SEARCH

SEARCH BY CITATION

Keywords:

  • Drosophila;
  • embryo;
  • Nudel;
  • glycosaminoglycan;
  • protease

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. Discussion
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

Dorsoventral patterning of the Drosophila embryo requires Nudel, a large mosaic protein with a protease domain. Previous studies have implicated Nudel's protease domain as the trigger of a proteolytic cascade that activates the Toll signaling pathway to establish dorsoventral polarity in the embryo. However, the function of other regions of Nudel has been unclear. By using two-dimensional gel electrophoresis and site-directed mutagenesis, we have obtained evidence that the N-terminal region of Nudel contains a site for glycosaminoglycan (GAG) attachment that is required for dorsoventral patterning. Disruption of this site blocks a disulfide-based association between N- and C-terminal Nudel polypeptides and proteolytic activation of Nudel's protease domain. We discuss how a GAG chain on Nudel might be required for Nudel protease activation. © 2002 Wiley-Liss, Inc.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. Discussion
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

Dorsoventral patterning of the Drosophila embryo is initiated by a signal that promotes ventral and lateral development. This signal is conveyed by an extracellular ligand that activates the receptor Toll, which in turn triggers intracellular signaling necessary to establish dorsoventral asymmetry within the embryo (Morisato and Anderson, 1995). The Toll ligand is produced by a proteolytic processing reaction that occurs at the end of a serine protease cascade and only on the ventral side of the embryo (Morisato and Anderson, 1994; Schneider et al., 1994). Failure to generate the Toll ligand, and thus to activate Toll signaling, results in a “dorsalized” embryo having only dorsal structures.

Activation of the Toll ligand is under spatial control of a ventrally restricted factor generated during oogenesis by follicle cells that surround the oocyte. Genetic studies have revealed that expression in ventral follicle cells of the pipe and windbeutel genes is required for synthesis of this factor (Nilson and Schüpbach, 1998). The Pipe protein is a homologue of heparan sulfate 2-O-sulfotransferase, an enzyme that modifies glycosaminoglycan (GAG) chains as found in proteoglycans of the extracellular matrix, whereas the Windbeutel protein is apparently required to localize Pipe to the Golgi apparatus (Konsolaki and Schüpbach, 1998; Sen et al., 1998, 2000). These findings have suggested the involvement of a specifically modified GAG chain, secreted by ovarian follicle cells, in restricting Toll ligand activation to the ventral side of the embryo.

The focus of this study is a third gene, nudel, which like pipe and windbeutel is expressed in ovarian follicle cells and required for local activation of the Toll ligand. The nudel gene encodes a large (>300 kDa) extracellular protein that is modular in structure, with a serine protease domain, multiple copies of a protein-binding motif known as the LDL-receptor type A repeat, and many sites for N- and O-linked glycosylation, including GAG attachment (Hong and Hashimoto, 1995). Previous studies have implicated the Nudel protease as the initiator of the protease cascade that produces the Toll ligand, as well as having a role in cross-linking of eggshell proteins (LeMosy et al., 1998; LeMosy and Hashimoto, 2000).

The function of Nudel protein regions outside the protease domain has been unclear. Early genetic studies suggested that Nudel is functionally modular, with the protease domain involved in dorsoventral patterning and another region required for structural integrity of the egg (Hong and Hashimoto, 1996). This conclusion was based in part on the finding that nudel mutations fall into two distinct phenotypic classes: class I alleles produce very fragile eggs as well as embryos that, perhaps as a consequence, fail to develop past the syncytial stage, whereas class II alleles produce dorsalized embryos like mutations in other genes required to activate Toll (Hong and Hashimoto, 1996). Moreover, complementation was observed for certain combinations of class I and II alleles, which could be due to proteins having defects in distinct regions and thus able in combination to provide full Nudel function. However, subsequent molecular studies revealed that class I mutations do not affect a common region of Nudel but rather cause a variety of defects that ultimately result in marked reduction of extracellular Nudel (LeMosy et al., 2000). On the other hand, almost all class II alleles, with one exception, have an alteration in the protease domain that is predicted to impair catalytic activity. The one exception, ndl9, is an alteration of a cysteine just N-terminal to the protease domain. If this cysteine is involved in a crucial disulfide bond between regulatory and catalytic domains as in other serine proteases, then its loss could also impair Nudel protease activity. Thus, although molecular studies of extant nudel mutations have provided evidence that the Nudel protease is involved in dorsoventral patterning, they have not defined a function for other regions of the Nudel protein.

Here we provide biochemical and genetic evidence that the N-terminal region of Nudel contains a GAG attachment site and that this site is specifically required for dorsoventral patterning. Disruption of this site blocks a disulfide-based association between N- and C-terminal Nudel polypeptides and Nudel protease activity. We propose that a GAG chain on Nudel is required for proper assembly of Nudel into an extracellular matrix structure, which in turn is necessary for Nudel protease activation.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. Discussion
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

Biochemical Evidence for Sulfated GAG on Nudel

The Nudel protein has many potential sites for the attachment of both N- and O-linked carbohydrates, including GAG chains (Hong and Hashimoto, 1995). It was previously demonstrated that the presumptive full-length Nudel protein of 350 kDa, detectable as a minor species in the ovary, contains approximately 30 kDa of N-linked carbohydrate (LeMosy et al., 2000). This polypeptide appears to be processed into two nonoverlapping fragments, an N-terminal polypeptide of 210 kDa lacking the protease domain and a C-terminal polypeptide of 250 kDa containing the protease domain, that represent the major forms of Nudel in the ovary (LeMosy et al., 1998). The composition of the additional mass (∼110 kDa), which accumulates after proteolytic processing, has been mysterious, although suspected to be due largely to glycosylation such as GAG addition. However, we did not detect by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) a change in the apparent molecular weight of ovarian Nudel after chemical (nitrous acid) or enzymatic (heparinase, heparitinase, and chondroitinase ABC) treatment known to destroy GAG structure (data not shown), suggesting that Nudel is not extensively modified by GAGs. In case Nudel contains a small amount of short GAG chains, we analyzed Nudel by two-dimensional (2D) -gel electrophoresis, with isoelectric focusing followed by SDS-PAGE, reasoning that this method could resolve polypeptides of similar size but differing in charge due to GAGs bearing negatively charged sulfates. This analysis revealed that the 210-kDa form of Nudel in the ovary is strikingly heterogeneous with respect to charge and can be resolved into approximately 12 different acidic species with a pI between 3 and 6 (Fig. 1A).

thumbnail image

Figure 1. Two-dimensional (2D) gel analysis of Nudel. Western blots of 2D-gels containing total ovary proteins separated first by isoelectric focusing and then by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) were probed with N-terminal Nudel antibody. Only the region of the 210-kDa form of Nudel is shown. A: In OR ovary, 210-kDa Nudel is represented by 12 distinct forms with a pI between 3 and 6. Arrowhead points to the most acidic species. B: After treatment of OR ovary extract with nitrous acid pH 1.5, which degrades sulfated glycosaminoglycan (GAG) chains, the most acidic species of the 210-kDa Nudel is absent (arrowhead). In addition, other acidic species decrease in intensity while less acidic species increase in intensity, consistent with loss of negatively charged sulfates. C: In ovary containing the GAG S-A protein, the most acidic species of the 210-kDa Nudel seen in OR ovary is absent (arrowhead). IEF, isoelectric focusing.

Download figure to PowerPoint

To test whether the heterogeneity of the 210-kDa Nudel polypeptide is due to GAGs, we treated ovary extracts with nitrous acid pH 1.5 at room temperature and then analyzed the treated extracts by 2D-gel electrophoresis. Such nitrous acid treatment is known to result in selective cleavage of glycosidic bonds involving N-sulfated saccharides in GAG chains and, thus, has been used to demonstrate the presence of GAGs of the heparin and heparan sulfate family (Guo and Conrad, 1989). We found that nitrous acid treatment resulted in a shift in the pattern of gel spots representing the 210-kDa Nudel polypeptide (Fig. 1B). Compared with the wild-type sample, the intensity of more acidic spots decreased while the intensity of less acidic spots increased, consistent with a reduction of negatively charged sulfates. The most conspicuous change after nitrous acid treatment was the absence of the most acidic spot. These results are consistent with the existence of sulfated GAG chains on the 210-kDa Nudel polypeptide.

Mutation of Potential GAG Addition Sites in Nudel Results in Dorsalized Embryos

In proteoglycans of the extracellular matrix, GAGs have been found attached to the protein at a serine within the sequence SG (Bourdon et al., 1987; Krueger et al., 1990). From these studies, (D/E)XSG and SGXG (with X being any amino acid) were defined as possible consensus sequences for GAG attachment. Nudel contains three matches to these two sequences within a short region just N-terminal to the protease domain (Hong and Hashimoto, 1995; Fig. 2A). By using site-directed mutagenesis, we mutated the three sequences in a nudel gene construct that provides wild-type function in transgenic rescue experiments. We changed every serine in the three sequences to an alanine, an alteration known to prevent GAG attachment (Langford et al., 1998), thereby generating the mutant gene designated GAG S-A.

thumbnail image

Figure 2. Mutation of potential glycosaminoglycan (GAG) attachment sites in Nudel results in dorsalized embryos. A: N-terminal region of Nudel contains three matches to two different consensus sequences for GAG addition (asterisk). Every serine in the three wild-type (WT) sequences was mutated to an alanine to generate the GAG S-A mutant. Nudel also contains a central serine protease domain (black box), four clusters of 11 LDL receptor type-A repeats (grey boxes), and two serine/threonine-rich regions that are potential sites for O-linked glycosylation (hatched boxes). N, PD, and C refer to the N-terminal, protease domain, and C-terminal antibodies, respectively, used in this study and the protein regions that they recognize (LeMosy et al., 1998). aa, amino acids. B: Cuticle of embryo from OR female shows normal dorsoventral pattern. C: Cuticle of embryo from ndl14/Df(3L)CH12 female carrying the GAG S-A transgene displays the classic dorsalized phenotype, characterized by twisted morphology and lack of ventral and lateral structures, which is rescued by wild-type transgene (LeMosy et al, 1998). The dorsalized phenotype indicates that the GAG S-A transgene rescued the fragile egg/early embryo arrest phenotype produced by ndl14/Df(3L)CH12 females, but did not provide full nudel function to restore the normal dorsoventral pattern of the embryo.

Download figure to PowerPoint

We assayed the activity of the GAG S-A transgene in females lacking full-length Nudel (see Experimental Procedures section). These females are sterile, because they produce fragile eggs and early arrest embryos, the null phenotype associated with class I nudel alleles. A wild-type nudel transgene was previously shown to rescue this phenotype as well as the dorsalized phenotype produced by class II nudel alleles, thereby restoring female fertility (LeMosy et al., 1998). We found that the GAG S-A transgene, in either one copy or two copies, partially rescued the fragile egg and early arrest phenotype but not the dorsalized phenotype (Fig. 2B,C; Table 1). Thus, the GAG S-A mutant is specifically defective in dorsoventral patterning.

Table 1. Rescue of nudel Function by GAG S-A Transgene
Transgene% Early arrest% Dorsalized% HatchNo. of lines
  • a

    Data from LeMosy et al., 1998. ND, not determined.

  • b

    Activity of GAG S-A transgene was assayed in the nudel null background ndl14/Df(3L)CH12. Over 200 embryos were examined for each line, and in the case of multiple lines, each value represents the average of the numbers for the different lines.

Nonea9730
Wild-typeaNDND883
GAG S-A (1 copy)b366403
GAG S-A (2 copies)b336601

The GAG S-A mutant behaves as a class II nudel allele, as it retains nudel activity required for egg structural integrity and embryonic development past the syncytial stage but not the activity required for embryonic dorsoventral patterning. Among the class II alleles, ndl9 and ndl46 are weak loss-of-function mutations that can complement certain class I alleles as well as each other, whereas the remainder, like ndl111, are stronger mutations that act antagonistically to ndl9 or ndl46 (Hong and Hashimoto, 1996; LeMosy et al., 2000). We found that the GAG S-A mutant, like the latter type of class II allele, did not complement any class I allele tested (ndl15, ndl17, and ndl18; data not shown); however, it strongly complemented both ndl9 and ndl46 (Table 2).

Table 2. Genetic Interaction of GAG S-A Transgene with Class II nudel Allelesa
 ndl9ndl46ndl111Df
  • a

    Each value represents percentage of embryos that hatched from eggs laid by females carrying the indicated combination of nudel alleles and the GAG S-A transgene. ND, not determined.

  • b

    Data from Hong and Hashimoto (1996) for eggs laid at 18° and at 22° (in parentheses).

  • c

    Over 200 eggs laid at 18° and at 25° (in parentheses) were examined. General genotype of flies was GAG S-A/+; ndl/Df(3L)CH12.

  • d

    Represents activity of GAG S-A in nudel null background ndl14/Df(3L)CH12.

ndl9bND80 (50)0 (2)24 (0)
ndl46b8 (0)0 (0)8 (0)
GAG S-Ac40 (11)70 (1)ND (0)0 (0)d

Mutation of Potential GAG Addition Sites in Nudel Blocks an Intermolecular Association and Protease Activity of Nudel

We investigated why the GAG S-A protein might be defective in dorsoventral patterning. First, we checked whether the GAG S-A protein was proteolytically processed during oogenesis like the wild-type protein. In ovary extracts containing the GAG S-A protein, we detected the presumptive full-length polypeptide of 350 kDa, as well as the N-terminal 210-kDa and the C-terminal 250-kDa polypeptides, the major processed forms normally seen in wild-type ovary extracts (Fig. 3, lanes 1–4). Processing of the mutant protein appeared to be less efficient and to give rise to some minor aberrant forms. Nonetheless, these results indicate that initial proteolytic processing of Nudel was not blocked by the alterations in the GAG S-A mutant.

thumbnail image

Figure 3. Processing of GAG S-A Nudel protein during oogenesis. Western blots of sodium dodecyl sulfate-polyacrylamide gel electrophoresis gels containing total ovary proteins were probed with N-terminal (N) or C-terminal (C) Nudel antibodies. The 170-, 210-, and 250-kDa polypeptides generated by proteolytic cleavage of presumptive full-length 350-kDa protein are detected in OR ovaries (lanes 1 and 3) and ovaries containing GAG S-A protein (lanes 2 and 4). Some minor aberrant forms are also detected in the latter. Females carrying only one copy of the GAG S-A transgene were used in this analysis; thus, the mutant protein may be present at a lower level than normal.

Download figure to PowerPoint

To examine whether any actual GAG attachment site was destroyed in the GAG S-A mutant, we analyzed the structure of the 210-kDa polypeptide derived from the mutant by 2D-gel electrophoresis of ovary extracts as described above. When compared with wild-type, the most conspicuous difference in the pattern produced by the GAG S-A mutant was the absence of the most acidic spot (Fig. 1C). A spot in the same position is missing after nitrous acid treatment to degrade GAG chains in the wild-type protein (Fig. 1B). These results suggest that a GAG chain is attached to one or more of the sites in Nudel mutated in the GAG S-A mutant.

We also examined the subcellular distribution of the GAG S-A Nudel protein during oogenesis. Immunostaining of sectioned ovaries revealed that the mutant protein was secreted from follicle cells and localized to the oocyte surface, albeit with a lower efficiency than the wild-type protein (Fig. 4A,B). Thus, the alterations in the GAG S-A mutant did not prevent localization of Nudel to the oocyte surface.

thumbnail image

Figure 4. Localization of GAG S-A Nudel protein. Confocal photomicrographs of ovaries probed with N-terminal Nudel antibody. A: OR egg chamber shows localization of secreted Nudel at oocyte plasma membrane (arrowhead) as well as some Nudel still in follicle cells surrounding the oocyte. B: Egg chamber expressing GAG S-A protein, at similar stage as shown in A, shows secreted Nudel at the oocyte surface (arrowhead). Level of Nudel staining in follicle cells is higher, suggesting less efficient secretion of GAG S-A protein than wild-type protein.

Download figure to PowerPoint

An earlier study suggested that secreted Nudel is stabilized by means of disulfide bonds within an extracellular matrix layer around the oocyte (LeMosy et al., 2000). Key evidence was the finding that the 210- and 250-kDa forms of Nudel could be coimmunoprecipitated under denaturing conditions from a low-speed pellet fraction of ovary extracts enriched in eggshell proteins. Although we were able to coimmunoprecipitate the polypeptides encoded by the wild-type gene as shown previously (Fig. 5, lanes 1 and 2), we failed to coimmunoprecipitate the 210- and 250-kDa polypeptides derived from the GAG S-A mutant (Fig. 5, lanes 3 and 4). Interestingly, we also could not coimmunoprecipitate the polypeptides encoded by the ndl9 allele, a mutation of a cysteine just N-terminal to the protease domain, thereby demonstrating that this cysteine is required for association between 210- and 250-kDa polypeptides (Fig. 5, lanes 5 and 6). These data suggest that the alterations in the GAG S-A mutant prevent a disulfide-based association between N- and C-terminal polypeptides of Nudel.

thumbnail image

Figure 5. Disruption of physical association between N- and C-terminal Nudel polypeptides. Nudel was immunoprecipitated (IP) under denaturing conditions from an ovary homogenate with N-terminal (N) or C-terminal (C) Nudel antibody, and the immunoprecipitates were analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and Western blotting by using either antibody. N-terminal 210-kDa and C-terminal 250-kDa forms of wild-type Nudel coimmunoprecipitate (lanes 1 and 2) but not corresponding forms encoded by GAG S-A transgene (lanes 3 and 4) or ndl9 allele (lanes 5 and 6).

Download figure to PowerPoint

Finally, we examined whether the alterations in the GAG S-A mutant caused a defect in Nudel protease activation, as has been observed for all class II alleles (LeMosy et al., 2000). Just before the start of embryogenesis, a self-cleavage mechanism generates the active Nudel protease, detectable in fertilized eggs as a 33-kDa polypeptide corresponding in size to Nudel's protease domain (LeMosy et al., 1998; Fig. 6, lane 1). In egg extracts containing the GAG S-A protein, the 33-kDa polypeptide was not detectable (Fig. 6, lane 2). Instead, the 38-kDa intermediate representing a pro-form of the Nudel protease was detected at a higher level than in wild-type. In addition, C-terminal fragments of Nudel generated by processing involving Nudel's own protease activity were also not detected, consistent with a loss of Nudel protease activity (Fig. 6, lanes 4 and 5). These results suggest that Nudel protease activation is defective in the GAG S-A mutant.

thumbnail image

Figure 6. Defect in Nudel protease activation. Western blots of sodium dodecyl sulfate-polyacrylamide gel electrophoresis gels containing total egg proteins were probed with protease domain (PD) or C-terminal (C) Nudel antibody. The 33-kDa polypeptide corresponding to Nudel's protease domain is detected in OR eggs (lane 1) but not in eggs containing GAG S-A protein (lane 2). In the latter, the 38-kDa “pro-form” of the Nudel protease (LeMosy et al., 1998) is detected at a four- to fivefold higher level than normal by densitometry analysis. Blot of egg proteins derived from ndl14/Df(3L)CH12 females, the genetic background in which the GAG S-A mutant was analyzed, illustrates the nonspecific bands (lane 3). The 50- and 60-kDa C-terminal polypeptides, representing further processing of Nudel dependent on Nudel's own protease activity (LeMosy et al., 1998), are seen in OR eggs (lane 4) but not in eggs containing GAG S-A protein (lane 5). In the latter, 110- and 130-kDa C-terminal polypeptides, which accumulate in the absence of Nudel protease (LeMosy et al., 1998), are more prominent (lane 5).

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. Discussion
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

We have presented biochemical and genetic evidence that the Nudel protein contains a GAG chain of the heparin and heparan sulfate family. By using 2D-gel electrophoresis, we were able to resolve the 210-kDa Nudel polypeptide in the ovary into approximately 12 distinct acidic species (Fig. 1). Nitrous acid treatment, under conditions that specifically destroy heparin and heparan sulfate, caused the loss of the most acidic species and a reduced level of many others, consistent with the presence of a sulfated GAG chain on the 210-kDa Nudel polypeptide. The remaining heterogeneity of this polypeptide could be due to residual GAG structures resistant to nitrous acid degradation and/or to other posttranslational modifications. By using site-directed mutagenesis, we generated the GAG S-A mutant of Nudel potentially lacking 3 GAG attachment sites in the N-terminus of the protein (Fig. 2). As with nitrous acid treatment, the mutations resulted in loss of the most acidic form of the 210-kDa polypeptide, thereby demonstrating that one or more of the mutated sites is normally used for GAG attachment (Fig. 1). The mutations did not appear to affect other acidic species that were perturbed by nitrous acid treatment, suggesting that Nudel contains other sites for GAG attachment. Near the three mutated sites, Nudel contains three copies of the sequence SG, which has been shown to be sufficient for GAG attachment (Brinkmann et al., 1997).

The GAG chain missing in the GAG S-A mutant is specifically required for Nudel's role in dorsoventral patterning of the embryo (Fig. 2; Table 1). It does not appear to be essential for Nudel's “structural” role, as the GAG S-A mutant rescues the fragile egg/early embryo arrest phenotype produced by class I nudel alleles (Table 1). The GAG S-A mutant, thus, behaves like a class II nudel allele that encodes a protein lacking dorsoventral patterning activity. This defect can be compensated by the proteins encoded by two other class II alleles, ndl9, a mutation of a cysteine just N-terminal to the protease domain, and ndl46, a mutation in the protease domain (Table 2; Hong and Hashimoto, 1996; LeMosy et al., 2000). If Nudel function requires protein dimerization, as previously hypothesized (Hong and Hashimoto, 1996), complementation between these mutants could result from the formation of mixed dimers composed of two different mutant polypeptides. The lack of complementation between the GAG S-A mutant and ndl111 may be due to the antagonistic effect that the latter and other strong class II alleles have on weaker class II alleles (Table 2; Hong and Hashimoto, 1996; LeMosy et al., 2000).

Surprisingly, the GAG S-A mutant appears to be defective in Nudel protease activation, like other class II nudel alleles (LeMosy et al., 2000; Fig. 6). How could a GAG chain on Nudel be important for activation of the protease domain? One possible clue is that the GAG S-A protein is also defective in a disulfide-based association between N- and C-terminal Nudel polypeptides normally found in an extracellular matrix-like fraction containing eggshell proteins (LeMosy et al., 2000; Fig. 5). An explanation compatible with our data is that the GAG chain is required for proper assembly of Nudel into an extracellular matrix structure that is stabilized by disulfide bond formation between Nudel polypeptides (LeMosy et al., 2000). The incorporation of Nudel within an extracellular matrix structure in turn could be necessary to activate the Nudel protease, which triggers Toll signaling necessary for dorsoventral patterning.

Our data do not preclude other possible functions for a GAG chain on Nudel. In fact, minor and subtle defects in processing and secretion of Nudel were observed in the GAG S-A mutant (Figs. 3, 4). These defects could also lead to aberrant glycosylation, thus partially explaining the distinct 2D-gel patterns of the wild-type and GAG S-A polypeptides (Fig. 1). Multiple functions for a GAG chain on Nudel would be consistent with structural, mechanical, and regulatory roles of GAGs that have been described (Jackson et al., 1991).

Previous studies have provocatively suggested that the pipe and windbeutel genes are required in ventral follicle cells to secrete a sulfated GAG that restricts proteolytic activation of the Toll ligand to the ventral side of the embryo (Nilson and Schüpbach, 1998; Konsolaki and Schüpbach, 1998; Sen et al., 1998, 2000). To address the possibility that this GAG is attached to Nudel, we examined by 2D-gel electrophoresis the 210-kDa Nudel polypeptide in ovary extracts from pipe and windbeutel mutants and from females in which pipe was expressed in all follicle cells by using the heat shock promoter (Sen et al., 2000). However, no significant change in the 2D-gel pattern of the 210-kDa polypeptide was observed in these experiments that would be consistent with a GAG chain on Nudel being modified by the action of pipe and windbeutel (data not shown). We also could not address whether the GAG missing in the GAG S-A mutant normally contains a modification that regulates Toll ligand activation, because this mutant is defective in Nudel protease activation, an upstream event. Nonetheless, our experiments suggest the importance of a GAG in regulating the Nudel protease and, thus, dorsoventral patterning of the Drosophila embryo.

EXPERIMENTAL PROCEDURES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. Discussion
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

Fly Stocks

Oregon R was used as the wild-type strain, and the w1118 strain was used to generate transgenic lines (Lindsley and Zimm, 1992). The nudel alleles and the deficiency Df(3L)CH12 used in this study have been described previously (Hong and Hashimoto, 1995, 1996).

Generation of Transgenes

A genomic nudel transgene was previously reported to rescue the sterility of ndl14/Df(3L)CH12 females (LeMosy et al., 1998). This transgene was used as a template to generate the GAG S-A mutant in which each of four serines contained in the three potential GAG attachment sites in Nudel was mutated to an alanine (residues 537, 538, 794, and 829 in Nudel sequence; Hong and Hashimoto, 1995). All four mutations were introduced by sequential application of a two-step polymerase chain reaction (PCR) mutagenesis procedure that altered each serine separately (Cormack, 1997). DNA sequencing of PCR-derived regions confirmed that only the desired mutations had been introduced.

Several independent lines bearing the GAG S-A transgene were generated by germline transformation (Spradling, 1986). Females were constructed that contain one copy or two copies of the GAG S-A transgene in a nudel null background, ndl14/Df(3L)CH12. In ovaries derived from ndl14/Df(3L)CH12 females, no full-length Nudel but rather only a low level of truncated Nudel is detected (LeMosy et al., 2000). To score for rescue of nudel function, embryos under oil and cuticle preparations were examined by light microscopy (Wieschaus and Nüsslein-Volhard, 1986).

Biochemical Methods

2D-gel electrophoresis was performed essentially as described previously (O'Farrell, 1975). Six ovaries were homogenized in 2D sample solution (9 M urea, 5% 3–10 ampholytes, 4% NP-40, and 2% β-mercaptoethanol). Proteins in the homogenate were separated in an isoelectric focusing gel, which was then divided into acidic and basic halves such that proteins in each half were resolved on separate 5% SDS-polyacrylamide gels. Western blots of these gels were prepared by standard techniques (Harlow and Lane, 1988) and probed with anti-Nudel antibodies (LeMosy et al., 1998). Of the major Nudel forms in the ovary, only the 210-kDa polypeptide resolved into multiple acidic species and, thus, was further analyzed for the presence of sulfated GAGs.

Nitrous acid degradation of sulfated GAGs was performed essentially as described (Guo and Conrad, 1989). Twenty microliters of nitrous acid pH 1.5 was added to a lyophilized preparation of six ovaries that had been homogenized in H2O. This sample was incubated for 10 minutes at 25°C and then brought to pH 8.5 with 1 M Na2CO3. After proteins in the sample were precipitated with trichloroacetic acid, they were solubilized in 2D sample solution and analyzed by 2D- gel electrophoresis as described above. Previously described procedures were followed to examine proteolytic processing of Nudel and to immunoprecipitate Nudel (LeMosy et al., 1998, 2000).

Immunolocalization

Cryostat sections (8 μm) of ovaries were prepared and stained with anti-Nudel antibodies essentially as described previously (LeMosy et al., 1998). The immunostained sections were examined with a LSM-510 confocal microscope (Zeiss, Thornwood, NY).

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. Discussion
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

We thank Qiu Xuan, Jack Valentijn, and Jim Jamieson for help with 2D-gel electrophoresis. We also thank Dave Stein for the heat shock-pipe line, and Ellen LeMosy and Mike Tiemeyer for helpful suggestions on the manuscript. C.H. received funding from the National Institutes of Health.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. Discussion
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  • Bourdon MA, Krusius T, Campbell S, Schwartz NB, Ruoslahti E. 1987. Identification and synthesis of a recognition signal for the attachment of glycosaminoglycans to proteins. Proc Natl Acad Sci U S A 84: 31943198.
  • Brinkmann T, Weilke C, Kleesiek K. 1997. Recognition of acceptor proteins by UDP-D-xylose proteoglycan core protein beta-D-xylosyltransferase. J Biol Chem 272: 1117111175.
  • Cormack B. 1997. Directed mutagenesis using the polymerase chain reaction. In: AusubelFM, editor. Current protocols in molecular biology. New York, NY: Greene Publishing Associates/Wiley Interscience. p 8.5.1–8.5.10.
  • Guo YC, Conrad HE. 1989. The disaccharide composition of heparins and heparan sulfates. Anal Biochem 176: 96104.
  • Harlow E, Lane D. 1988. Antibodies: a laboratory manual. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press.
  • Hong CC, Hashimoto C. 1995. An unusual mosaic protein with a protease domain, encoded by the nudel gene, is involved in defining embryonic dorsoventral polarity in Drosophila. Cell 82: 785794.
  • Hong CC, Hashimoto C. 1996. The maternal nudel protein of Drosophila has two distinct roles important for embryogenesis. Genetics 143: 16531661.
  • Jackson RL, Busch SJ, Cardin AD. 1991. Glycosaminoglycans: molecular properties, protein interactions, and role in physiological processes. Physiol Rev 71: 481539.
  • Konsolaki M, Schüpbach T. 1998. Windbeutel, a gene required for dorsoventral patterning in Drosophila, encodes a protein that has homologies to vertebrate proteins of the endoplasmic reticulum. Genes Dev 12: 120131.
  • Krueger RC Jr, Fields TA, Hildreth J, Schwartz NB. 1990. Chick cartilage chondroitin sulfate proteoglycan core protein. I. Generation and characterization of peptides and specificity for glycosaminoglycan attachment. J Biol Chem 265: 1207512087.
  • Langford JK, Stanley MJ, Cao D, Sanderson RD. 1998. Multiple heparan sulfate chains are required for optimal syndecan-1 function. J Biol Chem 273: 2996529971.
  • LeMosy EK, Hashimoto C. 2000. The Nudel protease of Drosophila is required for eggshell biogenesis in addition to embryonic patterning. Dev Biol 217: 352361.
  • LeMosy EK, Kemler D, Hashimoto C. 1998. Role of Nudel protease activation in triggering dorsoventral polarization of the Drosophila embryo. Development 125: 40454053.
  • LeMosy EK, Leclerc CL, Hashimoto C. 2000. Biochemical defects of mutant nudel alleles causing early developmental arrest or dorsalization of the Drosophila embryo. Genetics 154: 247257.
  • Lindsley DL, Zimm GG. 1992. The genome of Drosophila. New York, NY: Academic Press.
  • Morisato D, Anderson KV. 1994. The spätzle gene encodes a component of the extracellular signaling pathway establishing the dorsal-ventral pattern of the Drosophila embryo. Cell 76: 677688.
  • Morisato D, Anderson KV. 1995. Signaling pathways that establish the dorsal-ventral pattern of the Drosophila embryo. Annu Rev Genet 29: 371399.
  • Nilson LA, Schüpbach T. 1998. Localized requirements for windbeutel and pipe reveal a dorsoventral prepattern within the follicular epithelium of the Drosophila ovary. Cell 93: 253262.
  • O'Farrell PH. 1975. High resolution two-dimensional electrophoresis of proteins. J Biol Chem 250: 40074021.
  • Schneider DS, Jin Y, Morisato D, Anderson KV. 1994. A processed form of the Spätzle protein defines dorsal-ventral polarity in the Drosophila embryo. Development 120: 12431250.
  • Sen J, Goltz JS, Stevens L, Stein D. 1998. Spatially restricted expression of pipe in the Drosophila egg chamber defines embryonic dorsal-ventral polarity. Cell 95: 471481.
  • Sen J, Goltz JS, Konsolaki M, Schüpbach T, Stein D. 2000. Windbeutel is required for function and correct subcellular localization of the Drosophila patterning protein Pipe. Development 127: 55415550.
  • Spradling AC. 1986. P element-mediated transformation. In: RobertsDB, editor. Drosophila: a practical approach. Washington, DC: IRL Press. p 175197.
  • Wieschaus E, Nüsslein-Volhard C. 1986. Looking at embryos. In: RobertsDB, editor. Drosophila: a practical approach. Washington, DC: IRL Press. p 199227.