We have previously identified a gene,
iff-1, encoding a Caenorhabditis elegans eIF-5A homologue in a reverse genetic screen as one of the genes that function in germline development (Hanazawa et al., 2001). A further analysis was made on its role in germline development. We constructed a deletion mutant of
iff-1,
iff-1(
tm483), which hatches and develops into sterile adults similar to the RNAi-affected worms; small gonads with few nuclei with an appearance of mitotic interphase nuclei were observed. The fact that no excess dead cell was observed in the gonad shows that the reduction in the number of germ cells is not caused by ectopic cell deaths in the
iff-1(
tm483) mutant. To assess the possible role of
iff-1 for germline proliferation, we tested the effect of
iff-1 knock-down in the
gld-2(
q497)
gld-1(
q485) double mutant. The germline-tumorigenesis of the
gld-2 gld-1 mutant was suppressed by RNAi against
iff-1, supporting the idea that
iff-1 is essential for the germ cells to proliferate. Localization of
iff-1 mRNA and protein, as detected by in situ hybridization, Northern analysis and Western analysis using the anti-IFF-1 antibodies, also supported this idea.
iff-1 mRNA and protein could not be detected in the
glp-4 mutant, which has few germ cells. This is consistent with our original assumption that IFF-1 is expressed in a germline-specific manner. Upon in situ hybridization,
iff-1 mRNA was detected in the distal region of the gonads where germ cells actively proliferate. Although eIF-5A was originally isolated as a candidate translation initiation factor, recent studies have suggested roles of eIF-5A in many aspects of RNA metabolism, including nuclear export, cytoplasmic degradation and translation. In the C. elegans gonads, many RNA binding proteins are known to play various key roles in regulating germ cell development. Many of them are localized on P granules, which are located close to the nuclear pore complex. Because the eIF-5A homologue in Xenopus was reported to localize on the filaments extending from the nuclear pore, it is of interest to test the possibility that IFF-1 collaborates with these RNA binding proteins that function on P granules. As one way of tesing the interplay of IFF-1 and P granule components, we examined the localizatioin of PGL-1 and GLH-1 in the
iff-1 mutant by immunostaining. In the
iff-1 mutant, unlike in wild type, these P granule components did not localize at the nuclear pore but were dispersed in the cytoplasm. These results suggest that IFF-1 is necessary for some P granule components to localize normally, and point to a possible functional link among IFF-1, P granules, and RNA metabolism. There is another gene,
iff-2, encoding an eIF-5A homologue in C. elegans. The
iff-2(
tm393) deletion mutant grows slowly during larval stages and becomes sterile adults lacking a part of the somatic gonad structure. This phenotype and the results of expression analysis indicated that two eIF-5A homologues might take charge of part of roles: one for germ cells and the other for somatic cells. We are now working to clarify the targets of IFF-1 and the functional differences between these two homologues.