The twisted Gene Encodes Drosophila Protein O-Mannosyltransferase and Genetically Interacts With the rotated abdomen Gene Encoding Drosophi
http://www.100md.com
遗传学杂志 2006年第1期
ABSTRACT The family of mammalian O-mannosyltransferases includes two enzymes, POMT1 and POMT2, which are thought to be essential for muscle and neural development. Similar to mammalian organisms, Drosophila has two O-mannosyltransferase genes, rotated abdomen (rt) and DmPOMT2, encoding proteins with high homology to their mammalian counterparts. The previously reported mutant phenotype of the rt gene includes a clockwise rotation of the abdomen and defects in embryonic muscle development. No mutants have been described so far for the DmPOMT2 locus. In this study, we determined that the mutation in the twisted (tw) locus, tw1, corresponds to a DmPOMT2 mutant. The twisted alleles represent a complementation group of recessive mutations that, similar to the rt mutants, exhibit a clockwise abdomen rotation phenotype. Several tw alleles were isolated in the past; however, none of them was molecularly characterized. We used an expression rescue approach to confirm that tw locus represents DmPOMT2 gene. We found that the tw1 allele represents an amino acid substitution within the conserved PMT domain of DmPOMT2 (TW) protein. Immunostaining experiments revealed that the protein products of both rt and tw genes colocalize within Drosophila cells where they reside in the ER subcellular compartment. In situ hybridization analysis showed that both genes have essentially overlapping patterns of expression throughout most of embryogenesis (stages 8–17), while only the rt transcript is present at early embryonic stages (5 and 6), suggesting its maternal origin. Finally, we analyzed the genetic interactions between rt and tw using several mutant alleles, RNAi, and ectopic expression approaches. Our data suggest that the two Drosophila O-mannosyltransferase genes, rt and tw, have nonredundant functions within the same developmental cascade and that their activities are required simultaneously for possibly the same biochemical process. Our results establish the possibility of using Drosophila as a model system for studying molecular and genetic mechanisms of protein O-mannosylation during development.
O-linked carbohydrate modifications of proteins can play important roles during animal development. O-mannosylation represents a prominent example of such a post-translational protein modification that has recently been linked to mammalian muscular and neural development (reviewed in MICHELE and CAMPBELL 2003; MARTIN 2003). Abnormal O-mannosylation has been implicated in several congenital neuromuscular disorders (VAN REEUWIJK et al. 2005; reviewed in ENDO and TODA 2003; MARTIN and FREEZE 2003); while the targeted disruption of the Pomt1 gene encoding O-mannosyltransferase, an enzyme initiating the biosynthesis of O-mannosyl glycans, results in early embryonic lethality in mice (WILLER et al. 2004).
In mammalian organisms, one of the main targets of O-mannosylation is thought to be -dystroglycan (CHIBA et al. 1997), a cell adhesion glycoprotein important for the integrity of the muscle cells and neuronal migration during development (MARTIN 2003). Dystroglycan plays a central role in the dystrophin-associated glycoprotein complex (DGC) that provides a crucial link between the extracellular basal lamina and the cytoskeletal proteins (MICHELE and CAMPBELL 2003). The O-mannosyl glycans of -dystroglycan have been found to affect its binding to the extracellular matrix-associated ligand laminin, which indicates the importance of O-mannosylation for the functional integrity of DGC and its interaction with extracellular matrix (CHIBA et al. 1997; MICHELE et al. 2002).
Intriguingly, the Drosophila genome encodes homologs of all essential components necessary for the formation of DGC but with substantially less diversity (GREENER and ROBERTS 2000; DEKKERS et al. 2004), which presents an attractive possibility for using fruit flies as a model system for studying the biology of DGC and several human congenital neuromuscular diseases.
Protein O-mannosyltransferase enzymes are best characterized in yeast. A total of seven O-mannosyltransferases have been identified in Saccharomyces cerevisiae (GENTZSCH and TANNER 1996). They comprise three evolutionary conserved subfamilies of proteins, PMT1, PMT2, and PMT4 (WILLER et al. 2003). On the other hand, the family of mammalian O-mannosyltransferases includes two enzymes, POMT1 and POMT2 that belong to PMT4 and PMT2 subfamilies, respectively (JURADO et al. 1999; WILLER et al. 2002). These enzymes are thought to modify numerous serine and threonine residues of -dystroglycan with O-linked mannose, which initiates the biosynthesis of O-mannosyl glycans of dystroglycan (WILLER et al. 2003). Recently, it was demonstrated that elevated O-mannosyltransferase activity could be detected in extracts from cultured cells only if POMT1 and POMT2 were coexpressed in these cells (MANYA et al. 2004). Although this result indicated the possibility that these enzymes function as a heterocomplex, no evidence for physical interaction between POMT1 and POMT2 proteins have been obtained so far. Thus, molecular and genetic mechanisms governing the function of these proteins remain elusive.
Similar to higher animals, Drosophila has two O-mannosyltransferases, RT (encoded by the rotated abdomen gene) and DmPOMT2, which are evolutionary related to mammalian POMT1 and POMT2, respectively (MARTIN-BLANCO and GARCIA-BELLIDO 1996; WILLER et al. 2002; ICHIMIYA et al. 2004). Several mutations in the rotated abdomen gene (rt) were previously isolated and phenotypically characterized (BRIDGES and MORGAN 1923; LINDSLEY and ZIMM 1992; MARTIN-BLANCO and GARCIA-BELLIDO 1996). Two molecularly characterized mutants, rt2 and rtP, are semiviable recessive alleles associated with disruptions of the gene coding region by P-element insertion; they possibly represent null mutations (COOLEY et al. 1988; MARTIN-BLANCO and GARCIA-BELLIDO 1996). The phenotypes of these mutations include some larval and adult muscle abnormalities, as well as a prominent, up to 90° clockwise rotation of abdominal segments in adult flies (MARTIN-BLANCO and GARCIA-BELLIDO 1996). No mutations in DmPOMT2 gene have been reported so far, although the possibility that mutations in the twisted locus might represent DmPOMT2 mutants has been suggested (J. CRUCES, cited in WILLER et al. 2002). This hypothesis was further supported by "twisted abdomen" phenotype recently obtained in RNAi-mediated DmPOMT knockdown experiments (ICHIMIYA et al. 2004). Similar to their mammalian counterparts, RT and DmPOMT2 proteins have to be coexpressed to produce O-mannosyltransferase activity (ICHIMIYA et al. 2004).
The twisted alleles represent a complementation group of viable and semiviable recessive mutations on the X chromosome that also exhibit clockwise rotated abdominal segments in the adult (LINDSLEY and ZIMM 1992). Several tw mutant alleles had been isolated (DAVIS 1980; LINDSLEY and ZIMM 1992); however, none of them was characterized in detail and no molecular data are currently available for the tw locus. Most of these tw mutants have been lost and there is only one mutant allele, tw1, available from public collections (FlyBase information, DRYSDALE et al. 2005).
In this study, we have tested the hypothesis that tw corresponds to DmPOMT2. By using an expression rescue approach, we established that the tw locus represents the DmPOMT2 gene. We also characterized the molecular nature of tw1 mutation. Our immunostaining analysis of subcellular localization of tagged RT and DmPOMT2 (TW) proteins revealed their colocalization in the endoplasmic reticulum (ER) of Drosophila cells. The pattern of embryonic tw expression was analyzed by in situ hybridization and compared to the expression of rt. Our data showed striking overlap of tw and rt expression at different stages of embryogenesis. Moreover, we found a genetic interaction between tw and rt mutant alleles. All these results are consistent with the hypothesis that RT and TW, the two Drosophila protein O-mannosyltransferases, participate in the same developmental cascade and may collaborate at the molecular level, potentially functioning as an enzymatic heterocomplex.
MATERIALS AND METHODS
Mutant and transgenic Drosophila stocks:
The following Drosophila mutant alleles, chromosomal aberrations, and transgenic insertions used in the study were obtained from the Bloomington Drosophila Stock Center, Indiana University: tw1, rtP, rt2, Df(1)su(s)83, Dp(1;Y)y2sc, Act5C–GAL4-25 (second chromosome insertion), Act5C–GAL4-17 (third chromosome insertion), tubP–GAL4 (LL7), and C155–GAL4. The rt571 allele (also designated as EP(3)0571) was obtained from the Exelixis EP collection (Exelixis; San Francisco). In our experiments, we used three different rt alleles to exclude potential influence of genetic background and possible peculiarities of some alleles. All three alleles of rt include an independent P-element insertion in the first exon of the gene and likely represent null or strong hypomorphic mutations (see MARTIN-BLANCO and GARCIA-BELLIDO 1996; DRYSDALE et al. 2005; and Figure 6). PDI::GFP transgenic flies were kindly provided by Alain Debec (BOBINNEC et al. 2003). The following transgenic insertions were generated by P-element mediated transformation: UAS–DmPOMT2 and UAS–twRNAi-39 (on the second chromosome) and UAS–twRNAi-77 (on the third chromosome).
Sequencing of the DmPOMT2 gene in tw1 mutant flies:
The DNA fragment of DmPOMT2 locus was PCR amplified from the genomic DNA of tw1 homozygous flies using primers 5' CGTGGCCAGGATAACAACACTGGC 3' and 5' AACGTTGACAGGGTTGTGGGTGTGGT 3'. The amplified genomic region included the transcribed region [determined by the alignment of the DmPOMT2 cDNA (clone LP01681) with Drosophila genomic sequences (ADAMS et al. 2000)], along with the 2-kb upstream and 500-bp downstream fragments. The PCR product was sequenced and several deviations of the DmPOMT2 locus sequence from the corresponding wild-type genomic sequence determined by the Drosophila genome project (ADAMS et al. 2000) were uncovered. The alterations include: CA(249,250) AGGAT; C(472) T; T(2545) C (nucleotide numbers are relative to the beginning of coding region). The first alteration is predicted to affect the translated protein sequence [T(59) GS], while the two other mutations are silent.
DNA constructs for cell-culture expression and fly transformation
The cDNA clones were obtained from the Drosophila Genomics Resource Center at Indiana University. For cell culture expression experiments, the coding region of DmPOMT2 was PCR amplified from the DmPOMT2 cDNA (LP01681) using PCR primers containing BglII and XbaI restriction sites. Using these sites, the PCR product was inserted into the pMK33 vector (KOELLE et al. 1991). The final construct, pMK–DmPOMT2, also included a short DNA fragment encoding the HA tag (NIMAN et al. 1983); it was obtained by annealing two synthesized oligonucleotides and then introduced into DmPOMT2 cDNA using SanDI restriction site. The resulting construct encoded a DmPOMT2 (TW) protein with HA-tag inserted into nonconserved region immediately after G(650). The same construct was subcloned into the pUAST vector (BRAND et al. 1994) to obtain pUAST–DmPOMT2 for fly transformation and in vivo expression. Functionality of the construct was confirmed by Western-blot analysis (data not shown) and immunostaining of pMK–DmPOMT2-transfected S2 cells (Figure 4), as well as by an in vivo rescue assay (Figure 2). For the expression of rt in Drosophila cell culture, the coding region of rt was PCR amplified from the rt cDNA (clone RE30211) and inserted in the pRMHA3 vector (BUNCH et al. 1988). In the final expression construct, the rt coding region was modified by the addition of a short DNA fragment encoding two MYC tags (EVAN et al. 1985) immediately following the last amino acid-coding triplet of rt. As revealed by sequencing, the RE30211 clone includes a short unspliced intron preceding the last coding exon of the gene. Efficient expression of the construct and removal of the intron by in-cell splicing was confirmed using anti-MYC immunostaining of pRMHA3–rt-MYC-transfected cells (Figure 4) and Western-blot analysis of RT–MYC protein expressed in Drosophila S2 cells (data not shown). The functionality of the MYC-tagged RT protein was also confirmed in vivo by its ability to rescue the rt mutant phenotype (data not shown). The UAS–twRNAi construct was produced essentially according to the published strategy (LEE and CARTHEW 2003). Briefly, the two PCR-amplified DNA fragments of DmPOMT2 gene, both including the 6th exon (700 bp) and one also including the following intronal sequence, were ligated together in the opposite orientation. The resulting inverted repeat was cloned into the pUAST vector using PCR-introduced EcoRI and XbaI restriction sites. The UAS–twRNAi construct is predicted to produce 700 bp hairpin dsRNA with 12 unpaired bases of loop region.
Cell culture:
Drosophila S2 cells were maintained and transfected using the protocols described earlier (KOLES et al. 2004).
Immunostaining and epifluorescent microscopy:
Expression of the UAST–DmPOMT2 construct was induced in salivary glands using the C155–GAL4 driver. Third instar larvae were dissected, fixed, and stained as described earlier (PANIN et al. 1997). The following primary antibodies and corresponding dilutions were used for immunostaining: rabbit anti-LVA (1:2000) (a gift from John Sisson, University of Texas, Austin, TX); mouse and rabbit anti-HA (1:1000); mouse anti-MYC (1: 1000) (BabCo, Berkeley, CA). We used the following fluorescent secondary donkey antibodies: anti-mouse-Cy3 (1:250); anti-rabbit-FITC (1:150); and anti-mouse-Cy5 (1:150) (Jackson Laboratories). Digital images were obtained using Zeiss Axioplan 2 fluorescent microscope with the ApoTome module for optical sectioning.
In situ hybridization was performed as described earlier (KOLES et al. 2004) using the tw and rt cDNA clones LP01681 and RE30211, respectively, as templates for the synthesis of DIG-labeled probes. Every in situ hybridization experiment included a negative control staining with a probe transcribed from the corresponding antiparallel cDNA sequence (data not shown).
Phenotype analysis:
All flies were grown at 25°. The rotation of abdomen was scored in CO2-anesthetized flies 1 day after eclosion (to eliminate possible influence of aging or altered morphology in very young flies). The rotation angle was measured from posterior viewpoint using Nikon SMZ microscope with a protractor reticle objective.
RESULTS
The tw1 mutant allele contains a mutation within the DmPOMT2 gene region encoding the conserved PMT domain:
The novel Drosophila O-mannosyltransferase 2 gene was initially found by BLAST (ALTSCHUL et al. 1990) searches of Drosophila genomic sequences on the basis of the homology of its conceptual translation to the sequences of RT, yeast, and mammalian O-mannosyltransferases (WILLER et al. 2002; FlyBase information, DRYSDALE et al. 2005). The software prediction was further confirmed by the identification and sequencing of the corresponding full-length cDNA clone LP01681 from the Drosophila EST collection (WILLER et al. 2002; ICHIMIYA et al. 2004). We also identified a mosquito EST sequence (GenBank accession no. XM 312249) encoding a protein with high homology to DmPOMT2, which suggested that this sequence corresponds to the mosquito ortholog of DmPOMT2, AgPOMT2 (Figure 1). The similarity between the products of rt and DmPOMT2 (Figure 1), along with functional data on the homologous yeast PMT family members, allowed us to suggest that both Drosophila genes might function within the same biochemical pathway. This scenario would predict similar phenotypes for rt and DmPOMT2 mutations. Interestingly, we noticed that the chromosomal position of DmPOMT2 (1D4) maps to the cytogenetic region that also includes tw (1C3–D4), the complementation group of mutations causing a clockwise rotation of the adult abdomen (LINDSLEY and ZIMM 1992). This phenotype is strikingly similar to that of rt mutants (Figure 2, A–C). These data indicated the possibility that tw corresponds to DmPOMT2. This hypothesis has also been mentioned by other researchers (J. CRUCES, cited in WILLER et al. 2002) and supported by DmPOMT2 RNAi experiments that generated rt-like abdomen rotation phenotype (ICHIMIYA et al. 2004). Thus, we decided to sequence the genomic region of DmPOMT2 in tw1 mutant flies carrying a semiviable recessive twisted allele. The tw1 homozygous female and hemizygous male flies have a fully penetrant phenotype that manifests in the clockwise rotation of the abdomen (20°–40°) along the anteroposterior axis, when scored from a posterior viewpoint (DAVIS 1975; Figure 2C). Genetically, the tw1 mutation behaves as a hypomorphic allele since its phenotype is significantly enhanced in tw1-hemizygous females carrying a deficiency that includes the tw locus [tw1/Df(1)su(s)83], while the phenotype is completely rescued in males carrying a duplication of the tw locus on the Y chromosome [tw1/Dp(1;Y)y2sc] (Figure 2C and Figure 3). We sequenced the DmPOMT2 locus including the transcribed region together with 2-kb upstream and 500-bp downstream fragments. The sequencing revealed several deviations of the DmPOMT2 genomic sequence from the corresponding wild-type sequence determined by the Drosophila genome project (ADAMS et al. 2000). All but one of the sequence alterations likely represent polymorphisms or mutations neutral for the gene's function (see MATERIALS AND METHODS). Notably, one of the identified mutations maps to the coding region and results in the amino acid substitution/insertion [T(59) GS] within the conserved PMT domain of DmPOMT2 (Figure 1), which suggests an altered functionality of the DmPOMT2 protein in the tw1 mutants. This finding further supported the hypothesis that the phenotype of tw mutants results from mutations in DmPOMT2.
The phenotype of tw1 mutation can be rescued by DmPOMT2 expression:
To confirm that the DmPOMT2 mutation(s) causes the abdominal rotation phenotype in tw1 mutants, we carried out a rescue assay. To this end, we induced DmPOMT2 expression in the tw1 mutant background using the UAS/GAL in vivo expression system (BRAND et al. 1994).
We produced UAS–DmPOMT2 transgenic flies and crossed them to flies carrying a ubiquitous Act5C–GAL4-17 driver (see MATERIALS AND METHODS). There were no visible defects in abdominal morphology of the flies from either parental stocks or UAS–DmPOMT2/+; Act5C–GAL4-17/+ progeny of the cross (data not shown). The presence of the UAS–DmPOMT2 or Act5C–GAL4 transgene alone did not modify the phenotype of tw1 mutants (Figure 3). At the same time, the complete rescue of the abdomen rotation phenotype was observed in tw1 hemizygous male and homozygous female flies carrying both UAS–DmPOMT2 and Act5C–GAL4-17 together (Figure 3). Similar results were obtained when DmPOMT2 was expressed using Act5C–GAL4-25 and tubP–GAL4 drivers (data not shown). These rescue experiments proved that tw indeed represents the DmPOMT2 gene. Thus, we suggest changing the designation of the DmPOMT2 (CG12311) gene to "twisted" (tw), the name of the originally discovered mutants that is in an agreement with the traditional nomenclature of Drosophila genetics (LINDSLEY and ZIMM 1992). Below, we use "tw" and "TW" to designate the DmPOMT2 gene and its protein product, respectively.
Subcellular localization of RT and TW proteins:
Glycosyltransferases that modify secreted glycoproteins commonly function in the Golgi apparatus. Interestingly, yeast PMT family members localize to the ER subcellular compartment (WILLER et al. 2003). The localization of O-mannosyltransferase proteins in animal cells has been reported only for human POMT2 protein expressed in human culture cells (WILLER et al. 2002). In that study, the C-terminally tagged POMT2 protein was detected in the ER by immunostaining; however, the functionality of this fusion protein remains undetermined. Thus, we decided to investigate the subcellular localization of the TW protein, the Drosophila ortholog of mammalian POMT2, using the UAS–DmPOMT2 transgenic construct that was functional in our rescue assay (Figure 2). Immunostaining for TW protein expressed in vivo in the salivary gland cells of Drosophila larvae revealed its colocalization with an ER marker, PDI-GFP (BOBINNEC et al. 2003). At the same time, the localization of TW showed minimal overlap with a Golgi marker, the LVA protein (SISSON et al. 2000), when TW was expressed in Drosophila S2 cells (Figure 4, D–F). Double immunostaining of S2 cells expressing both RT and TW proteins demonstrated their colocalization inside the cell (Figure 4, G–I). Thus, we concluded that both RT and TW proteins reside in the ER subcellular compartment.
Analysis of tw and rt expression during Drosophila embryogenesis by in situ hybridization:
Detailed analysis of the spatial and temporal patterns of gene expression by in situ hybridization can provide important information about the functioning and regulation of the gene. The pattern of embryonic rt expression was previously analyzed by in situ hybridization (MARTIN-BLANCO and GARCIA-BELLIDO 1996). The pattern of tw (DmPOMT2) expression has been reported only for embryonic stage 10, while the expression at other stages was estimated only using a real-time PCR assay (ICHIMIYA et al. 2004). Thus, we decided to perform in situ hybridization analysis of tw expression and compare it with the pattern of rt expression at different embryonic stages. In agreement with previous reports (MARTIN-BLANCO and GARCIA-BELLIDO 1996; ICHIMIYA et al. 2004), we found no significant expression of tw at early embryonic stages, while rt mRNA was detected at stages 5 and 6, which probably indicates the presence of maternally provided transcript (Figure 5, A and D). This early rt expression decays quickly and it is not readily detectable after stage 7. Although we did detect some weak staining for tw and rt expression during the stages of germband extension (stages 10 and 11), this staining was diffuse and barely detectable above the background (data not shown). Thus, we decided to concentrate our analysis on tw and rt expression at later stages, when the expression of these genes is readily detectable. The prominent expression of both genes, tw and rt, appears at early stage 14 that corresponds to the period of active muscle differentiation (Figure 5, B and E). However, we have not detected significant expression of tw or rt in the developing somatic muscle cells. Instead, the expression of these genes appeared to be pronounced in other tissues, including certain developing epidermal cells, as well as hindgut and foregut regions. At that stage, the expression of tw is also present in the developing trachea, while rt expression appears in the tracheal cells slightly later, during stage 15. Thus, remarkably overlapping patterns of rt and tw expression are established by stages 14 and 15, and this expression appears to persist through the late embryogenesis (Figure 5, C and F).
Genetic interaction between tw and rt genes:
The similarity of mutant phenotypes (Figure 2) strongly suggests that both genes function in the same developmental cascade. To test this possibility, we assayed the genetic interaction between tw1 and three rt alleles, rt2, rtP, and rt571. Somewhat surprisingly, we found that the presence of tw1 significantly suppressed the abdomen rotation phenotype of rt mutants (Figure 6A). Notably, this effect is dominant since even one copy of tw1 mutation in females is sufficient for the suppression. In fact, the suppression of rt phenotype in tw1 homozygous or hemizygous flies appears to be nearly complete since the phenotype of tw1/tw1(or tw1/Y); rt/rt double mutants is virtually indistinguishable from the phenotype of tw1-homo-/hemizygous mutants alone (Figure 6B).
We considered two competing explanations of these results. First, the relative excess of TW as compared to RT in the rt mutants might have a negative effect on the O-mannosylation pathway. In this case, a potential decrease of TW might explain the suppression of rt phenotype in tw1 mutant background. Alternatively, the suppression of rt phenotype by tw1 might be explained by a special feature of the tw1 allele, which would not implicate the dependence of phenotype on relative concentrations of RT and TW. To discriminate between these two possibilities, we analyzed further the genetic interaction between rt and tw.
The first possibility would predict that an increase of tw activity would increase the severity of rt mutant phenotype, while a decrease of tw would result in rt mutant phenotype suppression. To test this prediction, we varied the level of tw in rt mutants by several alternative ways: (i) by adding an extra copy of the tw locus [using duplication Dp(1;Y)y2sc], (ii) by overexpressing TW using UAS–GAL4 system, and (iii) by decreasing the tw activity via UAS–twRNAi construct expression. Neither the increase of tw expression (Figure 7A) nor the decrease of tw activity (Figure 7B) revealed the predicted sensitivity of rt mutant phenotype to varied concentrations of TW. Hence, we interpret the dominant suppression of rt phenotype by the tw1 mutation as a special feature of the tw1 allele that somehow bypasses the requirement for rt activity in the genetic pathway. This interpretation is further supported by a synergistic effect of tw RNAi and heterozygous rt mutant background (Figure 8), which revealed a positive interaction between rt and tw, indicating their close collaboration within the pathway. These data also support the hypothesis that the O-mannosylation pathway requires the simultaneous activities of both rt and tw genes (ICHIMIYA et al. 2004). This simultaneity requirement can explain the absence of any significant effect of varied TW levels on the phenotype of tested rt mutants that represent very strong hypomorphic or amorphic mutations and presumably lack the RT protein (see MARTIN-BLANCO and GARCIA-BELLIDO 1996; rt phenotypes in Figure 6A).
DISCUSSION
The family of protein O-mannosyltransferases in Drosophila includes two members, RT and DmPOMT2, and exhibits obvious evolutionary relation to the mammalian POMT protein family (WILLER et al. 2003). While several mutations in the rt locus have been previously isolated and the rt gene has been molecularly characterized (LINDSLEY and ZIMM 1992; MARTIN-BLANCO and GARCIA-BELLIDO 1996), the novel DmPOMT2 gene (CG12311) has been described only recently (FlyBase; ICHIMIYA et al. 2004), and no mutations in DmPOMT2 have been reported so far. In this study, we have established the relationship between DmPOMT2 and twisted, the previously isolated complementation group of recessive mutations with the characteristic phenotype of a clockwise twisted abdomen. Our results demonstrate that the tw1 semiviable recessive allele is associated with a mutation in the coding region of the DmPOMT2 gene, which alters the amino acid sequence of the conserved PMT domain of DmPOMT2 protein. We have also found that this mutation is associated with the decrease of tw activity, since the tw1 phenotype is enhanced over deficiency for tw locus, and it can be completely rescued by genetic duplication including the tw gene or by ubiquitous ectopic expression of the UAS–DmPOMT2 construct. Thus, tw1 represents the first molecularly and genetically characterized mutant allele of the DmPOMT2 gene.
Interestingly, we found that tw1 could efficiently suppress the phenotype of three tested rt alleles (Figure 6). This finding was unexpected, since previously it was reported that RT and DmPOMT2 collaborate biochemically in vitro, and DmPOMT2 RNAi phenotype is enhanced in the rtP heterozygous background (ICHIMIYA et al. 2004). Yet the possibility existed that the suppression is the result of a decreased activity of tw in tw1 mutants. In this case, the suppression would suggest that an unbalanced relative increase of TW in rt mutants had a negative effect on the pathway [which, for instance, might result from a competition between nonproductive homomeric TW complexes and active RT–TW heterocomplexes (GIRRBACH and STRAHL 2003; ICHIMIYA et al. 2004)]. To test this possibility, we further analyzed the genetic interaction between tw and rt by varying the level of tw in the rt mutant background. We found that the phenotype of rt homozygous or heteroallelic mutants was neither significantly sensitive to an increase of tw expression by introducing a duplication of the tw locus or by ectopic expression of UAS–DmPOMT2 construct nor significantly sensitive to a decrease of tw activity by UAS–twRNAi expression (Figure 7 A and B). At the same time, we also confirmed the synergistic effect of the partial reduction of tw and rt activities on the mutant phenotype using UAS–twRNAi construct expression in rt heterozygotes (Figure 8). Thus, we ruled out the possibility of negative effect of tw in rt mutants and concluded that the suppression of rt phenotype is a special feature of tw1 allele.
Taken together, our results indicate that both tw and rt are involved in the same developmental pathway, where they execute nonredundant functions. Insensitivity of rt mutant phenotype to the varied level of tw expression formally characterizes rt as epistatic to tw. At the same time, the tw1 mutation could dominantly suppress the phenotype of strong (probably amorphic) rt mutations, thus revealing its epistatic position relative to rt (Figure 6). We interpret these mutually epistatic relationships between rt and tw as the evidence for possible functioning of their protein products within the same molecular complex or being involved in a regulatory interaction within the same biochemical pathway. This conclusion is in agreement with previously reported data on simultaneous requirement of RT and DmPOMT2 for their biochemical activity in vitro (ICHIMIYA et al. 2004). The conclusion is also consistent with our other observations presented here, including (i) essentially identical phenotypes of clockwise abdomen rotation in both rt and tw mutants (Figure 2), (ii) the subcellular colocalization of RT and TW proteins within the ER compartment in Drosophila cells (Figure 4), and (iii) the overlapping pattern of rt and tw expression during different stages of embryogenesis (Figure 5). Although the close relationship between RT and TW functioning is obvious from all these data, molecular events underlying this relationship remain to be elucidated. Further biochemical and cell biological experiments are necessary to discriminate between different possible molecular mechanisms, including stable physical interaction between RT and TW, their enzymatic modifications of one another, chaperone activity of these proteins, or yet other possibilities.
Our combined genetic and molecular characterization of tw1 mutant highlighted the functional importance of the conserved PMT domain of TW protein. We found that tw1 mutation should result in the expression of the TW protein with just a subtle alteration of amino acid sequence [T(59) GS] of the PMT domain (Figure 1). Despite the apparent subtleness of this mutation and the fact that T(59) is not well conserved between different species, this mutation causes decreased tw function and a pronounced rotated abdomen phenotype in tw1 homozygotes (Figures 2 and 3). In addition, the tw1 phenotype is insensitive to the decreased level of rt activity, thus indicating that TW1 mutant protein can bypass the requirement for RT activity that is obligatory for wild-type TW (Figures 6–8). On the basis of protein sequence alignment of TW with other members of the POMT family (Figure 1 and data not shown) and the predicted topology of HsPOMT1 and ScPmt1p proteins (GIRRBACH et al. 2000; WILLER et al. 2003), the alteration in TW1 protein sequence maps to the lumenal terminus of the first transmembrane domain of TW protein. It is tempting to speculate that this protein region might be involved in TW–RT regulatory interactions; however, other possible mechanisms could also explain the properties of the tw1 mutation. Further biochemical and genetic experiments are necessary to clarify the properties of TW1 protein.
Several features of rt expression detected in our experiments are new and different from the previously reported expression of rt (MARTIN-BLANCO and GARCIA-BELLIDO 1996). In contrast to the previous report, we could not detect significant expression in the somatic muscle precursors and the midgut at stages 13 and 14. At the same time, we found the presence of the rt transcript in the developing epidermis at stages 14–16 (Figure 5), which was not reported earlier (MARTIN-BLANCO and GARCIA-BELLIDO 1996). A possible explanation for these differences arises from the difference between the rt cDNAs used in these studies. The sequence of cDNA clone used in our in situ hybridization experiments matches precisely the predicted sequence of the rt transcript reported in the FlyBase. We have also used this clone for in vivo expression experiments and found that it was able to rescue rt mutant phenotype (data not shown), which confirmed its functionality. On the other hand, the cDNA used in the previous study deviates from our sequence by 90 nucleotides (MARTIN-BLANCO and GARCIA-BELLIDO 1996), thus possibly representing a fusion with an irrelevant DNA fragment, which may explain the different in situ hybridization pattern reported earlier.
In vertebrates, O-mannosylation of -dystroglycan plays an important role in muscle and neural development (MARTIN 2003; MICHELE and CAMPBELL 2003). Drosophila Dystroglycan (Dg) is a fly homolog of the vertebrate Dystroglycan gene (GREENER and ROBERTS 2000; DENG et al. 2003). The predicted product of this gene, Drosophila DG protein, is structurally related to its vertebrate counterpart (DENG et al. 2003), thus representing a potential molecular target of RT/TW O-mannosyltransferase activity. Embryonic expression of Dg was detected in a dynamic fashion in many tissues, including visceral and somatic muscles, epidermis, nervous system, gut, and tracheal pits (DEKKERS et al. 2004). This expression has only partial overlap with the expression of rt and tw determined in our experiments, indicating that putative O-mannose modification of Drosophila DG protein may be limited to just a subset of DG-expressing cells. Interestingly, the expression of rt and tw in the developing epidermis revealed in our study corresponds to the region of epidermal segment border cells that are known to participate in the development of muscle attachment sites and to influence patterning of larval somatic muscles (VOLK and VIJAYRAGHAVAN 1994). These results indicate the intriguing possibility that these genes function in the epidermal muscle attachment cells, which would provide a novel mechanism for the involvement of O-mannosylation in muscle development. Further detailed characterization of rt and tw mutant phenotypes should shed light on this possibility and help elucidate the functions of these genes in Drosophila development.
ACKNOWLEDGEMENTS
We acknowledge the help of Maria Shubina and Aaron Smith with preparing twRNAi construct and balancing several transgenic strains, respectively. We also thank Vera Paulson and Stacey Whitman for their help with genetic cleaning of some mutant chromosomes. We thank John Sisson for anti-LVA antibody, Alain Debec for PDI::GFP transgenic Drosophila, and Robert Haltiwanger for the stimulating discussion that encouraged us to initiate the project. We also thank the Bloomington Drosophila Stock Center for Drosophila stocks, and the Drosophila Genomics Resource Center (Indiana University) for cDNA clones. We thank Suma Datta and Jim Erickson for their comments on the manuscript. We are also grateful to the reviewers for the constructive comments on the manuscript. This work was supported in part by Basil O'Connor Starter Scholar Research Award grant from the March of Dimes Foundation and by the start-up funds from Texas A&M University to V.M.P.
FOOTNOTES
1 Present address: Department of Biochemical Physiology, Faculty of Biology and Institute of Biomembranes, Utrecht University, Utrecht, The Netherlands.
LITERATURE CITED
ADAMS, M. D., S. E. CELNIKER, R. A. HOLT, C. A. EVANS, J. D. GOCAYNE et al., 2000 The genome sequence of Drosophila melanogaster. Science 287: 2185–2195.
ALTSCHUL, S. F., W. GISH, W. MILLER, E. W. MYERS and D. J. LIPMAN, 1990 Basic local alignment search tool. J. Mol. Biol. 215: 403–410.
BOBINNEC, Y., C. MARCAILLOU, X. MORIN and A. DEBEC, 2003 Dynamics of the endoplasmic reticulum during early development of Drosophila melanogaster. Cell Motil. Cytoskeleton 54: 217–225.
BRAND, A. H., A. S. MANOUKIAN and N. PERRIMON, 1994 Ectopic expression in Drosophila. Methods Cell Biol. 44: 635–654.
BRIDGES, C. B., and T. H. MORGAN, 1923 The third-chromosome group of mutant characters of Drosophila melanogaster. Pub. Carnegie Inst. 327: 1–251.
BUNCH, T. A., Y. GRINBLAT and L. S. GOLDSTEIN, 1988 Characterization and use of the Drosophila metallothionein promoter in cultured Drosophila melanogaster cells. Nucleic Acids Res. 16: 1043–1061.
CHIBA, A., K. MATSUMURA, H. YAMADA, T. INAZU, T. SHIMIZU et al., 1997 Structures of sialylated O-linked oligosaccharides of bovine peripheral nerve alpha-dystroglycan. The role of a novel O-mannosyl-type oligosaccharide in the binding of alpha-dystroglycan with laminin. J. Biol. Chem. 272: 2156–2162.
COOLEY, L., R. KELLEY and A. SPRADLING, 1988 Insertional mutagenesis of the Drosophila genome with single P elements. Science 239: 1121–1128.
DAVIS, B. K., 1975 A developmental and meiotic mutant of Drosophila melanogaster. Genetics 80: s25.
DAVIS, B. K., 1980 A new twist on an old mutant. Drosoph. Inf. Serv. 55: 31–33.
DEKKERS, L. C., M. C. VAN DER PLAS, P. B. VAN LOENEN, J. T. DEN DUNNEN, G. J. VAN OMMEN et al., 2004 Embryonic expression patterns of the Drosophila dystrophin-associated glycoprotein complex orthologs. Gene Expr. Patterns 4: 153–159.
DENG, W. M., M. SCHNEIDER, R. FROCK, C. CASTILLEJO-LOPEZ, E. A. GAMAN et al., 2003 Dystroglycan is required for polarizing the epithelial cells and the oocyte in Drosophila. Development 130: 173–184.
DRYSDALE, R. A., M. A. CROSBY;and THE FLYBASE CONSORTIUM, 2005 FlyBase: genes and gene models. Nucleic Acids Res. 33: D390–D395. http://flybase.org/.
ENDO, T., and T. TODA, 2003 Glycosylation in congenital muscular dystrophies. Biol. Pharm. Bull. 26: 1641–1647.
EVAN, G. I., G. K. LEWIS, G. RAMSAY and J. M. BISHOP, 1985 Isolation of monoclonal antibodies specific for human c-myc proto-oncogene product. Mol. Cell. Biol. 5: 3610–3616.
GENTZSCH, M., and W. TANNER, 1996 The PMT gene family: protein O-glycosylation in Saccharomyces cerevisiae is vital. EMBO J. 15: 5752–5759.
GIRRBACH, V., and S. STRAHL, 2003 Members of the evolutionarily conserved PMT family of protein O-mannosyltransferases form distinct protein complexes among themselves. J. Biol. Chem. 278: 12554–12562.
GIRRBACH, V., T. ZELLER, M. PRIESMEIER and S. STRAHL-BOLSINGER, 2000 Structure-function analysis of the dolichyl phosphate-mannose: protein O-mannosyltransferase ScPmt1p. J. Biol. Chem. 275: 19288–19296.
GREENER, M. J., and R. G. ROBERTS, 2000 Conservation of components of the dystrophin complex in Drosophila. FEBS Lett. 482: 13–18.
ICHIMIYA, T., H. MANYA, Y. OHMAE, H. YOSHIDA, K. TAKAHASHI et al., 2004 The twisted-abdomen phenotype of drosophila POMT1 and POMT2 mutants coincides with their heterophilic protein O-mannosyltransferase activity. J. Biol. Chem. 279: 42638–42647.
JURADO, L. A., A. COLOMA and J. CRUCES, 1999 Identification of a human homolog of the Drosophila rotated abdomen gene (POMT1) encoding a putative protein O-mannosyl-transferase, and assignment to human chromosome 9q34.1. Genomics 58: 171–180.
KOELLE, M. R., W. S. TALBOT, W. A. SEGRAVES, M. T. BENDER, P. CHERBAS et al., 1991 The Drosophila EcR gene encodes an ecdysone receptor, a new member of the steroid receptor superfamily. Cell 67: 59–77.
KOLES, K., K. D. IRVINE and V. M. PANIN, 2004 Functional characterization of Drosophila sialyltransferase. J. Biol. Chem. 279: 4346–4357.
LEE, Y. S., and R. W. CARTHEW, 2003 Making a better RNAi vector for Drosophila: use of intron spacers. Methods 30: 322–329.
LINDSLEY, D. L., and G. G. ZIMM, 1992 The Genome of Drosophila melanogaster. Academic Press, San Diego, CA.
MANYA, H., A. CHIBA, A. YOSHIDA, X. WANG, Y. CHIBA et al., 2004 Demonstration of mammalian protein O-mannosyltransferase activity: coexpression of POMT1 and POMT2 required for enzymatic activity. Proc. Natl. Acad. Sci. USA 101: 500–505.
MARTIN, P. T., 2003 Dystroglycan glycosylation and its role in matrix binding in skeletal muscle. Glycobiology 13: 55R–66R.
MARTIN, P. T., and H. H. FREEZE, 2003 Glycobiology of neuromuscular disorders. Glycobiology 13: 67R–75R.
MARTIN-BLANCO, E., and A. GARCIA-BELLIDO, 1996 Mutations in the rotated abdomen locus affect muscle development and reveal an intrinsic asymmetry in Drosophila. Proc. Natl. Acad. Sci. USA 93: 6048–6052.
MICHELE, D. E., and K. P. CAMPBELL, 2003 Dystrophin-glycoprotein complex: post-translational processing and dystroglycan function. J. Biol. Chem. 278: 15457–15460.
MICHELE, D. E., R. BARRESI, M. KANAGAWA, F. SAITO, R. D. COHN et al., 2002 Post-translational disruption of dystroglycan-ligand interactions in congenital muscular dystrophies. Nature 418: 417–422.
NIMAN, H. L., R. A. HOUGHTEN, L. E. WALKER, R. A. REISFELD, I. A. WILSON et al., 1983 Generation of protein-reactive antibodies by short peptides is an event of high frequency: implications for the structural basis of immune recognition. Proc. Natl. Acad. Sci. USA 80: 4949–4953.
OKAJIMA, T., A. XU, L. LEI and K. D. IRVINE, 2005 Chaperone activity of protein O-fucosyltransferase 1 promotes notch receptor folding. Science 307: 1599–1603.
PANIN, V. M., V. PAPAYANNOPOULOS, R. WILSON and K. D. IRVINE, 1997 Fringe modulates notch-ligand interactions. Nature 387: 908–912.
SISSON, J. C., C. FIELD, R. VENTURA, A. ROYOU and W. SULLIVAN, 2000 Lava lamp, a novel peripheral golgi protein, is required for Drosophila melanogaster cellularization. J. Cell Biol. 151: 905–918.
THOMPSON, J. D., D. G. HIGGINS and T. J. GIBSON, 1994 CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res. 22: 4673–4680.
VAN REEUWIJK, J., M. JANSSEN, C. VAN DEN ELZEN, D. BELTRAN-VALERO DE BERNABE, P. SABATELLI et al., 2005 POMT2 mutations cause alpha-dystroglycan hypoglycosylation and Walker Warburg syndrome. J. Med. Genet.: May 13 [Epub ahead of print] jmg.2005.031963.
VOLK, T., and K. VIJAYRAGHAVAN, 1994 A central role for epidermal segment border cells in the induction of muscle patterning in the Drosophila embryo. Development 120: 59–70.
WILLER, T., W. AMSELGRUBER, R. DEUTZMANN and S. STRAHL, 2002 Characterization of POMT2, a novel member of the PMT protein O-mannosyltransferase family specifically localized to the acrosome of mammalian spermatids. Glycobiology 12: 771–783.
WILLER, T., M. C. VALERO, W. TANNER, J. CRUCES and S. STRAHL, 2003 O-mannosyl glycans: from yeast to novel associations with human disease. Curr. Opin. Struct. Biol. 13: 621–630.
WILLER, T., B. PRADOS, J. M. FALCON-PEREZ, I. RENNER-MULLER, G. K. PRZEMECK et al., 2004 Targeted disruption of the Walker-Warburg syndrome gene Pomt1 in mouse results in embryonic lethality. Proc. Natl. Acad. Sci. USA 101: 14126–14131.
Department of Biochemistry & Biophysics, Texas A&M University, College Station, Texas 77843-2128(Dmitry Lyalin, Kate Koles)
O-linked carbohydrate modifications of proteins can play important roles during animal development. O-mannosylation represents a prominent example of such a post-translational protein modification that has recently been linked to mammalian muscular and neural development (reviewed in MICHELE and CAMPBELL 2003; MARTIN 2003). Abnormal O-mannosylation has been implicated in several congenital neuromuscular disorders (VAN REEUWIJK et al. 2005; reviewed in ENDO and TODA 2003; MARTIN and FREEZE 2003); while the targeted disruption of the Pomt1 gene encoding O-mannosyltransferase, an enzyme initiating the biosynthesis of O-mannosyl glycans, results in early embryonic lethality in mice (WILLER et al. 2004).
In mammalian organisms, one of the main targets of O-mannosylation is thought to be -dystroglycan (CHIBA et al. 1997), a cell adhesion glycoprotein important for the integrity of the muscle cells and neuronal migration during development (MARTIN 2003). Dystroglycan plays a central role in the dystrophin-associated glycoprotein complex (DGC) that provides a crucial link between the extracellular basal lamina and the cytoskeletal proteins (MICHELE and CAMPBELL 2003). The O-mannosyl glycans of -dystroglycan have been found to affect its binding to the extracellular matrix-associated ligand laminin, which indicates the importance of O-mannosylation for the functional integrity of DGC and its interaction with extracellular matrix (CHIBA et al. 1997; MICHELE et al. 2002).
Intriguingly, the Drosophila genome encodes homologs of all essential components necessary for the formation of DGC but with substantially less diversity (GREENER and ROBERTS 2000; DEKKERS et al. 2004), which presents an attractive possibility for using fruit flies as a model system for studying the biology of DGC and several human congenital neuromuscular diseases.
Protein O-mannosyltransferase enzymes are best characterized in yeast. A total of seven O-mannosyltransferases have been identified in Saccharomyces cerevisiae (GENTZSCH and TANNER 1996). They comprise three evolutionary conserved subfamilies of proteins, PMT1, PMT2, and PMT4 (WILLER et al. 2003). On the other hand, the family of mammalian O-mannosyltransferases includes two enzymes, POMT1 and POMT2 that belong to PMT4 and PMT2 subfamilies, respectively (JURADO et al. 1999; WILLER et al. 2002). These enzymes are thought to modify numerous serine and threonine residues of -dystroglycan with O-linked mannose, which initiates the biosynthesis of O-mannosyl glycans of dystroglycan (WILLER et al. 2003). Recently, it was demonstrated that elevated O-mannosyltransferase activity could be detected in extracts from cultured cells only if POMT1 and POMT2 were coexpressed in these cells (MANYA et al. 2004). Although this result indicated the possibility that these enzymes function as a heterocomplex, no evidence for physical interaction between POMT1 and POMT2 proteins have been obtained so far. Thus, molecular and genetic mechanisms governing the function of these proteins remain elusive.
Similar to higher animals, Drosophila has two O-mannosyltransferases, RT (encoded by the rotated abdomen gene) and DmPOMT2, which are evolutionary related to mammalian POMT1 and POMT2, respectively (MARTIN-BLANCO and GARCIA-BELLIDO 1996; WILLER et al. 2002; ICHIMIYA et al. 2004). Several mutations in the rotated abdomen gene (rt) were previously isolated and phenotypically characterized (BRIDGES and MORGAN 1923; LINDSLEY and ZIMM 1992; MARTIN-BLANCO and GARCIA-BELLIDO 1996). Two molecularly characterized mutants, rt2 and rtP, are semiviable recessive alleles associated with disruptions of the gene coding region by P-element insertion; they possibly represent null mutations (COOLEY et al. 1988; MARTIN-BLANCO and GARCIA-BELLIDO 1996). The phenotypes of these mutations include some larval and adult muscle abnormalities, as well as a prominent, up to 90° clockwise rotation of abdominal segments in adult flies (MARTIN-BLANCO and GARCIA-BELLIDO 1996). No mutations in DmPOMT2 gene have been reported so far, although the possibility that mutations in the twisted locus might represent DmPOMT2 mutants has been suggested (J. CRUCES, cited in WILLER et al. 2002). This hypothesis was further supported by "twisted abdomen" phenotype recently obtained in RNAi-mediated DmPOMT knockdown experiments (ICHIMIYA et al. 2004). Similar to their mammalian counterparts, RT and DmPOMT2 proteins have to be coexpressed to produce O-mannosyltransferase activity (ICHIMIYA et al. 2004).
The twisted alleles represent a complementation group of viable and semiviable recessive mutations on the X chromosome that also exhibit clockwise rotated abdominal segments in the adult (LINDSLEY and ZIMM 1992). Several tw mutant alleles had been isolated (DAVIS 1980; LINDSLEY and ZIMM 1992); however, none of them was characterized in detail and no molecular data are currently available for the tw locus. Most of these tw mutants have been lost and there is only one mutant allele, tw1, available from public collections (FlyBase information, DRYSDALE et al. 2005).
In this study, we have tested the hypothesis that tw corresponds to DmPOMT2. By using an expression rescue approach, we established that the tw locus represents the DmPOMT2 gene. We also characterized the molecular nature of tw1 mutation. Our immunostaining analysis of subcellular localization of tagged RT and DmPOMT2 (TW) proteins revealed their colocalization in the endoplasmic reticulum (ER) of Drosophila cells. The pattern of embryonic tw expression was analyzed by in situ hybridization and compared to the expression of rt. Our data showed striking overlap of tw and rt expression at different stages of embryogenesis. Moreover, we found a genetic interaction between tw and rt mutant alleles. All these results are consistent with the hypothesis that RT and TW, the two Drosophila protein O-mannosyltransferases, participate in the same developmental cascade and may collaborate at the molecular level, potentially functioning as an enzymatic heterocomplex.
MATERIALS AND METHODS
Mutant and transgenic Drosophila stocks:
The following Drosophila mutant alleles, chromosomal aberrations, and transgenic insertions used in the study were obtained from the Bloomington Drosophila Stock Center, Indiana University: tw1, rtP, rt2, Df(1)su(s)83, Dp(1;Y)y2sc, Act5C–GAL4-25 (second chromosome insertion), Act5C–GAL4-17 (third chromosome insertion), tubP–GAL4 (LL7), and C155–GAL4. The rt571 allele (also designated as EP(3)0571) was obtained from the Exelixis EP collection (Exelixis; San Francisco). In our experiments, we used three different rt alleles to exclude potential influence of genetic background and possible peculiarities of some alleles. All three alleles of rt include an independent P-element insertion in the first exon of the gene and likely represent null or strong hypomorphic mutations (see MARTIN-BLANCO and GARCIA-BELLIDO 1996; DRYSDALE et al. 2005; and Figure 6). PDI::GFP transgenic flies were kindly provided by Alain Debec (BOBINNEC et al. 2003). The following transgenic insertions were generated by P-element mediated transformation: UAS–DmPOMT2 and UAS–twRNAi-39 (on the second chromosome) and UAS–twRNAi-77 (on the third chromosome).
Sequencing of the DmPOMT2 gene in tw1 mutant flies:
The DNA fragment of DmPOMT2 locus was PCR amplified from the genomic DNA of tw1 homozygous flies using primers 5' CGTGGCCAGGATAACAACACTGGC 3' and 5' AACGTTGACAGGGTTGTGGGTGTGGT 3'. The amplified genomic region included the transcribed region [determined by the alignment of the DmPOMT2 cDNA (clone LP01681) with Drosophila genomic sequences (ADAMS et al. 2000)], along with the 2-kb upstream and 500-bp downstream fragments. The PCR product was sequenced and several deviations of the DmPOMT2 locus sequence from the corresponding wild-type genomic sequence determined by the Drosophila genome project (ADAMS et al. 2000) were uncovered. The alterations include: CA(249,250) AGGAT; C(472) T; T(2545) C (nucleotide numbers are relative to the beginning of coding region). The first alteration is predicted to affect the translated protein sequence [T(59) GS], while the two other mutations are silent.
DNA constructs for cell-culture expression and fly transformation
The cDNA clones were obtained from the Drosophila Genomics Resource Center at Indiana University. For cell culture expression experiments, the coding region of DmPOMT2 was PCR amplified from the DmPOMT2 cDNA (LP01681) using PCR primers containing BglII and XbaI restriction sites. Using these sites, the PCR product was inserted into the pMK33 vector (KOELLE et al. 1991). The final construct, pMK–DmPOMT2, also included a short DNA fragment encoding the HA tag (NIMAN et al. 1983); it was obtained by annealing two synthesized oligonucleotides and then introduced into DmPOMT2 cDNA using SanDI restriction site. The resulting construct encoded a DmPOMT2 (TW) protein with HA-tag inserted into nonconserved region immediately after G(650). The same construct was subcloned into the pUAST vector (BRAND et al. 1994) to obtain pUAST–DmPOMT2 for fly transformation and in vivo expression. Functionality of the construct was confirmed by Western-blot analysis (data not shown) and immunostaining of pMK–DmPOMT2-transfected S2 cells (Figure 4), as well as by an in vivo rescue assay (Figure 2). For the expression of rt in Drosophila cell culture, the coding region of rt was PCR amplified from the rt cDNA (clone RE30211) and inserted in the pRMHA3 vector (BUNCH et al. 1988). In the final expression construct, the rt coding region was modified by the addition of a short DNA fragment encoding two MYC tags (EVAN et al. 1985) immediately following the last amino acid-coding triplet of rt. As revealed by sequencing, the RE30211 clone includes a short unspliced intron preceding the last coding exon of the gene. Efficient expression of the construct and removal of the intron by in-cell splicing was confirmed using anti-MYC immunostaining of pRMHA3–rt-MYC-transfected cells (Figure 4) and Western-blot analysis of RT–MYC protein expressed in Drosophila S2 cells (data not shown). The functionality of the MYC-tagged RT protein was also confirmed in vivo by its ability to rescue the rt mutant phenotype (data not shown). The UAS–twRNAi construct was produced essentially according to the published strategy (LEE and CARTHEW 2003). Briefly, the two PCR-amplified DNA fragments of DmPOMT2 gene, both including the 6th exon (700 bp) and one also including the following intronal sequence, were ligated together in the opposite orientation. The resulting inverted repeat was cloned into the pUAST vector using PCR-introduced EcoRI and XbaI restriction sites. The UAS–twRNAi construct is predicted to produce 700 bp hairpin dsRNA with 12 unpaired bases of loop region.
Cell culture:
Drosophila S2 cells were maintained and transfected using the protocols described earlier (KOLES et al. 2004).
Immunostaining and epifluorescent microscopy:
Expression of the UAST–DmPOMT2 construct was induced in salivary glands using the C155–GAL4 driver. Third instar larvae were dissected, fixed, and stained as described earlier (PANIN et al. 1997). The following primary antibodies and corresponding dilutions were used for immunostaining: rabbit anti-LVA (1:2000) (a gift from John Sisson, University of Texas, Austin, TX); mouse and rabbit anti-HA (1:1000); mouse anti-MYC (1: 1000) (BabCo, Berkeley, CA). We used the following fluorescent secondary donkey antibodies: anti-mouse-Cy3 (1:250); anti-rabbit-FITC (1:150); and anti-mouse-Cy5 (1:150) (Jackson Laboratories). Digital images were obtained using Zeiss Axioplan 2 fluorescent microscope with the ApoTome module for optical sectioning.
In situ hybridization was performed as described earlier (KOLES et al. 2004) using the tw and rt cDNA clones LP01681 and RE30211, respectively, as templates for the synthesis of DIG-labeled probes. Every in situ hybridization experiment included a negative control staining with a probe transcribed from the corresponding antiparallel cDNA sequence (data not shown).
Phenotype analysis:
All flies were grown at 25°. The rotation of abdomen was scored in CO2-anesthetized flies 1 day after eclosion (to eliminate possible influence of aging or altered morphology in very young flies). The rotation angle was measured from posterior viewpoint using Nikon SMZ microscope with a protractor reticle objective.
RESULTS
The tw1 mutant allele contains a mutation within the DmPOMT2 gene region encoding the conserved PMT domain:
The novel Drosophila O-mannosyltransferase 2 gene was initially found by BLAST (ALTSCHUL et al. 1990) searches of Drosophila genomic sequences on the basis of the homology of its conceptual translation to the sequences of RT, yeast, and mammalian O-mannosyltransferases (WILLER et al. 2002; FlyBase information, DRYSDALE et al. 2005). The software prediction was further confirmed by the identification and sequencing of the corresponding full-length cDNA clone LP01681 from the Drosophila EST collection (WILLER et al. 2002; ICHIMIYA et al. 2004). We also identified a mosquito EST sequence (GenBank accession no. XM 312249) encoding a protein with high homology to DmPOMT2, which suggested that this sequence corresponds to the mosquito ortholog of DmPOMT2, AgPOMT2 (Figure 1). The similarity between the products of rt and DmPOMT2 (Figure 1), along with functional data on the homologous yeast PMT family members, allowed us to suggest that both Drosophila genes might function within the same biochemical pathway. This scenario would predict similar phenotypes for rt and DmPOMT2 mutations. Interestingly, we noticed that the chromosomal position of DmPOMT2 (1D4) maps to the cytogenetic region that also includes tw (1C3–D4), the complementation group of mutations causing a clockwise rotation of the adult abdomen (LINDSLEY and ZIMM 1992). This phenotype is strikingly similar to that of rt mutants (Figure 2, A–C). These data indicated the possibility that tw corresponds to DmPOMT2. This hypothesis has also been mentioned by other researchers (J. CRUCES, cited in WILLER et al. 2002) and supported by DmPOMT2 RNAi experiments that generated rt-like abdomen rotation phenotype (ICHIMIYA et al. 2004). Thus, we decided to sequence the genomic region of DmPOMT2 in tw1 mutant flies carrying a semiviable recessive twisted allele. The tw1 homozygous female and hemizygous male flies have a fully penetrant phenotype that manifests in the clockwise rotation of the abdomen (20°–40°) along the anteroposterior axis, when scored from a posterior viewpoint (DAVIS 1975; Figure 2C). Genetically, the tw1 mutation behaves as a hypomorphic allele since its phenotype is significantly enhanced in tw1-hemizygous females carrying a deficiency that includes the tw locus [tw1/Df(1)su(s)83], while the phenotype is completely rescued in males carrying a duplication of the tw locus on the Y chromosome [tw1/Dp(1;Y)y2sc] (Figure 2C and Figure 3). We sequenced the DmPOMT2 locus including the transcribed region together with 2-kb upstream and 500-bp downstream fragments. The sequencing revealed several deviations of the DmPOMT2 genomic sequence from the corresponding wild-type sequence determined by the Drosophila genome project (ADAMS et al. 2000). All but one of the sequence alterations likely represent polymorphisms or mutations neutral for the gene's function (see MATERIALS AND METHODS). Notably, one of the identified mutations maps to the coding region and results in the amino acid substitution/insertion [T(59) GS] within the conserved PMT domain of DmPOMT2 (Figure 1), which suggests an altered functionality of the DmPOMT2 protein in the tw1 mutants. This finding further supported the hypothesis that the phenotype of tw mutants results from mutations in DmPOMT2.
The phenotype of tw1 mutation can be rescued by DmPOMT2 expression:
To confirm that the DmPOMT2 mutation(s) causes the abdominal rotation phenotype in tw1 mutants, we carried out a rescue assay. To this end, we induced DmPOMT2 expression in the tw1 mutant background using the UAS/GAL in vivo expression system (BRAND et al. 1994).
We produced UAS–DmPOMT2 transgenic flies and crossed them to flies carrying a ubiquitous Act5C–GAL4-17 driver (see MATERIALS AND METHODS). There were no visible defects in abdominal morphology of the flies from either parental stocks or UAS–DmPOMT2/+; Act5C–GAL4-17/+ progeny of the cross (data not shown). The presence of the UAS–DmPOMT2 or Act5C–GAL4 transgene alone did not modify the phenotype of tw1 mutants (Figure 3). At the same time, the complete rescue of the abdomen rotation phenotype was observed in tw1 hemizygous male and homozygous female flies carrying both UAS–DmPOMT2 and Act5C–GAL4-17 together (Figure 3). Similar results were obtained when DmPOMT2 was expressed using Act5C–GAL4-25 and tubP–GAL4 drivers (data not shown). These rescue experiments proved that tw indeed represents the DmPOMT2 gene. Thus, we suggest changing the designation of the DmPOMT2 (CG12311) gene to "twisted" (tw), the name of the originally discovered mutants that is in an agreement with the traditional nomenclature of Drosophila genetics (LINDSLEY and ZIMM 1992). Below, we use "tw" and "TW" to designate the DmPOMT2 gene and its protein product, respectively.
Subcellular localization of RT and TW proteins:
Glycosyltransferases that modify secreted glycoproteins commonly function in the Golgi apparatus. Interestingly, yeast PMT family members localize to the ER subcellular compartment (WILLER et al. 2003). The localization of O-mannosyltransferase proteins in animal cells has been reported only for human POMT2 protein expressed in human culture cells (WILLER et al. 2002). In that study, the C-terminally tagged POMT2 protein was detected in the ER by immunostaining; however, the functionality of this fusion protein remains undetermined. Thus, we decided to investigate the subcellular localization of the TW protein, the Drosophila ortholog of mammalian POMT2, using the UAS–DmPOMT2 transgenic construct that was functional in our rescue assay (Figure 2). Immunostaining for TW protein expressed in vivo in the salivary gland cells of Drosophila larvae revealed its colocalization with an ER marker, PDI-GFP (BOBINNEC et al. 2003). At the same time, the localization of TW showed minimal overlap with a Golgi marker, the LVA protein (SISSON et al. 2000), when TW was expressed in Drosophila S2 cells (Figure 4, D–F). Double immunostaining of S2 cells expressing both RT and TW proteins demonstrated their colocalization inside the cell (Figure 4, G–I). Thus, we concluded that both RT and TW proteins reside in the ER subcellular compartment.
Analysis of tw and rt expression during Drosophila embryogenesis by in situ hybridization:
Detailed analysis of the spatial and temporal patterns of gene expression by in situ hybridization can provide important information about the functioning and regulation of the gene. The pattern of embryonic rt expression was previously analyzed by in situ hybridization (MARTIN-BLANCO and GARCIA-BELLIDO 1996). The pattern of tw (DmPOMT2) expression has been reported only for embryonic stage 10, while the expression at other stages was estimated only using a real-time PCR assay (ICHIMIYA et al. 2004). Thus, we decided to perform in situ hybridization analysis of tw expression and compare it with the pattern of rt expression at different embryonic stages. In agreement with previous reports (MARTIN-BLANCO and GARCIA-BELLIDO 1996; ICHIMIYA et al. 2004), we found no significant expression of tw at early embryonic stages, while rt mRNA was detected at stages 5 and 6, which probably indicates the presence of maternally provided transcript (Figure 5, A and D). This early rt expression decays quickly and it is not readily detectable after stage 7. Although we did detect some weak staining for tw and rt expression during the stages of germband extension (stages 10 and 11), this staining was diffuse and barely detectable above the background (data not shown). Thus, we decided to concentrate our analysis on tw and rt expression at later stages, when the expression of these genes is readily detectable. The prominent expression of both genes, tw and rt, appears at early stage 14 that corresponds to the period of active muscle differentiation (Figure 5, B and E). However, we have not detected significant expression of tw or rt in the developing somatic muscle cells. Instead, the expression of these genes appeared to be pronounced in other tissues, including certain developing epidermal cells, as well as hindgut and foregut regions. At that stage, the expression of tw is also present in the developing trachea, while rt expression appears in the tracheal cells slightly later, during stage 15. Thus, remarkably overlapping patterns of rt and tw expression are established by stages 14 and 15, and this expression appears to persist through the late embryogenesis (Figure 5, C and F).
Genetic interaction between tw and rt genes:
The similarity of mutant phenotypes (Figure 2) strongly suggests that both genes function in the same developmental cascade. To test this possibility, we assayed the genetic interaction between tw1 and three rt alleles, rt2, rtP, and rt571. Somewhat surprisingly, we found that the presence of tw1 significantly suppressed the abdomen rotation phenotype of rt mutants (Figure 6A). Notably, this effect is dominant since even one copy of tw1 mutation in females is sufficient for the suppression. In fact, the suppression of rt phenotype in tw1 homozygous or hemizygous flies appears to be nearly complete since the phenotype of tw1/tw1(or tw1/Y); rt/rt double mutants is virtually indistinguishable from the phenotype of tw1-homo-/hemizygous mutants alone (Figure 6B).
We considered two competing explanations of these results. First, the relative excess of TW as compared to RT in the rt mutants might have a negative effect on the O-mannosylation pathway. In this case, a potential decrease of TW might explain the suppression of rt phenotype in tw1 mutant background. Alternatively, the suppression of rt phenotype by tw1 might be explained by a special feature of the tw1 allele, which would not implicate the dependence of phenotype on relative concentrations of RT and TW. To discriminate between these two possibilities, we analyzed further the genetic interaction between rt and tw.
The first possibility would predict that an increase of tw activity would increase the severity of rt mutant phenotype, while a decrease of tw would result in rt mutant phenotype suppression. To test this prediction, we varied the level of tw in rt mutants by several alternative ways: (i) by adding an extra copy of the tw locus [using duplication Dp(1;Y)y2sc], (ii) by overexpressing TW using UAS–GAL4 system, and (iii) by decreasing the tw activity via UAS–twRNAi construct expression. Neither the increase of tw expression (Figure 7A) nor the decrease of tw activity (Figure 7B) revealed the predicted sensitivity of rt mutant phenotype to varied concentrations of TW. Hence, we interpret the dominant suppression of rt phenotype by the tw1 mutation as a special feature of the tw1 allele that somehow bypasses the requirement for rt activity in the genetic pathway. This interpretation is further supported by a synergistic effect of tw RNAi and heterozygous rt mutant background (Figure 8), which revealed a positive interaction between rt and tw, indicating their close collaboration within the pathway. These data also support the hypothesis that the O-mannosylation pathway requires the simultaneous activities of both rt and tw genes (ICHIMIYA et al. 2004). This simultaneity requirement can explain the absence of any significant effect of varied TW levels on the phenotype of tested rt mutants that represent very strong hypomorphic or amorphic mutations and presumably lack the RT protein (see MARTIN-BLANCO and GARCIA-BELLIDO 1996; rt phenotypes in Figure 6A).
DISCUSSION
The family of protein O-mannosyltransferases in Drosophila includes two members, RT and DmPOMT2, and exhibits obvious evolutionary relation to the mammalian POMT protein family (WILLER et al. 2003). While several mutations in the rt locus have been previously isolated and the rt gene has been molecularly characterized (LINDSLEY and ZIMM 1992; MARTIN-BLANCO and GARCIA-BELLIDO 1996), the novel DmPOMT2 gene (CG12311) has been described only recently (FlyBase; ICHIMIYA et al. 2004), and no mutations in DmPOMT2 have been reported so far. In this study, we have established the relationship between DmPOMT2 and twisted, the previously isolated complementation group of recessive mutations with the characteristic phenotype of a clockwise twisted abdomen. Our results demonstrate that the tw1 semiviable recessive allele is associated with a mutation in the coding region of the DmPOMT2 gene, which alters the amino acid sequence of the conserved PMT domain of DmPOMT2 protein. We have also found that this mutation is associated with the decrease of tw activity, since the tw1 phenotype is enhanced over deficiency for tw locus, and it can be completely rescued by genetic duplication including the tw gene or by ubiquitous ectopic expression of the UAS–DmPOMT2 construct. Thus, tw1 represents the first molecularly and genetically characterized mutant allele of the DmPOMT2 gene.
Interestingly, we found that tw1 could efficiently suppress the phenotype of three tested rt alleles (Figure 6). This finding was unexpected, since previously it was reported that RT and DmPOMT2 collaborate biochemically in vitro, and DmPOMT2 RNAi phenotype is enhanced in the rtP heterozygous background (ICHIMIYA et al. 2004). Yet the possibility existed that the suppression is the result of a decreased activity of tw in tw1 mutants. In this case, the suppression would suggest that an unbalanced relative increase of TW in rt mutants had a negative effect on the pathway [which, for instance, might result from a competition between nonproductive homomeric TW complexes and active RT–TW heterocomplexes (GIRRBACH and STRAHL 2003; ICHIMIYA et al. 2004)]. To test this possibility, we further analyzed the genetic interaction between tw and rt by varying the level of tw in the rt mutant background. We found that the phenotype of rt homozygous or heteroallelic mutants was neither significantly sensitive to an increase of tw expression by introducing a duplication of the tw locus or by ectopic expression of UAS–DmPOMT2 construct nor significantly sensitive to a decrease of tw activity by UAS–twRNAi expression (Figure 7 A and B). At the same time, we also confirmed the synergistic effect of the partial reduction of tw and rt activities on the mutant phenotype using UAS–twRNAi construct expression in rt heterozygotes (Figure 8). Thus, we ruled out the possibility of negative effect of tw in rt mutants and concluded that the suppression of rt phenotype is a special feature of tw1 allele.
Taken together, our results indicate that both tw and rt are involved in the same developmental pathway, where they execute nonredundant functions. Insensitivity of rt mutant phenotype to the varied level of tw expression formally characterizes rt as epistatic to tw. At the same time, the tw1 mutation could dominantly suppress the phenotype of strong (probably amorphic) rt mutations, thus revealing its epistatic position relative to rt (Figure 6). We interpret these mutually epistatic relationships between rt and tw as the evidence for possible functioning of their protein products within the same molecular complex or being involved in a regulatory interaction within the same biochemical pathway. This conclusion is in agreement with previously reported data on simultaneous requirement of RT and DmPOMT2 for their biochemical activity in vitro (ICHIMIYA et al. 2004). The conclusion is also consistent with our other observations presented here, including (i) essentially identical phenotypes of clockwise abdomen rotation in both rt and tw mutants (Figure 2), (ii) the subcellular colocalization of RT and TW proteins within the ER compartment in Drosophila cells (Figure 4), and (iii) the overlapping pattern of rt and tw expression during different stages of embryogenesis (Figure 5). Although the close relationship between RT and TW functioning is obvious from all these data, molecular events underlying this relationship remain to be elucidated. Further biochemical and cell biological experiments are necessary to discriminate between different possible molecular mechanisms, including stable physical interaction between RT and TW, their enzymatic modifications of one another, chaperone activity of these proteins, or yet other possibilities.
Our combined genetic and molecular characterization of tw1 mutant highlighted the functional importance of the conserved PMT domain of TW protein. We found that tw1 mutation should result in the expression of the TW protein with just a subtle alteration of amino acid sequence [T(59) GS] of the PMT domain (Figure 1). Despite the apparent subtleness of this mutation and the fact that T(59) is not well conserved between different species, this mutation causes decreased tw function and a pronounced rotated abdomen phenotype in tw1 homozygotes (Figures 2 and 3). In addition, the tw1 phenotype is insensitive to the decreased level of rt activity, thus indicating that TW1 mutant protein can bypass the requirement for RT activity that is obligatory for wild-type TW (Figures 6–8). On the basis of protein sequence alignment of TW with other members of the POMT family (Figure 1 and data not shown) and the predicted topology of HsPOMT1 and ScPmt1p proteins (GIRRBACH et al. 2000; WILLER et al. 2003), the alteration in TW1 protein sequence maps to the lumenal terminus of the first transmembrane domain of TW protein. It is tempting to speculate that this protein region might be involved in TW–RT regulatory interactions; however, other possible mechanisms could also explain the properties of the tw1 mutation. Further biochemical and genetic experiments are necessary to clarify the properties of TW1 protein.
Several features of rt expression detected in our experiments are new and different from the previously reported expression of rt (MARTIN-BLANCO and GARCIA-BELLIDO 1996). In contrast to the previous report, we could not detect significant expression in the somatic muscle precursors and the midgut at stages 13 and 14. At the same time, we found the presence of the rt transcript in the developing epidermis at stages 14–16 (Figure 5), which was not reported earlier (MARTIN-BLANCO and GARCIA-BELLIDO 1996). A possible explanation for these differences arises from the difference between the rt cDNAs used in these studies. The sequence of cDNA clone used in our in situ hybridization experiments matches precisely the predicted sequence of the rt transcript reported in the FlyBase. We have also used this clone for in vivo expression experiments and found that it was able to rescue rt mutant phenotype (data not shown), which confirmed its functionality. On the other hand, the cDNA used in the previous study deviates from our sequence by 90 nucleotides (MARTIN-BLANCO and GARCIA-BELLIDO 1996), thus possibly representing a fusion with an irrelevant DNA fragment, which may explain the different in situ hybridization pattern reported earlier.
In vertebrates, O-mannosylation of -dystroglycan plays an important role in muscle and neural development (MARTIN 2003; MICHELE and CAMPBELL 2003). Drosophila Dystroglycan (Dg) is a fly homolog of the vertebrate Dystroglycan gene (GREENER and ROBERTS 2000; DENG et al. 2003). The predicted product of this gene, Drosophila DG protein, is structurally related to its vertebrate counterpart (DENG et al. 2003), thus representing a potential molecular target of RT/TW O-mannosyltransferase activity. Embryonic expression of Dg was detected in a dynamic fashion in many tissues, including visceral and somatic muscles, epidermis, nervous system, gut, and tracheal pits (DEKKERS et al. 2004). This expression has only partial overlap with the expression of rt and tw determined in our experiments, indicating that putative O-mannose modification of Drosophila DG protein may be limited to just a subset of DG-expressing cells. Interestingly, the expression of rt and tw in the developing epidermis revealed in our study corresponds to the region of epidermal segment border cells that are known to participate in the development of muscle attachment sites and to influence patterning of larval somatic muscles (VOLK and VIJAYRAGHAVAN 1994). These results indicate the intriguing possibility that these genes function in the epidermal muscle attachment cells, which would provide a novel mechanism for the involvement of O-mannosylation in muscle development. Further detailed characterization of rt and tw mutant phenotypes should shed light on this possibility and help elucidate the functions of these genes in Drosophila development.
ACKNOWLEDGEMENTS
We acknowledge the help of Maria Shubina and Aaron Smith with preparing twRNAi construct and balancing several transgenic strains, respectively. We also thank Vera Paulson and Stacey Whitman for their help with genetic cleaning of some mutant chromosomes. We thank John Sisson for anti-LVA antibody, Alain Debec for PDI::GFP transgenic Drosophila, and Robert Haltiwanger for the stimulating discussion that encouraged us to initiate the project. We also thank the Bloomington Drosophila Stock Center for Drosophila stocks, and the Drosophila Genomics Resource Center (Indiana University) for cDNA clones. We thank Suma Datta and Jim Erickson for their comments on the manuscript. We are also grateful to the reviewers for the constructive comments on the manuscript. This work was supported in part by Basil O'Connor Starter Scholar Research Award grant from the March of Dimes Foundation and by the start-up funds from Texas A&M University to V.M.P.
FOOTNOTES
1 Present address: Department of Biochemical Physiology, Faculty of Biology and Institute of Biomembranes, Utrecht University, Utrecht, The Netherlands.
LITERATURE CITED
ADAMS, M. D., S. E. CELNIKER, R. A. HOLT, C. A. EVANS, J. D. GOCAYNE et al., 2000 The genome sequence of Drosophila melanogaster. Science 287: 2185–2195.
ALTSCHUL, S. F., W. GISH, W. MILLER, E. W. MYERS and D. J. LIPMAN, 1990 Basic local alignment search tool. J. Mol. Biol. 215: 403–410.
BOBINNEC, Y., C. MARCAILLOU, X. MORIN and A. DEBEC, 2003 Dynamics of the endoplasmic reticulum during early development of Drosophila melanogaster. Cell Motil. Cytoskeleton 54: 217–225.
BRAND, A. H., A. S. MANOUKIAN and N. PERRIMON, 1994 Ectopic expression in Drosophila. Methods Cell Biol. 44: 635–654.
BRIDGES, C. B., and T. H. MORGAN, 1923 The third-chromosome group of mutant characters of Drosophila melanogaster. Pub. Carnegie Inst. 327: 1–251.
BUNCH, T. A., Y. GRINBLAT and L. S. GOLDSTEIN, 1988 Characterization and use of the Drosophila metallothionein promoter in cultured Drosophila melanogaster cells. Nucleic Acids Res. 16: 1043–1061.
CHIBA, A., K. MATSUMURA, H. YAMADA, T. INAZU, T. SHIMIZU et al., 1997 Structures of sialylated O-linked oligosaccharides of bovine peripheral nerve alpha-dystroglycan. The role of a novel O-mannosyl-type oligosaccharide in the binding of alpha-dystroglycan with laminin. J. Biol. Chem. 272: 2156–2162.
COOLEY, L., R. KELLEY and A. SPRADLING, 1988 Insertional mutagenesis of the Drosophila genome with single P elements. Science 239: 1121–1128.
DAVIS, B. K., 1975 A developmental and meiotic mutant of Drosophila melanogaster. Genetics 80: s25.
DAVIS, B. K., 1980 A new twist on an old mutant. Drosoph. Inf. Serv. 55: 31–33.
DEKKERS, L. C., M. C. VAN DER PLAS, P. B. VAN LOENEN, J. T. DEN DUNNEN, G. J. VAN OMMEN et al., 2004 Embryonic expression patterns of the Drosophila dystrophin-associated glycoprotein complex orthologs. Gene Expr. Patterns 4: 153–159.
DENG, W. M., M. SCHNEIDER, R. FROCK, C. CASTILLEJO-LOPEZ, E. A. GAMAN et al., 2003 Dystroglycan is required for polarizing the epithelial cells and the oocyte in Drosophila. Development 130: 173–184.
DRYSDALE, R. A., M. A. CROSBY;and THE FLYBASE CONSORTIUM, 2005 FlyBase: genes and gene models. Nucleic Acids Res. 33: D390–D395. http://flybase.org/.
ENDO, T., and T. TODA, 2003 Glycosylation in congenital muscular dystrophies. Biol. Pharm. Bull. 26: 1641–1647.
EVAN, G. I., G. K. LEWIS, G. RAMSAY and J. M. BISHOP, 1985 Isolation of monoclonal antibodies specific for human c-myc proto-oncogene product. Mol. Cell. Biol. 5: 3610–3616.
GENTZSCH, M., and W. TANNER, 1996 The PMT gene family: protein O-glycosylation in Saccharomyces cerevisiae is vital. EMBO J. 15: 5752–5759.
GIRRBACH, V., and S. STRAHL, 2003 Members of the evolutionarily conserved PMT family of protein O-mannosyltransferases form distinct protein complexes among themselves. J. Biol. Chem. 278: 12554–12562.
GIRRBACH, V., T. ZELLER, M. PRIESMEIER and S. STRAHL-BOLSINGER, 2000 Structure-function analysis of the dolichyl phosphate-mannose: protein O-mannosyltransferase ScPmt1p. J. Biol. Chem. 275: 19288–19296.
GREENER, M. J., and R. G. ROBERTS, 2000 Conservation of components of the dystrophin complex in Drosophila. FEBS Lett. 482: 13–18.
ICHIMIYA, T., H. MANYA, Y. OHMAE, H. YOSHIDA, K. TAKAHASHI et al., 2004 The twisted-abdomen phenotype of drosophila POMT1 and POMT2 mutants coincides with their heterophilic protein O-mannosyltransferase activity. J. Biol. Chem. 279: 42638–42647.
JURADO, L. A., A. COLOMA and J. CRUCES, 1999 Identification of a human homolog of the Drosophila rotated abdomen gene (POMT1) encoding a putative protein O-mannosyl-transferase, and assignment to human chromosome 9q34.1. Genomics 58: 171–180.
KOELLE, M. R., W. S. TALBOT, W. A. SEGRAVES, M. T. BENDER, P. CHERBAS et al., 1991 The Drosophila EcR gene encodes an ecdysone receptor, a new member of the steroid receptor superfamily. Cell 67: 59–77.
KOLES, K., K. D. IRVINE and V. M. PANIN, 2004 Functional characterization of Drosophila sialyltransferase. J. Biol. Chem. 279: 4346–4357.
LEE, Y. S., and R. W. CARTHEW, 2003 Making a better RNAi vector for Drosophila: use of intron spacers. Methods 30: 322–329.
LINDSLEY, D. L., and G. G. ZIMM, 1992 The Genome of Drosophila melanogaster. Academic Press, San Diego, CA.
MANYA, H., A. CHIBA, A. YOSHIDA, X. WANG, Y. CHIBA et al., 2004 Demonstration of mammalian protein O-mannosyltransferase activity: coexpression of POMT1 and POMT2 required for enzymatic activity. Proc. Natl. Acad. Sci. USA 101: 500–505.
MARTIN, P. T., 2003 Dystroglycan glycosylation and its role in matrix binding in skeletal muscle. Glycobiology 13: 55R–66R.
MARTIN, P. T., and H. H. FREEZE, 2003 Glycobiology of neuromuscular disorders. Glycobiology 13: 67R–75R.
MARTIN-BLANCO, E., and A. GARCIA-BELLIDO, 1996 Mutations in the rotated abdomen locus affect muscle development and reveal an intrinsic asymmetry in Drosophila. Proc. Natl. Acad. Sci. USA 93: 6048–6052.
MICHELE, D. E., and K. P. CAMPBELL, 2003 Dystrophin-glycoprotein complex: post-translational processing and dystroglycan function. J. Biol. Chem. 278: 15457–15460.
MICHELE, D. E., R. BARRESI, M. KANAGAWA, F. SAITO, R. D. COHN et al., 2002 Post-translational disruption of dystroglycan-ligand interactions in congenital muscular dystrophies. Nature 418: 417–422.
NIMAN, H. L., R. A. HOUGHTEN, L. E. WALKER, R. A. REISFELD, I. A. WILSON et al., 1983 Generation of protein-reactive antibodies by short peptides is an event of high frequency: implications for the structural basis of immune recognition. Proc. Natl. Acad. Sci. USA 80: 4949–4953.
OKAJIMA, T., A. XU, L. LEI and K. D. IRVINE, 2005 Chaperone activity of protein O-fucosyltransferase 1 promotes notch receptor folding. Science 307: 1599–1603.
PANIN, V. M., V. PAPAYANNOPOULOS, R. WILSON and K. D. IRVINE, 1997 Fringe modulates notch-ligand interactions. Nature 387: 908–912.
SISSON, J. C., C. FIELD, R. VENTURA, A. ROYOU and W. SULLIVAN, 2000 Lava lamp, a novel peripheral golgi protein, is required for Drosophila melanogaster cellularization. J. Cell Biol. 151: 905–918.
THOMPSON, J. D., D. G. HIGGINS and T. J. GIBSON, 1994 CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res. 22: 4673–4680.
VAN REEUWIJK, J., M. JANSSEN, C. VAN DEN ELZEN, D. BELTRAN-VALERO DE BERNABE, P. SABATELLI et al., 2005 POMT2 mutations cause alpha-dystroglycan hypoglycosylation and Walker Warburg syndrome. J. Med. Genet.: May 13 [Epub ahead of print] jmg.2005.031963.
VOLK, T., and K. VIJAYRAGHAVAN, 1994 A central role for epidermal segment border cells in the induction of muscle patterning in the Drosophila embryo. Development 120: 59–70.
WILLER, T., W. AMSELGRUBER, R. DEUTZMANN and S. STRAHL, 2002 Characterization of POMT2, a novel member of the PMT protein O-mannosyltransferase family specifically localized to the acrosome of mammalian spermatids. Glycobiology 12: 771–783.
WILLER, T., M. C. VALERO, W. TANNER, J. CRUCES and S. STRAHL, 2003 O-mannosyl glycans: from yeast to novel associations with human disease. Curr. Opin. Struct. Biol. 13: 621–630.
WILLER, T., B. PRADOS, J. M. FALCON-PEREZ, I. RENNER-MULLER, G. K. PRZEMECK et al., 2004 Targeted disruption of the Walker-Warburg syndrome gene Pomt1 in mouse results in embryonic lethality. Proc. Natl. Acad. Sci. USA 101: 14126–14131.
Department of Biochemistry & Biophysics, Texas A&M University, College Station, Texas 77843-2128(Dmitry Lyalin, Kate Koles)