InteractiveFly: GeneBrief

liquid facets-Related: Biological Overview | References

Epsin and epsin-Related (epsinR) are multi-modular proteins that stimulate clathrin-coated vesicle formation. Epsin promotes endocytosis at the plasma membrane, and epsinR functions at the Golgi and early endosomes for trans-Golgi network/endosome vesicle trafficking. In Drosophila, endocytic epsin is known as Liquid facets, and it is essential specifically for Notch signaling. By generating and analyzing loss-of-function mutants in the liquid facets-Related (lqfR) gene of Drosophila, this study investigated the function of Golgi epsin in a multicellular context. LqfR was found to be a Golgi protein, and like liquid facets, lqfR is essential for Drosophila viability. In addition, primarily by analyzing mutant eye discs, it was found that lqfR is required for cell proliferation, insulin-independent cell growth, and cell patterning, consistent with a role in one or several signaling pathways. Epsins in all organisms share an ENTH (epsin N terminal homology) domain, which binds phosphoinositides enriched at the plasma membrane or the Golgi membrane. The epsinR ENTH domain is also the recognition element for particular cargos. By generating wild-type and mutant lqfR transgenes, it was found that all apparent LqfR functions are independent of its ENTH domain. These results suggest that LqfR transports specific cargo critical to one or more signaling pathways, and lays the foundation for identifying those proteins (Lee, 2009).

Epsins are multi-modular membrane-associated proteins that function in endosome trafficking in yeast and metazoans. The distinctive feature of all epsins is an ENTH domain that binds membrane phosphoinositides. There are two classes of epsins: endocytic epsins, known as vertebrate epsin-1, yeast ent1p and ent2p, and Liquid facets (Lqf) in Drosophila and nematodes, and Golgi-associated epsins, known as yeast ent3p and ent5p (Duncan, 2003a), and vertebrate epsin-Related (epsinR), also known as enthoprotin or Clint (Kalthoff, 2002; Wasiak, 2002; Hirst, 2003; Mills, 2003). Endocytic or Golgi epsin ENTH domains prefer to bind the phosophoinositides enriched in the plasma membrane and Golgi membranes, respectively (Duncan, 2003b; Legendre-Guillemin, 2004). Both endocytic and Golgi epsins have a variety of motifs C-terminal to their ENTH domains. Endocytic epsins have motifs for interaction with ubiquitin, clathrin, the clathrin adapter complex AP-2, and EH-domain containing endocytic factors. Golgi epsins have clathrin-binding motifs, and also motifs for binding the Golgi-associated clathrin adapter proteins AP-1 and Gga (Lee, 2009).

Endocytic epsins have been studied more intensively than Golgi epsins. Most of the available data supports a model where endocytic epsin promotes clathrin-dependent endocytosis, acting either as a clathrin adapter, or as an accessory factor for the AP-2 adapter complex. As a clathrin adapter, through its UIMs, epsin binds transmembrane proteins that use ubiquitin as an internalization signal, and recruits clathrin and other endocytic factors to the plasma membrane. As an accessory factor, in order to facilitate internalization of transmembrane proteins whose endocytic signals are amino acid motifs in their intracellular domains that bind the AP-2 adapter, epsin bound to AP-2 would bring clathrin and other proteins to the plasma membrane. Epsin's ENTH domain may also promote vesicle formation by inducing membrane curvature. Yeast epsin ENTH domains also coordinate actin cytoskeleton rearrangement with endocytosis. In yeast and also in vertebrate cell culture, endocytic epsin functions in internalization of a variety of different cargos. Although this is probably also the case in Drosophila, the only apparent requirement for Lqf is for endocytosis of Notch ligands, which is essential for Notch receptor activation. Thus at least in Drosophila, endocytic epsin plays a pivotal role in Notch signaling, and epsin is therefore critical for virtually all aspects of cell determination and differentiation during development (Lee, 2009).

Golgi epsins promote vesicular trafficking mainly between endosomes and the trans-Golgi network (TGN)(reviewed in Duncan, 2003b; Legendre-Guillemin, 2004). In higher organisms, epsinR promotes clathrin-coated vesicle formation and trafficking between the TGN and early endosomes, in both directions. One type of cargo known to be transported in an epsinR-dependent manner from the TGN to the early endosome is lysosomal proteins bound to mannose-6-phosphate receptors on their way to the lysosome. Also, epsinR-dependent retrograde trafficking from the early endosome to the TGN retrieves mannose-6-phosphate receptors, and also other resident Golgi membrane proteins (Mills, 2003; see also Hirst, 2003). Yeast ent3p and ent5p are required for trafficking of carboxypeptidase S from the TGN to the vacuole, where it is processed into active form, and also for endosome-to-TGN transport of Kex2p, a protease required for α-factor mating pheromone maturation (Duncan, 2003). Another type of Golgi epsin cargo in both yeast and higher eukaryotes is SNARE proteins. In yeast, SNAREs required for vesicle fusion at late endosomes are transported from the TGN to the late endosome in an ent3p-dependent manner (Chidambaram, 2008). In mammalian cells, although the biological rationale for this is unclear, SNAREs that function at late endosomes appear to depend on epsinR for early endosome-to-TGN transport (Chidambaram, 2008). Yeast ent3p and ent5p also functions in sorting proteins within the multi-vesicular body, the late endosome whose internal vesicles eventually fuse with the vacuole. Like yeast endocytic epsins, ent3p is also a factor in actin cytokeletal organization (Lee, 2009 and references therein).

The mechanism of Golgi epsin function is probably similar to that of endocytic epsin (Duncan, 2003b; Legendre-Guillemin, 2004), with at least one notable difference. EpsinR is likely the clathrin adapter for SNARE cargo, but unlike endocytic epsin which recognizes ubiquitin internalization signals via its UIMs, EpsinR binds SNARES directly with its ENTH domain (Hirst, 2004; Chidambaram, 2004; Miller, 2007). In the transport of other cargos, EpsinR and also ent3p/ent5p may function either as the key clathrin adapter, or as accessory proteins for the AP-1 clathrin adapter complex or Gga adapters (Lee, 2009).

No analysis of Golgi epsin function in a developmental context has been reported in a multicellular organism. Only through genetic characterization of the Drosophila endocytic epsin gene, lqf, was it revealed that endocytic epsin plays a critical and specific role in Notch signaling. This study examined whether Drosophila Golgi epsin is also essential, and if it functions in cell patterning. This paper reports an analysis of the mutant phenotype of flies with weak and strong loss-of-function alleles of the Golgi epsin gene, which is called liquid facets-Related (lqfR). It was found that lqfR⁺ is indeed essential for Drosophila viability, and that more specifically, lqfR⁺ is required for cell proliferation, insulin-independent cell growth, and cell fate determination in the developing eye. In addition, a transgene was used to generate flies that express only LqfR protein specifically lacking the ENTH domain. It was found that all of the functions of lqfR⁺ that were detected are independent of the ENTH domain (Lee, 2009).

Endocytic epsin (Lqf) functions without its ENTH domain, but overexpression of the ENTH-less protein may be required. This study found that even when expressed by its own promoter, an ENTH-less lqfR gene can substitute for all apparent functions of endogenous lqfR. Like Lqf, LqfR is able to localize to the appropriate (Golgi) membrane without the ENTH domain. Also, if the LqfR ENTH domain induces curvature of clathrin-coated vesicles as proposed for the ENTH domain of endocytic epsin, this function is not essential to the major role of LqfR in Drosophila. The Golgi epsin ENTH domain also serves as the recognition element for SNARE cargos (Miller, 2007), and thus it can be inferred that SNARES are not the cargo relevant to the critical function of LqfR (Lee, 2009).

It is thought that lqfR is likely required for cell signaling because signaling pathways control both cell proliferation/growth and patterning, and in the absence of lqfR, there are defects in both of these aspects of cell behavior. Also, the observation that expression levels of the transcription factor Rough are altered in lqfR mutants is consistent with a signaling pathway defect. Virtually all signaling pathways have an endosomal component, and moreover, Wnt signaling is known to require endosome-to-TGN trafficking specifically. As a single signaling pathway may control cell proliferation and growth and also patterning, it seems possible, although by no means necessary, that all aspects of the lqfR mutant phenotype could be due to misregulation of a single pathway (Lee, 2009).

Wnt signaling (Wingless [Wg] in Drosophila) requires the Golgi transmembrane protein Wntless. Wntless is a Golgi protein that promotes secretion of the ligand Wg, and which accompanies Wg to the plasma membrane. Wntless is subsequently endocytosed, and sorted for recycling back to the TGN through the action of the retromer complex. Protein sorting by retromer in the early endosome may be followed by epsinR-dependent clathrin-coated vesicle formation (Popoff, 2007). Thus, it seems possible that along with Wntless and the retromer proteins, LqfR might promote Wg signaling (Lee, 2009).

Can the lqfR mutant eye phenotype be explained by wg loss-of-function? Early in eye disc development, Wg and Hh signaling collaborate in the dorsal half of the eye disc to set up the D/V axis. Later, by diffusing from the dorsal and ventral disc margins, Wg organizes the gradients of Dachsous and Four-jointed proteins that in part define planar cell polarity in the eye. Dampening of Wg signaling could lead to patterning disruptions at the equator in weak lqfR mutants. Wg emanating from the dorsal and ventral margins also positions the morphogenetic furrow centrally by repressing expression of decapentaplegic (dpp), and promotes proliferation of cells anterior to the morphogenetic furrow that will become head capsule. Reduction of Wg activity in the eye disc results in initiation of ectopic furrows at the lateral margins of the eye disc, and a larger eye at the expense of head tissue. Essentially the opposite is observed in lqfR mutant eyes: halting of the furrow and a tiny eye. Also, the cell size defect observed in adult eyes is cell autonomous. Thus, some, but not all aspects of the lqfR eye phenotype can be explained easily by weak Wg signaling alone (Lee, 2009).

One hypothesis that may explain most features of the lqfR mutant eye phenotype is overactive Hh signaling. The key observation in support of this idea is that rough is overexpressed in lqfR eye discs. Excessive Hh signaling results in arrest of the morphogenetic furrow through overexpression of rough. Morphogenetic furrow progression is controlled mainly by Hh and Dpp. Posterior to the furrow, differentiating R-cells express Hh, which activates Dpp expression, and both ligands diffuse anteriorly. In cells anterior to the furrow, Dpp arrests the cell cycle in G1, and Hh subsequently initiates R-cell differentiation. Hh signaling is able to arrest the cell cycle also, but as Dpp diffuses faster than Hh, Dpp normally plays this role. As the R-cells differentiate, they start to express Hh, and this cycle moves the furrow forward. Anterior to the furrow, through Notch, Hh activates expression of the proneural transcription factor Atonal, which is required for subsequent Hh expression by R-cells posterior to the furrow. Rough expression is also activated by Hh, and it blocks Atonal expression in some cells. The ro^DOM allele appears to be hypersensitive to Hh activity, and thus Rough is overexpressed and blocks Atonal in too many cells, leading to loss of Hh expression posterior to the furrow, and the furrow stops. Thus, overactivity of Hh in lqfR discs could result in rough overexpression and halting of the furrow. Third, the D/V axis in the eye disc is defined by a stripe of cells in which Notch is activated. Setting up of the Notch stripe requires Hh, which is expressed only dorsally and defines the dorsal half of the eye disc. Hh overactivity sometimes results in enlargement of the dorsal area of the eye, but not always, and as in lqfR, the D/V boundary is still present. As Hh can negatively regulate the cell cycle anterior to the furrow, it seems that Hh overactivity could account for the cell proliferation defect (Lee, 2009).

By what mechanism could LqfR negatively regulate Hh signaling? Hh signaling is transduced by the transmembrane protein Smoothened (Smo), which is negatively regulated (through mechanisms that are not entirely clear) by the Hh receptor Patched (Ptc) when it is not bound to Hh. Binding of Hh to Ptc relieves negative regulation of Smo. Hh/Ptc induces Smo phosphorylation, thereby inducing Smo to switch into active conformation. Hh/Ptc also binds the glypican Dally-like (Dlp), which serves as an endocytosis signal. Hh/Ptc endocytosis results in Smo translocation to the plasma membrane and signaling. LqfR could regulate Smo negatively through recycling of Ptc to the plasma membrane, or by promoting the transport of Smo or Dlp away from the plasma membrane, possibly to the lysosome instead. This role for LqfR would be similar to that of yeast ent3p/ent5p in the multi-vesicular body (Lee, 2009).

Further experiments are required to determine if LqfR plays any role at all in Wntless endosome-to-TGN recycling, in Hh signaling, or in one or several other pathways. The mutants and phenotypic characterization described here suggest strongly that like Lqf, LqfR is involved in signaling, and provide the tools for testing these and other hypotheses (Lee, 2009).

Liquid facets-related (lqfR) is required for egg chamber morphogenesis during Drosophila oogenesis

Clathrin interactor 1 [CLINT1] (also called enthoprotin/EpsinR) is an Epsin N-terminal homology (ENTH) domain-containing adaptor protein that functions in anterograde and retrograde clathrin-mediated trafficking between the trans-Golgi network and the endosome. Removal of both Saccharomyces cerevisiae homologs, Ent3p and Ent5p, result in yeast that are viable, but that display a cold-sensitive growth phenotype and mistrafficking of various vacuolar proteins. Similarly, either knock-down or overexpression of vertebrate CLINT1 in cell culture causes mistrafficking of proteins. This study has characterized Drosophila CLINT1, liquid-facets Related (lqfR). LqfR is ubiquitously expressed throughout development and is localized to the Golgi and endosome. Strong hypomorphic mutants generated by imprecise P-element excision exhibit extra macrochaetae, rough eyes and are female sterile. Although essentially no eggs are laid, the ovaries do contain late-stage egg chambers that exhibit abnormal morphology. Germline clones reveal that LqfR expression in the somatic follicle cells is sufficient to rescue the oogenesis defects. Clones of mutant lqfR follicle cells have a decreased cell size consistent with a downregulation of Akt1. While total Akt1 levels were found to be increased, there is also a significant decrease in activated phosphorylated Akt1. Taken together, these results show that LqfR function is required to regulate follicle cell size and signaling during Drosophila oogenesis (Leventis, 2011).

A strong hypomophic mutant of Drosophila CLINT1 (liquid facets-Related), lqfR^D66, displays defects in the development of the eye and macrochaetae. These phenotypes are highly variable and are dependent on genetic background. In contrast, lqfR^D66 homozygous mutants are 100% female sterile. They have defects in oogenesis in the eggshell and dorsal appendages, a cell-non-autonomous defect in nurse cell dumping, and impaired cell growth of the follicle cells. The results further show that all of these phenotypes are due to a requirement for LqfR in somatic follicle cells during oogenesis (Leventis, 2011).

The lqfR gene structure is somewhat different than what is seen in other species. The final 924 aa of the larger product, which is encoded by exon 6, is homologous to the S. cerevisiae telomere length regulating protein Tel2p and the related Caenorhabditis elegans CLK-2, which affects lifespan and may also play a role in regulating telomere length. While a similar genetic structure of lqfR with the putative telomere binding exon is found in other Drosophilids, the CLINT1 and Tel2p homologs are found on different parts of the genome in S. cerevisiae, C. elegans and in vertebrates. The nature of this protein is unclear. It is predicted to share the N-terminal 492 aa with the shorter LqfR protein and was recognized by anti-LqfR antibodies. By immunofluorescence no significant nuclear staining was observed, as might be expected with a telomere binding protein. However, a functional GFP fusion of C. elegans CLK-2 has a cytoplasmic localization Benard (2001). Determining the function and significance of this larger product will require further study (Leventis, 2011).

The results with respect to the shorter lqfR product, which is predicted to be like vertebrate CLINT1, are consistent with its role in TGN-endosome trafficking. It has the expected domain structure to allow it to interact with other components of the TGN-endosome machinery and GST-LqfR did indeed bind to clathrin, γ-adaptin and GGA. Although the antibodies recognized both the short and long products, LqfR was ubiquitously expressed throughout development and co-localized with appropriate TGN and endosome markers (Leventis, 2011).

The lqfR^D66 mutant, which behaves genetically as a strong hypomorph and affects both products, displayed a wide variety of phenotypes which together suggest that LqfR has pleiotropic effects. The mutants were not lethal, although homozygotes were produced in fewer numbers than would be predicted by Mendelian ratios. During these studies a null mutant of lqfR was characterized (Lee, 2009) which was homozygous lethal as third instar larvae, further supporting the lqfR^D66 mutant as a strong hypomorph. Generation of a strong hypomorph that can produce homozygous adults has allowed the characterization of the essential role of lqfR during oogenesis (Leventis, 2011).

Since LqfR likely functions in clathrin-dependent TGN-endosome trafficking, the fact that a severe hypomorphic mutation is still viable suggests several mutually compatible possibilities. First, the reduction of the amount of protein produced in lqfR^D66 homozygotes is able to maintain levels of trafficking at levels sufficient to prevent cell death. Second, another protein or proteins could be functionally redundant with LqfR. Third, proteins dependent upon LqfR for trafficking may reach their appropriate destination by way of a different pathway. This could either be directly, such as via the AP-3 pathway, or indirectly perhaps via the secretory pathway to the plasma membrane and then to the endosome and lysosome via endocytosis (Leventis, 2011).

Trafficking of lysosomal proteins via the cell surface may explain some of the observed phenotypes. This shunting to the plasma membrane occurs in S. cerevisiae missing both homologs of LqfR. When ent3Δent5Δ yeast are grown on plates containing milk, the mistrafficking of a portion of vacuolar enzymes causes a breakdown of milk proteins surrounding the yeast cells. Similarly, in vertebrate cell culture either knockdown by RNAi or overexpression of CLINT1 leads to a large increase in the amount of the lysosomal protein cathepsin D at the plasma membrane (Hirst, 2003; Mills, 2003; Leventis, 2011 and references therein).

If in the lqfR^D66 mutants lysosomal proteases were at the cell surface or secreted in small amounts due to mistrafficking it is possible that they could disrupt the chorionic structure. Mutants that exhibit similar defects to lqfR^D66 in the morphology of the dorsal appendages show defects in the ultrastructural organization of the eggshell. Two of the mutants, humpty dumpty and Origin recognition complex subunit 2, disrupt genes encoding nuclear proteins that promote gene amplification of the chorion genes. The products of the chorion genes are secreted by the follicle cells to create the eggshell, including the dorsal appendages. A defect in the chorion is consistent with both the collapsed dorsal appendages seen in the stage 14 egg chambers and the flaccid oocytes that are laid (Leventis, 2011).

There were several phenotypes in the nurse cells and oocytes of lqfR^D66 mutant ovaries that were absent from the germline clones, suggesting a role for LqfR in signaling from the follicle cells to the nurse cells and oocyte. A change was observed in the anterior actin cytoskeleton of the developing oocyte, a decrease in the relative size of the oocyte in the egg chamber, and an apparent delay in nurse cell dumping in stage 11 egg chambers. The presentation of these phenotypes is unlike what is seen in other mutants. It is possible that these could be a result of interactions between LqfR and the cytoskeleton, which is known to be important for nurse cell dumping. LqfR interacts with Cheerio, the Drosophila homolog of the actin binding protein filamin, in yeast two-hybrid screens. Cheerio is required for proper formation of the ring canals interconnecting the nurse cells and oocyte and maintenance of these structures is critical for nurse cell dumping. Although the ring canals appeared normal in lqfR mutants, there could be subtle defects that slow dumping. Another possibility is that LqfR is directly involved in nurse cell dumping via interactions with microtubules. The ENTH domain of rat epsins 1 and 2 can bind to microtubules by way of a different part of the ENTH domain than binds phospholipids. If the ENTH domain of Drosophila liquid facets-Related were also able to bind microtubules LqfR might have a direct role in dumping. Arguing against the actin and microtubule binding is the germline clone experiment, which showed that LqfR is definitively required only in the follicle cells to rescue all of the observed phenotypes. This suggests that the dumping and actin phenotypes are due to impaired signaling from the follicle cells to the germline, rather than due to LqfR-cytoskeleton interactions in the germline (Leventis, 2011).

The regulation of follicle cell size is known to require proper Akt1 signaling, and lqfR mutants have defects in follicle cell growth. In lqfR mutants, the levels of activated Akt1 were decreased leading to a decrease in Akt1 signaling and a subsequent decrease in cell size. At the same time, total Akt1 levels were elevated. The increased amount of total Akt1 could be due to an autoregulatory feedback loop. Together, these data suggest that the effect of LqfR in Akt1 signaling is upstream of Akt1 activation. However, the effect is much milder than is seen in Akt1 hypomorphs, which are lethal except in clones (Leventis, 2011).

How Akt1 signaling could be affected in the lqfR mutants remains unclear. Members of the pathway are either integral plasma membrane proteins, or cytosolic plasma membrane associated proteins. Neither of these classes of protein would be expected to require LqfR for their localization. Cytosolic Akt1 requires recruitment to the plasma membrane by the local production of PIP₃ before it can be activated. An attractive model would be that PIP₃ precursor phospholipids could be sequestered away from the cell surface in the mutant. However, endogenous expression levels of lqfR lacking the ENTH domain can rescue all apparent functions in an lqfR null mutant (Lee, 2009) suggesting that PIP binding by the ENTH domain is nonessential in Drosophila. Another possibility is the mistrafficking hypothesis raised earlier with respect to dorsal appendage formation. Akt1 signaling is instigated by insulin-like ligand binding to the InR at the plasma membrane. It is possible that in the mutant some quantity of lysosomal proteases are mistrafficked to the cell surface where they are then able to degrade some portion of either the ligand or the receptor (or both) leading to the observed decrease in Akt1 activation. Clearly this unexpected result opens up intriguing avenues for further research (Leventis, 2011).

Search PubMed for articles about Drosophila Liquid facets Related

Benard, C., et al. (2001). The C. elegans maternal-effect gene clk-2 is essential for embryonic development, encodes a protein homologous to yeast Tel2p and affects telomere length. Development 128: 4045-4055. PubMed ID: 11641227

Chidambaram, S., Zimmermann, J. and Fischer von Mollard, G. (2008). ENTH domain proteins are cargo adaptors for multiple SNARE proteins at the TGN endosome. J. Cell Sci. 121: 329-338. PubMed ID: 18198191

Duncan, M. C., Castaguta, G. and Payne, G. S. (2003a). Yeast epsin-related proteins required fro Golgi-endosome traffic define a γ-adaptin ear-binding motif. Nat. Cell Biol. 5: 77-81. PubMed ID: 12483220

Duncan, M. C. and Payne, G. S. (2003b). ENTH/ANTH domains expand to the Golgi. Trends Cell Biol. 13: 211-215. PubMed ID: 12742163

Hirst, J., Motley, A., Harasaki, K., Chew, S. Y. P. and Robinson, M. S. (2003). EpsinR: an ENTH domain-containing protein that interacts with AP-1. Mol. Biol. Cell 14: 625-641. PubMed ID: 12589059

Hirst, J., Miller, S. E., Taylor, M. J., Fischer von Mollard, G. and Robinson, M. S. (2004). EpsinR is an adaptor for the SNARE protein Vti1b. Mol, Biol. Cell 15: 5593-5602. PubMed ID: 15371541

Kalthoff, C., Groos, S., Kohl, R., Mahrhold, S. and Ungewickell, E. J. (2002). Clint: A novel clathrin-binding ENTH-domain protein at the Golgi. Mol. Biol. Cell 13: 4060-4073. PubMed ID: 12429846

Lee, J. H., Overstreet, E., Fitch, E., Fleenor, S. and Fischer, J. A. (2009). Drosophila liquid facets-Related encodes Golgi epsin and is an essential gene required for cell proliferation, growth, and patterning. Dev. Biol. 331(1): 1-13. PubMed ID: 19376106

Legendre-Guillemin, V., Wasiak, S., Hussain, N. K., Angers, A. and McPherson P. S. (2004). ENTH/ANTH proteins and clathrin-mediated membrane budding. J. Cell Sci. 117: 9-18. PubMed ID: 14657269

Leventis, P. A., et al. (2011). Liquid facets-related (lqfR) is required for egg chamber morphogenesis during Drosophila oogenesis. PLoS One 6(10): e25466. PubMed ID: 22043285

Miller, S. E., Collins, B. M., McCoy, A. J., Robinson, M. S. and Owen, D. J. (2007). A SNARE-adaptor interaction is a new mode of cargo recognition in clathrin-coated vesicles. Nature 450: 570-574. PubMed ID: 18033301

Mills, I. G., et al. (2003). EpsinR: an AP1/clathrin interacting protein involved in vesicle trafficking. J. Cell Biol. 160: 213-222. PubMed ID: 12538641

Popoff, V., et al. (2007). The retromer complex and clathrin define an early enodsomal retrograde exit site. J. Cell Sci. 120: 2022-2031. PubMed ID: 1755097

Wasiak, S., et al. (2002). Enthoprotein: a novel clathrin-associated protein identified through subcellular proteomics. J. Cell Biol. 158: 855-862. PubMed ID: 12213833

date revised: 10 August 2012

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

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