The Caenorhabditis elegans Skp1-related (skr) gene family: diverse functions in cell proliferation, morphogenesis, and meiosis

Sudhir Nayak1, Fernando Santiago2, Hui Jin2, Debbie Lin1, Tim Schedl1, Edward T. Kipreos2
1
Department of Genetics, Washington University School of Medicine, St. Louis, MO 63110
2
Department of Cellular Biology, University of Georgia, Athens, GA 30602

The SCF complex is a multi-subunit ubiquitin-ligase (E3) that mediates the degradation of a variety of cellular proteins such as cell cycle regulators and transcription factors. There are four SCF subunits: Skp1, a cullin (orthologs of metazoan CUL-1 or yeast Cdc53), Rbx1/Roc1/Hrt1, and an F-box protein. The cullin subunit acts as a scaffold to link the E2 (Ubc3/Cdc34) to the E3 complex and this association is facilitated by the Rbx1/Roc1/Hrt1 subunit. The cullin scaffold also binds to Skp1, and Skp1 binds to the F-box protein through a direct interaction with the F-box motif. In the SCF complex, the F-box protein selectively binds phosphorylated proteins to bring them into close proximity to the associated E2 that covalently transfers ubiquitin to the substrate. Poly-ubiquitinated substrates are then subsequently degraded by the 26S proteasome.

Here we present data indicating that the cullin (cul), F-box, and Skp1-related (skr) gene families have expanded in C. elegans. However, in contrast with the relatively ancient expansion of the cullin genes, the skr and F-box gene families have expanded more recently and to a greater extent in the C. elegans genome. We identified 21 skr genes in the C. elegans genome compared to a single SKP1 gene in fission yeast, budding yeast, and humans. Similarly, numerous F-box proteins have been identified primarily based on gene predictions including 11 in budding yeast, 22 in Drosophila, at least 38 in humans, and a remarkable 325-350 in C. elegans; however, the vast majority have not been characterized. In the context of the current SCF paradigm, any given cullin can form the base to an SCF with unique substrate specificity conferred by the selection of a Skp1 family member and an F-box. The staggering potential for combinatorial interactions between core components is believed to increase the repertoire of substrates that can be recognized by the core complex. We probed potential interactions with cullins using the two-hybrid system and found that multiple C. elegans SKRs have the ability to interact with CUL-1 and at least one SKR is able to interact with CUL-6, a close paralog of CUL-1. Thus, there is the potential to generate many distinct SCF and SCF-like complexes from combinatorial SKR/CUL interactions. Finally, using RNAi we have identified roles of the skr gene family in regulating multiple cellular processes including restricting cell proliferation, morphogenesis, and the pachytene stage of meiosis. From comparisons of skr RNAi phenotypes with cul-1 mutant and RNAi phenotypes we conclude that the skr gene family has cul-1 dependent and cul-1 independent functions.

Early 2002 Midwest Worm Meeting abstract 443982

Mis-regualtion of mRNA targets of GLD-1 and tumor formation

Min-Ho Lee, Mitsue Ohmachi, Tim Schedl

Department. of Genetics, Washington University School of Medicine, St. Louis, MO 63117

During germ line development, translational control emerges as a heavily utilized mechanism by which gene expression is regulated. The germ line thus provides an excellent model for studying translational regulation. Recently, many RNA binding proteins have been identified as essential components governing germ line development and early embryogenesis. A comprehensive understanding of how these RNA binding proteins control development remains largely unknown. Identification of their RNA targets is a necessary first step. Subsequent work can provide information on the normal function of each target and the consequences of its mis-regulation in the mutant/disease state.

GLD-1 is a germ line specific, maxi-KH motif containing RNA binding protein that acts as a tumor suppressor and regulates multiple aspects of C. elegans germ cell development, suggesting that it regulates multiple RNAs. GLD-1 is a member of a family of proteins, including mouse/human Quaking and Drosophila How, that share not only a maxi-KH motif but sequences surrounding the maxi-KH motif. At the distal end of C. elegans germ line, there are proliferating germ cells that, as they move proximally, initiate meiotic development and then undergo gametogenesis. GLD-1 is abundant in the cytoplasm of early meiotic prophase germ cells but absent in developing oocytes. GLD-1 was proposed to spatially restrict the accumulation of certain proteins by regulating the translation and/or the stability of a subset of maternal mRNAs that function in oocyte differentiation, maturation/ovulation and/or early embryogenesis. The tumorous mutant phenotype could result from mis-regulation of mRNA targets during early meiotic prophase.

We have identified multiple in vivo mRNA targets of GLD-1 by their ability to interact with GLD-1 in cytosol extracts. These target mRNAs are preferentially expressed in the germline and as expected, several of them have essential functions in oocyte differentiation, maturation/ovulation and early embryogenesis. Analysis of three mRNA targets (rme-2, oma-1 and oma-2), reveals that GLD-1 acts as a translational repressor. Antibody staining of wild-type hermaphrodite germlines show that the corresponding proteins for these three mRNA targets are absent from the distal region where GLD-1 is abundant, while they increase in abundance in growing oocytes in the proximal region where GLD-1 levels fall precipitously. Consistent with GLD-1 functioning as a translational repressor, they prematurely accumulate in the distal region of gld-1 null hermaphrodites. These data, as well as RNA in situ analysis, imply that GLD-1 is likely acting as a translational repressor for most mRNA targets.

lin-45 RAF was identified as one of the mRNA targets, suggesting that GLD-1 may modulate the activation of RAS/MAP kinase pathway and mis-activation of this pathway may contribute to the tumor formation in gld-1 mutant animals. Consistent with this idea, we have found that the gld-1 tumorous phenotype is partly suppressed by loss of lin-45, as well as downstream components mek-2 and mpk-1. In addition, MPK-1 is ectopically activated in transition zone and early pachytene stage germ cells in gld-1 null germlines. This suggests that mis-regulation of the lin-45 RAF kinase mRNA leads to the mis-activation of MPK-1, which partly contributes to tumor formation in gld-1 null germlines.

Early 2002 Midwest Worm Meeting abstract 486214

GLD-2 and NOS-3 Act in Opposition to GLP-1/Notch Signaling in Regulating the Accumulation of GLD-1 for Meiotic Development

Dave Hansen, Thanh Dang, Laura Wilson-Berry, Diana Duffy, Tim Schedl

Department of Genetics, Washington University School of Medicine, St. Louis, MO 63110

The distal ends of both male and hermaphrodite germlines are populated by proliferating stem cells that initiate meiotic development as they move proximally, away from the distal tip cell (DTC). We are interested in how the distal germ cells exit from a mitotic proliferative state to meiosis and differentiation. Previous work has shown that a GLP-1/Notch signaling pathway, with the LAG-2 ligand expressed by the DTC and the GLP-1 receptor expressed by the germ cells, functions to promote proliferation (Austin and Kimble, 1987; Lambie and Kimble, 1991). We have shown genetically that gld-1, encoding a KH domain containing RNA binding protein, promotes entry into meiosis or inhibits proliferation (Francis et al., 1995). Kadyk and Kimble (1998) further demonstrated that gld-1functions redundantly with gld-2 for this role such that loss of the activities of both genes results in a germ line tumor of proliferating cells in males and hermaphrodites, similar to that seen when glp-1 is constitutively activated (Berry et al., 1997). Kadyk and Kimble (1998) also demonstrated that the gld-2(0) gld-1(0) tumor is epistatic to glp-1(0), suggesting that GLP-1/Notch signaling inhibits the activities of gld-1 and gld-2. GLD-1 levels are low in the distal end but increase proximally until reaching maximum levels around 21 cell diameters from the DTC, approximately where germ cells appear to initiate meiotic development. gld-1(oz10gf) single mutants, which have increased GLD-1 accumulation in the distal end, have a smaller proliferative zone, and gld-1(oz10) enhances the weak glp-1(bn18) allele, both supporting that GLD-1 spatial regulation is important in regulating the entry into meiosis decision. Crittenden et al. (2002) have also suggested that GLD-1 levels in the distal end are important for regulating entry into meiosis based on analysis of FBF regulation of GLD-1 accumulation. Three lines of evidence suggest that the GLP-1/Notch signaling pathway is involved in inhibiting GLD-1 accumulation in the distal end. First, in the absence of GLP-1 activity GLD-1 levels are increased in the distal end. The lack of GLP-1 activity normally results in a very small germline, therefore we looked at GLD-1 accumulation in gld-2(0) gld-1(q361) mutants with and without GLP-1 activity (in the gld-1(q361) allele protein is made but is non functional). In gld-2(0) gld-1(q361) animals GLD-1 protein levels are low in the distal end but increase gradually (same as wild-type), however in gld-2(0) gld-1(q361); glp-1(0) animals, GLD-1 accumulation is dramatically increased in the distal end. Second, the rise in GLD-1 level is delayed (more proximal) in animals with increased GLP-1 activity in the distal end. Third, GLD-1 levels are low or absent in animals with tumorous germlines due to constitutively activated GLP-1.

In order to identify other genes that regulate the entry into meiosis decision, possibly through regulating GLD-1 levels, we conducted a genetic screen looking for mutations that are synthetic tumorous with gld-2(0). From this, we identified five loss-of-function mutations in a Drosophila Nanos homologue, nos-3, which was previously identified for its ability to bind FBF (Kraemer et al., 1999). While GLD-1 levels are roughly equivalent to wild-type in single nos-3(0) and gld-2(0) mutants, they are dramatically reduced in gld-2(0); nos-3(0) double mutants suggesting that gld-2 and nos-3 function redundantly to promote GLD-1 accumulation. Therefore gld-2 and nos-3 function in opposition to Glp-1/Notch signaling in regulating GLD-1 accumulation in the distal germline.

Early 2002 Midwest Worm Meeting abstract 524424

Genetic Studies of the Proliferation versus Meiotic Development Decision

Jessica Amrozowicz, Dave Hansen, Shawn Shafer, Tim Schedl

Washington University, 4566 Scott Ave., Campus Box 8232, Saint Louis, MO 63110

GLP-1 is a member of the Notch family of transmembrane receptors.  Expressed in the germline of C. elegans, it functions to promote proliferation of germ cells.  Spatial regulation of GLP-1 receptor activation is controlled, at least in part, by localizing the transmembrane ligand, LAG-2, to the distal tip cell, which caps the distal end of the germline.  Binding of LAG-2 to GLP-1 induces cleavage of the intracellular domain of GLP-1, where the activated GLP-1(INTRA) is presumed to translocate to the nucleus of the germ cell and bind LAG-1, a Notch downstream effector and putative transcriptional regulator. These three components of the GLP-1 pathway work together to produce a population of proliferating germ cells that extends ~20 cell diameters from the distal tip.  Disruption of this pathway, through a loss of function glp-1 mutation, for example, causes nearly all proliferating germ cells to prematurely enter meiosis.  Constitutive activation by the glp-1(oz112 gf) mutation, conversely, results in the formation of a germline tumor where proliferative cells are found throughout the gonad.  Weaker glp-1(gf) alleles (eg, ar202) also cause an over-proliferation phenotype, characterized by proximal proliferation and the formation of a late onset tumor (Pepper et al, 13th Int. C. elegans Meeting).  These results indicate that the activity state of the GLP-1 signaling pathway determines whether germ cells remain in the proliferative state or enter meiotic development. 

            We have taken a genetic approach to identify genes that regulate or are targets of GLP-1 signaling.  A number of genes have been isolated that enhance the glp-1(gf) phenotype.  teg-1, for tumorous enhancer of glp-1,  enhances a very weak glp-1(gf) allele by causing the formation of a late onset tumor.  This result suggests that teg-1 is a negative regulator of glp-1 signaling (Hansen et al, 2000 Midwest Worm Meeting).  However, to date, little is know about the specific function of TEG-1. 

            In a genetic screen to find enhancers of teg-1, we have isolated a glp-1 double mutant, oz264oz270, that forms a tumor with teg-1.  In a teg-1(+) background, the double mutant displays a weak over-proliferation phenotype.  Another weak glp-1 (gf) allele, ar202, also displays a weak over-proliferation phenotype, however it is stronger than that of glp-1(oz264oz270).  The glp-1(oz264) lesion lies in a residue adjacent to glp-1(ar202) (Hubbard et al, personal communication), suggesting that glp-1(oz264) may also act as a glp-1(gf) mutant by causing some level of constitutive activation.   But in the teg-1(-) background, the tumor formed with glp-1(ar202) is apparently weaker than that with glp-1(oz264oz270).  These conflicting results could be explained by the other lesion, glp-1(oz270), acting with both partial loss and partial gain of function characteristics. Experiments are in progress to try to understand the nature of the glp-1(oz264oz270) double mutant. 

            We are also conducting other genetic screens to identify novel interactors in the GLP-1 pathway.  One mutation, oz272, causes the formation of a germline tumor, but does not behave like glp-1(gf) alleles, or the synthetic tumorous alleles previously isolated, suggesting that it is either a mutation in a novel gene, or a new allele in a previously identified gene involved in the proliferation versus meiotic development decision.

Early 2002 Midwest Worm Meeting abstract 446341

MPK-1 ERK signaling is necessary for male germline sex determination

Mitsue Ohmachi1, Min-Ho Lee1, Eric Lambie2, Ross Francis1, Tim Schedl1
1
Department of Genetics, Washington University School of Medicine, St. Louis MO
2
Department of Biological Sciences, Dartmouth College, Hanover, NH

MPK-1 ERK signaling is necessary for pachytene progression and oocyte meiotic maturation (Church et al., 1994; Lee et al., 13th International C. elegans Meeting, 2001). Strong loss-of-function/null mutations in lin-45 RAF, mek-2 MAPKK and mpk-1 ERK result in hermaphrodites that fail to produce either sperm or oocytes and instead germline nuclei are arrested in pachytene. In the distal meiotic prophase germline of wild-type animals, germ cells (nucleus with surrounding plasma membranes except on the side facing the central cytoplasmic core) are packed in a hexagonal pattern on the surface of the gonadal tube. In lin-45, mek-2 and mpk-1 mutants, the hexagonal germ cell packing is disrupted and instead pachytene arrested nuclei and surrounding plasma membranes are found in clumps, often in the center of the gonadal tube. We call this a Pac phenotype, for pachytene arrest and clumped nuclei and membranes. A similar Pac phenotype is observed in mutant males.

Under certain partial loss-of-function conditions, where pachytene progression and oocyte differentiation is normal, we observed hermaphrodites with feminized germlines (proximal germ cells developing as oocytes instead of sperm). The partial loss-of-function conditions were: a) viable escapers from lin-45 or mpk-1 RNAi; b) mek-2(n1859) (Kornfeld et al., 1995); and c) trans-heterozygotes of mpk-1 null and mpk-1(ts) (Lackner and Kim, 1998) at the permissive temperature. These results raise the possibility that the Pac phenotype observed in males and in the proximal germline of hermaphrodites for lin-45, mek-2 and mpk-1 strong loss-of-function/null mutants is the result of the combined effect of a sexual fate transformation followed by pachytene arrest during oogenesis. In situ hybridization using a probe to rme-2 mRNA, which encodes the oogenesis specific yolk receptor, was used to test this possibility. In wild-type L4 hermaphrodites, germ cells in the proximal gonad, most of which are in pachytene, are undergoing spermatogenesis and lack rme-2 mRNA while germ cells in the proximal third of the distal gonad contain rme-2 mRNA. By contrast, germ cells throughout the proximal gonad and into the distal gonad have rme-2 mRNA in L4 mpk-1 null hermaphrodites. In mpk-1 null males, rme-2 mRNA is found throughout the region showing the Pac phenotype, unlike in wild-type males. Thus, mpk-1 is required for the male germ cell fate in both hermaphrodites and males. Additional experiments indicate that mpk-1 is functioning in the germline, not in the soma, for male germline sex determination and that mpk-1 activity is required continuously. Epistasis experiments indicate that mpk-1 is acting downstream or in parallel to tra-2 in the sex determination pathway, as tra-2 null; mpk-1 null double mutant germlines are feminized. mpk-1 may not be required for male somatic sex determination as sexual transformation in the adult soma (yolk synthesis) has not been observed.

Accumulation of MPK-1 and the appearance of activated MPK-1 is sexually dimorphic. An affinity purified C-terminal fraction of the SC94 (Santa Cruz) antisera is specific for germline MPK-1 as germ cells do not stain in the null mutant. MPK-1 is found at high levels throughout the germline of wild-type adult hermaphrodites. By contrast, in wild-type males, MPK-1 is found at significantly lower levels and only in the distal most germ cells, up to about the distal third of the pachytene region. ERK MAP kinases are activated by phosphorylation on a specific Thr and Tyr motif, which can be detected by the MAPKYT mAb (Sigma; Miller et al., 2000). In hermaphrodites, activated MPK-1 is found at high levels in the 2 to 3 most proximal oocytes and at moderate levels in the proximal half of pachytene. In males, activated MPK-1 is only found at very low levels in the transition zone and the first few pachytene germ cells, a region where activated MPK-1 is not observed in adult hermaphrodites. Thus, MPK-1 is activated sex specifically and to different extents in three spatially distinct regions of the C. elegans germline.