InteractiveFly: GeneBrief

uninflatable: Biological Overview | References

BIOLOGICAL OVERVIEW

Cell fate decision during asymmetric division is mediated by the biased partition of cell fate determinants during mitosis. In the case of the asymmetric division of the fly sensory organ precursor cells, directed Notch signaling from pIIb to the pIIa daughter endows pIIa with its distinct fate. Previous studies have shown that Notch/Delta molecules internalized in the mother cell traffic through Smad anchor for receptor activationSmad anchor for receptor activation (Sara) endosomes and are directed to the pIIa daughter. This study shows that the receptor Notch itself is required during the asymmetric targeting of the Sara endosomes to pIIa. Notch binds Uninflatable, and both traffic together through Sara endosomes, which is essential to direct asymmetric endosomes motility and Notch-dependent cell fate assignation. The data uncover a part of the core machinery required for the asymmetric motility of a vesicular structure that is essential for the directed dispatch of Notch signaling molecules during asymmetric mitosis (Loubery, 2014).

The Notch signaling pathway plays multiple roles in organisms ranging from flies and worms to mammals. A powerful model system to elucidate the cell biology of Notch signaling is the Drosophila sensory organs. Each sensory organ precursor (SOP) cell divides asymmetrically to produce a pIIa cell and a pIIb daughter cell, which perform directed Notch signaling: pIIb signals to pIIa. Four independent endocytic mechanisms control asymmetric signaling in the SOP. These include asymmetric endocytic events mediated by the E3 ubiquitin ligase Neuralized, recycling endosomes, and the endocytic adaptors α- and γ-adaptin together with Numb (Loubery, 2014).

During SOP cytokinesis, a fourth mechanism involves a population of endosomes marked by the adaptor protein Sara. Sara endosomes contain as cargo a pool of endocytosed Notch and Delta molecules. Notch and Delta reach the Sara endosome 20 min after their endocytosis in the SOP; this pool is dispatched into pIIa during cytokinesis. In contrast, the pools of Notch in endosomal populations upstream (Rab5 early endosomes) or downstream (Rab7 late endosomes) of Sara endosomes are segregated symmetrically. The specific pool of Notch in Sara endosomes is relevant for signaling: it is cleaved in a ligand- and gamma-secretase-dependent manner to release the transcriptionally active Notch intracellular domain (NICD) in pIIa (Loubery, 2014).

A key question is what machineries control the asymmetric targeting of these endosomes. Is the cargo (the ligand Delta or its receptor Notch) playing a role on the specific targeting of these endosomes? To unravel the machinery regulating the behavior of Sara endosomes during SOP mitosis, candidate factors from previously reported proteomics approaches or genetic screens were tested for Notch signaling. Thus, Uninflatable was identified as a factor involved in the asymmetric dynamics of Sara endosomes (Loubery, 2014).

MARCM homozygous mutant clones were generated for a null allele of Uninflatable (Uif2B7) an the trafficking of Delta, Notch, and the Notch effector Sanpodo through Sara endosomes was monitored. To look at the motility of the endogenous population of Sara endosomes, the cohort of internalized Delta molecules 20 min after its endocytosis was followed in the SOP by means of a pulse-chase antibody uptake assay. Delta, Notch, and Sanpodo traffic normally through Sara endosomes in the absence of Uif, and these endosomes are targeted to the cleavage plane (the central spindle) in cytokinesis (Loubery, 2014).

In Uif mutants or RNAi knockdown conditions, iDl20'/Sara endosomes fail to be asymmetrically dispatched to pIIa after their targeting to the central spindle. These results indicate that Uif is not required to bring Notch to the Sara endosomes or to target the endosomes to the central spindle. However, once in the spindle, Uif is essential for the specific dispatch of Sara endosomes from the spindle into the pIIa cell (Loubery, 2014).

This function of Uninflatable is specific to the asymmetric segregation of Sara endosomes. To gain mechanistic insights into the mechanism of action of Uif, this study has analyzed the density of microtubules in the central spindle and has shown that Uninflatable does not regulate the organization of the microtubular cytoskeleton. In contrast, it was found that Uif controls the residence time of Sara endosomes on the central spindle: in control SOPs, Sara endosomes depart from the central spindle with a decay time of 103 ± 21 s, whereas upon Uif downregulation this decay time goes up to 175 ± 42 s. These data indicate that Uif is not involved in the organization of the spindle, but rather in the motility properties of the endosomes, particularly their last step of departing from the central spindle and end up in pIIa (Loubery, 2014).

Consistent with the role of Uif in the asymmetric targeting of Sara endosomes, Uif contributes to Notch-dependent cell fate assignation in the SOP lineage. To address this, the composition of SOP lineages was examined in homozygous Uif2B7 MARCM clones or upon Uif RNAi. In wild-type animals, the SOP lineage consists of four different cells: two external cells (the shaft and the socket) originating from pIIa and two internal cells (the sheath and the neuron) from pIIb, which can be identified by immunostaining. In Uif mutant clones, instead of a sheath and a neuron per SOP lineage, two sheath cells can be frequently observed in the notum, indicating a symmetric division in the pIIb lineage. Similarly, upon Uif downregulation in the postorbital SOPs, duplications of sockets were observed, which is diagnostic of symmetric divisions in the pIIa lineage. These data uncover a role for Uninflatable in Notch-dependent asymmetric cell fate assignation that is mediated by the asymmetric dispatch of the Sara endosomes (Loubery, 2014).

The Uif phenotype during asymmetric endosomal targeting and cell fate assignation prompted us to look whether Uif is a cargo of Sara endosomes. To detect the endogenous protein, anti-Uif antibodies were generated. To look at Uif trafficking in vivo, transgenic flies were generated expressing a Uif-GFP protein, which can provide activity to rescue the lethality of a Uif lethal mutation at least partly (Loubery, 2014).

Uif-GFP is strongly colocalized with both Sara-GFP and iDelta20'. Since a cargo of Sara endosomes is Notch itself (73% ± 2.7% of the vesicular population of Notch molecules is in Sara endosomes), the presence of Notch cargo was examined in Uif vesicles: 44% ± 4.7% of Uif-positive vesicular structures contain Notch. Therefore, a population of Uninflatable and Notch traffics through Sara endosomes during SOP asymmetric mitosis (Loubery, 2014).

The fact that Uninflatable controls the asymmetric dispatch of the Sara endosomes, which contain internalized Notch and Uninflatable, prompted a look at a possible molecular interaction between Uninflatable and Notch. Uif- and Notch-expressing plasmids were cotransfected in S2 cells and immunoprecipitation experiments were performed by using anti-Uif-coupled beads, followed by immunoblotting with a clean anti-Notch antibody that was purified from a hybridoma cell line (DSHB #C17.9C6). Uif was shown to immunoprecipitate Notch. This coimmunoprecipitation can be reproduced from lysates of S2 cells expressing Notch and Uif tagged with the PC peptide tag and anti-PC-coupled beads; as a control, other transmembrane proteins such as Tkv-GFP are not coimmunoprecipitated with Uif-PC. Together, these results indicate a specific molecular interaction between Notch and Uif (Loubery, 2014).

Uninflatable is a transmembrane protein that, like Notch, contains an array of epidermal growth factor (EGF) repeats. It has been shown that Notch is engaged in protein-protein interactions through its EGF repeats with other factors containing EGF repeats. These include its ligand Delta, but also a number of noncanonical Notch ligands, secreted or membrane proteins lacking the DSL domain characteristic of canonical Notch ligands (Dlk-1, Dlk-2, DNER, Trombospondin, LRP1, EGFL7, and Weary). Consistently, it has recently been reported that a synergistic genetic interaction between Uif and Notch depends on Notch EGF repeats. Therefore, studies were performed to discover which EGF repeats of Uif could be involved in the molecular interaction with Notch. A coimmunoprecipitation experiment was performed in S2 cells coexpressing Notch and an N-terminal, truncated form of Uif tagged with PC (UifΔCter-PC) that lacks the four EGF domains flanking the transmembrane domain but still contains the other 17 EGF repeats and other extracellular domains. While full-length Uif-PC coimmunoprecipitates Notch, UifΔCter-PC does not. This indicates that the interaction between Uif and Notch may be mediated by the four EGF domains of Uif flanking its transmembrane domain (Loubery, 2014).

Although Uif binds and colocalizes with Notch, it does not play a role in core Notch signaling: embryos deprived of maternal and zygotic Uif in germline clones do not show a Notch signaling phenotype, whereas they display loss of inflation of the trachea as previously reported. Consistently, loss of Uif in wing mosaics does not cause a defect of Notch-dependent expression of Wingless at the wing margin. This indicates that Uninflatable interaction with Notch is not essential during core Notch signaling, but rather during the asymmetric dispatch of Notch-containing Sara endosomes during asymmetric cell division. This prompted the possibility that Notch itself is required for the asymmetric motility of the endosomes (Loubery, 2014).

To study whether Notch plays a role during the asymmetric dispatch of Sara endosomes, the trafficking was studied of a Notch-GFP fusion expressed at endogenous levels. The idea was to confirm previous observations using a Notch antibody uptake assay to follow Notch expressed at endogenous levels. To achieve this, a reporter transgenic fly strain was set up in which Notch-GFP fusion is driven by the Notch endogenous promoter and is expressed at endogenous levels. In this fusion, GFP is inserted in the middle of the Notch-intra domain. Since in protein fusions GFP is frequently cleaved out, whether the fusion protein is intact was examined. This would be particularly important in this case, since a cleavage event would lead to a truncated Notch-intra peptide (Loubery, 2014).

In these transgenic Notch-GFP flies, GFP is very efficiently cleaved out (74% of total GFP is cleaved, leading to truncated Notch-intra peptides that can only partially support Notch function and thereby cause a highly penetrant mutant phenotype. This precludes the usage of this reagent as a bona fide marker for Notch. In particular, the cytosolic GFP signal cannot be used as a readout of signaling as previously reported: a nuclear accumulation of the GFP signal in these flies does not solely reflect the accumulation of Notch-intra-GFP, but rather the overall accumulation of different GFP-containing fragments (Loubery, 2014).

Whether, in these conditions, the pool of membrane associated GFP-Notch traffics through Sara endosomes and is asymmetrically dispatched to the pIIa cell was studied. Only 11% ± 1.3% of the total GFP signal in these flies is membrane associated (plasma membrane and intracellular vesicular structures). The rest, representing the vast majority (89%), corresponds to cytosolic and nuclear cleaved GFP (Loubery, 2014).

In Notch-GFP flies, 3.1% of the total GFP signal is associated with intracellular vesicular structures. These correspond to various intracellular vesicular compartments, including Notch in the secretory pathway, as well as in early endosomes, Sara endosomes, recycling endosomes, and late endosomes. To measure the size of the specific pool of Notch in Sara endosomes, a Notch antibody internalization assay was performed, and internalized Notch was chased 20 min after its endocytosis (iNotch20'). As previously established, 73% ± 2.7% of Notch-GFP vesicles are positive for iNotch20'. Of this iNotch20'-positive pool, 79% would be targeted to pIIa . This is consistent with only 65% ± 3.1% of the total pool of Notch-GFP being dispatched to pIIa (Loubery, 2014).

Whether Notch itself plays a role on the asymmetric targeting of Sara endosomes was addressed. Notch was depleated in the SOP by expressing a previously validated Notch dsRNA, and the behavior of Sara endosomes was examined. Upon Notch knockdown in the SOP, iDl20'/Sara endosomes are still targeted to the central spindle, but the subsequent directed dispatch to pIIa is defective. This indicates that Notch itself contributes to the endosomal recruitment of the machinery that endows the Sara endosomes with their asymmetric behavior (Loubery, 2014).

It has been shown that the targeting of Notch to Sara endosomes does not depend on Uninflatable; it was then determined whether the recruitment of Uninflatable on Sara endosomes depended on Notch. Interestingly, it was found that, conversely, the targeting of Uif to Sara endosomes is not controlled by Notch. This implies that these two molecules use different machineries to get to the endosome, where they can interact and are both required for the asymmetric motility of the endosome (Loubery, 2014).

Since the Notch receptor itself is required for the asymmetric targeting of Sara endosomes, it was asked whether Notch signaling plays a role in the process. Notch signaling was blocked by inactivating the ligand Delta through overexpression of Tom in the SOP cell; Tom overexpression leads to inactivation of the Ubiquitin ligase Neuralized and thereby blocks endocytosis-dependent activation of Delta. In the absence of Notch signaling, targeting of Sara endosomes to the central spindle and their asymmetric dispatch to the pIIa cell remains intact. This indicates that although the Notch receptor is essential for the asymmetric targeting of Sara endosomes, Notch signaling is not (Loubery, 2014).

This report has started to unravel the machinery that mediates asymmetric endosome motility during asymmetric cell division. Both Notch and Uninflatable were shown to play a key role in the last step of the asymmetric motility of endosomes: the final, specific stride of the Sara endosomes from the central spindle into the anterior pIIa cell. This is based on the following four key sets of observations (Loubery, 2014).

First, it was confirmed that a functional Notch-GFP fusion expressed at endogenous level does traffic through Sara endosomes, which are indeed dispatched asymmetrically during SOP mitosis. Second, Notch binds Uninflatable, and both colocalize in Sara endosomes. Third, neither Notch nor Uninflatable is essential for the targeting of Notch/Delta/Uif to the Sara endosomes or the targeting of those endosomes to the central spindle, but they are essential for the final dispatch from the central spindle into the pIIa cell. Although Notch is necessary for this process, Notch signaling is not. Fourth, Uninflatable is not an integral component of the Notch signaling pathway, but it plays a role during asymmetric Notch signaling in the SOP, and therefore mutant Uif conditions lead to a lineage identity phenotype. It remains to be elucidated what machineries downstream of Notch/Uninflatable implement the control of the final step toward pIIa and what is asymmetrical in the cytoskeleton so that this final step occurs toward pIIa and not pIIb (Loubery, 2014).

Uninflatable and Notch control the targeting of Sara endosomes during asymmetric division

During asymmetric division, fate assignation in daughter cells is mediated by the partition of determinants from the mother. In the fly sensory organ precursor cell, Notch signalling partitions into the pIIa daughter. Notch and its ligand Delta are endocytosed into Sara endosomes in the mother cell and they are first targeted to the central spindle, where they get distributed asymmetrically to finally be dispatched to pIIa. While the processes of endosomal targeting and asymmetry are starting to be understood, the machineries implicated in the final dispatch to pIIa are unknown. This study shows that Sara binds the PP1c phosphatase and its regulator Sds22. Sara phosphorylation on three specific sites functions as a switch for the dispatch: if not phosphorylated, endosomes are targeted to the spindle and upon phosphorylation of Sara, endosomes detach from the spindle during pIIa targeting (Loubery, 2017).

Asymmetric cell division plays many roles in development. In particular, stem cells divide asymmetrically to self-renew while also forming differentiated cells. Asymmetric cell division involves the specific partitioning of cell fate determinants (RNA, proteins or organelles) in one of the two sibling daughter cells. The Sensory Organ Precursor cells (SOPs) of the Drosophila notum are a model system of choice to unravel the molecular mechanisms of asymmetric cell division (Loubery, 2017).

The division of each SOP gives rise to a pIIa and a pIIb daughter cell and, after two more rounds of asymmetric cell divisions, to the four cells of the sensory organ: the outer cells (shaft and socket) are progeny of the pIIa, while the pIIb forms the inner cells (sheath and neuron) and a glial cell that rapidly undergoes apoptosis. The Notch signalling pathway controls cell fate determination in this system: a signalling bias between the pIIa-pIIb sibling cells is essential to obtain a correct lineage (Loubery, 2017).

The asymmetric dispatch of cell fate determinants during SOP division is governed by the polarity of the dividing cell. The Par complex (composed by the aPKC, Par-3 and Par-6 proteins) is the master regulator of the establishment of this polarity. Downstream the Par complex, Notch signalling is regulated by endocytosis and endosomal trafficking through four independent mechanisms: (1) The E3 Ubiquitin ligase Neuralized is segregated to the pIIb cell, where it induces the endocytosis and thereby the activation of the Notch ligand Delta; (2) Recycling endosomes accumulate in the perinuclear region of the pIIb cell, in which they enhance the recycling and activation of Delta; (3) The endocytic proteins α-adaptin and Numb are segregated to the pIIb cell, where they inhibit the Notch activator Sanpodo; (4) During SOP mitosis, Sara endosomes transport a signalling pool of Notch and Delta to the pIIa cell, where Notch can be activated. Asymmetric Sara endosomes have also been shown to operate in the larval neural stem cells (Coumailleau, 2009) as well as in the adult intestinal stem cells in flies, where they also play a role during asymmetric Notch signalling. In fish, Sara endosomes mediate asymmetric cell fate assignation mediated by Notch during the mitosis of neural precursor of the spinal cord (Loubery, 2017).

Sara endosomes are a subpopulation of Rab5-positive early endosomes characterised by the presence of the endocytic protein Sara. Sara directly binds the lipid phosphatidyl-inositol-3-phosphate and both molecules are found at the surface of these endosomes. A pulse-chase antibody uptake assay has been established to monitor the trafficking of endogenous internalised Notch and Delta and showed that both Notch and Delta traffic through Sara endosomes. Furthermore, it was shown that Sara endosomes are specifically targeted to the pIIa cell during SOP division, mediating thus the transport of a pool of Notch and Delta that contribute to the activation of Notch in the pIIa. The Notch cargo and its Uninflatable binding partner are required for this asymmetric dispatch. Targeting of Sara endosomes to the central spindle is mediated by a plus-end-directed kinesin, Klp98A. The asymmetric distribution of endosomes at the central spindle results from a higher density of microtubules in pIIb with their plus ends pointed towards pIIa15 (Loubery, 2017).

This study shows that the Sara protein itself controls both the targeting and the final dispatch of Sara endosomes to the pIIa daughter cell. Sara binds and is a target of the PP1 phosphatase complex. The phosphorylation state of Sara functions as a switch that enables the targeting of Sara endosomes to the central spindle of the dividing SOP, and their subsequent detachment from the central spindle, which is necessary to allow their movement to the pIIa daughter cell (Loubery, 2017).

Previous work has shown that a subpopulation of Rab5 early endosomes positive for Sara are asymmetrically dispatched into the pIIa daughter cell during cytokinesis of the SOP. This was monitored by following in vivo either GFP-Sara or internalized Delta or Notch, which reach the Sara endosomes 20 min after their endocytosis in the mother cell. These vesicles were termed iDelta^20' endosomes. In contrast, the pools of Notch in endosomal populations upstream or downstream of the Sara endosomes (that is, the Rab5 early endosomes with low Sara levels and the Rab7 late endosomes, respectively) were segregated symmetrically. Rab5 endosomes show different levels of Sara signal: by a progressive targeting of Sara to the Rab5 endosomes, Rab5 early endosomes mature into Sara endosomes. This prompts the question whether the levels of Sara in endosomes correlate indeed with their asymmetric behaviour (Loubery, 2017).

To study the relationship between the levels of Sara in endosomes and their targeting to the spindle, Matlab codes were written to perform automatic 3D-tracking of the Sara endosomes. Sara endosomes were detected by monitoring a GFP-Sara fusion, which was overexpressed through the UAS/Gal4 system. This way, the position of the endosomes, their displacement towards and away from the central spindle was monitored as well as the levels of Sara. In addition, the position was detected automatically of the Pon cortical crescent, which forecasts the side of the cell that will become the pIIb cell (Loubery, 2017).

The localization of endosomes was studied with respect to a 2 μm-wide box centered in the central spindle during SOP mitosis. The enrichment was measured of endosomes in this central spindle as a function of time. Two phases were observed in the movement of the endosomes during mitosis: (1) targeting to the central spindle and (2) departure into the pIIa cell. The endosomes are progressively accumulating in the central spindle area from the end of metaphase (~450 s before abscission) through anaphase and during cytokinesis until they are enriched at the central spindle by about 10-fold at 250 s before abscission (Loubery, 2017).

Subsequently, the endosomes depart from the central spindle area into the pIIa cell. By fitting an exponential decay to the profile of abundance of the endosomes at the central spindle, the characteristic residence time of the endosomes at the central spindle was measured after the recruitment phase: after recruitment, endosomes remain at the central spindle 98±9.8 s before they depart into one of the daughter cells, preferentially the pIIa cell (Loubery, 2017).

To address a potential role of Sara on central spindle targeting and asymmetric segregation, the behaviour was tracked and quantified of the endosomes in a Sara loss of function mutant (Sara¹²) and in conditions of Sara overexpression in the SOP (Neur-Gal4; UAS-GFP-Sara). In Sara¹² SOPs, targeting of iDelta^20' endosomes to the cleavage plane is severely impaired. Consistent with the fact that the asymmetric dispatch of endosomes to pIIa requires first their targeting to the central spindle as previously shown, in Sara¹² SOPs the dispatch to the pIIa daughter is strongly affected. A slight bias (60% pIIa targeting) is, however, retained in the mutant, consistent with a previous report (Loubery, 2017).

Conversely, overexpression of Sara increases targeting to the central spindle. In these conditions, Sara is found not only in Rab5 endosomes, but also in Rab7 late endosomes as well as in the Rab4 recycling endosomes. Correlating with this, Rab4, Rab5 and Rab7 endosomes, which are not all recruited to the central spindle in wild-type conditions, are now targeted to the central spindle upon Sara overexpression and are asymmetrically targeted (Loubery, 2017).

Furthermore, consistent with the correlation that is observed between the levels of Sara at the endosomes and their displacement towards the cleavage plane, quantification of central spindle targeting of the Sara endosomes upon its overexpression shows that targeting of the endosomes to the cleavage plane is increased by a factor of 2.5 in these conditions. These observations indicate that Sara plays a crucial role on the targeting of the endosomes to the spindle and the subsequent dispatch of the Notch/Delta containing endosomes to pIIa. Does this play a role during Notch-dependent asymmetric cell fate assignation? (Loubery, 2017).

Sara function contributes to cell fate assignation through asymmetric Notch signalling, but this activity is redundantly covered by Neuralized. Neuralized E3 Ubiquitin ligase does play an essential role during the endocytosis and activation of the Notch ligand Delta. Therefore, during larval development, Neuralized is essential for Notch-mediated lateral inhibition in the proneural clusters, which leads to the singling-out of SOP cells from the proneural clusters. Later, during pupal development, Neuralized appears as a cortical crescent in the pIIb side of the dividing SOPs, thereby biasing Delta activation in the pIIb cell and asymmetric activation of Notch in pIIa6 (Loubery, 2017).

Consistently, a partial loss of function of Neuralized by RNAi interference in the centre of the notum (Pnr>Neur^RNAi Control) showed lateral inhibition defects in the proneural clusters, causing the appearance of supernumerary SOPs as well as asymmetric Notch signalling defects in the SOP lineage, leading to supernumerary neurons and loss of the external shaft/socket cells in the lineage. The remaining Neuralized activity in this partial loss of function condition allows many sensory organs (more than forty in the centre of the notum) to perform asymmetric cell fate assignation and to develop, as in wild type, into structures containing at least the two external cells (Loubery, 2017).

In Pnr>Neur^RNAi, Sara¹²/Df(2R)48 transheterozygote mutants, the number of supernumerary SOPs is increased by 35% with respect to the Pnr>Neur^RNAi controls (668±38 versus 498±52). This indicates that during lateral inhibition, Sara endosomes contributes to Notch signalling. This general role of Sara is uncovered when the Neuralized activity during Notch signalling is compromised (Loubery, 2017).

In the case of Neuralized, its localization to the anterior cortex biases Notch signalling to be elicited in the pIIa cell. This is the same in the case of Sara endosomes: asymmetric dispatch of Sara endosomes also biases Notch signalling to pIIa10. Indeed, in Pnr>Neur^RNAi, Sara¹²/Df(2R)48 transheterozygote mutants, the number of bristles (external shaft/socket cells) in the notum is strongly reduced at the expense of supernumerary neurons compared to the Pnr>Neur^RNAi controls. This indicates that Notch-dependent asymmetric cell fate assignation in the SOP lineage is synergistically affected in the Sara/Neuralized mutant. This implies that the SOP lineages which still could generate bristles with lower levels of Neuralized function in Pnr>Neur^RNAi need Sara function to perform asymmetric cell fate assignation: in Pnr>Neur^RNAi, Sara¹²/Df(2R)48 and Pnr>Neur^RNAi, Sara¹²/Sara1 transheterozygote mutants, these lineages failed to perform asymmetric signalling, causing the notum to be largely bald. Therefore, Sara contributes to Notch signalling and asymmetric cell fate assignation, as observed in conditions in which other redundant systems for asymmetric Notch signalling are compromised (Loubery, 2017).

Both Neuralized and Sara play general roles in Notch signalling: they are both involved in lateral inhibition at early stages and, at later stages, in asymmetric cell fate assignation. Indeed, both Neuralized and Sara mutants show early defects in lateral inhibition and, accordingly, they show supernumerary SOPs. In addition, Neuralized and Sara mutant conditions also show defective Notch signalling during cell fate assignation in the SOP lineage and therefore cause the transformation of the cells in the lineage into neurons. In this later step, Notch signalling is asymmetric. The possibility that both Sara and Neuralized play key roles in ensuring the asymmetric nature of this signalling event is only correlative: in the case of Neuralized, it is enriched in the anterior cortex of the cell, which will give rise to pIIb; in the case of Sara, (1) both Delta and Notch are cargo of these endosomes, (2) cleaved Notch is seen in the pIIa endosomes and (3) Sara endosomes are dispatched asymmetrically to pIIa10. It is tantalizing to conclude that the asymmetric localization of these two proteins mediate the asymmetric nature of Notch signalling in the SOP lineage, but further assays will be necessary to unambiguously address this issue. Clonal analysis is unfortunately a too slow assay to sort out the specific requirement of these cytosolic factors (Sara and Neuralized) in the pIIa versus the pIIb cell (Loubery, 2017).

Sara mediates the targeting of Notch/Delta containing endosomes to the central spindle and could contributes to Notch-mediated asymmetric signalling in the SOP lineage. What machinery controls in turn the Sara-dependent targeting of endosomes to the central spindle? Previous proteomic studies uncovered bona fide Sara-binding factors, including the Activin pathway R-Smad, Smox17 and the beta subunit of the PP1c serine-threonine phosphatase (PP1β(9C)). In an IP/Mass Spectrometry approach, those interactions were confirmed and in addition to PP1β(9C), two of the other three Drosophila isoforms of PP1c: PP1α(87B) and PP1α(96A) were found. Furthermore, the PP1c regulatory subunit Sds22 was found, suggesting that Sara binds the full serine-threonine PP1 phosphatase complex. The interaction with Sds22 was confirmed by immunoprecipitation of overexpressed Sds22-GFP and western blot detection of endogenous Sara in the immunoprecipitate (Loubery, 2017).

Prompted by these results, whether the PP1 complex plays a role in the asymmetric targeting of the Sara endosomes was explored by manipulating the activity of Sds22, the common regulatory unit in all the complexes containing the different PP1 isoforms. Sds22 was overexpressed specifically during SOP mitosis, by driving Sds22-GFP under the Neur-Gal4 driver with temporal control by the Gal80ts system. In SOPs where PP1-dependent dephosphorylation is enhanced by overexpressing Sds22, the Sara endosomes fail to be dispatched asymmetrically toward the pIIa daughter cell (Loubery, 2017).

The role of PP1-dependent dephosphorylation in the SOP was examined by knocking down Sds22 (through a validated Sds22-RNAi). Loss of function Sds22 did also affect the asymmetric targeting of endosomes. These data uncover a key role for phosphorylation and PP1-dependent dephosphorylation as a switch that contributes to the asymmetric targeting of Sara during asymmetric cell division (Loubery, 2017).

The observations raise the question of which is the step in the asymmetric dispatch of the endosomes that is controlled by the levels of phosphorylation: central spindle targeting, central spindle detachment or targeting to the pIIa cell? PP1/Sds22-dependent dephosphorylation controls a plethora of mitotic events, including mitotic spindle morphogenesis, cortical relaxation in anaphase, epithelial polarity and cell shape, Aurora B activity and kinetochore-microtubule interactions as well as metabolism, protein synthesis, ion pumps and channels. Therefore, to establish the specific event during the asymmetric dispatch of Sara endosomes that is controlled by PP1/Sds22 dephosphorylation, focus was placed on the phosphorylation state of Sara itself and its previously identified phosphorylation sites. This allowed specific interference with this phosphorylation event and thereby untangle it from other cellular events also affected by dephosphorylation (Loubery, 2017).

PP1/Sds22 was shown to bind Sara. It has previously been shown that mammalian Sara itself is phosphorylated at multiple sites and that the level of this Sara phosphorylation is independent on the level of TGF-beta signalling. Three phosphorylation sites have been identified at position S636, at position S709, and at position S774 in Sara protein and these sites were confirmed by Mass Spectrometry of larval tissue expressing GFP-Sara. Phosphorylation of Sara had been previously reported to be implicated in BMP signalling during wing development. However, the role of these three phosphorylation sites during asymmetric division are to date unknown (Loubery, 2017).

ProQ-Diamond phospho-staining of immunoprecipitated GFP-Sara confirmed that Sara is phosphorylated. To test whether PP1/Sds22 controls the phosphorylation state of Sara, ProQ-Diamond stainings of GFP-Sara were performed with and without down-regulation of Sds22. Downregulating Sds22 induced a 40%-increase in the normalized quantity of phosphorylated Sara, showing that PP1/Sds22 does control the phosphorylation state of Sara (Loubery, 2017).

To study the role of Sara phosphorylation during asymmetric targeting of the endosomes, the mitotic behaviour of the endosomes was analyzed in conditions of overexpression of mutant versions of Sara where (1) the three phosphorylated Serines (at position S636, S709, and S774) were substituted by Alanine (phosphorylation defective: GFP-Sara3A) or (2) the PP1 interaction was abolished by an F678A missense mutation in the PP1 binding domain (hyper-phosphorylated: GFP-SaraF678A). Neither mutation affects the general levels of abundance of the Sara protein in SOPs, the targeting of Sara itself to the endosomes, nor the residence time of Sara in endosomes as determined by FRAP experiments. Also, the targeting dynamics of internalized Delta to endosomes are not affected in these mutants (Loubery, 2017).

Upon overexpression of GFP-Sara3A in SOPs, the rate of targeting of the endosomes to the central spindle is greatly increased. In addition, GFP-Sara3A shows impaired departure from the spindle: while the residence time of Sara endosomes at the central spindle after their recruitment is around 100 s in wild type, GFP-Sara3A endosomes stay at the spindle significantly longer (151±21 s). In GFP-Sara3A endosomes, impaired departure leads to defective asymmetric targeting to the pIIa cell while, in wild type, departure from the central spindle occurs well before abscission, in the GFP-Sara3A condition, endosomes that did not depart are caught at the spindle while abscission occurs. These data indicate that the endosomal targeting to the central spindle is greatly favoured when these three sites in Sara are dephosphorylated and suggest that the departure from the microtubules of the central spindle requires that the endosomes are disengaged by phosphorylation of Sara (Loubery, 2017).

Loss of Sara phosphorylation in these sites impairs disengagement from the central spindle. Conversely, impairing Sara binding to the PP1 phosphatase results in defective targeting to the central spindle. Indeed, when binding of Sara to the PP1/Sds22 phosphatase is impaired in the GFP-SaraF678A overexpressing SOP mutants, Sara endosomes fail to be targeted to the spindle. Mistargeted away from the central spindle, the GFP-SaraF678A endosomes fail thereby to be asymmetrically targeted to the pIIa cell. Loss and gain of function phenotypes of the Phosphatase regulator Sds22 during endosomal spindle targeting support the role of Sara phosphorylation during targeting to the central spindle microtubules suggested by the GFP-Sara3A and GFP-SaraF678A experiments (Loubery, 2017).

What are the functional consequences on signalling of impaired phosphorylation/dephosphorylation in Sara mutants? The presence of Sara in endosomes is itself essential for Notch signalling. Sara loss of function mutants show a phenotype in SOP specification (supernumerary SOPs) as well as during fate determination within the SOP lineage (all cells in the lineage acquire a neural fate). In addition, this study showed that Sara is also essential for the targeting of endosomes to the spindle: in the absence of Sara, endosomes fail to move to the spindle in the SOP. They are therefore dispatched symmetrically, but those endosomes do not mediate Notch signalling. As a consequence, both daughters fail to perform Notch signalling in sensitized conditions in which Neuralized is compromised. The result is a Notch loss of function phenotype: the whole lineage differentiates into neurons (Loubery, 2017).

In both Sara3A and SaraF678A mutants, because of reasons that are different in the two cases (either they do not go to the spindle or their departure from the spindle is impaired), functional Sara endosomes are dispatched symmetrically (Fig. 6a,b,e). In contrast to the situation in the Sara loss of function mutant, those endosomes are functional Sara signalling endosomes, which can mediate Notch signalling in both cells. Therefore, these mutations are consistently shown to cause a gain of function Sara signalling phenotype: supernumerary sockets are seen in the lineages (88% of the lineages for Sara3A and 82% of the lineages for SaraF678A). A milder version of this phenotype can be also seen by overexpressing wild-type Sara (34% of the lineages) consistent again with some gain of function Notch signalling phenotype when Sara concentrations are elevated. In summary, this implies that the 3A and F678A mutations impair the phosphorylation state of Sara (with consequences in targeting), but not its function in Notch signalling (Loubery, 2017).

These results indicate that Sara itself plays a key, rate limiting role on the asymmetric targeting of the endosomes by controlling the targeting to the spindle and its departure. Maturation of the early endosomes by accumulating PI(3)P leads to accumulation of the PI(3)P-binding protein Sara to this vesicular compartment. At the endosome, the phosphorylation state of Sara indeed determines central spindle targeting and departure: in its default, dephosphorylated state, Sara is essential to engage the endosomes with the mitotic spindle. Phosphorylation of Sara disengages the endosomes from the central spindle allowing the asymmetric departure into the pIIa cell (Loubery, 2017).

Uif, a large transmembrane protein with EGF-like repeats, can antagonize Notch signaling in Drosophila

Notch signaling is a highly conserved pathway in multi-cellular organisms ranging from flies to humans. It controls a variety of developmental processes by stimulating the expression of its target genes in a highly specific manner both spatially and temporally. The diversity, specificity and sensitivity of the Notch signaling output are regulated at distinct levels, particularly at the level of ligand-receptor interactions. This paper reports that the Drosophila gene uninflatable (uif), which encodes a large transmembrane protein with eighteen EGF-like repeats in its extracellular domain, can antagonize the canonical Notch signaling pathway. Overexpression of Uif or ectopic expression of a neomorphic form of Uif, Uif*, causes Notch signaling defects in both the wing and the sensory organ precursors. Further experiments suggest that ectopic expression of Uif* inhibits Notch signaling in cis and acts at a step that is dependent on the extracellular domain of Notch. these results suggest that Uif can alter the accessibility of the Notch extracellular domain to its ligands during Notch activation. This study shows that Uif can modulate Notch activity, illustrating the importance of a delicate regulation of this signaling pathway for normal patterning (Xie, 2012).

The canonical Notch signaling pathway is one of a limited group of pathway modules that transduce signals from outside the cell to alter gene expression inside the nucleus. These pathways together orchestrate the developmental processes that can be dauntingly complex. Yet it is the same modules that are used repeatedly, not only in different organisms, but also in vastly different processes within an organism. Thus, how these pathway modules are activated in a specific manner, with regard to not only space and time but also the quantity of their signaling output, represents a fundamental question in developmental biology. This study describes a newly characterized protein, Uif, which can antagonize the canonical Notch signaling pathway in a neomorphic manner. These findings underscore the importance of the precise tuning of Notch activity in normal patterning (Xie, 2012).

EGF-like repeats are a common feature of Notch receptors, ligands and co-ligands. While Uif was originally characterized for its role in tracheal development, its EGF-like repeats suggest a possible role in Notch signaling. The current results are consistent with a model where ectopically expressed Uif may modulate the accessibility of the extracellular domain of Notch to its ligands during activation. It is possible that the EGF-like repeats of Uif directly interact with the extracellular domain of Notch to exert its inhibitory effect in a manner similar to the cis inhibition by Notch ligands themselves. The finding that Uif* acts on Notch through a cis inhibitory mechanism is supportive of this possibility. In the current experiments, Uif* is more effective than wt Uif in antagonizing Notch, and this difference may be attributed to the difference in their expression levels. These results suggest that ectopically expressed Uif* and wt Uif have a similar neomorphic function in regulating Notch signaling (Xie, 2012).

A proposed neomorphic function of Uif* and Uif in Notch signaling is consistent with the results of loss of function analysis of uif. Knockdown (assayed for adult wing phenotypes and Notch target gene expression using independent RNAi lines) or knockout (assayed for Notch target gene expression in somatic mutant clones) of uif revealed neither Notch loss of function nor gain of function phenotypes. However, it remains formally possible that the endogenous uif gene has a native role in regulating Notch signaling in tissues or cells (other than those examined in this study) at a time during Drosophila development. Further studies are required to investigate this possibility (Xie, 2012).

The biological activities of Uif are not restricted to regulating Notch signaling. The fact that Uif was originally characterized for its role in tracheal inflation underscores the complexity of its biological activities. In addition to the EGF-like repeats, Uif also contains several other domains that may have important biological functions. These domains include a C-type lectin-like (CLECT) domain, three CUB domains, eight complement control protein (CCP) domains, two coagulation factor 5/8 C-terminal (FA58C) domains and three hyaline repeat (HYR) domains. Both CLECT and FA58C domains are putative carbohydrate binding domains known to play important roles in many diverse processes. The CUB domain is an evolutionary conserved protein domain found almost exclusively in extracellular and plasma membrane-associated proteins. HYR is an immunoglobulin fold domain likely involved in cell adhesion. The CCP domains, also known as the Sushi domains or Short Consensus Repeats (SCR), exist in a wide variety of complement and adhesion proteins. These domains suggest that Uif may also play a role in cell adhesion. Indeed, in a recent genetic modifier screen, uif was identified as a regulator (Mod29) of the Drosophila Dystroglycan-Dystrophin Complex, a specialized cell adhesion complex (Kucherenko, 2008). Mod29/Uif was suggested to play roles in multiple developmental processes, including wing vein formation, muscle and photoreceptor axon development, and oogenesis (Kucherenko, 2008). Although it remains to be investigated whether Uif, a large regulator with multiple conserved protein domains, may functionally connect distinct cellular processes, unpublished data offer some speculative insights. In particular, the blistering wing phenotype caused by knockdown of Dl or Ser can be fully rescued by depletion of uif, suggesting that Uif may functionally extend the role of Notch ligands to cell adhesion. Uif is an N-glycosylated protein, a modification shared by several proteins known to play a role in the formation of large protein complexes. Understanding the full spectrum of the biological functions of Uif during development and, importantly, its potential role in harmonizing different cellular processes, represents future challenges (Xie, 2012).

uninflatable encodes a novel ectodermal apical surface protein required for tracheal inflation in Drosophila

The tracheal system of Drosophila has proven to be an excellent model system for studying the development of branched tubular organs. Mechanisms regulating the patterning and initial maturation of the tracheal system have been largely worked out, yet important questions remain regarding how the mature tubes inflate with air at the end of embryogenesis, and how the tracheal system grows in response to the oxygen needs of a developing larva that increases nearly 1000-fold in volume over a four day period. This study describes the cloning and characterization of uninflatable (uif), a gene that encodes a large transmembrane protein containing carbohydrate binding and cell signaling motifs in its extracellular domain. Uif is highly conserved in insect species, but does not appear to have a true ortholog in vertebrate species. uif is expressed zygotically beginning in stage 5 embryos, and Uif protein localizes to the apical plasma membrane in all ectodermally derived epithelia, most notably in the tracheal system. uif mutant animals show defects in tracheal inflation at the end of embryogenesis, and die primarily as larvae. Tracheal tubes in mutant larvae are often crushed or twisted, although tracheal patterning and maturation appear normal during embryogenesis. uif mutant larvae also show defects in tracheal growth and molting of their tracheal cuticle (Zhang, 2009).

This study has described the initial characterization of a novel gene, uninflatable, which encodes a single-pass type I transmembrane protein with several different carbohydrate binding motifs and epidermal growth factor repeats in its extracellular domain. This gene is highly conserved in insect species, but does not appear to have a true ortholog in mammals or other vertebrate organisms. Uninflatable is expressed on the apical surface of ectodermally-derived epithelial cells including the epidermis, trachea, salivary gland and fore- and hindgut, although it is most highly expressed in tracheal epithelia starting as the tracheal placode invaginates and persisting throughout its development. Three mutant alleles of uif were isolated, and the predominant defects associated with loss of uif were found to be defective air inflation at the end of embryogenesis, and tracheal growth and tracheal molting defects in larvae (Zhang, 2009).

The most obvious defects observed in uif mutant late embryos and newly hatched first instar larvae were incomplete inflation of the trachea coupled with tracheal tubes that often appeared pinched or stretched. Live imaging of individual uif mutant embryos and larvae revealed that these tracheal inflation defects result from the inability to completely inflate the trachea at the end of embryogenesis rather than a defect in the maintenance of an inflated tracheal system after hatching. Interestingly, tracheae in uif mutant embryos are normally patterned and appear to mature in a manner indistinguishable to tracheae in wild type animals. Thus this tracheal inflation defect likely results from a requirement for uif function late during embryogenesis (Zhang, 2009).

Several genes have been characterized whose defects include incomplete tracheal inflation. Included in these genes are those that function at the end of embryogenesis to clear the trachea of solid luminal material. This process requires clathrin-mediated endocytosis and includes the proteins clathrin heavy chain, the GTPase dynamin (encoded by shibire), and a clathrin binding transmembrane protein encoded by wurst. Notably, mutations in all of these genes also result in elongated tracheal tubes suggesting a defect in tracheal tube size control. Gas filling defects were also observed for mutations in serp and verm, two putative matrix chitin deacytlases that function to regulate tracheal tube length. Tracheal maturation appears normal in uif mutant embryos, including the secretion of chitin into the tracheal lumen and the subsequent uptake of luminal solids near the end of embryogenesis. In addition, uif mutant tracheae do not show diameter or length defects during embryogenesis, suggesting that the luminal chitin cylinder forms normally and that Serp and Verm are correctly functioning during the maturation process. Taken together these results suggest that the air filling defects observed in uif mutant embryos occurs through a mechanism independent of that linked to endocytosis of solid luminal material (Zhang, 2009).

A second class of proteins that likely function in proper air inflation is the epithelial sodium channels (ENaCs). In mammals, ENaC proteins are required to help remove liquid from embryonic lung prior to birth. There are 16 ENaC genes (also referred to as pickpocket or ppk genes) in the Drosophila genome, of which 9 are expressed in the embryonic tracheal system. It is thought that the influx of sodium through these channels drives water from the lumen into the epithelial cells, and that degassing of this liquid inflates the trachea. Although there are no loss of function mutations in ENaCs that result in air filling defects in Drosophila, RNA interference mediated knockdown of ppk4 and ppk11 results in partially liquid-filled tracheae in larvae. In addition, inhibiting these channels using amiloride also results in fluid filled tracheae in larvae. Although both of these examples affected air filling after a molt, it seems likely that ENaC proteins may also contribute to air filling at the end of embryogenesis. At this point it is not possible to exclude the possibility that uif may regulate the expression or activity of ENaC encoded proteins in tracheal epithelia (Zhang, 2009).

A third possibility, and one that is favored, is that uif is playing primarily a structural role in embryonic tracheal maturation. Two pieces of evidence support this notion. First, in early stage 17 mutant embryos the tracheae are of normal length and diameter and have a stereotypic appearance that is indistinguishable from that of wild type animals, whereas in newly hatched mutant larvae the tubes are often crushed or twisted. Second, while dissecting tracheae from wild type and uif mutant third instar larvae it was clear that the mutant tracheae were more brittle. Wild type tracheae have an elastic property that makes them difficult to break, whereas we had to be very careful dissecting uif tracheae in order to get a section that included more than one metameric unit. Together these observations suggested a model in which the mechanical properties of uif mutant tracheae are compromised, thereby allowing the tracheal cuticle to fail when the embryo initiates violent muscular contractions prior to hatching. As noted in the live imaging, tracheal inflation initiates well after the embryo begins these dramatic muscular contractions. It is predicted that a combination of crushed tubes and small cracks in the tracheal cuticle prevent complete inflation (Zhang, 2009).

uif mutant larvae that survive to second or third instar show striking defects in tracheal growth and tracheal molting. The tracheae in these uif mutant larvae have short dorsal trunks that are well out of proportion to the body length of the animal. It is difficult to accurately assess the diameter of the dorsal trunk in these animals, however, because the mutant animals fail to completely shed their tracheal cuticle, and often only the first instar lumen is inflated. This tracheal molting defect was nearly completely penetrant in all third instar uif^2B7 and uif^1A15 mutant larvae examined, but the severity seemed to vary along the anterior-posterior axis, with some middle and anterior sections showing no molting defects, whereas the sections near the posterior spiracles showed at least two and sometimes three tracheal cuticles. A previous study carried out morphometric analyses of larval tracheal growth and observed that tracheae lengthen in a continuous fashion, whereas the diameter increases in a stepwise manner coincident with the molt, suggesting that tube length is controlled independently from tube diameter. The current results indicate that uif plays a role in regulating the growth of the tracheae along their length (Zhang, 2009).

How might uif regulate larval tracheal tube length? One possibility is that Uif might regulate the interface between the tracheal epithelium and the cuticle, possibly by providing lubrication through its hyalin domains. Loss of uif would therefore result in a situation where the epidermis is too tightly bound to the overlying cuticle to allow for growth along the long axis of the tracheae. This type of mechanism could also account for the tracheal molting defects and thereby couple these phenotypes. An earlier study identified mutations in the Matrix Metalloproteinase encoded by Mmp1; similar defects were found in tracheal elongation. It was speculated that the Mmp1 tracheal defects might be caused by the inability of the tracheal epithelial cells to loosen their attachment to the cuticle. Since uif^2B7 mutant larvae have no full length Uif protein, it is unlikely that Uif is a primary target of MMP1 in tracheal cells. In addition, this study observed no genetic interaction between uif and Mmp1 by second-site noncomplementation (Zhang, 2009).

An alternative hypothesis is that Uif might serve as a receptor or co-receptor for a systemic signal that couples tracheal growth with larval growth. The extracellular domain of Uif contains multiple EGF domains and a laminin G domain, and the cytoplasmic domain is remarkably conserved in all insect species. A combination of structure/function analysis and genetic and biochemical approaches to identify Uif interacting proteins should help to shed light on this functions of uif. Understanding how Uif couples tracheal growth to the growth of the larva may serve as an important paradigm for similar couplings of organ growth to organismal growth in other species (Zhang, 2009).

Finally, hypoxia may account for all the other larval phenotypes associated with loss of uif, including early larval lethality, slow growth, developmental arrest and failure to pupariate. Most uif mutant larvae die as first instars, and those that die early almost invariably have the most severe tracheal inflation defects. uif mutant larvae that survive to second or third instar grow much slower than their heterozygous siblings. btl>uifRNAi larvae that have only lost uif function in their tracheae show an identical slow growth phenotype. This study consistently observed uif mutant larvae wandering away from food, suggesting that they were experiencing hypoxia. Not surprisingly these larvae accumulate less fat and have a transparent appearance. Examination of the tracheae in these mutant animals revealed that the inflated portion of the trachea was often just through the first instar tracheal lumen, and therefore these animals were likely oxygen starved as well. Consistent with this notion, additional tracheal branching was observed in these mutant animals suggesting that the hypoxia induced factor pathway had been engaged. Thus, hypoxia could explain the growth defects observed in uif mutant larvae. These growth defects may have then resulted in the observed developmental delays. For example, mutant third instar larvae may not have reached a critical weight threshold needed for pupariation, and thus would not have been able to pupariate even if they experienced the metamorphic pulse of ecdysone. Interestingly, many uif mutant animals even failed to advance to third instar, but this did not reflect a defect in epidermal molting or ecdysis, as no evidence was found for an extra set of head skeleton or epidermal cuticle. Rather the animals just arrested as first or second instars. It is possible that hypoxia-induced growth defects contributed to this phenotype as well, since larvae have to attain a critical size to be competent for molting, just as they do for pupariation (Zhang, 2009).

Search PubMed for articles about Drosophila Uninflatable

Kucherenko, M. M., Pantoja, M., Yatsenko, A. S., Shcherbata, H. R., Fischer, K. A., Maksymiv, D. V., Chernyk, Y. I. and Ruohola-Baker, H. (2008). Genetic modifier screens reveal new components that interact with the Drosophila dystroglycan-dystrophin complex. PLoS One 3: e2418. PubMed ID: 18545683

Loubery, S., Seum, C., Moraleda, A., Daeden, A., Furthauer, M. and Gonzalez-Gaitan, M. (2014). Uninflatable and Notch control the targeting of Sara endosomes during asymmetric division. Curr Biol 24: 2142-2148. PubMed ID: 25155514

Loubery, S., Daeden, A., Seum, C., Holtzer, L., Moraleda, A., Damond, N., Derivery, E., Schmidt, T. and Gonzalez-Gaitan, M. (2017). Sara phosphorylation state controls the dispatch of endosomes from the central spindle during asymmetric division. Nat Commun 8: 15285. PubMed ID: 28585564

Xie, G., Zhang, H., Du, G., Huang, Q., Liang, X., Ma, J. and Jiao, R. (2012). Uif, a large transmembrane protein with EGF-like repeats, can antagonize Notch signaling in Drosophila. PLoS One 7: e36362. PubMed ID: 22558447

Zhang, L. and Ward, R. E. t. (2009). uninflatable encodes a novel ectodermal apical surface protein required for tracheal inflation in Drosophila. Dev Biol 336: 201-212. PubMed ID: 19818339

date revised: 28 December 2015

Home page: The Interactive Fly © 2011 Thomas Brody, Ph.D.