The simple cellular composition and array of distally pointing hairs has made the Drosophila wing a favored system for studying planar polarity and the coordination of cellular- and tissue-level morphogenesis. The developing hairs are filled with F-actin and microtubules and the activity of these cytoskeletons is important for hair morphogenesis. On the basis of mutant phenotypes several genes have been identified as playing a key role in stimulating hair formation. Mutations in shavenoid (sha) (also known as kojak) result in a delay in hair morphogenesis and in some cells forming no hair and others several small hairs. We report here the molecular identification and characterization of the sha gene and protein. sha encodes a large novel protein that has homologs in other insects, but not in more distantly related organisms. The Sha protein accumulated in growing hairs and bristles in a pattern that suggested that it could directly interact with the actin cytoskeleton. Consistent with this mechanism of action we found that Sha and actin co-immunopreciptated from wing disc cells. The morphogenesis of the hair involves temporal control by sha and spatial control by the genes of the frizzled planar polarity pathway. We found a strong genetic interaction between mutations in these genes consistent with their having a close but parallel functional relationship.

---

THE Drosophila wing has been one of the prime model systems for studying morphogenesis at both the cellular and the tissue level. The wing is the largest Drosophila appendage and a great deal has been learned about the genetic basis for wing patterning and the regulation of wing cell proliferation (e.g., EDGAR 1999; TELEMAN and COHEN 2000; DE CELIS 2003; MARTIN et al. 2004). In addition, the flat simple structure of both the pupal and the adult cuticular wing has made it a favored system for studies of cellular morphogenesis and planar polarity (ADLER 2002; EATON 2003). Most wing blade cells differentiate a single distally pointing cuticular hair. The extension that forms the hair contains both actin filaments and microtubules and the function of both cytoskeletons is required for normal morphogenesis (WONG and ADLER 1993; EATON et al. 1996; TURNER and ADLER 1998). The distal polarity of hairs is regulated by the frizzled (fz) tissue polarity pathway (WONG and ADLER 1993). The timing of hair initiation is at least indirectly regulated by the ecdysone cascade, but relatively little is known about how temporal aspects of wing cell differentiation are controlled. Among the genes previously implicated as having a role in regulating the time of hair initiation are grainy head (LEE and ADLER 2004), shavenoid (sha; also known as kojak) (HE and ADLER 2002), and ovo/svb (DELON et al. 2003). Mutations in sha and ovo/svb often lead to the failure of a cell to form a hair.

Mutations in sha also affect the differentiation of two additional types of extensions of epidermal cells. The long, thin laterals found on the arista (the distal-most segment of the antenna) are the product of single epidermal cells and in a sha mutant the laterals are bothbranched, multipled, and shorter than normal (HE and ADLER 2002). In vivo observation of the development of laterals in sha mutants revealed that lateral initiation was delayed 6 hr and the subsequent growth was also slower than normal. Electron microscopy thin sections showed that the distribution of actin filament bundles was abnormal in mutant laterals. sha mutations also result in a reduction in the number of larval denticles and those that are present are shorter and thinner than normal (NUSSLEIN-VOLHARD et al. 1984). Interestingly, sha does not display a mutant phenotype in sensory bristles, which share many characteristics with arista laterals (HE and ADLER 2002).

We report here the molecular characterization of the sha gene and protein. Previous work in our lab had mapped sha to a 60-kb region in 47F (HE 2001). In a separate study of gene expression in pupal wings, we identified one annotated gene in this region (CG13209) whose expression increased 11-fold from 24 to 32 hr, suggesting that it could be sha (REN et al. 2005). We confirmed this by identifying the sequence changes associated with six EMS/gamma-ray-induced sha alleles, by identifying a P-insertion allele, and by transformation rescue. Somewhat surprisingly, we found that the even expression of sha from a transgene was sufficient to rescue the mutant phenotype; thus the temporal change in expression level was not essential. The sha gene encodes a 179-kDa protein that is conserved in other insects. We found that the Sha protein accumulated close to the plasma membrane in growing hairs, suggesting that it functions directly in the hair to promote cytoskeletal-mediated outgrowth. When expressed in bristles, the Sha protein appeared to localize between the large bundles of actin filaments found in these cells and the plasma membrane. We further found that Sha and actin could be co-immunoprecipitated from wing disc cells, consistent with Sha acting directly on the cytoskeleton. To determine if Sha was sufficient to activate the cytoskeleton to initiate hair morphogenesis, we examined the effects of driving sha expression at other developmental stages. We failed to see any effects of sha expression on the actin cytoskeleton in third instar wing discs or in young pupal wings. Hence, sha is necessary but not sufficient to activate the cytoskeleton to drive hair formation. We also found strong genetic interactions between mutations in sha and mutations in genes of the frizzled planar polarity pathway, consistent with these two systems respectively controlling temporal and spatial aspects of hair morphogenesis.

Fly stocks:

FRT-, FLP-, GFP-expressing, mutant-, and deficiency-carrying chromosomes were obtained from the Drosophila Stock Center in Bloomington.

Clonal analysis:

Somatic clones were generated using the FRT/FLP system (XU and RUBIN 1993). Pupal wing clones were marked by the loss of GFP. Unmarked clones were detected by mutant phenotypes.

Cytological techniques:

White prepupae were collected and aged until dissection. Immunostaining was done by standard techniques after fixation with paraformaldehyde (HE et al. 2005). Fluorescent secondary antibodies and fluorescent phalloidin for staining the actin cytoskeleton were obtained from Molecular Probes (Eugene, OR). Confocal images were obtained on an ATTO CARV confocal unit attached to a Nikon microscope. In situ hybridization on pupal wings was done as described previously using digoxygenin-labeled probes (GENG et al. 2000).

Generating antibodies:

A 516-bp fragment that encodes the amino terminal 172 amino acids of Sha was subcloned into the pET28a vector. Expression of fusion protein was induced with IPTG and the fusion protein was purified using the his6 tag provided by the vector. Fusion protein was injected into rats and guinea pigs at Spring Valley Laboratories (Sykesville, MD). The sera was used without further purification.

Immunoprecipitation and Western blotting:

For Western blotting 20 third instar wing discs were dissected in cold PBS and homogenized in SDS sample buffer. The extract was heated at 100° for 5 min and then fractionated on 8% SDS–PAGE gels (Invitrogen, San Diego) and blotted to Millipore (Bedford, MA) Immobilon-P transfer membrane (Sigma, St. Louis). The blot was probed with the desired primary antibody and detected using Supersignal West Pico reagents (Pierce, Rockford, IL).

For immunoprecipitation experiments, 20–40 wing discs were dissected from third instar larvae and homogenized in prechilled lysis buffer (50 mM Tris–HCl, pH 7.5, 150 mM NaCl, 5 mM EDTA). The extract was spun at 12,000 x g at 2°–8° for 15 min to remove cellular debris. To reduce the background caused by nonspecific adsorption of irrelevant cellular proteins, a preclearing step was done by adding protein A agarose beads (Roche) to the sample and incubating at 2°–8° for 3 hr. The beads were pellet by centrifugation at 12,000 x g for 30 sec. Ten microliters of anti-Sha antibody was added to the supernantants, and the samples were incubated at 2°–8° for 2 hr and then added to 50 µl of protein A agarose (Roche). The samples were rotated gently at 2°–8° overnight. Beads were then washed three times with wash buffer 1 (50 mM Tris–HCl, pH 7.5, 150 mM NaCl, 1% nonidet P40, 0.5% sodium deoxycholate, 1 protease inhibitor tablet from Roche), wash buffer 2 (50 mM Tris–HCl, pH 7.5, 500 mM NaCl, 1% nonidet P40, 0.05% sodium deoxycholate), and wash buffer 3 (10 mM Tris–HCl, pH 7.5, 0.1% nonidet P40, 0.05% sodium deoxycholate) and treated with protein sample/gel-loading buffer (Sigma) and proteins denatured by heating to 100° for 3 min. The protein samples were stored at –20° and then analyzed by Western blotting as noted above.

Fractionation of the Sha protein:

Wing discs were dissected from Drosophila third instar larvae and homogenized in prechilled lysis buffer (50 mM Tris–HCl, pH 7.5, 150 mM NaCl, 5 mM EDTA) and the extract was spun at 500 x g for 10 min by microcentrifuge to remove particulate materials. The supernatant was again spun at 16,000 x g for 10 min, and the soluble fraction (S1) and the pellet fraction (P1) were analyzed by Western blotting.

Construction of the sha cDNA:

Because the BDGP did not recover a sha cDNA clone, we made one using RT–PCR. Because of the length of the sha ORF, the amplification was done in two pieces. One fragment covered exons 1–5 and the other exons 5–8. The 5' fragment was inserted into pBluescript as a NotI–HindIII fragment. The NotI site was put into the primer and the HindIII site is present in the endogenous gene. The 3' fragment was subcloned into this fragment using the HindIII site and an added KpnI site, resulting in a complete cDNA. The appropriate NotI–KpnI fragment 4881-bp sha coding region was inserted into the pUAST vector (BRAND and PERRIMON 1993). The GFP tag was inserted as a KpnI–XbaI fragment into sha–pUAST.

The sha mutant phenotype:

As described previously, sha mutations lead to some wing cells forming no hair, others forming multiple small hairs, and some cells forming relatively normal hairs (Figure 1) (HE and ADLER 2002). The severity of the phenotype varied across the wing. In general the phenotype was stronger on the ventral vs. dorsal surface and in medial and proximal wing regions than in peripheral and distal regions. The regions with the most severe phenotypes were ones where hair initiation occurred last, suggesting a connection between the time of hair initiation and the mutant phenotype. Even in null alleles, rare cells formed relatively normal hairs; thus sha was not absolutely essential for building a hair. In this way, it differed from genes such as crinkled (ck), which encodes a cellular myosin as all hairs in a ck mutant showed a mutant phenotype (TURNER and ADLER 1998; KIEHART et al. 2004). The epidermal hair phenotype of sha was not restricted to the wing and was seen in all regions of the adult epidermis. The strength of the phenotype varied from one tissue to another. It was strong on the notum (dorsal thorax) (Figure 2, A and B) and legs (data not shown), which lost essentially all hairs, and weaker on the abdomen, where hairs were lost in a patchy pattern (Figure 2, C and D). The sha mutant phenotype was not restricted to the adult epidermis and was also seen in larvae, where it caused a loss and reduction in size of denticles (Figure 2, E and F).

View larger version (101K): In this window In a new window Download PPT slide   FIGURE 1.— sha adult wing phenotype. All micrographs are of adult wings and proximal is to the left and distal is to the right. A, B, D, E, G, and H are all equivalent images (same magnification and wing region) [dorsal surface, center of the E region (most posterior region of the wing)]. C, F, and I are a second set [the ventral surface in the proximal part of the C region (the central region of the wing)]. (A) A wild-type wing (Oregon-R). (B) A shaVAI51/Df wing (strong, phenotypic null allele). Almost all of the cells in this region of the shaVAI51/Df wing form several small hairs. (D) A shaVB13/shaVB13 wing from an animal raised at 25°. This allele is a temperature-sensitive hypomorphic allele that has a very weak phenotype in the dorsal E region. The hairs are slightly finer and are more elevated, so that they appear shorter than they are. (E) An unmarked shaVB13 clone in the dorsal E region. The asterisk indicates the clone. Note that the clone cells do not produce a hair. Note how this phenotype is stronger than that in D even though the cells have the same genotype. This shows how enhanced the sha phenotype is in clones. (G) A region from the dorsal E region from a fly that carried one copy of the dominant-negative shaFK2871 allele. The arrow points to a cell that did not form a hair. (H) The same region from a fly hemizygous for shaFK2871. Note that the phenotype is stronger than that seen with the phenotypic null (B). (C) From the ventral surface of a wild-type wing (Oregon-R). F is from shaVB13/shaVB13. Note that in this region of the wing the phenotype is strong. I is from the same region of a shaVB13/shaVB13; act–Gal4 UAS–sha–GFP wing. Note the rescue by the transgene.

 

View larger version (112K): In this window In a new window Download PPT slide   FIGURE 2.— The sha phenotype in other body regions. (A and B) From the dorsal notum (thorax) of wild type (Oregon-R) (A) and shaVAI51/Df (B). Note that the bristles appear normal but the epidermal hairs are lost in the mutant. C (Oregon-R) and D (shaVAI51/Df) are from the dorsal abdomen (fourth abdominal segment). Once again, note that the bristles are normal in the mutant but the small epidermal hairs are reduced. E (shaVAI51/Df) and F (Oregon-R) are from the central region of the dorsal third abdominal segment from a third instar larvae. Note how the denticles are smaller and less numerous in the mutant.

 
In genetic mosaics, sha acted largely cell autonomously. Neighboring wild-type cells differentiated normally and mutant cells at the clone edges were not rescued (Figure 3). However, the mutant phenotype of sha was enhanced in mutant clones relative to entirely mutant wings (Figure 1, D and E). Thus, most mutant cells in clones failed to produce any hair, even when located in wing regions that showed only a weak phenotype in an entirely mutant wing. This did not appear to be correlated with clone size and was true even for very small clones. We suggest that this apparent lack of perdurance was due to the dramatic increase in sha expression just prior to hair initiation. The enhancement of the mutant phenotype in clones indicated that in some way neighboring wild-type cells functioned to inhibit the differentiation of mutant clone cells.

View larger version (106K): In this window In a new window Download PPT slide   FIGURE 3.— The sha mutant phenotype in pupal wings. (A–C) A shaVAI51 clone in the wing marked by a loss of GFP (C). (A) The wing hairs labeled by staining the actin cytoskeleton using Alexa 568 phalloidin. (B) The merged image. Note the lack of hair formation by the mutant cells. (D–F) A shaVAI51 clone in the wing marked by a loss of GFP. The wing was stained using an anti-Starry night (aka flamingo) antibody. Note the typical zig-zag pattern seen for planar polarity proteins and that this is not altered by the shaVAI51 clone. Similar results were obtained for In (data not shown).

 

Molecular identification and characterization of sha:

Genetic screens in this laboratory resulted in the isolation of 18 sha alleles. Four of these were on chromosomes that had a rearrangement in the 47F region. RFLPs on these chromosomes mapped to >50 kb and a dozen genes and did not pinpoint the sha transcription unit (HE 2001). It is likely that multiple hits were generated during the mutagen treatment. We found the expression of one gene in this region, CG13209, increased 11-fold from 24 to 32 hr after white prepupae (REN et al. 2005). We sequenced CG13209 from six cytologically normal sha mutations (and a parental chromosome for three of these). We identified mutations associated with all of these alleles (Figure 4). Three of the alleles were missense mutations. One of these was the temperature-sensitive hypomorphic sha^VB13 allele. This resulted from a mutation of pro 195 to leu. This proline is conserved in all of the insect species where we have identified sha homologs and is in the center of a highly conserved region (Figure 4). A second missense mutation was the strong sha^VAI51 allele that resulted from a change of arg 314 to ser. This residue is also conserved in all insect species where we have identified sha homologs (Figure 4). The third missense mutation was the strong sha^XO51 allele that resulted in a change of asn 446 to lys. This asn is conserved in all of the dipteran sha genes (Figure 4). Two of the strong alleles were associated with nonsense mutations in CG13209 (Figure 4). One (sha^VAG11) was predicted to produce a 378-amino-acid protein and the other (sha^VE41) a 966-amino-acid protein. The sixth allele was associated with a single-base-pair insertion that results in 8 new amino acids being added after amino acid 1260, followed by a termination codon (Figure 4). Interestingly, this allele (sha^FK2871) acted as a weak dominant negative (Figure 1, G and H). While these experiments were going on, we became aware that the BDGP had isolated a P insertion 1 kb upstream of CG13209, which was a semilethal (KG08508). We found that the mutant escapers were phenotypically sha. The sha^P allele failed to complement members of our collection of sha alleles, consistent with it being an insertion into sha. We confirmed that the insertion was responsible for the sha mutation as the mobilization of the P insert reverted the mutation.

View larger version (21K): In this window In a new window Download PPT slide   FIGURE 4.— Molecular characterization of sha. (Top) The sha protein and the location of six different sha alleles. (Bottom) The amino acid sequences for insect Sha proteins surrounding the regions where the three missense mutations are located. The location of the mutation is indicated by an asterisk. Residues that are identical to the D. melanogaster sequence are in boldface type. Below are the consensus sequences for each 30-amino-acid region. Residues that are completely conserved are in boldface type. Note that in several instances there appears to be a small deletion or insertion. We were unable to identify the region surrounding residue 445 in Tribolium.

 
The BDGP predicted that CG13209 encoded a 4881-nucleotide mRNA from eight exons. However, no sha cDNA clone was recovered by the BDGP, leaving some doubt about the accuracy of the transcript prediction. We confirmed the existence of all of the predicted splice sites by sequencing RT–PCR products (data not shown). We constructed a sha cDNA that contained all of the predicted coding sequences, fused it to GFP, subcloned this into pUAST, and generated transgenic flies. When the expression of this transgene was driven by act–GAL4, it provided almost complete sha rescue activity in the wing (Figure 1, C, F, and I) and arista (data not shown) (a few cells in the rescued wings produced small hairs). Thus, the expression of the predicted Sha–GFP fusion protein was sufficient for essentially normal sha function in the two tissues where it had principally been studied.

Clones homozygous for sha mutations showed a cell-autonomous wing hair phenotype in all regions of the wing, suggesting that sha was expressed and functioned in all wing cells. We examined the expression of sha by in situ hybridization and found, as expected, that it was expressed at relatively even levels across the wing (Figure 5E).

View larger version (74K): In this window In a new window Download PPT slide   FIGURE 5.— Expression of sha. (A) A Western blot comparing act–Gal4 UAS sha–GFP larval and Oregon-R larval extracts. The blot was probed with anti-Sha polyclonal antibody. Note the strong signal in the transgene-driven lane. (B) A Western blot probed with anti-Sha antibody that compares extracts from 24- and 32-hr pupal wings. Note the signal present in the 32-hr sample that is missing in the 24-hr sample (arrow). (A and B, bottom) An unknown protein that was routinely detected by the secondary antibody on this set of Western blots and serves as a loading control. (C) A Western blot of act–Gal4 UAS sha–GFP wing disc proteins probed with anti-Sha antibody. The wing discs were homogenized in a low-salt buffer. An aliquot of this is shown in the total lane. The sample was then centrifuged and the pellet and supernatant fractions were applied to the gel. Note that the Sha protein is soluble in low-salt buffer. (D) The results of an immunoprecipitation experiment. act–Gal4 UAS sha–GFP and Oregon-R wings were homogenized and immunoprecipitated with anti-Sha antibody. As a control, no antibody was added to one sample. The precipitated material was then assayed by a Western blot using either anti-sha or anti-actin antibodies. Note that the anti-Sha antibody was able to pull down actin. (E) An in situ hybridization to 32-hr pupal wings using CG13209 antisense (top) and sense (bottom) probes. We conclude that sha is expressed in all regions of the pupal wing.

 
It is worth noting that the dramatic increase in sha expression seen in our gene chip experiments did not appear to be essential for sha function as act–Gal4 is expected to drive transgene expression in a relatively uniform fashion. To determine if sha expression was sufficient to stimulate hair initiation, we examined ptc–GAL4 UAS–sha–GFP pupal wings from 24 to 32 hr after white pupae (Figure 6, G–I). We did not see any signs of premature hair initiation nor any abnormal accumulation of F-actin. We also failed to see any effect on the actin cytoskeleton of third instar ptc–GAL4 UAS–sha–GFP wing discs (data not shown). We concluded that sha was not sufficient to activate the cytoskeleton. Perhaps Sha needed to act with another protein to stimulate hair outgrowth.

View larger version (80K): In this window In a new window Download PPT slide   FIGURE 6.— Sha protein accumulates in growing hairs and bristles. (A) Sha–GFP in pupal head epidermis. Longitudinal stripes of GFP (arrow) are seen in developing bristles, and epidermal hairs (arrowhead) also stain brightly. (B) A section of a pupal wing expressing sha–GFP. The hairs stain brightly. (C) A high-magnification image of several hairs oriented perpendicularly to the plane of focus. The peripheral staining of GFP can be seen and this is brighter on the distal side of the hair (arrow). (D–F) Bristles on the marginal row of the wing that express Sha:GFP (D, green) and are stained for actin (F, red). (E) The merged image. Note that the green staining extends beyond the red (arrow). (G–I) A 28-hr wing that expresses sha:GFP driven by ptc–GAL4. Actin is in red (G) and GFP is in green (I); the merge is shown in H. Note that the expression of sha does not appear to influence the distribution of actin. (J–L) A sha clone in the wing marked by the loss of GFP. The pupal wing is stained with anti-sha antibody (red). Note the loss of Sha staining in the clone (asterisk) and the staining of the hairs outside of the clone (arrow).

 
We examined the sequence of sha homologs in other Drosophila species where genome sequence is available (http://species.flybase.net/). The intron–exon structure of sha is conserved in the melanogaster subgroup. The fifth exon in Drosophila melanogaster is split into two exons in D. mojavensis and D. virilis and the eighth exon in D. melanogaster is split into two exons in D. virilis.

The Sha protein:

Conceptual translation of sha yielded a protein of 1626 amino acids with a predicted molecular weight of 179,311 Da. The protein did not contain any informative motifs. Some, but not all, programs tested predicted a single transmembrane domain. To determine if Sha was an integral membrane protein, we homogenized Sha-expressing wing discs in a low-salt buffer in the absence of any detergent. The Sha protein, as assayed by Western blotting, was soluble under these conditions, arguing that it is not an integral membrane protein (Figure 5C).

sha has clear homologs in other Drosophila species. In species such as D. erecta, D. virilis, D. yakuba, and D. pseudoobscura, the protein is well conserved over essentially the entire sequence. In comparisons between members of the melanogaster subgroup, we found 96% identity and 97% similarity. In more distantly related species such as D. virilis, we found 73% identity and 80% similarity to D. melanogaster sha. Homologs of sha can also be found in other insects. In Anopheles gambiae and Aedes aegypti, we could construct conceptual homologs that were conserved over >1400 amino acids. The percentage identity fell to 33% and 30% and the percentage similarity to 46% and 41%, respectively. We could also identify regions of similarity in other insects such as Bombyx mori, Apis mellifera, and Tribolium castaneum, although for these organisms we could find homology to at most 950 amino acids of sha. The percentage identity here ranged from 44% to 38%. That these identities were higher than those seen for the mosquitoes was likely due to only the most highly conserved regions being identified. The completion of the sequencing and annotation of their genomes will be needed to determine if these animals contain homologs that contain most if not all of D. melanogaster sha. We did not find any homologs in noninsect species. Using the reiterative psi-blast program (ALTSCHUL et al. 1997), we detected weak similarity to a number of proteins in more distantly related organisms. These included the Caenorhabditis elegans UNC89, which encodes a very large protein required for myofibril organization in C. elegans. This protein contains a myosin light-chain kinase domain (BENIAN et al. 1996; SMALL et al. 2004).

Sha accumulates in developing hairs:

We took two approaches in examining the subcellular distribution of the Sha protein, both of which gave similar results. In one we examined the distribution of the transgene-encoded Sha–GFP fusion protein in pupal wing cells, taking advantage of the GFP tag. We found the sha protein in developing hairs consistent with it acting at the level of the cytoskeleton (Figure 6, B and C). In developing hairs, Sha–GFP appeared to accumulate to somewhat higher levels proximally than distally. In hairs viewed perpendicularly to the long axis, we could see that Sha was peripheral (Figure 6C), suggesting that it could be mediating the attachment of cytoskeletal elements to the plasma membrane. The peripheral localization of Sha often appeared more extensive on the distal side of the hair. In developing bristles, we observed Sha–GFP in longitudinal stripes (Figure 6, A, D, E, and F). In bristles stained for both Sha and actin, these stripes of Sha were largely colocalized with the large actin bundles that are juxtaposed to the plasma membrane in bristles. The colocalization was not complete as the Sha protein routinely appeared to be more peripheral than the actin (Figure 6E), consistent with Sha connecting the actin bundles to the plasma membrane.

As an alternative approach, we generated an anti-Sha antibody. When we stained pupal wings using this antibody, we also found anti-Sha staining in the developing hairs although the background was often high (data not shown). Only weak staining (essentially background) was detected prior to hair formation as expected from the expression profile of the gene. As a control we examined pupal wings that contained sha clones and found a loss of staining in the clones (Figure 6, J–L). We examined the accumulation of Sha using our polyclonal antibody to probe Western blots of pupal wing proteins. As a first test of the usefulness of this antibody for Western blots, we examined extracts of whole wild-type larvae and larvae where Sha expression was driven by Act–GAL4. A strong high molecular weight (>170 kDa) signal was detected in transgene-driven extracts (Figure 5A). We also examined extracts of 24- and 32-hr pupal wings. We did not detect any sha protein in the 24-hr sample but a signal was detected in the 32-hr sample (Figure 5B), consistent with the marked increase in sha RNA levels seen.

The location of the Sha protein next to the plasma membrane and actin bundles at sites of hair and bristle outgrowth suggested the hypothesis that sha functioned in hair development by interacting directly with one or more components of the actin cytoskeleton to promote actin polymerization and hair outgrowth. To test this possibility, we immunoprecipitated sha protein from wing discs and then analyzed the precipitate by Western blotting. We found that actin was co-immunoprecipitated with Sha, consistent with the possibility that it interacted directly with the actin cytoskeleton (Figure 5D).

sha with planar polarity genes:

In experiments where we used sha as a cuticular marker in mosaic experiments with planar polarity genes, we generated chromosomes that were mutant for sha and planar polarity genes such as pk, Vang, and stan (TAYLOR et al. 1998; WOLFF and RUBIN 1998; CHAE et al. 1999; GUBB et al. 1999; USUI et al. 1999; ADLER et al. 2000; TREE et al. 2002). To maximize the viability of the mutant animals, we used the hypomorphic sha^VB13 allele in many of these experiments. Because of the enhanced sha phenotype seen in clones, even a weak allele serves as a reliable cuticular marker. Doubly mutant animals that carried the hypomorphic sha^VB13 allele showed a dramatically enhanced sha mutant phenotype (Figure 7, Table 1). All of the planar polarity mutants that we examined (fz, in, fy, frtz, pk, Vang, stan, and dsh) (VINSON and ADLER 1987; PARK et al. 1996; COLLIER and GUBB 1997; TAYLOR et al. 1998; CHAE et al. 1999; GUBB et al. 1999; AXELROD 2001; SHIMADA et al. 2001; TREE et al. 2002; COLLIER et al. 2005) acted as enhancers (Figure 7, C, E, and F). In the stereomicroscope, the adult flies had bowed, wet-looking wings. In the compound microscope, we could see that the doubly mutant animals had large wing regions where cells formed no hair. In this way the phenotype was substantially stronger than that of a null sha allele (Figure 1). For several of the double-mutant combinations we quantified the fraction of the wing where cells formed no or only a very small hair and this increased dramatically in the double mutants (Table 1). We also examined flies that were doubly mutant for a fz pathway mutation and a null sha allele (Figure 7D). The null sha phenotype was also enhanced in these wings but it was difficult to quantify as most of the wings were at least partially folded and difficult to score (Figure 7, B and D). Since the phenotype of a null sha allele was enhanced, we can rule out the possibility that the fz pathway functions upstream of sha. In regions where hairs formed the polarity and multiple hair cell phenotypes of the fz pathway, mutants did not seem to be altered (data not shown).

View larger version (125K): In this window In a new window Download PPT slide   FIGURE 7.— sha interacts with the fz pathway. (A) A region from the dorsal E region of the wing from a shaVB13 fly raised at 25°. The phenotype is very weak in this region on such flies. (C–F) VangTBS42 shaVB13, frtz2 shaVB13, and shaVB13; in double mutants, respectively. Note how the sha hair-loss phenotype is strongly enhanced by each of the planar polarity mutants. (B) the same region from a shaVAI51/Df wing. (F) From a pk shaVAI51 double mutant. Note that the pk mutant enhanced the phenotype of the phenotypic sha null allele (shaVAI51).

 

View this table: In this window In a new window   TABLE 1 Planar polarity genes enhance shavenoid hair phenotype

 
On the basis of observations that suggested that the fz pathway regulated the site for hair initiation and that sha regulated the timing of hair initiation (HE and ADLER 2002; WONG and ADLER 1993), it seemed likely that the two pathways functioned in parallel. The observations noted above argue that the fz pathway does not function upstream of sha. To test if sha could be functioning upstream of the planar polarity genes, we examined the distribution of two planar polarity proteins (Stan/Fmi and In) in sha clones in pupal wings. The Stan/Fmi protein accumulates at both the proximal and distal sides of wing cells (USUI et al. 1999) and the In protein accumulates at the proximal side (ADLER et al. 2004). stan function is required for the function and asymmetric accumulation of all of the planar polarity proteins, while in functions as a downstream component of the fz pathway. Both proteins accumulated asymmetrically in sha clones, indicating that sha function is not required for fz pathway function (Figure 3, D–F; data for in not shown).

sha—a key player in hair morphogenesis:

Mutations in sha produce perhaps the most dramatic wing hair phenotype of any known Drosophila gene: loss of the hair. Previously we found that sha mutations resulted in a delay in both wing hair and arista lateral initiation (HE and ADLER 2002). Temperature-shift experiments with the temperature-sensitive sha^VB13 allele were consistent with sha principally functioning at or just prior to hair and lateral initiation (HE and ADLER 2002). On the basis of these observations we suspected that sha expression might be linked to hair initiation. We found that the expression of CG13209 sharply increased from 24 to 32 hr (REN et al. 2005) and, combined with previous mapping data, it suggested that sha could be CG13209. We showed that this was correct both by identifying sequence changes associated with six sha alleles and by transformation rescue. It is interesting that the ubiquitous expression of sha provided rescue activity even though the expression of the endogenous sha gene is strongly modulated in vivo. It appears that sha is essential but not instructive with respect to hair initiation. Consistent with this hypothesis, we saw no evidence for an alteration in the actin cytoskeleton nor an effect on the timing of hair initiation from premature expression of sha. In principle, sha could produce its dramatic mutant phenotype by interacting either directly or indirectly to activate either the actin or the microtubule cytoskeletons. Direct activation could entail providing a site on the membrane for the polymerization of actin while indirect activation could work by regulating the expression of hair-forming genes. Consistent with sha playing a direct role in hair outgrowth, we found that the Sha protein accumulated at the periphery of growing hairs and that Sha and actin co-immunoprecipitated from wing disc cells. When expressed in developing bristles, we found that Sha accumulated in longitudinal stripes that largely colocalized with the large bundles of actin filaments seen in these cells. These large bundles are in close association with the plasma membrane and Sha appeared to be displaced slightly on the peripheral side of the actin bundles. This suggests that Sha is localized on the membrane. On the basis of these observations we suggest that Sha serves to activate/organize actin polymerization by recruiting it to the plasma membrane. Developing arista laterals contain actin bundles that are similar to but smaller than those seen in bristles. Thin sections of mutant sha laterals showed abnormal actin bundles (HE and ADLER 2002). The bundles were irregular in shape and often not juxtaposed to the plasma membrane, consistent with our hypothesis that Sha regulates/organizes actin at the plasma membrane.

The Sha protein sequence does not provide any "smoking guns" as to Sha biochemical function. However, a weak similarity was detected in several proteins that are known to interact with the actin cytoskeleton (e.g., UNC89), consistent with our finding that Sha and actin can be co-immunoprecipitated from disc cells. This interaction suggests that Sha mechanistically acts directly on the actin cytoskeleton to promote hair morphogenesis. In vitro studies will be needed to determine the biochemical function of sha. For example, Does Sha promote actin polymerization or bundling? Unfortunately, Sha is rather large, which makes in vitro studies difficult.

sha functions in parallel to planar polarity genes:

The fz planar polarity pathway regulates hair morphogenesis in at least two ways. First, it controls polarity by controlling the subcellular location for hair initiation, a process that requires downstream planar polarity effectors such as Inturned, Fuzzy, and Frtz (LEE and ADLER 2002; ADLER et al. 2004; COLLIER et al. 2005). Second, the fz planar polarity pathway activates the cytoskeleton for hair morphogenesis via RhoA-mediated Rho kinase activation of Spaghetti Squash (Drosophila myosin regulatory light chain) (WINTER et al. 2001). We found that mutations in fz pathway genes act as strong enhancers of the sha mutant hair phenotype. This enhancement is seen with null sha alleles; thus the fz pathway appears to act in parallel to sha. This implies that the regulation of the location and timing of hair initiation converge on a common target. This could be achieved either by a common target that integrates these inputs and then regulates the cytoskeleton or by pathways independently regulating the cytoskeleton.

Cell autonomy of sha:

By the classical genetic mosaic tests, sha mutations act completely cell autonomously. All sha mutant cells show a mutant phenotype and all neighboring wild-type cells differentiate normally. However, the phenotype of sha mutant cells in a clone is substantially stronger than that seen when the entire wing is mutant. Thus, in some way the presence of wild-type cells influences the differentiation of the mutant clone cells. This could be due either to a diffusible signal from another tissue or to an influence by neighboring wild-type cells in the wing epithelium. We suggest that this effect is a consequence of the delayed and slower morphogenesis of sha mutant cells (HE and ADLER 2002). When the entire animal is mutant, the process of metamorphosis may be delayed in a way that allows the mutant wing cells more time to build a hair. This could result in mutant cells being able to form small multiple hairs. In a mosaic animal, the overall timing of differentiation may be closer to normal. Mutant cells in a clone would have less time to elaborate a hair prior to an irreversible block to further hair morphogenesis. This could result in a cell that did not form any obvious hair. If this is the case, what could be the nature of the "block"? Such a block could be hormonal and/or due to cuticle deposition and crosslinking. Cuticle starts to secrete during the process of hair outgrowth (GUILD et al. 2005). Perhaps in an entirely mutant animal, cuticle secretion and/or crosslinking, as well as other aspects of epidermal development, is delayed. In sha clone cells, the timing of cuticle deposition and/or crosslinking could be driven by wild-type cells elsewhere in the animal. The deposition and/or crosslinking of cuticle prior to the outgrowth of the hair could prevent any subsequently delayed hair outgrowth by mutant cells.

Conservation of cytoskeleton regulatory genes:

Many genes that are important for development are highly conserved across a wide spectrum of organisms. This is true for many genes that play an important role in the morphogenesis of the array of hairs that cover the adult Drosophila epidermis. The elaboration of hairs involves the actin and microtubule cytoskeletons and many genes that play an important role in cytoskeleton function in other organisms have a similar role in Drosophila (BAUM 2002). Thus, mutations in numerous genes that encode conserved cytoskeleton interacting proteins result in strong wing hair phenotypes. A classic example of such a gene is singed, which encodes a fascin that functions in bundling actin (BRYAN et al. 1993; CANT et al. 1994). The hairs of sn mutants are curved and bent. Mutations in genes that encode cytoskeleton regulatory proteins, such as the Tricornered kinase that results in strong hair phenotypes, are also often highly conserved (GENG et al. 2000; HE et al. 2005). Mutations in tricornered give rise to wing cells that can form as many as a dozen small hairs instead of a single long one. Thus, it is somewhat surprising that sha is not conserved beyond the insects. Within the genus Drosophila sha is moderately well conserved. D. melanogaster and D. virilis are thought to have diverged 40 million years ago and the sha proteins from these two species remain 73% identical and 80% similar over 1663 amino acids. Compared to other genes known to be important for hair morphogenesis, we find this is similar to pawn (72% identical and 78% similar), higher than inturned (63% identical and 75% similar), and much lower than singed (99% identical and 99.8% similar) or tricornered (97% identical and 98% similar) (BRYAN et al. 1993; CANT et al. 1994; PARK et al. 1996; GENG et al. 2000; ARRUDA and DOLPH 2003). sha may be an example of a gene whose primary structure changes during evolution to facilitate morphological changes.

This work was supported by grants from the National Institute of General Medical Sciences to P.N.A. Core facilities used in carrying out the research were supported by the University of Virginia Cancer Center. Fly stocks were obtained from the Drosophila Stock Center at Indiana University in Bloomington.

¹ Present address: Department of Surgery, University of California, San Francisco, 513 Parnassus Ave., San Francisco, CA 94143-0104.

² Present address: Center for Oral Health Research, College of Dentistry, University of Kentucky, Lexington, KY 40536.

ADLER, P. N., 2002 Planar signaling and morphogenesis in Drosophila. Dev. Cell 2: 525–535.[CrossRef][Medline]

ADLER, P. N., J. TAYLOR and J. CHARLTON, 2000 The domineering non-autonomy of frizzled and van Gogh clones in the Drosophila wing is a consequence of a disruption in local signaling. Mech. Dev. 96: 197–207.[CrossRef][Medline]

ADLER, P. N., C. ZHU and D. STONE, 2004 Inturned localizes to the proximal side of wing cells under the instruction of upstream planar polarity proteins. Curr. Biol. 14: 2046–2051.[CrossRef][Medline]

ALTSCHUL, S. F., T. L. MADDEN, A. A. SCHAFFER, J. ZHANG, Z. ZHANG et al., 1997 Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res. 25: 3389–3402.[Abstract/Free Full Text]

ARRUDA, S. E., and P. J. DOLPH, 2003 Molecular cloning of the pawn locus from Drosophila melanogaster. Gene 310: 169–173.[Medline]

AXELROD, J., 2001 Unipolar membrane association of Dishevelled mediates Frizzled planar cell polarity signaling. Genes Dev. 15: 1182–1187.[Abstract/Free Full Text]

BAUM, B., 2002 Winging it—actin on the fly. Dev. Cell 2: 125–126.[CrossRef][Medline]

BENIAN, G. M., T. L. TINLEY, X. TANG and M. BORODOVSKY, 1996 The Caenorhabditis elegans gene unc-89, required fpr muscle M-line assembly, encodes a giant modular protein composed of Ig and signal transduction domains. J. Cell Biol. 132: 835–848.[Abstract/Free Full Text]

BRAND, A. H., and N. PERRIMON, 1993 Targeted gene expression as a means of altering cell fates and generating dominant phenotypes. Development 118: 401–415.[Abstract]

BRYAN, J., R. EDWARDS, P. MATSUDAIRA, J. OTTO and J. WULFKUHLE, 1993 Fascin, an echinoid actin-bundling protein, is a homolog of the Drosophila singed gene product. Proc. Natl. Acad. Sci. USA 90: 9115–9119.[Abstract/Free Full Text]

CANT, K., B. A. KNOWLES, M. S. MOOSEKER and L. COOLEY, 1994 Drosophila singed, a fascin homolog, is required for actin bundle formation during oogenesis and bristles extension. J. Cell Biol. 125: 369–380.[Abstract/Free Full Text]

CHAE, J., M. J. KIM, J. H. GOO, S. COLLIER, D. GUBB et al., 1999 The Drosophila tissue polarity gene starry night encodes a member of the protocadherin family. Development 126: 5421–5429.[Abstract]

COLLIER, S., and D. GUBB, 1997 Drosophila tissue polarity requires the cell-autonomous activity of the fuzzy gene, which encodes a novel transmembrane protein. Development 124: 4029–4037.[Abstract]

COLLIER, S., H. LEE, R. BURGESS and P. ADLER, 2005 The WD40 repeat protein Fritz links cytoskeletal planar polarity to Frizzled subcellular localization in the Drosophila epidermis. Genetics 169: 2035–2045.[Abstract/Free Full Text]

DE CELIS, J. F., 2003 Pattern formation in the Drosophila wing: the development of the veins. BioEssays 25: 443–451.[CrossRef][Medline]

DELON, I., H. CHANUT-DELALANDE and F. PAYRE, 2003 The Ovo/Shavenbaby transcription factor specifies actin remodelling during epidermal differentiation in Drosophila. Mech. Dev. 120: 747–758.[CrossRef][Medline]

EATON, S., 2003 Cell biology of planar polarity transmission in the Drosophila wing. Mech. Dev. 120: 1257–1264.[CrossRef][Medline]

EATON, S., R. WEPF and K. SIMONS, 1996 Roles for Rac1 and Cdc42 in planar polarization and hair outgrowth in the wing of Drosophila. J. Cell Biol. 135: 1277–1289.[Abstract/Free Full Text]

EDGAR, B. A., 1999 From small flies come big discoveries about size control. Nat. Cell Biol. 1: E191–E193.[CrossRef][Medline]

GENG, W., B. HE, M. WANG and P. N. ADLER, 2000 The tricornered gene, which is required for the integrity of epidermal cell extensions, encodes the Drosophila nuclear DBF2-related kinase. Genetics 156: 1817–1828.[Abstract/Free Full Text]

GUBB, D., C. GREEN, D. HUEN, D. COULSON, G. JOHNSON et al., 1999 The balance between isoforms of the prickle LIM domain protein is critical for planar polarity in Drosophila imaginal discs. Genes Dev. 13: 2315–2327.[Abstract/Free Full Text]

GUILD, G. M., P. S. CONNELLY, L. RUGGIERO, K. A. VRANICH and L. G. TILNEY, 2005 Actin filament bundles in Drosophila wing hairs: hairs and bristles use different strategies for assembly. Mol. Biol. Cell 16: 3620–3631.[Abstract/Free Full Text]

HE, B., 2001 Genetic control of cell extensions. Ph.D. Thesis, University of Virginia, Charlottesville, VA.

HE, B., and P. N. ADLER, 2002 The genetic control of arista lateral morphogenesis in Drosophila. Dev. Genes Evol. 212: 218–229.[CrossRef][Medline]

HE, Y., X. FANG, K. EMOTO, Y. N. JAN and P. N. ADLER, 2005 The tricornered Ser/Thr protein kinase is regulated by phosphorylation and interacts with furry during Drosophila wing hair development. Mol. Biol. Cell 16: 689–700.[Abstract/Free Full Text]

KIEHART, D. P., J. D. FRANKE, M. K. CHEE, R. A. MONTAGUE, T. L. CHEN et al., 2004 Drosophila crinkled, mutations of which disrupt morphogenesis and cause lethality, encodes fly myosin VIIA. Genetics 168: 1337–1352.[Abstract/Free Full Text]

LEE, H., and P. N. ADLER, 2002 inturned and fuzzy are essential for the function of the frizzled pathway in the wing. Genetics 160: 1535–1547.[Abstract/Free Full Text]

LEE, H., and P. N. ADLER, 2004 The grainy head transcription factor is essential for the function of the frizzled pathway in the Drosophila wing. Mech. Dev. 121: 37–49.[CrossRef][Medline]

MARTIN, F. A., A. PEREZ-GARIJO, E. MORENO and G. MORATA, 2004 The brinker gradient controls wing growth in Drosophila. Development 131: 4921–4930.[Abstract/Free Full Text]

NUSSLEIN-VOLHARD, C., E. WIESCHAUS and H. KLUDING, 1984 Mutations affecting the pattern of the larval cuticle in Drosophila melanogaster. I. Zygotic loci on the second chromosome. Roux's Arch. Dev. Biol. 193: 267–282.

PARK, W. J., J. LIU, E. J. SHARP and P. N. ADLER, 1996 The Drosophila tissue polarity gene inturned acts cell autonomously and encodes a novel protein. Development 122: 961–969.[Abstract]

REN, N., C. ZHU, H. LEE and P. N. ADLER, 2005 Gene expression during Drosophila wing morphogenesis and differentiation. Genetics 171: 625–638.[Abstract/Free Full Text]

SHIMADA, Y., T. USUI, S. YANAGAWA, M. TAKEICHI and T. UEMURA, 2001 Asymmetric colocalization of Flamingo, a seven-pass transmembrane cadherin, and Dishevelled in planar cell polarization. Curr. Biol. 11: 859–863.[CrossRef][Medline]

SMALL, T. M., K. M. GERNERT, D. B. FLAHERTY, K. B. MERCER, M. BORODOVSKY et al., 2004 Three new isoforms of Caenorhabditis elegans UNC-89 containing MLCK-like protein kinase domains. J. Mol. Biol. 342: 91–108.[CrossRef][Medline]

TAYLOR, J., N. ABRAMOVA, J. CHARLTON and P. N. ADLER, 1998 Van Gogh: a new Drosophila tissue polarity gene. Genetics 150: 199–210.[Abstract/Free Full Text]

TELEMAN, A. A., and S. M. COHEN, 2000 Dpp gradient formation in the Drosophila wing imaginal disc. Cell 103: 971–980.[CrossRef][Medline]

TREE, D. R. P., J. M. SHULMAN, R. ROUSSET, M. P. SCOTT, D. GUBB et al., 2002 Prickle mediates feedback amplification to generate asymmetric planar cell polarity signaling. Cell 109: 1–11.[CrossRef][Medline]

TURNER, C. M., and P. N. ADLER, 1998 Distinct roles for the actin and microtubule cytoskeletons in the morphogenesis of epidermal hairs during wing development in Drosophila. Mech. Dev. 70: 181–192.[CrossRef][Medline]

USUI, T., Y. SHIMA, Y. SHIMADA, S. HIRANO, R. W. BURGESS et al., 1999 Flamingo, a seven-pass transmembrane cadherin, regulates planar cell polarity under the control of Frizzled. Cell 98: 585–595.[CrossRef][Medline]

VINSON, C. R., and P. N. ADLER, 1987 Directional non-cell autonomy and the transmission of polarity information by the frizzled gene of Drosophila. Nature 329: 549–551.[CrossRef][Medline]

WINTER, C. G., B. WANG, A. BALLEW, A. ROYOU, R. KARESS et al., 2001 Drosophila rho-associated kinase (drok) links frizzled-mediated planar cell polarity signaling to the actin cytoskeleton. Cell 105: 81–91.[CrossRef][Medline]

WOLFF, T., and G. RUBIN, 1998 Strabismus, a novel gene that regulates tissue polarity and cell fate decisions in Drosophila. Development 125: 1149–1159.[Abstract]

WONG, L. L., and P. N. ADLER, 1993 Tissue polarity genes of Drosophila regulate the subcellular location for prehair initiation in pupal wing cells. J. Cell Biol. 123: 209–221.[Abstract/Free Full Text]

XU, T., and G. M. RUBIN, 1993 Analysis of genetic mosaics in developing and adult Drosophila tissues. Development 117: 1223–1237.[Abstract]

Communicating editor: K. V. ANDERSON

This article has been cited by other articles:

				N. Ren, J. Charlton, and P. N. Adler The flare Gene, Which Encodes the AIP1 Protein of Drosophila, Functions to Regulate F-Actin Disassembly in Pupal Epidermal Cells Genetics, August 1, 2007; 176(4): 2223 - 2234. [Abstract] [Full Text] [PDF]

				S. Chung, S. Kim, J. Yoon, P. N. Adler, and J. Yim The Balance Between the Novel Protein Target of Wingless and the Drosophila Rho-Associated Kinase Pathway Regulates Planar Cell Polarity in the Drosophila Wing Genetics, June 1, 2007; 176(2): 891 - 903. [Abstract] [Full Text] [PDF]

International Access Link

Online ISSN: 1943-2361  Print ISSN: 0016-6731
Copyright 2009 by the Genetics Society of America
phone: 412-268-1812   fax: 412-268-1813   email: genetics-gsa{at}andrew.cmu.edu