The CDM superfamily protein MBC directs myoblast fusion through a mechanism that requires phosphatidylinositol 3,4,5-triphosphate binding but is independent of direct interaction with DCrk

Abstract

Myoblast city (mbc), a member of the CDM superfamily, is essential in the Drosophila melanogaster embryo for fusion of myoblasts into multinucleate fibers. Using germ line clones in which both maternal and zygotic contributions were eliminated and rescue of the zygotic loss-of-function phenotype, we established that mbc is required in the fusion-competent subset of myoblasts. Along with its close orthologs Dock180 and CED-5, MBC has an SH3 domain at its N terminus, conserved internal domains termed DHR1 and DHR2 (or "Docker"), and C-terminal proline-rich domains that associate with the adapter protein DCrk. The importance of these domains has been evaluated by the ability of MBC mutations and deletions to rescue the mbc loss-of-function muscle phenotype. We demonstrate that the SH3 and Docker domains are essential. Moreover, ethyl methanesulfonate-induced mutations that change amino acids within the MBC Docker domain to residues that are conserved in other CDM family members nevertheless eliminate MBC function in the embryo, which suggests that these sites may mediate interactions specific to Drosophila MBC. A functional requirement for the conserved DHR1 domain, which binds to phosphatidylinositol 3,4,5-triphosphate, implicates phosphoinositide signaling in myoblast fusion. Finally, the proline-rich C-terminal sites mediate strong interactions with DCrk, as expected. These sites are not required for MBC to rescue the muscle loss-of-function phenotype, however, which suggests that MBC's role in myoblast fusion can be carried out independently of direct DCrk binding.

FIG. 1.

Analysis of embryos lacking both maternal and zygotic mbc. The muscle pattern was visualized by immunostaining of late stage 15 or early stage 16 embryos with antisera directed against muscle myosin (myosin heavy chain). All panels are lateral views, with anterior to the left and dorsal to the top. (A) Wild-type embryo. The segmentally repeating array of multinucleate muscle fibers is apparent. (B and D) Low and high magnification views, respectively, of embryos that lack zygotic expression of mbc. As previously described (14, 39), no myoblast fusion occurs in the absence of zygotic mbc. (C and E) Low- and high-magnification views, respectively, of embryos lacking both the maternal and zygotic mbc transcripts. No myoblast fusion is observed, and the mutant phenotype is approximately as severe as that observed in panels B and D. Scale bar = 10 μm.

FIG. 2.

MBC is required in the fusion-competent myoblasts. Embryos were immunostained as in Fig. 1. All panels are lateral views of stage 16 embryos, with anterior to the left and dorsal to the top. (A) mbc mutant embryo in which expression of the mbc cDNA is directed to mesodermal tissues using twi-GAL4. An apparently normal pattern of muscles is evident. (B) mbc mutant embryo in which expression of UAS-mbcHA is directed to mesodermal cells under control of 24B-GAL4. The muscle pattern is similar to that seen in the wild type (Fig. 1) and panel A. (C) mbc mutant embryo in which expression of UAS-mbcHA is provided exclusively to the founder cells under control of rP298-GAL4. Muscle myotubes are absent and have been replaced by mononucleate myoblasts. Thus, the embryo is phenotypically mutant. Scale bar = 13 μm.

FIG. 3.

The MBC SH3 and Docker domains are essential for MBC function but not membrane localization. (A) Domains of MBC in which mutations were generated. The MBC-SH3W47K mutant has an altered consensus SH3 domain amino acid. The MBC-DockerF6.4 mutant recapitulates a mutation found in the mbc^F6.4 EMS-induced missense allele. An HA tag has been added to the C terminus of all three constructs. (B) Embryos were immunostained as in Fig. 1, with the genotypes indicated. MBC transgenes with point mutations that disrupt either the SH3 domain or the Docker domain are unable to rescue myoblast fusion in mbc mutant embryos. Scale bar = 10 μm. (C) Immunoblot from Drosophila embryo cytoplasmic (C) and membrane (M) lysates in which the MBC SH3 or Docker domain mutations were under control of the mesodermal mef2-GAL4 driver. We conclude that all forms of MBC are expressed in embryos and enriched at the membrane.

FIG. 4.

Interaction with DCrk is dependent on proline-rich sites in the carboxy terminus of MBC. (A) Mutations that disrupt putative DCrk-binding sites in MBC are shown in red. Also shown are the two identified isoforms of DCrk (21). The MBC-CBS construct contains a mutation in the consensus DCrk-binding site. The MBCΔ1807 construct has a deletion of amino acids 1807 to 1970. The MBC-NPXXP mutation alters the N-terminal PXXP motif. The MBC-NPXXPΔ1807 construct was generated by combining the previous two constructs. MBCΔPRM has a deletion of amino acids 1662 to 1970 (21). An HA tag has been added to the C terminus of all MBC constructs except MBC-CBS. (B) Yeast two-hybrid assay, with all MBC constructs fused in frame to the DNA binding domain (BD) and DCrk constructs fused in frame to the activation domain (AD). Growth on selective plates is shown on the right. Clearly, both DCrkSH2L and DCrkSH2S exhibit substantial interaction with full-length MBC, whereas that observed with MBC-NPXXPΔ1807 or MBCΔPRM is severely reduced. (C) Yeast two-hybrid assay with constructs described for panel B. Growth rates were analyzed by drop tests using serial dilutions of mid-log-phase cultures. Fourfold dilutions of the cell culture are shown from left to right. U, undiluted. (D) The yeast two-hybrid interactions were quantified by assaying β-galactosidase activity in liquid cultures using ONPG. β-Galactosidase activity in liquid cultures is expressed in Miller units as the mean ± standard deviation of eight independent assays.

FIG. 5.

The MBC C terminus mediates strong interaction with DCrk. Immunoprecipitations (IP) are from lysates of Drosophila S2 cells cotransfected with HA-tagged MBC, MBC-NPXXPΔ1807, or MBCΔPRM and FLAG-tagged DCrk. Anti-FLAG or anti-HA antibodies were used for immunoprecipitation and immunoblotting as indicated. (A) Full-length (FL) MBC, but neither MBC-NPXXPΔ1807 nor MBCΔPRM, is immunoprecipitated by interaction with DCrk. (B) DCrk is immunoprecipitated through an interaction with full-length MBC but not with MBC-NPXXPΔ1807.

FIG. 6.

MBC C-terminal truncations are expressed and present in the embryo membrane. (A) Immunoblot of whole-cell lysates from embryos aged 8 to 15 h AEL, in which the HA-tagged full-length (FL) MBC or MBC-NPXXPΔ1807 and MBCΔPRM truncations were under control of the later mesodermal 24B-GAL4 driver (lanes 1 to 3) or the early mesodermal driver twi-GAL4 (lanes 4 to 6). Lanes 1 and 4, both endogenous and GAL4-directed MBC migrate at the same position and are detected by antisera directed against the MBC N terminus; lanes 2, 3, 5, and 6, truncated MBC proteins as indicated correspond to the lower migrating form in each lane, while the endogenous protein corresponds to the more slowly migrating form. While there are some differences in signal intensity between the samples from 24B-GAL4 (lanes 1 to 3) and those from twi-GAL4 (lanes 4 to 6), the level of the truncated MBC proteins is lower than endogenous full-length MBC when 24B-GAL4 is used. (B) Immunoblot of embryo cytoplasmic (C) and membrane (M) lysates in which the HA-tagged MBC-NPXXPΔ1807 or MBCΔPRM truncations were under control of the mesodermal mef2-GAL4 driver. Truncated proteins were detected by anti-HA antisera and correspond to the anticipated sizes. All forms of MBC are present at the membrane, with a significant membrane enrichment of full-length MBC and MBC-NPXXPΔ1807. Cytoplasmic α-tubulin served as a control.

FIG. 7.

The MBC DCrk-binding sites are not essential for myoblast fusion. Embryos were immunostained as in Fig. 1. All panels are lateral views of stage 16 embryos, with genotypes as indicated. In panels A, C, E, G, and I, expression of the indicated MBC truncations is driven early in muscle development using twi-GAL4. In panels B, D, F, H, and J, expression of the indicated MBC truncations is driven later and at lower levels during muscle formation using 24B-GAL4. (A and B) The MBC consensus DCrk-binding site is not essential to rescue the mbc mutant muscle phenotype. (C and D) The MBC region carboxy terminal to amino acid 1807 is not essential to rescue the mbc mutant muscle phenotype. (E and F) The amino-terminal DCrk-binding site is not essential to rescue the mbc mutant muscle phenotype. (G and H) Neither the amino- nor carboxy-terminal DCrk-binding sites direct binding of DCrk to MBC. (I and J) The MBCΔPRM C-terminal truncation described by Ishimaru et al. (21), which does not interact with DCrk, is able to rescue the MBC mutant muscle phenotype. Scale bar = 10 μm.

FIG. 8.

The MBC DHR1 domain binds PtdIns(3,4,5)P₃ (PIP3) and is essential for MBC function in myoblasts. (A) Schematics of DHR1 deletions in MBC. (B) Immunoprecipitations from lysates of Drosophila S2 cells transfected with HA-tagged MBC (lanes 1 and 6), MBC-NPXXPΔ1807 (lanes 2 and 7), MBCΔPRM (lanes 3 and 8), MBCΔDHR1 (lanes 4 and 9), and MBC-NPXXPΔDHR1:1807 (lanes 5 and 10) as described in Materials and Methods. Only those forms of MBC that lacked the DHR1 domain were not immunoprecipitated by interaction with PtdIns(3,4,5)P₃ (lanes 9 and 10). (C) Immunoblot from Drosophila embryo cytoplasmic (C) and membrane (M) lysates in which expression of the MBCΔDHR1 deletion alone or the DHR1 deletion in a background lacking DCrk-binding sites was under control of the mesodermal mef2-GAL4 driver. The top panel was probed with anti-HA to detect the tagged MBC protein, and the bottom panel provides a control for loading and membrane fractionation. We conclude that all forms of MBC are expressed in embryos and present at the membrane. (D) Embryos were immunostained as in Fig. 1, and the genotypes are indicated. MBC transgenes lacking the PtdIns(3,4,5)P₃ binding region DHR1 are unable to rescue myoblast fusion in mbc mutant embryos.

FIG. 9.

Sequence lesions in EMS-induced mbc mutant alleles. A total of 17 EMS-induced mbc mutant alleles were isolated in previously published studies (14), obtained in other screens, or generously provided by other investigators. The sequence lesions in these mutant alleles were determined as described in Materials and Methods. (A) Schematic representation of the MBC protein, with protein domains and the relative positions of EMS-induced mutations indicated. (B) The mutant allele, the amino acid (aa) affected by the mutation, and the nature of the change are listed. (C) Homology alignment of the locations of the four Docker domain missense mutations within the Dock180-related subgroup of CZH Rho-GEFs described by Meller et al. (34). Alignment was carried out using Vector NTI 9.0.0 (Informax) and shaded using Boxshade. The following are GenBank GenInfo Identifier numbers: for D. melanogaster MBC, 7511969; for human Dock180/Dock1, 4503355; for human Dock2, 31377468; for human Dock3, 23297197; for human Dock4, 29568109; for human Dock5, 45439362; for D. melanogaster Sponge/Dock4, 28381487; for C. elegans CED-5, 7511497; for Dictyostelium discoideum (Dicty) DocA, 66801748; for Dictyostelium discoideum DocB, 60474615; for Dictyostelium discoideum DocC, 66801673; for Dictyostelium discoideum DocD, 66809741; for Neurospora crassa Dock, 32418746; and for Saccharomyces cerevisiae Dock, 6323454.