Distinct roles of two eIF4E isoforms in the germline of Caenorhabditis elegans

Abstract

Germ cells use both positive and negative mRNA translational control to regulate gene expression that drives their differentiation into gametes. mRNA translational control is mediated by RNA-binding proteins, miRNAs and translation initiation factors. We have uncovered the discrete roles of two translation initiation factor eIF4E isoforms (IFE-1, IFE-3) that bind 7-methylguanosine (m7G) mRNA caps during Caenorhabditiselegans germline development. IFE-3 plays important roles in germline sex determination (GSD), where it promotes oocyte cell fate and is dispensable for spermatogenesis. IFE-3 is expressed throughout the germline and localizes to germ granules, but is distinct from IFE-1 and PGL-1, and facilitates oocyte growth and viability. This contrasts with the robust expression in spermatocytes of IFE-1, the isoform that resides within P granules in spermatocytes and oocytes, and promotes late spermatogenesis. Each eIF4E is localized by its cognate eIF4E-binding protein (IFE-1:PGL-1 and IFE-3:IFET-1). IFE-3 and IFET-1 regulate translation of several GSD mRNAs, but not those under control of IFE-1. Distinct mutant phenotypes, in vivo localization and differential mRNA translation suggest independent dormant and active periods for each eIF4E isoform in the germline.

Keywords: 4EBP; Gametogenesis; Germline sex determination; Polysomes; eIF4E; mRNA translational control.

Fig. 1.

IFE-1 and IFE-3 localization in germ cells and gametes. (A) Differential translational regulation of mRNAs involved in gamete maturation (red), sperm/oocyte fate (green) and housekeeping mRNAs (black) by germline eIF4E isoforms (IFE-1, -2 and -3). Mutually exclusive binding partners for eIF4Es (4EBPs and eIF4G) regulate mRNA selection by forming a repression-to-activation node after nuclear export. The model, based on this and other studies, is that IFE-1 and IFE-3 use distinct 4EBPs (x and y) for differential spatial and temporal localization in germ granule messenger ribonucleoproteins (mRNPs) that prevent translation of their cargo mRNAs. During germ cell development, each mRNP is remodeled independently. IFE-bound mRNAs are released and recruited by eIF4G to actively translating ribosomal complexes. In this model, direct recruitment by IFE-1 or IFE-3 following nuclear export is unlikely (dashed line arrow). Consequently, the IFEs and their 4EBPs play cooperative roles in both mRNA repression and activation. IFE-2 is a fully soluble germ cell eIF4E not found in granules, and it recruits a different subset of mRNAs (Song et al., 2010). (B–D) IFE-1 (red) and IFE-3 (green) co-expression in distal germ cells shows that they localize to adjacent yet distinct perinuclear foci (insets, arrowheads). Cytosolic IFE-1 and IFE-3 overlap substantially (yellow) throughout distal germ cells. (E–G) Co-expression in the gonad core (dashed line boxes) shows that IFE-1 is soluble and localized diffusely, whereas IFE-3 associates with extensive lattice-like structures. (H–K) Co-expression in spermatocytes shows that IFE-1 is upregulated and becomes enriched in perinuclear foci in primary spermatocytes (1°), but then becomes soluble in secondary spermatocytes (2°). Primary, secondary and spermatid differentiation regions are indicated by the dashed lines. IFE-1 is deposited into residual bodies (rb) after spermatid (sp) budding. IFE-3 does not form perinuclear foci and is diminished during the latter stages of spermatogenesis. (L–O) Co-expression in oocytes shows that IFE-1 forms perinuclear foci that move away from the nucleus (arrowheads) as oocytes approach the spermatheca, whereas IFE-3 is soluble and diffuse. The last oocyte prior to fertilization is indicated by -1. CRISPR/Cas9 fluorescently tagged mKate2::TEV::3xmyc::IFE-1 and GFP::TEV::3xflag::IFE-3 were used to evaluate expression of endogenous genes. Gonads were dissected, fixed and counterstained (DAPI, blue) for nuclear morphology. Scale bars: 50 µm.

Fig. 2.

Characterization of ife-3(ok191) mutants. (A) Schematic of ife-3 on chromosome V (left). The ok191 deletion removes the promoter and exon 1. The gel (right) shows whole-worm genomic PCR at the ife-3 locus from wild-type (+/+), heterozygous (+/−) and homozygous (−/−) animals. (B) Wild-type hermaphrodite gonad showing sperm in the spermatheca stained for major sperm protein (MSP; red), and oocytes, showing LIN-41 expression (green). (C) Wild-type male gonad showing only MSP-stained sperm and no LIN-41. (D) ife-3(−/−) hermaphrodite gonad showing masculinization phenotype (Mog) at 20°C, as indicated by overabundance of MSP-stained sperm throughout the gonad, and complete lack of LIN-41-expressing oocytes. (E) ife-3(−/−) hermaphrodite gonad showing both MSP-stained sperm and small LIN-41-expressing oocytes at 25°C. (F) Fertility and embryonic viability in ife-3(+/−) and ife-3(−/−) mothers at both 20°C and 25°C. Genotype of each individual was confirmed by genomic PCR. Germlines shown (B–E) carry lin-41(tn1541[gfp::tev::s::lin-41])I to visualize oocytes and were dissected, fixed, stained with anti-MSP antibody and counterstained (DAPI, blue). Scale bar: 50 µm.

Fig. 3.

Epistatic analysis of ife-3 and other regulators of GSD. (A) Germline sex determination (GSD) in C. elegans is a series of positive and negative genetic switches that rely on mRNA translational control. Factors that drive sperm fate (red) or oocyte fate (green) are shown. (B) Epistatic outcomes from ife-3(RNAi) in mutant backgrounds depicted in a bar graph that shows the distributions of phenotypes for each group. (C) Control(RNAi)-treated wild-type hermaphrodite gonad showing sperm (MSP, red) and oocytes, as identified by nuclear morphology (DAPI, blue). (D) ife-3(RNAi)-treated wild-type hermaphrodite gonad showing masculinization phenotype (Mog) as indicated by an overabundance of sperm and lack of oocytes. (E) Control(RNAi)-treated fog-2(q71) gonad showing feminization phenotype (Fog), as indicated by lack of MSP-stained sperm and presence of stacked oocytes. (F) ife-3(RNAi)-treated fog-2(q71) gonad showing Mog phenotype, demonstrating reversion of fog-2(q71)-induced feminization. (G) Control(RNAi)-treated fem-3(e2006) gonad showing Fog phenotype. (H) ife-3(RNAi)-treated fem-3(e2006) gonad showing Fog phenotype with small germ cells. No sperm were ever observed in fem-3(e2006) worms (B), demonstrating failure to revert the feminization phenotype (asterisk). The resulting small germ cells were later found to be LIN-41-expressing oocytes (Fig. 4A). (I) Control(RNAi)-treated fbf-1(ok91) gonad showing sperm and oocytes. (J) ife-3(RNAi)-treated fbf-1(ok91) gonad showing Mog phenotype. fbf-1(ok91) mutants treated with ife-3(RNAi) displayed enhanced penetrance of Mog phenotype relative to wild-type animals. Germlines of the indicated genotype were dissected, fixed, stained for MSP to visualize sperm and counterstained (DAPI, blue). Gonads are outlined (dashed lines) for clarity. Fog phenotype is denoted by asterisks. Scale bar: 50 µm.

Fig. 4.

Translational efficiency of GSD mRNAs is altered by IFE-3 depletion. (A) Schematic of the experimental design for high-resolution polysome analysis. Wild-type animals were grown on control(RNAi) (blue) or ife-3(RNAi) (red). Micrographs depict representative phenotypes of control(RNAi) or ife-3(RNAi) worms. As before, IFE-3 depletion resulted in a minority (22%) of Mog animals; the majority (78%) remained oogenetic. Absorbance (A₂₅₄) profiles for gradient fractionation of extracts from control(RNAi) (blue) and ife-3(RNAi)-treated (red) worms depict the polysome content. The highest resolved peak is the monosome (80S position is indicated by arrowhead) and each peak to the right represents the addition of one ribosome (polysomes). (B–H) Individual graphs depict normalized mRNA content across the gradient by qRT-PCR for fem-3 (B), fog-1 (C), fbf-1 (D), gld-1 (E), tra-2 (F), gpd-3 (G) and mex-1 (H) mRNAs. Normalization by total RNA content across the gradient corrects for slight differences in polysome yield, allowing direct comparison of the partitioning of each mRNA. Error bars indicate the s.d. of triplicate qRT-PCR measurements. Similar profiles were obtained for fem-3, fog-1 and gpd-3 mRNAs in two to three independent experiments. (I) Relative mRNA abundance in each gradient (normalized to gpd-3) shows that in ife-3(RNAi)-treated animals, fog-1 is increased 2.5-fold, while all other mRNAs assayed (tra-2, mex-1, fem-3, fbf-1, fbf-2 and gld-1) were decreased by 25–50% compared to control(RNAi)-treated animals. Germlines shown (A) carry lin-41(tn1541[gfp::tev::s::lin-41])I to visualize oocytes and were dissected, fixed, stained for MSP to visualize sperm and counterstained (DAPI, blue). Scale bar: 50 µm.

Fig. 5.

Steady-state levels of fem-3 and fog-1 mRNAs in ife-3(ok191) hermaphrodites. (A) Relative fem-3 mRNA expression in early L4 hermaphrodites shows no statistically significant change (P=0.328) between ife-3(−/−) (gray) and wild-type (white) animals. Relative fem-3 mRNA expression in older L4 hermaphrodites shows a statistically significant (P=0.018) 2-fold decrease in ife-3(−/−) animals. (B) Relative fog-1 mRNA expression in early L4 hermaphrodites shows a 9-fold increase (P=0.007) in ife-3(−/−) animals (gray) compared to wild type (white). Relative fog-1 mRNA expression in older L4 hermaphrodites shows a similar 6-fold increase (P=0.019) in ife-3(−/−) animals. Animals were synchronized and staging of L4 larvae was confirmed via vulva morphology. Triplicate qRT-PCR measurements from three independent biological replicates are shown, with error bars indicating s.d. Quantifications were normalized to tubulin (tbb-2) mRNA. Student's t-test was used to determine statistical significance between wild-type and mutant animals (*P<0.05). a.u., arbitrary unit.

Fig. 6.

IFE-1 and IFE-3 localization in the germline relative to PGL-1. (A–C) IFE-1 (red) and PGL-1 (green) co-expression in distal germ cells shows that IFE-1 localizes to P granules (insets, arrowheads) with perfect overlap. (D–F) Co-expression in an L4 hermaphrodite gonad shows that IFE-1 is upregulated during spermatogenesis and becomes enriched in PGL-1 granules in primary spermatocytes (1°). Upon PGL-1 disappearance from secondary spermatocytes (2°), IFE-1 is lost from granules and becomes soluble. (G–I) Co-expression in an adult hermaphrodite gonad shows that IFE-1 localizes to PGL-1 granules that move toward the cortex as oocytes approach the spermatheca (arrowheads). There is also substantial soluble IFE-1 in oocytes. (J–L) IFE-3 (green) and PGL-1 (red) co-expression in distal germ cells shows that IFE-3 perinuclear granules are adjacent yet distinct from PGL-1 granules (insets, arrowheads). Multiple IFE-3 foci surrounding a singular PGL-1 focus were frequently observed (J–L, top arrowhead). (M–O) Co-expression in an L4 hermaphrodite shows that IFE-3 expression is relatively constant and soluble throughout spermatogenesis, not enriched in granules. (P–R) Co-expression in an adult hermaphrodite gonad shows that IFE-3 is completely soluble in oocytes, not localized to PGL-1 granules (arrowheads). The boundary between primary and secondary spermatocytes is indicated by a dashed line. The last oocyte prior to fertilization is indicated by -1. CRISPR/Cas9 fluorescently tagged mKate2::TEV::3xmyc::IFE-1 and GFP::TEV::3xflag::IFE-3 were used to evaluate expression. CRISPR-tagged PGL-1 of the reciprocal color was used to track expression of endogenous PGL-1. Gonads were imaged in immobilized live animals. Scale bars: 50 µm.

Fig. 7.

IFE-3:IFET-1 localization following depletion of each subunit. (A) IFE-3 (green) expression in animals treated with control(RNAi) shows that IFE-3 forms perinuclear granules in distal germ cells (inset, arrowheads). (B) IFE-3 expression in animals treated with ifet-1(RNAi) shows that perinuclear IFE-3 granules are disrupted (inset). (C) IFE-3 expression in animals treated with control(RNAi) shows that IFE-3 also forms lattice-like structures in the gonad core (inset). (D) IFE-3 expression in animals treated with ifet-1(RNAi) shows that IFE-3 lattice-like structures in the gonad core are disrupted (inset). (E) IFET-1 (green) expression in animals treated with control(RNAi) shows that IFET-1 forms perinuclear granules in distal germ cells (inset, arrowheads). (F) IFET-1 expression in animals treated with ife-3(RNAi) shows that perinuclear IFET-1 granules remain intact (inset, arrowheads). (G) IFET-1 expression in animals treated with control(RNAi) shows that IFET-1 also forms lattice-like structures in the gonad core (inset). (H) IFET-1 expression in animals treated with ife-3(RNAi) shows that IFET-1 lattice-like structures remain intact (inset). (I) IFE-1 (red) expression in animals treated with control(RNAi) shows that IFE-1 forms perinuclear granules in distal germ cells (inset, arrowheads). (J) IFE-1 expression in animals treated with ifet-1(RNAi) shows that perinuclear IFE-1 granules remain intact (inset, arrowheads). (K) IFE-1 expression in animals treated with control(RNAi) shows that IFE-1 does not form structures in the gonad core (inset). (L) IFE-1 expression in animals treated with ifet-1(RNAi) shows that IFE-1 accumulates into aggregates in the gonad core (inset, arrowheads). We found these IFE-1 aggregates to be coincident with PGL-1 (Fig. S3H). CRISPR/Cas9 fluorescently tagged mKate2::TEV::3xmyc::IFE-1 and GFP::TEV::3xflag::IFE-3 were used to evaluate expression. Gonads were dissected, fixed, and counterstained (DAPI, blue). Scale bars: 50 µm.

Fig. 8.

Loss of IFE-3 or IFET-1 causes identical changes in GSD mRNA translation. (A) Absorbance (A₂₅₄) profiles for matched gradient fractionation of extracts from control(RNAi) (blue), ife-3(RNAi) (red) and ifet-1(RNAi)-treated (yellow) worms depict the polysome content. The highest resolved peak is the monosome (80S position is indicated by arrowhead) and each peak to the right represents the addition of one ribosome (polysomes). (B–F) Individual graphs depict normalized mRNA content across three matched gradients as measured by qRT-PCR for fog-1 (B), fem-3 (C), daz-1 (D), gpd-3 (E) and mex-1 (F) mRNAs. Normalization by total RNA content across the gradient corrects for slight differences in polysome yield, allowing direct comparison of the partitioning of each mRNA in polysomal and non-polysomal complexes. Error bars indicate the s.d. of triplicate qRT-PCR measurements.