BIOLOGICAL OVERVIEW

Syntaxin 5 is a Golgi-localized SNARE protein that has been shown to be required for ER-Golgi traffic in yeast (Dascher, 1994) and Golgi reassembly following cell division in mammalian cells (Rabouille, 1998). The Drosophila ortholog, Syx5, like its mammalian and yeast counterparts is localized to the Drosophila Golgi and binds to alpha-SNAP. Null mutations in Syx5 are larval lethal and demonstrate impaired transport of vesicles through the secretory pathway. A hypomorphic allele of Syx5 results in impenetrant lethality, and escaping adult flies are male sterile. The male sterility results both from failure of germ cells to complete cytokinesis and from defects in spermatid elongation and maturation. Together, these results show that Syx5 is required for the proper function of the Golgi apparatus and that an efficiently functioning Golgi apparatus is required for the steps leading to the completion of cytokinesis and formation of mature sperm (H. Xu, 2002).

Transport of membranes and membrane proteins within the cell requires the interaction of SNARE proteins on the transport vesicle with their cognate SNARE partners on the target membrane, and this binding appears necessary to achieve membrane fusion. The SNARE proteins, so named because they form a stable complex that acts as the receptor for the soluble NSF attachment protein alpha-SNAP (hence SNAP Receptors), are composed of three protein families: VAMP, syntaxin, and SNAP-25. The formation of this strong coiled-coil complex is thought to draw the vesicle and target membranes into apposition, and in so doing may provide the energy needed to cause membrane fusion (H. Xu, 2002 and references therein).

Although the role of SNARE proteins in secretion is generally accepted, a large number of different SNARE protein isoforms have been identified that appear to have specific subcellular distributions. This has led to the idea that membrane fusion at each step in the secretory pathway may be mediated by the interactions of a unique set of SNAREs. A recent comparison of the partially completed human and Drosophila genomes with those of Caenorhabditis elegans and Saccharomyces cerevisiae has revealed that, while yeast, worms, and flies have roughly the same number of SNAREs, humans have significantly more SNARE proteins (Bock, 2001). This suggests that mammals may have evolved unique forms of the SNAREs for specialized purposes. However, certain ancestral forms of the SNAREs, such as syntaxin 5 (also known as Sed5p), have orthologs that are present in all organisms (H. Xu, 2002).

Sed5 was first identified by Hardwick and Pelham (1992) as a multicopy suppressor of the lethal phenotype that arose from the lack of the yeast HDEL receptor ERD2. Their studies revealed that Sed5p participates in vesicular traffic between the ER and Golgi. The mammalian ortholog, called syntaxin 5, was found to have a cis-Golgi distribution (Bennett, 1993) consistent with a role in this compartment. Similarly, the Drosophila ortholog, when expressed in mammalian cells, also localized to a perinuclear compartment (Banfield, 1994). Subsequent studies revealed that overexpression of a truncated form of mammalian syntaxin 5 lacking the transmembrane domain (Dascher, 1994) or microinjection of syntaxin 5-specific antibodies (Rowe, 1998) blocked the transport of vesicular stomatitis virus glycoprotein in a pre-Golgi intermediate compartment. Together, these results suggest that syntaxin 5 is required for the fusion of the carrier vesicles at the cis-face of the Golgi complex. In addition, syntaxin 5 has been implicated (Rabouille, 1998) in the reassembly of the Golgi apparatus from mitotic fragments following cell division (H. Xu, 2002).

In single cell systems, there is clear evidence that Sed5/syntaxin 5 (Syx5) functions in the Golgi complex, but the overall function of Syx5 in the development of a multicellular organism is not known. The Drosophila Syx5 locus has been characterized, and, like its mammalian and yeast counterparts, Syx5 is localized to the Golgi complex. As in mammals, Syx5 binds to alpha-SNAP in both two-hybrid and in vitro assays. As well, Syx5 interacts genetically with N-ethylmaleimide sensitive factor (NSF). A null mutation within the Syx5 gene has been characterized; the absence of Syx5 protein causes lethality early during the first larval instar. Moreover, using the polarized epithelial cells in the embryonic salivary gland as a model, it has been found that null mutations display defects in apical transport. Interestingly, hypomorphic combinations of Syx5 mutations lead to male sterility, and this is due to a failure in both cytokinesis and sperm maturation. Together, these results show that Syx5 is required for normal membrane protein transport and development. Moreover, these results suggest that cytokinesis is dependent on a functional Golgi complex and that this process is particularly sensitive to levels of Syx5 (H. Xu, 2002).

To search for novel SNARE proteins in Drosophila, advantage was taken of the fact that members of the syntaxin family of SNARE proteins are able to bind directly to alpha-SNAP. Therefore a two-hybrid screen of an ovarian cDNA library was performed using Drosophila alpha-SNAP as bait. In this screen, five independent, strongly positive clones were identified that were identical to the Drosophila homolog of syntaxin 5, Syx5. All five cDNAs isolated were partial and all contained sequences corresponding to the helical region of the protein (termed H3 in syntaxin family members) closest to the transmembrane domain. The H3 helical domain is involved in the formation of coiled-coil interactions with partner SNAREs and in mammalian syntaxin-1 is the region to which alpha-SNAP binds (H. Xu, 2002).

Null mutations in Syx5 cause an accumulation of membrane proteins in intracellular compartments and are larval lethal, providing evidence for a required role for membrane traffic during early stages of development. This suggests that Syx5 is important for transport of membrane proteins and, further, that blockade of the ER-Golgi traffic by Syx5 deficiency likely results in the accumulation of secretory proteins in ER-derived transport vesicles. Moreover, the early larval lethality probably arises due to depletion of the maternally supplied Syx5, leading to the arrest of a variety of signaling pathways and physiological controls that require membrane protein synthesis and transport. Unexpectedly, however, these studies have also revealed a role for Syx5 function in animal cell cytokinesis and spermatid differentiation (H. Xu, 2002).

The process of spermatogenesis in Drosophila has been the subject of extensive analysis and mutations have been identified that affect many stages in sperm development. During spermatogenesis in Drosophila, germ cells undergo four mitotic divisions and two meiotic divisions, each with incomplete cytokinesis. These divisions led to the formation of syncytial cysts of 64 spermatids connected to each other by 63 ring canals. Each testis contains a number of cysts at different stages in maturation (H. Xu, 2002 and references therein).

Within each syncytium, the spermatids form bundles that elongate the length of the testis, but then must acquire their own cell membranes in a process called individualization. Individualization occurs when a fiber-rich structure called the investment cone surrounds each elongated spermatid and progresses toward the caudal end of the spermatids, excluding all of the organelles into a so-called waste bag and encasing each spermatid in its own membrane (H. Xu, 2002).

By analyzing the phenotype of hypomorphic alleles of Syx5, it has been found that the developmental steps most sensitive to the proper function of Syx5 are the processes of cytokinesis and sperm maturation within the male germline. Syx5, an EP line [EP(2)2313] was obtained that resulted from an insertion into the Syx5 promoter region. The combination of the EP allele with a amber mutant dSyx5^AR113 functions as a hypomorph, which allows for adult escapers. The EP(2)2313/dSyx5^AR113 hypomorph together with flies expressing increasing levels of Syx5 from the EP chromosome (driven by da-GAL4 and act-GAL4) represent, in effect, an allelic series that demonstrates a requirement for Syx5 at multiple stages of spermatogenesis. EP(2)2313/dSyx5^AR113 flies are defective in cytokinesis, the formation of elongated spermatid bundles, and sperm individualization. Expression of Syx5 under da-GAL4 control rescues both cytokinesis and bundle elongation, yet mature individualized sperm fail to form. Expression of Syx5 at the highest level (from act-GAL4) results in the production of motile sperm. These results imply that different levels of Syx5-dependent secretion are required for at least three distinct processes in this tissue: meiotic cytokinesis, outgrowth of membranes to form elongated spermatid cysts, and individualization of sperm. Interestingly, in situ hybridization studies revealed that Syx5 expression is highest in primary spermatocytes, suggesting that the proteins are produced at sufficient levels to persist throughout individualization (H. Xu, 2002).

Spermatid bundle elongation and individualization represent two aspects of male germ cell development that are accompanied by extensive plasma membrane remodeling. During the elongation of spermatids, a specialized part of the germ cell cytoplasm called the fusome passes through the intercellular bridges, or ring canals, which are composed of the actin-binding proteins anillin and the septins. The fusome consists of highly branched membranous structures, probably contiguous with ER, and is marked by the presence of membrane skeletal proteins alpha-spectrin and adducin. It is possible that the fusome may provide membrane for bundle elongation. Alternatively, a source for this membrane may be the many Lava lamp-positive vesicles present along the elongating spermatid tails that are disrupted in Syx5 mutants. Since septins play roles in secretion, perhaps they coordinate fusion of these Golgi-derived vesicles at the site of membrane growth during bundle formation (H. Xu, 2002 and references therein).

Separation of syncytial spermatids into individual sperm is achieved by the action of an individualization complex (IC) composed of F-actin-rich investment cones that form around the nuclei of elongated spermatids and traverse the length of the spermatid bundles. This process invests each cell with its own plasma membrane and simultaneously strips away excess cytoplasmic material not needed by mature sperm. The discarded material is deposited in so-called waste bags that contain, among other components, investment cones, ring canals, and fusome material. Proper IC function requires a large number of genes, including jaguar, which encodes an unconventional myosin. ICs formed and progressed at least part way down the bundles of da-GAL4;EP(2)2313/dSyx5^AR113 spermatids. However, few waste bags were observed and no mature sperm formed. Flies expressing higher levels of Syx5 (from the act-GAL4 driver) have motile sperm and are fertile. Syx5 is thus the first secretory protein shown to play a role in the elaborate membrane remodeling events required to produce individual sperm (H. Xu, 2002).

While a requirement for membrane addition during animal cell cytokinesis is only an emerging concept, it has long been known that cytokinesis in plants requires membrane traffic for the formation of a membrane plate, called the phragmoplast, to separate the daughter cells (for a review, see Verma, 2001). Unlike animal cells in which fission is mediated in part by an actomyosin-based constrictive ring, the rigid cell wall precludes the action of constriction to divide the cytoplasm. In Arabidopsis, two proteins called KNOLLE and KEULE interact to promote vesicle fusion during cytokinesis. KNOLLE is a cytokinesis-specific syntaxin homolog, and KEULE is related to the syntaxin-binding protein Sec1, suggesting that KNOLLE proteins may serve as the target SNARE proteins during membrane addition. Interestingly, centrifugal growth of the phragmoplast plate is thought to occur by the addition of membranes derived from the Golgi complex, since the fungal metabolite brefeldin A inhibits it, apparently by eliminating the supply of vesicles from the Golgi complex (H. Xu, 2002 and reference therein).

To date, the best evidence of a role for membrane addition during the division of animal cells has been the demonstration that plasma membrane syntaxin proteins in sea urchin and C. elegans are required for this process. In sea urchins, introduction of Botulinum C1 neurotoxin, a protease that cleaves certain mammalian syntaxins, blocked cytokinesis, although the specific target for this toxin was not identified (Conner, 1999). In C. elegans, double-strand RNA inhibition (dsRNAi) was used to test the function of syntaxin isoforms, and it was found that inhibition of syntaxin 4, the worm ortholog of mammalian syntaxin 1, led to multinucleated cells (Jantsch-Plunger, 1999). Interestingly, of the eight syntaxins present in C. elegans, dsRNAi injections of only two leads to embryonic lethality -- syntaxin 4 and syntaxin 3 (the C. elegans ortholog of Syx5). However, the latter had a complex and pleiotropic phenotype and it was not analyzed for its effect on cytokinesis (H. Xu, 2002).

Further supporting the requirement of efficient Golgi function for spermatocyte mitosis and maturation comes from a screen for mutations in spermatogenesis. Null mutations in four way stop (fws), the Drosophila ortholog of cog5, give a virtually identical phenotype to the hypomorphic mutation in syntaxin 5. Cog5 appears to act as a molecular tether to enhance Golgi transport in vitro. The Drosophila Cog5 homolog is required for cytokinesis, polarized cell growth, and assembly of specialized Golgi architecture during spermatogenesis. Loss-of-function mutations in fws causes failure of cleavage furrow ingression in dividing spermatocytes and failure of cell elongation in differentiating spermatids and disrupts the formation and/or stability of the Golgi-based spermatid acroblast. Consistent with the lack of a growth defect in yeast lacking Cog5, adult flies lacking fws function were viable, although males were sterile. Fws protein localizes to Golgi structures throughout spermatogenesis. It is proposed that Fws may directly or indirectly facilitate efficient vesicle traffic through the Golgi to support rapid and extensive increases in cell surface area during spermatocyte cytokinesis and polarized elongation of differentiating spermatids (Farkas, 2003).

Taken together, these results suggest that SNARE-mediated membrane addition is required for cytokinesis in animal cells and support the notion that syntaxin proteins play an important role as target SNAREs for this new membrane, and specifically that a well characterized Golgi protein, Syntaxin 5 is involved in this process (H. Xu, 2002).

The source of the membranes involved in cytokinesis, however, has been less clear. As indicated above, phragmoplast formation occurs through the fusion of vesicles derived from the Golgi apparatus. Similarly, cellularization in Drosophila is also achieved through vesicles from the Golgi (Sisson, 2000). Recent studies in C. elegans have shown that incubation of embryos in brefeldin A led to an inhibition of the completion of cytokinesis. In this case, the cleavage furrow ingressed properly, but stalled and finally regressed (Skop, 2001). Interestingly, using FM1-43 labeling methods, the authors were able to demonstrate the accumulation of vesicles near the cleavage furrow in normal embryos, but showed an absence of such vesicles following brefeldin A treatment, suggesting that they may be derived from the Golgi. Finally, the recent discovery that a kinesin-like protein (Rab6-KIFL) that binds Rab6, a Golgi-localized Rab protein necessary for intra-Golgi transport (Hill, 2000), is localized to the narrow bridge linking dividing HeLa cells during late telophase, has implicated Golgi membranes in mammalian cell division (H. Xu, 2002).

Why would the Golgi complex represent an ideal source of membranes to carry out the task of building a membrane plate between dividing cells? (1) Golgi-derived vesicles can fuse with other Golgi complex compartments, a type of homotypic fusion necessary to generate a membrane de-novo. Indeed, during mitosis the Golgi complex is dissociated into mitotic vesicles that partition between the daughter cells and subsequently reassemble into new Golgi complexes. Interestingly, Syx5 is important in this reassembly, as antibodies against syntaxin 5 block reassembly in vitro (Rabouille, 1998). (2) Since the Golgi membranes must be partitioned between the two cells, their movement must be linked to the cell cycle. In fact, Rab6- KIFL may be responsible for partitioning the Golgi into the daughter cells and to the midbody between the daughter cells in telophase. (3) The Golgi complex is itself a plate-like structure and the formation of a sheet of membranes may be ideally achieved with vesicles of the lipid composition that makes up the flat stacks of the Golgi complex. The homotypic fusion of Golgi-derived vesicles at the narrow, constricted midbody region could contribute to furrow ingression, or form a sheetlike stucture akin to the phragmoplast that would sever the two cells and resolve cytokinesis. Future studies will be aimed at determining how cytokinesis is regulated and which other membrane trafficking proteins are involved (H. Xu, 2002).

GENE STRUCTURE

Drosophila Syx5 gene resides on chromosome 2, band 35E5. The Syx5 gene is quite compact with a single small intron of 68 nucleotides that splits the codon for amino acid 67 (H. Xu, 2002).

cDNA clone length - 1861

Bases in 5' UTR - 275

Exons - 3

Bases in 3' UTR - 182

PROTEIN STRUCTURE

Amino Acids - 310

Structural Domains

See t-SNARE coiled-coil homology domain profile for information from PROSITE about Syx5 structural domains.

The yeast Sed5 protein, which is required for vesicular transport between ER and Golgi complex, is a membrane protein of the syntaxin family. These proteins are thought to provide the specific targets that are recognized by transport vesicles. The mechanism by which Sed5 protein is itself localized has been investigated. Expression of epitope-tagged versions of the yeast, Drosophila and rat Sed5 homologs in COS cells results in a perinuclear distribution; immuno-EM reveals that the majority of the protein is in a tubulo-vesicular compartment on the cis side of the Golgi apparatus. A similar distribution was obtained with a chimeric molecule consisting of a plasma membrane syntaxin with the Drosophila Sed5 transmembrane domain. This indicates that the membrane-spanning domain contains targeting information, as is the case with resident Golgi enzymes. However, alterations to the transmembrane domain of Drosophila Sed5 itself did not result in its mistargeting, implying that an additional targeting mechanism exists which involves only the cytoplasmic part of the protein. This was confirmed by modifying the transmembrane domain of the yeast Sed5 protein: substitution with the corresponding region from the Sso1 protein (a plasma membrane syntaxin homolog) did not affect yeast Sed5 function in vivo (Banfield, 1994).

date revised: 20 May 2003

Home page: The Interactive Fly © 1995, 1996 Thomas B. Brody, Ph.D.

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