BIOLOGICAL OVERVIEW

Smallminded is a member of a large superfamily of related protein: the ATPases Associated with diverse cellular Activities (AAA). Members share a single or duplicated highly conserved nucleotide binding domain that defines their superfamily, but otherwise show considerable structural heterogeneity beyond these regions. Even the members of the subfamily to which Smallminded belongs appear to have a diversity of functions that makes it difficult to begin to even guess the biological function(s) of Smallminded. Members of the Cdc48p/VPC subfamily, the group to which Smallminded belongs, have been convincingly associated with programmed cell death, protein degradation, membrane fusion and cell division. The AAA server is a web site describing members of the AAA superfamily, their structural relationships and their biological functions. This essay will concern itself with Drosophila Smallminded and C. elegans MAC-1, structurally and possibly functionally the closest known Smallminded-related protein in another species. Yeast and mammalian Smid-related proteins will be dealt with briefly as well.

Smallminded was isolated in a screen of GAL4 enhancer-trap lines designed to identify those genes in which reporter gene expression is restricted to sensory neurons. In addition to a neural expression pattern, this line also has a mutant phenotype associated with the insert. Larvae homozygous for the insertion hatch normally but exhibit poor locomotion and become developmentally arrested as second instar larvae. Homozygous larvae can survive as second instar larvae for up to 8 weeks without further development before dying. Over the course of larval development both wild-type and C161 nervous systems increase in size but the rate and extent of the growth of the wild-type CNS is far greater than its C161 equivalent. As early as 36 h after hatching the divergence is such that the wild-type CNS is significantly larger than the equivalently aged CNS from C161 larvae (Long, 1998a). In smid mutants, the majority of preparations show no detectable signs of S-phase cells in the CNS. In those that do show evidence of S-phase cells, the number of NBs undergoing DNA synthesis is significantly less than wild-type, and the number of labelled cells associated with each NB is also reduced. Abnormal levels of cell death by apoptosis occur in smid mutants. Apoptosis is accompanied by heightened expression of reaper and head involution defective. The observation that the CNS of some smid individuals still contains a number of enlarged, non-mitotic NBs five days into larval development suggests that only mutant NBs arrested during mitosis become committed to apoptosis, and that the reduction in S-phases observed in mutant NBs results directly from cell cycle arrest rather than from ectopic cell death (Long, 1998b).

Several yeast and mammalian proteins related to Smid have a demonstrable role in the cell cycle. Disruption of the yeast CDC48 locus produces a phenotype characterized by a late arrest mitosis (Frohlich, 1991). Cdc48p is essential for homotypic membrane fusion of the ER (Latterich, 1995), suggesting that the observed mitotic phenotype is a result of the malfunction of the homotypic membrane fusion machinery essential for the progression of the cell cycle. This idea is reinforced by the observation that the vertebrate homolog of Cdc48p, VCP (or p97), functions in homotypic membrane fusion required for the post-mitotic reassembly of Golgi cisternae (Ribouille, 1995). cdc48 mutants also display a failure to duplicate the spindle pole body, thus implicating Cdc48p in the formation of the spindle apparatus. The differential localization of Cdc48p and VCP during respective yeast and mammalian cell cycles is also indicative of such a role (Madeo, 1998). The subcellular localization of both is regulated in a cell cycle-dependent fashion: both are present on the cytosolic surface of the ER at all times; however Cdc48p also enters the nucleus at the end of G1, when duplication of the spindle pole body occurs, whereas VCP is mobilized to the centrosome throughout mitosis. The correlation of subcellular localization with spindle/centrosome localion, i.e. cytosolic for the centrosome and nuclear membrane-associated for the spindle pole body, indicates conservation of this function but divergence of target localization during evolution. Thus it appears that each of these proteins may perform multiple roles during the cell cycle encompassing organelle enlargement prior to mitosis, spindle body/centrosome duplication, nuclear membrane fusion in yeast, or post-mitotic organelle assembly in higher eukaryotes (Long, 1998b and references).

More closely related to Smallminded than the proteins described above is the C. elegans protein MAC-1. MAC-1 was identified as a protein that binds the C. elegans effector of apoptosis, CED-4. The discovery of physical interactions between CED-3, CED-4 and CED-9 helped elucidate the molecular basis of cell death. Wild-type CED-9 interacts with CED-4 and prevents cell death. CED-9 might function by binding and inactivating CED-4. CED9 is homologous to the mammalian Bcl-2 and Bcl-XL proteins, which both promote survival. CED-4 is homologous to the mammalian protein Apaf-1 (Drosophila homolog: Apaf-1-related-killer), which also promotes cell death. This similarity between CED-4 and Apaf-1 is restricted to their amino-termini, which each contain a caspase recruitment domain similar to that found in the prodomains of the C. elegans caspase CED-3 and mammalian caspases like caspase-9. Thus, homophilic interactions between caspase recruitment domains might allow Apaf-1 to interact with caspase-9, and CED-4 with CED-3. Once bound, Apaf-1 promotes the processing and activation of caspase-9, just as CED-4 promotes the activation of CED-3. Immunoprecipitation studies confirm that MAC-1 interacts with CED-4, and also with Apaf-1 (Wu, 1999 and references). MAC-1 can form a multi-protein complex that also includes CED-3 or CED-9. A MAC-1 transgene under the control of a heat shock promoter prevents some natural cell deaths in C. elegans. Because it is premature to draw conclusions as to the function of MAC-1 in cell death in C. elegans (Wu, 1999), this discussion will new turn to the developmental biology of MAC-1.

Inactivation of mac-1 by RNA-mediated interference, a phenomenon that resembles mutation in its effects, causes animals to arrest as L2 larvae. This arrest is similar to that observed in smallminded mutants, but is not related to the ability of MAC-1 to bind CED-4, since it still occurs in ced-3 or ced-4 null mutants. Although most worms arrest permanently, some eventually complete development, perhaps because the double-stranded RNA that inactivates mac-1 is degraded. These mature animals show additional defects, such as the formation of vacuoles in the intestine, and abnormal vulval and gonadal development. It is not known if these problems reflect a widespread requirement for mac-1 in larval development, or are a side effect of slowed development. However, animals that fail to develop because of starvation have not been seen to exhibit these problems. These results show that the requirement for mac-1 during larval development is unlikely to involve the regulation of programmed cell death, since mutations in ced-3 and ced-4 do not suppress the arrest of mac-1(dsRNAi) animals. Since the screen that identified mac-1 was capable of finding proteins that are essential, or which have pleiotropic effects on development, it should not be surprising that MAC-1 appears to have more than one function, and that its major function might be unrelated to programmed cell death (Wu, 1999). In this respect, several members of the AAA family of ATPases, including mammalian VCP, are known to be involved in diverse cellular functions, including cycle regulation, endocytosis, membrane transport and proteasome regulation (Wu, 1999 and references).

What then is the function of Smallminded? Mutants are defective in larval mitoses, and the expression of Smallminded in cells that undergo late embryonic or larval mitosis suggests a general role in promoting mitosis. Whether Drosophila Smallminded will have other roles in cell biology related to apoptosis, membrane dynamics or protein degradation awaits further exploration into the protein and pathway interactions of the Smallminded protein.

GENE STRUCTURE

Exons - 9

Bases in 3' UTR - 285

PROTEIN STRUCTURE

Amino Acids - 943

Structural Domains

Analysis of the predicted primary structure of the protein encoded by smid indicates that it is a member of the AAA superfamily of proteins and contains a tandem duplication of the characteristic AAA module, each of which is conserved within a region of 170 amino acid residues (290-459 and 702-873). Alignment of these regions shows that they share 38% identical and 5% similar residues. The N-terminal region of each domain contains Motif-A (GPPGCGKT in both cases) and Motif B ( VLFIDE and VIFFDE) nucleotide binding consensus sequences. This is followed by the AAA protein family signature. The first domain contains two deviations from the consensus, whereas the second domain meets consensus in full. A search for related sequences shows Smid to be a new member of the Cdc48p/VCP subfamily of AAA proteins. It is most closely related to the final AAA family member to be identified in S. cerevisiae, L0919-chroXII; the SAV, CdcH and mCdc48p proteins from the archaebacteria S. acidocaldarus, H. salinarum and Methanococcus jannaschii, respectively; porcine VCP and its homologs from mouse, rat, Xenopus Arabidopsis, Glycine max, and S. cerevisiae. Alignment of Smid with these sequences shows that they all contain a duplication of the characteristic AAA module. which is the definitive feature of this subfamily. Overall, Smid shares 33% identical residues with its closest relative, L0919-chrXII, and 25%-29% for other related AAA proteins (Long, 1998).

A newly discovered member of the AAA family in C. elegans is more closely related to Smid than previously described AAA proteins. The percentage identity between Drosophila Smid and C. elegans MAC-1 is particularly high within two regions of MAC-1: amino acids 204-455 and 507-757. These regions correspond to the conserved AAA modules characteristic of this family of ATPases. Overall Smid and MAC-1 share 42.3% identity, with identity rising to 49.1% and 62.7% in the first and second nucleotide binding motifs (Wu, 1999).

The most striking difference between Smid and the other members of the AAA protein family, with the exception of MAC-1 is found in the region separating its AAA modules. Most family members have around 100 aa separating the conserved domains. The only significant deviations from this are found in L0919-chrXII, which has 156 aa and Smid, with 239 aa. Here Smid has a region of 150 aa (478-634) with little relation to any of the other family members except L0919-chrXII, which shows some sequence similarity but few identical residues. Smid also contains four putative bipartite nuclear localization signal in regions 2-19, 160-177, 406-426 and 550-568. The most likely candidate for a functional signal is that found within the region 550-568, KKATNGNSSIKSPQKTPKK, since this also contains four consensus sequences (TPKK, SPQK, TPKK and SAEK) for cyclin-dependent kinase, all of which are present within or overlap with the localization signal itself. The phosphorylation by Cdc28p of such a site adjacent to the nuclear localization signal of the yeast transcription factor SW15 is known to be a means of regulating its entry into the nucleus. Two such potential signal sequences have also been noted in the N-terminal region of Atcdc48p, and a nuclear localization has been reported for Atcdc48p, Cdc48p and p97. If, like Cdc48p and Atcdc48p, Smid performs a mitotic role, one would expect its presence in the nucleus to be tightly regulated. The presence of four potential sites of phosphorylation by cyclin-dependent kinase within this region of Smid makes it a strong candidate as a functional signal for the cell-cycle regulation of the protein. A definitive assignment, however, will require both a confirmation of its phosphorylation by cyclin-dependent kinase and its mutagenesis. As reported for other AAA proteins, numerous potential phosphorylation sites of the calmodulin kinase II, casein kinase II and protein kinase C types are also present in the Smid sequence. The amino acid composition indicates that it is a hydrophiic protein with a predicted pI of 5.13, lacking potential membrane spanning regions or signal sequences for ER import (Long, 1998a).

date revised: 18 April 99

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.