingert, 2015).

The specification of hub cells from a pool of somatic gonadal precursors (SGPs) occurs via Notch activation (Kitadate, 2010; Okegbe, 2011). Around mid-embryogenesis, the germ cells travel through the gut, where they coalesce with SGPs. At this time, the SGPs are briefly exposed to an endodermally derived Delta ligand which activates Notch in some SGPs (Okegbe, 2011). Whereas Notch signaling is required early, cell surface and gene expression markers of hub cell fate are not detected until many hours later, at the end of embryogenesis. Thus, the hub only becomes identifiable in late-stage embryonic gonads after hub cells have sorted to an anterior position, accumulated significant levels of epithelial adhesion proteins and adopted cobblestone morphology. At this time, gene regulatory changes result in hub-specific expression of unpaired (Wingert, 2015).

The delay in differentiation suggests that there are unknown intermediate effectors in hub cell specification downstream of Notch. The transcription factor Bowl and receptor tyrosine kinase (RTK) signaling have also been implicated in hub cell fate determination; however, their relationship to Notch activation has not been elucidated. This work focuses on the role of the large Maf transcription factor Traffic jam (Tj) during hub cell specification. Previous work has shown that tj is expressed in SGPs and is required in gonad morphogenesis (Li, 2003). Additionally, Tj can suppress accumulation of the septate junction protein Fasciclin III (FasIII) and its RNA in the somatic cells of the adult testis and ovary (Li, 2003). This study shows that (1) Tj represses markers of hub cell fate and niche signaling, (2) Tj is downregulated in a subset of anterior gonadal cells and (3) Notch is required for this suppression. Finally, it was shown that Tj acts along one arm of a pathway downstream of Notch activation for hub cell specification, whereas Bowl acts in a parallel arm to direct hub cell assembly (Wingert, 2015).

Previous genetic and lineage tracing data demonstrated that a subset of SGPs is Notch-activated early during gonadogenesis (Okegbe, 2011). Several hours later, these cells assemble into a compact niche that expresses various factors required for stem cell recruitment, attachment and self-renewal. This study has dissected the pathway downstream of Notch activation in hub cell specification. Specifically, Notch activation downregulates Traffic jam, thus relieving repression of unpaired and allowing for the accumulation of multiple adhesion proteins in these cells. In a parallel arm, Bowl is activated and regulates anterior assembly of hub cells. It is likely that hedgehog is also activated by this parallel arm. Therefore, this study is the first to look at the individual inputs into hub cell fate and uncovers a branched pathway downstream of Notch specification of niche fate (Wingert, 2015).

Previously, the earliest effect of Notch activation detected in SGPs was the induction many hours later of hub-specific unpaired expression. Now, this study has demonstrated that Tj downregulation is visible prior to this time, and loss-of-function data suggest that this is controlled by Notch. Although some work suggests that Notch is activated in all SGPs, relatively few cells induced for Notch (Okegbe, 2011). This is consistent with downregulation of Tj in the few hub cells that are specified. Furthermore, forced activation of Notch only moderately increases the number of hub cells (Kitadate, 2010), and these extra hub cells also exhibited reduced Tj accumulation. Although this might suggest direct control of Tj by Notch, the situation is more complex. The fact that Notch activation is insufficient for hub cell fate, combined with the fact that a significant period exists between the requirement for Notch and a detectable reduction of Tj protein (∼6 h), suggests that Notch repression of Tj probably occurs through intermediate effectors. It was recently shown that robust Tj accumulation in early-stage SGPs requires the gene midline, which encodes a T-box 20 transcription factor (Tripathy, 2014). In other tissues, Midline can antagonize Notch signaling and is repressed in Notch-activated cells (Das, 2013). Therefore, it would be interesting to determine whether midline is regulated by Notch in SGPs (Wingert, 2015).

Notch also acts in specifying the female germline stem cell niche in Drosophila. As these niche cells also express traffic jam, it is worth investigating whether Traffic jam mediates Notch signaling during germarium niche specification. Furthermore, the regulatory relationship this study identified in the male gonad between Notch and the Maf factor Tj might also apply in mammals. Interestingly, both c-Maf and MafB are expressed in somatic cells intermingled with the germ line in the developing mammalian gonad (DeFalco, 2011). Additionally, Notch signaling restricts Leydig cell differentiation within the interstitial compartment at the same developmental timepoint. The current data suggest that it would be reasonable to test whether Notch and Maf factors function together in specifying mammalian gonadal cell types (Wingert, 2015).

Tj depletion relieves repression of unpaired in SGPs, allowing them to activate Jak/STAT, thus bypassing one role for Notch. Recent work showed that Tj functions in border cell migration, in which it also modulates the Jak/STAT pathway. In that case, Tj enhances expression of the Jak/STAT pathway antagonist Suppressor of cytokine signaling E (Socs36E – Flybase.org) (Gunawan, 2013). Whereas this study found that Tj functions by repressing expression of the ligand, both studies support a role in which Tj attenuates STAT signaling. Notably, in T helper cells, a large Maf factor also regulates expression of cell type-specific cytokines that activate STAT signaling (Ho, 1996; Kim, 1999). In this system, c-Maf activates expression of the ligand rather than repressing it; however, large Mafs can activate or repress transcription depending on context (Kataoka, 2007). These examples highlight recurring evidence of cooperation between Maf factors and cytokine signaling (Wingert, 2015).

Proteins representing different adhesive complexes accumulate on hub cells, including FasIII, E-cadherin and N-cadherin. However, depleting one or several subsets of these factors has little effect on hub cell assembly or aggregation at the anterior. This suggests either significant redundancy in hub cell adhesion or that a yet unidentified factor is responsible for mediating the proper aggregation and assembly of hub cells. In tj mutants, ectopic hub cells were enriched for multiple, different epithelial complexes (FasIII and E-cadherin or FasIII and N-cadherin), yet these cells did not exhibit compact, anterior assembly. This suggests a requirement for an unknown factor. As activating Bowl was sufficient to significantly rescue hub cell assembly, perhaps Bowl regulates this factor. Bowl and Drumstick, both members of the Odd-skipped family of zinc-finger proteins, regulate morphological changes in the developing Drosophila leg. Overexpression of another family member, Odd, results in cell-autonomous and non-autonomous morphological changes and increases in F-actin. Therefore, examining changes in F-actin enrichment and identifying actin regulators downstream of Bowl would reveal the mechanism for hub assembly (Wingert, 2015).

This study has not clarified the epistatic relationship between bowl and Notch, which can be complex. In leg development, bowl expression is induced by Notch signaling. Furthermore, Bowl can repress delta in the Notch-activated cell, thus stabilizing a Notch signaling interface at leg segment boundaries. In the wing, however, Bowl modulates Notch signaling by reducing availability of the Notch co-repressor Groucho. In the gonad, mutants for either Notch or bowl have fewer hub cells, although the Notch phenotype is more severe. Due to this and to the early requirement for Notch in SGPs, it is suspected that Notch functions upstream of Bowl, forming a pathway parallel to Notch and Tj. Whether the Notch-Bowl arm results in hedgehog expression is unresolved. Similar to the complex relationship between Bowl and Notch, Hedgehog signaling can either promote Bowl accumulation (in the epidermis) or Bowl can induce hedgehog expression (in retinogenesis). Whereas hhlacZ induction was not detected in gonads activated for Bowl, inviability prevented examining later timepoints. Previously, hhlacZ induction was detected five days after clonal activation of Bowl in adult CySCs. Given that CySCs are derived from SGPs, the idea is favored that Bowl can induce hedgehog (Wingert, 2015).

The current model suggests that Tj and Bowl mediate many aspects of hub cell fate downstream of Notch signaling. As noted above, a fraction of hub cells are still specified in bowl mutants, and these appear to assemble anteriorly. This could suggest that only a subset of hub cells require Bowl for proper assembly (perhaps centrally located SGPs that need to move anteriorly). Alternatively, there could be an additional, unknown Notch effector that can compensate for bowl. Indeed, there is room for some complexity as RTK signaling represses hub cell fate in posterior SGPs and its interplay with the Notch pathway is yet unclear. In this regard, it is intriguing that tj mutant SGPs with forced activation of Bowl continue to cycle, whereas endogenous hub cells are quiescent (Fig. 6D and data not shown). Interestingly, EGFR pathway activity was detected in some ectopic hub cells. EGF signaling mediates the proliferation of Drosophila intestinal and gastric stem cells. Perhaps the ectopic hub cells in tj mutants remain cycling due to an inability to repress EGF signaling. With the pathway for hub cell specification now more clearly delineated, future work can address the intersecting cell biological and gene expression targets of multiple pathways required for hub cell fate (Wingert, 2015).

Transcriptional regulation of rod photoreceptor homeostasis revealed by in vivo NRL targetome analysis

A stringent control of homeostasis is critical for functional maintenance and survival of neurons. In the mammalian retina, the basic motif leucine zipper transcription factor NRL (Drosophila homolog: Traffic Jam ) determines rod versus cone photoreceptor cell fate and activates the expression of many rod-specific genes. This study reports an integrated analysis of NRL-centered gene regulatory network by coupling chromatin immunoprecipitation followed by high-throughput sequencing (ChIP-Seq) data with global expression profiling and in vivo knockdown studies. Approximately 300 direct NRL target genes were identified. Of these, 22 NRL targets are associated with human retinal dystrophies, whereas 95 mapped to regions of as yet uncloned retinal disease loci. In silico analysis of NRL ChIP-Seq peak sequences revealed an enrichment of distinct sets of transcription factor binding sites. Specifically, genes involved in photoreceptor function include binding sites for both NRL and homeodomain protein CRX, an Orthodenticle homolog. Evaluation of 26 ChIP-Seq regions validated their enhancer functions in reporter assays. In vivo knockdown of 16 NRL target genes resulted in death or abnormal morphology of rod photoreceptors, suggesting their importance in maintaining retinal function. Histone demethylase Kdm5b (Drosophila homolog: Little imaginal discs) was identified as a novel secondary node in NRL transcriptional hierarchy. Exon array analysis of flow-sorted photoreceptors in which Kdm5b was knocked down by shRNA indicated its role in regulating rod-expressed genes. These studies identify candidate genes for retinal dystrophies, define cis-regulatory module(s) for photoreceptor-expressed genes and provide a framework for decoding transcriptional regulatory networks that dictate rod homeostasis (Hao, 2012).

Dimeric combinations of MafB, cFos and cJun control the apoptosis-survival balance in limb morphogenesis

Apoptosis is an important mechanism for sculpting morphology. However, the molecular cascades that control apoptosis in developing limb buds remain largely unclear. This study showed that MafB was specifically expressed in apoptotic regions of chick limb buds, and MafB/cFos (see Drosophila Kayak) heterodimers repressed apoptosis, whereas MafB/cJun (see Drosophila Jun) heterodimers promoted apoptosis for sculpting the shape of the limbs. Chromatin immunoprecipitation sequencing in chick limb buds identified potential target genes and regulatory elements controlled by Maf and Jun. Functional analyses revealed that expression of p63 and p73, key components known to arrest the cell cycle, was directly activated by MafB and cJun. The data suggest that dimeric combinations of MafB, cFos and cJun in developing chick limb buds control the number of apoptotic cells, and that MafB/cJun heterodimers lead to apoptosis via activation of p63 and p73 (Suda, 2014).

Search PubMed for articles about Drosophila Traffic jam

Aravin, A. A., Hannon, G. J. and Brennecke, J. (2007). The Piwi-piRNA pathway provides an adaptive defense in the transposon arms race. Science 318: 761-764. PubMed ID: 17975059

Brennecke, J., et al. (2007). Discrete small RNA-generating loci as master regulators of transposon activity in Drosophila. Cell 128: 1089–1103. PubMed ID: 17346786

Blank, V. and Andrews, N.C. (1997). The Maf transcription factors: regulators of differentiation. Trends Biochem. Sci. 22: 437–441. PubMed ID: 9397686

Cooke, J., et al. (2001). Eph signalling functions downstream of Val to regulate cell sorting and boundary formation in the caudal hindbrain. Development 128: 571–580. PubMed ID: 11171340

Das, S., Chen, Q. B., Saucier, J. D., Drescher, B., Zong, Y., Morgan, S., Forstall, J., Meriwether, A., Toranzo, R. and Leal, S. M. (2013). The Drosophila T-box transcription factor Midline functions within the Notch-Delta signaling pathway to specify sensory organ precursor cell fates and regulates cell survival within the eye imaginal disc. Mech Dev 130: 577-601. PubMed ID: 23962751

DeFalco, T., Takahashi, S. and Capel, B. (2011). Two distinct origins for Leydig cell progenitors in the fetal testis. Dev Biol 352: 14-26. PubMed ID: 21255566

Dlakic, M., Grinberg, A. V., Leonard, D. A. and Kerppola, T. K. (2001). DNA sequence-dependent folding determines the divergence in binding specificities between Maf and other bZIP proteins. EMBO J. 20: 828–840. PubMed ID: 11179227

Gunawan, F., Arandjelovic, M. and Godt, D. (2013). The Maf factor Traffic jam both enables and inhibits collective cell migration in Drosophila oogenesis. Development 140: 2808-2817. PubMed ID: 23720044

Hao, H., Kim, D. S., Klocke, B., Johnson, K. R., Cui, K., Gotoh, N., Zang, C., Gregorski, J., Gieser, L., Peng, W., Fann, Y., Seifert, M., Zhao, K. and Swaroop, A. (2012). Transcriptional regulation of rod photoreceptor homeostasis revealed by in vivo NRL targetome analysis. PLoS Genet 8: e1002649. PubMed ID: 22511886

Ho, I. C., Hodge, M. R., Rooney, J. W. and Glimcher, L. H. (1996). The proto-oncogene c-maf is responsible for tissue-specific expression of interleukin-4. Cell 85: 973-983. PubMed ID: 8674125

Ishizu, H., Iwasaki, Y. W., Hirakata, S., Ozaki, H., Iwasaki, W., Siomi, H. and Siomi, M. C. (2015). Somatic primary piRNA biogenesis driven by cis-acting RNA elements and trans-acting Yb. Cell Rep 12: 429-440. PubMed ID: 26166564

Jenkins, A. B., McCaffery, J. M. and Van Doren, M. (2003). Drosophila E-cadherin is essential for proper germ cell–soma interaction during gonad morphogenesis. Development 130: 4417–4426. PubMed ID: 12900457

Kataoka, K., Nishizawa, M. and Kawai, S. (1993) Structure– function analysis of the maf oncogene product, a member of the b-Zip protein family. J. Virol. 67: 2133–2141. PubMed ID: 8383235

Kataoka, K. (2007). Multiple mechanisms and functions of maf transcription factors in the regulation of tissue-specific genes. J Biochem 141: 775-781. PubMed ID: 17569705

Kawashima, T., Nakamura, A., Yasuda, K. and Kageyama, Y. (2003). Dmaf, a novel member of Maf transcription factor family is expressed in somatic gonadal cells during embryonic development and gametogenesis in Drosophila. Gene Expr. Patterns 3(5): 663-7. PubMed ID: 12972003

Kerppola, T. K. and Curran, T. (1994). A conserved region adjacent to the basic domain is required for recognition of an extended DNA binding site by Maf–Nrl family proteins. Oncogene 9: 3149–3158. PubMed ID: 7936637

Kim, J. I., Ho, I. C., Grusby, M. J. and Glimcher, L. H. (1999). The transcription factor c-Maf controls the production of interleukin-4 but not other Th2 cytokines. Immunity 10: 745-751. PubMed ID: 10403649

Kitadate, Y. and Kobayashi, S. (2010). Notch and Egfr signaling act antagonistically to regulate germ-line stem cell niche formation in Drosophila male embryonic gonads. Proc Natl Acad Sci U S A 107: 14241-14246. PubMed ID: 20660750

Kurschner, C. and Morgan, J. I. (1995). The maf proto-oncogene stimulates transcription from multiple sites in a promoter that directs Purkinje neuron-specific gene expression. Mol. Cell Biol. 15: 246–254. PubMed ID: 7799931

Li, C., et al. (2009). Collapse of germline piRNAs in the absent of Argonaute3 reveals somatic piRNAs in flies. Cell 137: 509–521. PubMed ID: 19395009

Li, M. A., Alls, J. D., Avancini, R. M., Koo, K. and Godt D. (2003). The large Maf factor Traffic Jam controls gonad morphogenesis in Drosophila. Nat. Cell Biol. 5(11): 994-1000. PubMed ID: 14578908

Malone, C. et al. (2009). Specialized piRNA pathways act in germline and somatic tissues of the Drosophila ovary. Cell 137: 522–535. PubMed ID: 19395010

Okegbe, T. C. and DiNardo, S. (2011). The endoderm specifies the mesodermal niche for the germline in Drosophila via Delta-Notch signaling. Development 138: 1259-1267. PubMed ID: 21350008

Panchal, T., Chen, X., Alchits, E., Oh, Y., Poon, J., Kouptsova, J., Laski, F. A. and Godt, D. (2017). Specification and spatial arrangement of cells in the germline stem cell niche of the Drosophila ovary depend on the Maf transcription factor Traffic jam. PLoS Genet 13(5): e1006790. PubMed ID: 28542174

Robine, N., et al. (2009). A broadly conserved pathway generates 3'UTR-directed primary piRNAs. Curr. Biol. 19(24): 2066-76. PubMed ID: 20022248

Sadl, V., et al. (2002). The mouse Kreisler (Krml1/MafB) segmentation gene is required for differentiation of glomerular visceral epithelial cells. Dev. Biol. 249: 16–29. PubMed ID: 12217315

Saito, K., et al. (2009). A regulatory circuit for piwi by the large Maf gene traffic jam in Drosophila. Nature 461(7268): 1296-9. PubMed ID: 19812547

Schüpbach, T. and Wieschaus, E. (1991). Female sterile mutations on the second chromosome of Drosophila melanogaster. II. Mutations blocking oogenesis or altering egg morphology. Genetics 129: 1119–1136. PubMed ID: 1783295

Siomi, H. and Siomi, M. C. (2009). On the road to reading the RNA-interference code. Nature 457: 396–404. PubMed ID: 19158785

Slattery, M., Voutev, R., Ma, L., Negre, N., White, K. P. and Mann, R. S. (2013). Divergent transcriptional regulatory logic at the intersection of tissue growth and developmental patterning. PLoS Genet 9: e1003753. PubMed ID: 24039600

Suda, N., Itoh, T., Nakato, R., Shirakawa, D., Bando, M., Katou, Y., Kataoka, K., Shirahige, K., Tickle, C. and Tanaka, M. (2014). Dimeric combinations of MafB, cFos and cJun control the apoptosis-survival balance in limb morphogenesis. Development 141: 2885-2894. PubMed ID: 25005477

Tripathy, R., Kunwar, P. S., Sano, H. and Renault, A. D. (2014). Transcriptional regulation of Drosophila gonad formation. Dev Biol 392: 193-208. PubMed ID: 24927896

Veraksa, A., McGinnis, N., Li, X., Mohler, J. and McGinnis, W. (2000). Cap 'n' collar B cooperates with a small Maf subunit to specify pharyngeal development and suppress deformed homeotic function in the Drosophila head. Development 127: 4023–4037. PubMed ID: 10952900

Wingert, L. and DiNardo, S. (2015). Traffic jam functions in a branched pathway from Notch activation to niche cell fate. Development 142: 2268-2277. PubMed ID: 26092848

date revised: 8 May 2022

Home page: The Interactive Fly © 2009 Thomas Brody, Ph.D.

The Interactive Fly resides on the
Society for Developmental Biology's Web server.