Introduction

Germ cells are distinct from somatic cells in their immortality, totipotency, and ability to undergo meiosis. Germ cells of various organisms contain distinctive cytoplasmic organelles called germ granules, which are made of RNAs and RNA-binding proteins1,2. Because of their presence in diverse organisms, germ granules are considered to play pivotal roles in germline development possibly through conferring special characteristics on germ cells. P granules, the C. elegans germ granules3,4, are required for proper postembryonic germline development5,6,7. We previously reported that germline apoptosis increased by loss of PGL-1 and PGL-3, members of a family of constitutive P-granule components5,6, from germ cells in adult hermaphrodite gonads8.

In synthetic multivulva class B (synMuv B) mutants, somatic cells acquire germline-like characteristics and express germline-specific proteins including PGL-1 and PGL-3 especially at elevated temperatures9,10,11. It was previously reported that a number of synMuv B mutants showed decreased somatic cell death by an unknown mechanism12. Since we found that PGL-1 and PGL-3 repress apoptosis in the germline8, we hypothesized that ectopic expression of PGL-1 and PGL-3 in synMuv B-mutant somatic cells might antagonize programmed cell death in the soma. Here we show that somatic apoptosis decreases by ectopic expression of PGL-1 and PGL-3 in somatic cells, indicating that PGL-1 and PGL-3 can serve as negative regulators of apoptosis not only in the germline but also in the soma in C. elegans. Our finding brings up a fascinating possibility that expression of a germline program in somatic cells can provide the soma with partial germline immortality at a cellular level.

Results and Discussion

Somatic apoptosis was reduced in hpl-2 mutants due to ectopic expression of PGL-1 in somatic cells

To measure somatic apoptosis, we scored the number of persistent cell corpses in the heads of newly hatched L1 larvae using Nomarski differential interference contrast (DIC) microscopy in ced-1 mutant background, in which engulfment of cell corpses is blocked, hence a sensitive condition to score cell corpses is provided13,14. Among synMuv B-class mutants, we mainly examined hpl-2 mutants15. First, we confirmed that both PGL-1 and PGL-3 were ectopically expressed in somatic tissues, including the head part, in newly hatched L1 larvae of ced-1; hpl-2 double, but not ced-1 single, mutants at an elevated temperature, 25 °C (Fig. 1a). We found that the number of cell corpses in the L1 head was significantly reduced in ced-1; hpl-2 double mutants compared to ced-1 single mutants at 25 °C but not at 20 °C (Fig. 1b,c). We consider that the reason why the number of cell corpses was more decreased at 25 °C than at 20 °C in ced-1; hpl-2 double mutants was because there were more PGL-1 and PGL-3 proteins at 25 °C than at 20 °C in hpl-2 mutant somatic cells, as previously described9. Furthermore, this reduced somatic apoptosis in ced-1; hpl-2 double mutants was significantly recovered in ced-1; hpl-2; pgl-1 triple mutants at 25 °C (Fig. 1b,c). These results suggest that somatic apoptosis is reduced in hpl-2 mutants due to ectopic expression of PGL-1 in somatic cells.

Figure 1: Somatic apoptosis was reduced in synMuv B mutants due to ectopic expression of P-granule component genes in somatic cells.
figure 1

(a) Ectopic expression of PGL-1 and PGL-3, members of a family of germline-specific P-granule components, in the L1 larval head of ced-1(e1735); hpl-2(tm1489) double, but not ced-1(e1735) single, mutants grown at 25 °C. Differential interference contrast (DIC) and immunofluorescence images double-immunostained with anti-PGL-1 (green) and anti-PGL-3 (red) antibodies along with TO-PRO-3 DNA staining (blue). Scale bar: 20 μm. (b) L1 larval heads of ced-1(e1735) single, ced-1(e1735); hpl-2(tm1489) double, and ced-1(e1735); hpl-2(tm1489); pgl-1(ct131) triple mutants displayed different numbers of persistent cell corpses at 25 °C. Some of the cell corpses are marked with yellow arrowheads. Scale bar: 20 μm. (c) Mean ± s.d. numbers of persistent cell corpses per L1 head (n > 50) in ced-1(e1735) single, ced-1(e1735); hpl-2(tm1489) double, and ced-1(e1735); hpl-2(tm1489); pgl-1(ct131) triple mutants grown at either 20 °C (white bars) or 25 °C (black bars). *P < 0.05. (d) Mean ± s.d. numbers of persistent cell corpses per L1 head at 25 °C (n > 50) in ced-1(tm2420) mutants after RNAi depletion of lin-9, lin-35, lin-37, and lin-54 gene, respectively, with mock RNAi control. *P < 0.05 compared with mock RNAi. (e) Mean ± s.d. numbers of persistent cell corpses per L1 head at 25 °C (n > 50) in ced-1(tm2420); hpl-2(tm1489) double mutants after RNAi depletion of pgl-1, pgl-3, glh-1, glh-4, cgh-1, and car-1 gene, respectively, with mock RNAi control. *,#P < 0.05 compared with mock RNAi.

Full size image

Somatic apoptosis was reduced by RNAi depletion of synMuv B-class genes, and reduced somatic apoptosis in hpl-2 mutants was recovered by RNAi depletion of a subset of P-granule component genes

To examine whether depletion of other synMuv B-class genes also causes a reduction in somatic apoptosis, we scored the number of cell corpses in the L1 heads of ced-1 mutants after RNAi depletion of several synMuv B-class genes at 25 °C (Fig. 1d). We found that the level of somatic apoptosis was significantly reduced, as in ced-1; hpl-2 double mutants, after RNAi depletion of lin-9, lin-35, lin-37, and lin-54 (Fig. 1d), all of which are synMuv B-class genes, and depletion of them was previously shown to cause ectopic expression of PGL-1 and PGL-3 in the soma9. These results indicate that reduction of somatic apoptosis is a common characteristic among synMuv B-class mutants. We also examined whether depletion of other P-granule component genes recovers the reduced somatic apoptosis in ced-1; hpl-2 double mutants, as did pgl-1 mutation (Fig. 1b,c,e). We found that the reduced somatic apoptosis in ced-1; hpl-2 double mutants was significantly recovered after RNAi depletion of pgl-1, pgl-3, glh-116, and glh-47, but not after RNAi depletion of cgh-117 and car-118 (Fig. 1e). These results indicate that reduced somatic apoptosis in synMuv B-class mutants is recovered by depletion of a subset of P-granule component genes. We confirmed that our RNAi treatments indeed diminished the expression or activity of target genes through the observation that all the RNAi treatments successfully phenocopied previously reported loss-of-function mutant phenotypes of these genes (Figs S1 and S3c).

CED-4 expression was reduced in hpl-2-mutant somatic cells due to ectopic expression of PGL-1

To understand underlying mechanisms that contribute to a reduction in somatic apoptosis in hpl-2 mutants, we examined whether expression level of any essential apoptosis regulator was affected in hpl-2 mutants. In a previous study, we found that the protein level of CED-4, an adaptor protein similar to mammalian apoptotic protease-activating factor-1 (Apaf-1)19, increased in pgl-1 mutant gonads, indicating that PGL-1 is a negative regulator of CED-4 expression in the germline8. We hypothesized that ectopic expression of PGL-1 in synMuv B-mutant somatic cells might repress CED-4 expression in the soma. First, we confirmed that somatic apoptosis in ced-1; hpl-2; pgl-1 triple mutants was dependent on the activities of ced-3 and ced-4, the core apoptotic machinery, because somatic apoptosis in the triple mutants was significantly suppressed by RNAi depletion of ced-3 and ced-4 (Fig. 2a). Next, we examined expression level of a ced-4::gfp transgene, opIs219, the transcription of which is controlled by its own ced-4 promoter20, in ca. 300-cell-stage embryos in wild-type, hpl-2 single-mutant, and hpl-2; pgl-1(RNAi) double-mutant backgrounds at 25 °C (Fig. 2b,c). We found that in hpl-2-mutant background, although PGL-1 was ectopically expressed in the somatic blastomeres, subcellular localizations of CED-4::GFP and PGL-1 were not overlapping. CED-4::GFP displayed a weblike pattern, suggesting that CED-4::GFP most likely localized to mitochondria, whereas PGL-1 displayed a granular pattern in the embryonic cells (Fig. 2b). There may exist an inverse correlation between localization of CED-4 and PGL-1 because when this ced-4::gfp transgene was expressed in wild-type background, the signal was substantially weaker in the primordial germ cells, Z2 and Z3, to which PGL-1 specifically localized, than in the neighboring somatic blastomeres in the transgenic embryos (Fig. S2a). Notably, CED-4::GFP overall expression was significantly reduced in hpl-2-mutant embryos compared to wild-type embryos (Fig. 2b,c). Reduction of CED-4::GFP expression compared to wild type was also observed in another synMuv B-class, lin-13-mutant embryos (Fig. S2b). This reduction of ced-4 expression in hpl-2 (and lin-13) mutants seems to occur at a post-transcriptional level, because mRNA levels of ced-4 were not significantly different between wild-type N2 and hpl-2 mutants at 25 °C (Fig. S2c; p = 0.83). Furthermore, CED-4::GFP expression was recovered to a higher level by RNAi depletion of pgl-1 in hpl-2-mutant embryos compared to mock RNAi-treated hpl-2-mutant embryos (Fig. 2b,c). These results indicate that ectopic expression of PGL-1 in hpl-2-mutant somatic cells was responsible for the reduction in CED-4 expression in the soma, which likely contributes to repression of somatic apoptosis in hpl-2 and other synMuv B-class mutants.

Figure 2: CED-4::GFP expression was reduced in hpl-2-mutant embryos due to ectopic expression of PGL-1.
figure 2

(a) Mean ± s.d. numbers of persistent cell corpses per L1 head at 25 °C (n > 50) in ced-1(e1735); hpl-2(tm1489); pgl-1(ct131) triple mutants with or without RNAi depletion of ced-3 or ced-4. *P < 0.05 compared with mock RNAi control. (b) Expression of a ced-4::gfp transgene, opIs219, which is driven by its own ced-4 promoter, in ca. 300-cell-stage embryos at 25 °C in wild type (wt) and hpl-2(tm1489) mutants with or without RNAi depletion of pgl-1. Expressions of PGL-1 (red) and CED-4::GFP (green) were detected by immunofluorescence staining with anti-PGL-1 (OIC1D4) and anti-GFP antibodies, respectively. Note the ectopic expression of PGL-1 in somatic blastomeres in the mock RNAi-treated hpl-2 mutant embryo. Also note that PGL-1 signals were mostly eliminated in the pgl-1 RNAi-treated hpl-2 mutant embryo. Arrows: primordial germ cells, Z2 and Z3, which contained PGL-1-positive P granules. Scale bar: 20 μm. (c) Expression levels of CED-4::GFP in the opIs219[Pced-4::ced-4::gfp] transgenic embryos at 25 °C in wild-type (wt) and hpl-2(tm1489)-mutant backgrounds with or without RNAi depletion of pgl-1, which were digitally quantified from immunofluorescence images and shown as mean ± s.d. pixel intensities per embryo using an arbitrary unit. More than 50 embryos were quantified for each RNAi condition. *P < 0.05 compared with mock RNAi-treated hpl-2(tm1489); opIs219[Pced-4::ced-4::gfp] embryos.

Full size image

Somatic apoptosis in hpl-2 mutants increased following DNA damage in a sir-2.1-dependent manner

In germ cells, apoptosis increases following DNA damage compared to physiological conditions, which is termed DNA-damage-induced germline apoptosis21. By contrast, in wild-type somatic cells, the apoptosis level does not change before and after DNA damage21. It was shown that sir-2.1, which encodes a C. elegans Sirtuin homolog22, was essential for inducing a higher level of germline apoptosis following DNA damage23. During apoptosis, SIR-2.1 changes its subcellular localization from the nucleus to the cytoplasm and transiently colocalizes with CED-4, possibly to activate CED-323. In a previous study, we identified that sir-2.1 was essential for high levels of germline apoptosis in pgl-1 and pgl-3 mutants8. We also uncovered that cytoplasmic translocation of SIR-2.1 was substantially suppressed in wild-type germ cells by perinuclearly localized PGL-1 and PGL-3 to repress excessive germline apoptosis8. Therefore, we examined whether SIR-2.1 also plays a pivotal role in somatic apoptosis in synMuv B mutants.

We found that somatic apoptosis increased significantly following DNA damage generated by UV irradiation24, compared to non-irradiated conditions in ced-1; hpl-2 double and ced-1; hpl-2; pgl-1 triple mutants, but not in ced-1 single, ced-1; sir-2.1(RNAi) double, ced-1; hpl-2; sir-2.1(RNAi) triple, and ced-1; hpl-2; pgl-1 sir-2.1(RNAi) quadruple mutants (Fig. 3a; compare pairs of white and black bars). Increase in somatic apoptosis following DNA damage was observed not only in hpl-2 mutants but also in the other synMuv B mutants tested (Fig. S3a,b). These results strongly suggest that, in contrast to wild-type somatic apoptosis, somatic apoptosis in synMuv B mutants is DNA-damage inducible as is germline apoptosis. Furthermore, because the apoptosis levels failed to increase after DNA damage in ced-1; hpl-2; sir-2.1(RNAi) triple, ced-1; hpl-2; pgl-1 sir-2.1(RNAi) quadruple, and ced-1; lin-13; sir-2.1(RNAi) triple mutants (Fig. 3a and Fig. S3b), the increase in somatic apoptosis following DNA damage in synMuv B mutants is dependent on the activity of sir-2.1, as is DNA-damage-induced germline apoptosis. In addition, higher levels of somatic apoptosis in ced-1; hpl-2; pgl-1 triple mutants than in ced-1; hpl-2 double mutants under both physiological and DNA-damaged conditions (Fig. 3a) suggest that pgl-1 is involved in suppression of sir-2.1 activity during the course of somatic apoptosis in synMuv B mutants. From these results, we conclude that somatic apoptosis in synMuv B mutants, which is repressed under physiological conditions, increases following DNA damage in a sir-2.1-dependent manner, as does germline apoptosis.

Figure 3: Somatic apoptosis in hpl-2 mutants increased following UV irradiation in a sir-2.1-dependent manner.
figure 3

(a) Mean ± s.d. numbers of persistent cell corpses per L1 head at 25 °C (n > 50) in ced-1(tm2420) single, ced-1(tm2420); hpl-2(tm1489) double, and ced-1(tm2420); hpl-2(tm1489); pgl-1(ct131) triple mutants, which were treated or not treated with sir-2.1 RNAi and irradiated (black bars) or not irradiated (white bars) with 400 J/m2 of UV. *P < 0.05. #P > 0.05. (b) Subcellular localization of SIR-2.1 in N2 and hpl-2 mutant embryos, which were irradiated (400 J/m2) or not irradiated (0 J/m2) with UV. Percent SIR-2.1 cytoplasmic translocation in each category of embryos is shown at the bottom. Scale bar: 20 μm. (c) Schematic diagram of a sir-2.1 transgene introduced into hpl-2(tm1489) mutant animals, which is translationally fused with gfp and controlled by the elt-2 promoter and unc-54 3′UTR. (d) Schematic diagram depicting anterior-end part of intestine in C. elegans adult hermaphrodite. The anterior end of the intestine consists of four int1 cells and two int2 cells26,27, and is connected to the pharynx via the pharyngeal-intestinal valve (p.i.v.). Circles: nuclei of int1 and int2 cells. Arrow: cytoplasm of intestinal cells. (e) Subcellular distribution of SIR-2.1::GFP expressed in intestinal cells under the control of the elt-2 promoter in hpl-2(tm1489); Ex[Pelt-2::sir-2.1::gfp] transgenic animals at 25 °C, which were treated or not treated with pgl-1 RNAi and irradiated or not irradiated with 400 J/m2 of UV. Arrows: cytoplasmically translocated SIR-2.1::GFP signals. Scale bar: 20 μm. (f) Mean ± s.d. percent of intestinal cells that showed cytoplasmic translocation of SIR-2.1::GFP in hpl-2(tm1489); Ex[Pelt-2::sir-2.1::gfp] transgenic animals at 25 °C with or without pgl-1 RNAi depletion and with or without 400 J/m2 of UV irradiation.

Full size image

In search for molecular basis that increases somatic apoptosis following DNA damage in synMuv B mutants but not in wild type, we UV-irradiated or not irradiated wild-type N2 and hpl-2 mutant animals, and immunostained their embryos with anti-SIR-2.1 antibody (Fig. 3b). We found that even without UV irradiation, SIR-2.1 was more significantly translocated to the cytoplasm in N2 embryos (90.6 ± 0.1%, n = 216) than in hpl-2 mutant embryos (10.1 ± 0.2%, n = 261), when they developed at 25 °C (Fig. 3b). With UV irradiation, the frequency of SIR-2.1 cytoplasmic translocation significantly increased in hpl-2 mutant embryos (86.8 ± 0.1%, n = 229), while this frequency stayed high in N2 embryos (91.1 ± 0.2%, n = 223) (Fig. 3b). These observations suggest that somatic apoptosis in hpl-2 mutants was maintained at a low level under physiological conditions (Fig. 3a), at least partly by substantial nuclear retention of SIR-2.1 (Fig. 3b). The level of somatic apoptosis in ced-1; hpl-2 double mutants was increased to a higher level following DNA damage (Fig. 3a), due to significant cytoplasmic translocation of SIR-2.1 (Fig. 3b). By contrast, in wild type (ced-1 single mutants), the level of somatic apoptosis was already high under physiological conditions (Fig. 3a), due to significant cytoplasmic translocation of SIR-2.1 without DNA damage (Fig. 3b). The level of somatic apoptosis did not change (stayed high) following DNA damage in wild type (ced-1 single mutants) (Fig. 3a), because the frequency of SIR-2.1 cytoplasmic translocation stayed high following DNA damage (Fig. 3b). We consider that in wild-type (non-synMuv B-mutant) somatic cells, SIR-2.1 is not a critical factor that affects the level of developmental apoptosis, because the majority of SIR-2.1 is consistently translocated to the cytoplasm with or without DNA damage as observed (Fig. 3b). Currently, it is not clear why the levels of somatic apoptosis were higher in ced-1; sir-2.1(RNAi) double mutants than in ced-1; hpl-2; sir-2.1(RNAi) triple or ced-1; hpl-2; pgl-1 sir-2.1(RNAi) quadruple mutants, although sir-2.1 was depleted in all the three mutants (Fig. 3a). Indeed, the levels of somatic apoptosis in ced-1; sir-2.1(RNAi) double mutants were not significantly different from those in ced-1 single mutants, as previously reported (Fig. 3a)23. One possible hypothesis is that although SIR-2.1 is one of the most critical factors that determine the levels of apoptosis in synMuv B-mutant backgrounds, other factors such as CED-4, which was shown to be more abundantly expressed in wild type than in synMuv B mutants (Fig. 2b,c and Fig. S2b), can be more crucial than SIR-2.1 in non-synMuv B-mutant backgrounds to determine the levels of apoptosis. In conclusion, the above results strongly suggest that SIR-2.1 plays a critical role in the control of somatic apoptosis in synMuv B mutants but not in wild type. These results also suggest that somatic apoptosis in synMuv B mutants is controlled by a mechanism that mimics DNA-damage-induced germline apoptosis, in which PGL-1 and SIR-2.1 play critical roles.

Cytoplasmic translocation of SIR-2.1 was substantially suppressed by ectopically expressed PGL-1 in hpl-2-mutant intestinal cells

To further investigate underlying mechanisms, we expressed a sir-2.1::gfp transgene in hpl-2-mutant intestinal cells under the control of the elt-2 promoter (Fig. 3c). elt-2 is specifically expressed in the intestine throughout the animal’s life25. We expressed SIR-2.1::GFP in the intestine because intestinal cells are large somatic cells26,27; hence, differences in SIR-2.1::GFP subcellular localization are readily distinguishable (Fig. 3d). In addition, in hpl-2-mutant intestinal cells, PGL-1 and PGL-3 were often observed as perinuclear granules, as are P granules in germ cells, at elevated temperatures9. We found that cytoplasmic translocation of SIR-2.1::GFP occurred only in a small proportion of hpl-2-mutant intestinal cells without UV irradiation (Fig. 3e,f; 12.2 ± 3.4%, n = 278). However, after UV irradiation, which was previously shown to cause elimination of PGL-1 and PGL-3 from germ cells8, the proportion of intestinal cells that showed SIR-2.1::GFP cytoplasmic translocation substantially increased (Fig. 3e,f; 58.9 ± 4.1%, n = 258). Further, when pgl-1 was depleted by RNAi in hpl-2 mutants, a substantial proportion of intestinal cells showed SIR-2.1::GFP cytoplasmic translocation even without UV irradiation (Fig. 3e,f; 59.7 ± 3.1%, n = 226). Moreover, after UV irradiation, cytoplasmic translocation of SIR-2.1::GFP occurred in the majority of intestinal cells in pgl-1 RNAi-treated hpl-2 mutants (Fig. 3e,f; 88.2 ± 2.0%, n = 407). These results indicate that cytoplasmic translocation of SIR-2.1 was substantially suppressed by ectopically expressed PGL-1 in hpl-2-mutant intestinal cells. We assume that this suppression also occurs in somatic cells that are programmed to die in hpl-2 mutants. We consider that suppression of SIR-2.1 cytoplasmic translocation by ectopically expressed PGL-1 is one of the key mechanisms for repression of somatic apoptosis in hpl-2 mutants. Therefore, failure of this suppression in ced-1; hpl-2; pgl-1 triple mutants or by UV irradiation leads to increased somatic apoptosis compared to ced-1; hpl-2 double mutants under physiological conditions (Fig. 3a).

Somatic apoptosis was reduced by ectopic expression of pgl-1 and pgl-3 transgenes in wild-type somatic cells

Finally, we asked whether ectopically expressed PGL-1 and PGL-3 repress somatic apoptosis only in a synMuv B-mutant background, in which various germline-specific proteins are co-expressed in addition to PGL-1 and PGL-39, or also in a wild-type background, in which no other germline-specific proteins are co-expressed. To assess this, we ectopically expressed gfp::pgl-1 and gfp::pgl-3 transgenes in ced-1-mutant somatic cells under the control of the sur-5 promoter (Fig. 4a). sur-5 is strongly expressed in most of the somatic cells throughout C. elegans development28. We found that GFP::PGL-1 and GFP::PGL-3 were ectopically expressed in most of the somatic cells in transgenic animals throughout development, including the embryonic stage (Fig. 4b), during which the majority of programmed cell death takes place29. We also observed that, in some unidentified somatic cells in embryos and larvae, GFP::PGL-1 and GFP::PGL-3 were present as perinuclear granules, as were P granules in Z2 and Z3 primordial germ cells (Fig. 4b; yellow arrowheads vs. white arrows). Intriguingly, we found that somatic apoptosis levels were significantly reduced in the transgenic animals compared to the control ced-1-mutant animals even at a permissive temperature, 20 °C (Fig. 4c,d). We consider that observed self-association of GFP::PGL-1 and GFP::PGL-3 into perinuclear granules in somatic cells in the transgenic animals possibly bypassed the requirement of other germline proteins, such as GLH-1 and GLH-4, for suppression of apoptosis in the soma. It was previously shown that GLH-1 and GLH-4 were required for localization of PGL-1 and PGL-3 to P granules in the germline5,8. On the other hand, an ability of PGL-1 and PGL-3 to self-associate into granules in the absence of other germline proteins was demonstrated using a cultured mammalian cell system30. Therefore, if GFP::PGL-1 or GFP::PGL-3 can self-associate into perinuclear granules in somatic cells as observed (Fig. 4b), the functions of GLH-1 and GLH-4 are no longer necessary and these germline proteins would become dispensable for suppressing apoptosis in the soma. In addition, we found that ectopically expressed GFP::PGL-1 substantially suppressed cytoplasmic translocation of SIR-2.1 in the transgenic embryos as in hpl-2 mutant embryos (Fig. 3b and Fig. S4). Although in the absence of pgl-1 RNAi (mock), only 29% (n = 27) of the transgenic embryos showed SIR-2.1 cytoplasmic translocation, after pgl-1 RNAi depletion, as much as 88% (n = 26) of the transgenic embryos showed substantial SIR-2.1 cytoplasmic translocation (Fig. S4). These results suggest that overlapping molecular mechanisms operate between the ectopic pgl-1 (or pgl-3) transgenic animals and synMuv B mutants for the control of somatic apoptosis. These results indicate that somatic ectopic expression of PGL-1 and/or PGL-3 represses somatic apoptosis not only in synMuv B mutants but also in a wild-type background. Therefore, we conclude that PGL-1 and PGL-3 are both necessary and sufficient to repress apoptosis in C. elegans cells.

Figure 4: Somatic apoptosis in ced-1 mutants was reduced by ectopic expression of pgl-1 and pgl-3 transgenes in the soma.
figure 4

(a) Schematic diagram of pgl-1 and pgl-3 transgenes introduced into ced-1(tm2420) mutant animals, which are translationally fused with gfp and controlled by the sur-5 promoter and unc-54 3′UTR. (b) The sur-5 promoter-driven gfp::pgl-1 transgene was expressed in most somatic cells throughout development in transgenic animals. The transgenic animal’s embryos at around the 300-cell stage, comma stage, 3-fold stage, and an L1 larva, were co-immunostained with anti-GFP (green) and anti-PGL-3 (red) antibodies together with TO-PRO-3 DNA staining (blue). Scale bars in whole images: 20 μm. Scale bar in enlarged images: 2 μm. Yellow arrowheads in the enlarged images indicate some of GFP::PGL-1-positive granules localized to the perinuclear region of unidentified somatic cells. White arrows indicate primordial germ cells, Z2 and Z3, which contained PGL-3-positive perinuclear P granules. (c) Mean ± s.d. numbers of persistent cell corpses per L1 head (n > 50) in ced-1(tm2420) mutants and ced-1(tm2420) mutants that contained a transgene Ex[Psur-5::gfp::pgl-1], Ex[Psur-5::gfp::pgl-3], or Ex[Psur-5::gfp::pgl-1 + Psur-5::gfp::pgl-3], which were grown at 20 °C. *P < 0.05. (d) L1 larval heads of ced-1(tm2420) mutant animal and ced-1(tm2420) mutant animal that contained the Ex[Psur-5::gfp::pgl-1] transgene displayed different numbers of persistent cell corpses at 20 °C. Yellow arrowheads in DIC images indicate some of the cell corpses. Fluorescence signal of GFP::PGL-1 in the head part of the transgenic animal is also shown. Scale bar: 20 μm.

Full size image

In a previous study, it was reported that long-lived C. elegans daf-2 mutants ectopically expressed several germline-specific genes including pgl-1 and pgl-3 in their somatic cells31. Intriguingly, in that study, RNAi depletion of pgl-1 and pgl-3 resulted in a decrease in lifespan in daf-2 mutants, suggesting that somatic ectopic expression of PGL-1 and PGL-3 contributed to the lifespan extension in daf-2 mutants31. However, in a recent follow-up study, Knutson et al.32 found no evidence of P-granule protein expression in daf-2 mutant somatic cells, or of extended longevity in synMuv B mutants32. Therefore, currently it is controversial whether somatic expression of PGL-1 or PGL-3 really contributes to lifespan extension of C. elegans individuals. In this study, we demonstrated that although PGL-1 and PGL-3 are originally designed to serve as germline-specific apoptosis repressors8, they could also function as negative regulators of programmed cell death when introduced into the soma. This result raises a fascinating possibility that expression of a germline program in somatic cells can provide the soma with partial germline immortality at a cellular level. If this is the case, our finding may provide a novel clue to develop a treatment for diseases that arise from excess cell death such as Parkinson’s or Alzheimer’s diseases33,34.

Methods

C. elegans strains and maintenance

All strains were maintained at either 15 or 20 °C on Nematode Growth Medium (NGM) agar plates seeded with Escherichia coli OP50, according to standard protocols35. When synMuv B-mutant phenotype was examined, gravid adult hermaphrodites were shifted from 20 °C to 25 °C, and their progeny developed at 25 °C were analyzed for their phenotype. The following mutant alleles were used. LG I: unc-13(e51), ced-1(e1735), ced-1(tm2420); LG III: hpl-2(tm1489), lin-13(ok838); LG IV: pgl-1(ct131), him-3(e1147), opIs219[Pced-4::ced-4::gfp]. The following extrachromosomal arrays were used. kkuEx10[Psur-5::gfp::pgl-1 + rol-6(su1006)], kkuEx11[Psur-5::gfp::pgl-3 + rol-6(su1006)], kkuEx12[Psur-5::gfp::pgl-1 + Psur-5::gfp::pgl-3 + rol-6(su1006)], kkuEx13[Pelt-2::sir-2.1::gfp + rol-6(su1006)]. Cell corpses were counted in ced-1(e1735, tm2420) mutant background. Therefore, either ced-1(e1735) single or ced-1(tm2420) single mutant was used as a wild-type control for the analysis. For the other analyses, the strain N2 was used as a wild-type control. Some strains were obtained from the Caenorhabditis Genetics Center (CGC) or from National BioResource Project (NBRP).

Construction of transgenic animals

To construct the Pelt-2::sir-2.1::gfp::unc-54 3′UTR vector, pMAA1, first, sir-2.1 genomic sequence was fused with gfp sequence by fusion PCR36. Second, mCherry sequence in pOLB1872 vector (a gift from Olaf Bossinger), which was flanked by elt-2 promoter and unc-54 3′UTR sequences in the vector, was replaced with this sir-2.1::gfp fused sequence using In-Fusion HD cloning kit (Clontech) to construct pMAA1. pMAA1 was then co-microinjected with pRF4, which expresses dominant roller gene, rol-6(su1006), into N2 adult hermaphrodite gonads to establish a heritable Roller transgenic line expressing Pelt-2::sir-2.1::gfp, as previously described37. Then, this transgenic line was crossed with hpl-2(tm1489) mutants to construct the transgenic line, YHS140: hpl-2(tm1489) III; kkuEx13[Pelt-2::sir-2.1::gfp + rol-6(su1006)]. To construct the Psur-5::gfp::pgl-1::unc-54 3′UTR vector, pMAA2, and the Psur-5::gfp::pgl-3::unc-54 3′UTR vector, pMAA3, first, luciferase sequence was removed from the plasmid, pSLGCV (Addgene, No. 49862), which contains sur-5 promoter, gfp sequence, and unc-54 3′UTR38. Then, pgl-1 and pgl-3 genomic sequences were inserted into the junction between gfp sequence and unc-54 3′UTR of pSLGCV using In-Fusion HD cloning kit (Clontech) to construct pMAA2 and pMAA3, respectively. Then, pMAA2 and/or pMAA3 were co-microinjected with pRF4 into ced-1(tm2420) mutant adult hermaphrodite gonads to establish the following heritable Roller transgenic lines. YHS133: ced-1(tm2420) I; kkuEx10[Psur-5::gfp::pgl-1 + rol-6(su1006)]. YHS134: ced-1(tm2420) I; kkuEx11[Psur-5::gfp::pgl-3 + rol-6(su1006)]. YHS135: ced-1(tm2420) I; kkuEx12[Psur-5::gfp::pgl-1 + Psur-5::gfp::pgl-3 + rol-6(su1006)].

Cell corpse counting

Numbers of persistent cell corpses in the head region of newly-hatched L1 larvae were scored as previously described13,14. Briefly, synchronized L1 larvae, which were newly hatched in the absence of food, were anesthetized in M9 buffer containing 0.2 mM tetramisole (Sigma), mounted on a 2% agarose pad on a glass slide, covered with a coverslip, and sealed with vaseline. Then, the numbers of persistent cell corpses in the head region were counted under Nomarski DIC microscopy. Apoptotic cells have a characteristic button-like morphology and are easily distinguishable from living cells29. Throughout this study, we counted the number of cell corpses under ced-1 mutant background, because in ced-1 mutants, engulfment of cell corpses is blocked, and therefore, a more sensitive condition to count cell corpses is provided than N213,14. When the numbers of cell corpses were counted in animals that contained a synMuv B mutation, their mothers were shifted from 20 °C to 25 °C, and their L1 progeny grown at 25 °C through the development were scored for their cell corpse numbers.

UV irradiation

L4 mother hermaphrodites were pre-cultured at 20 °C for 24 h, irradiated or not irradiated with 400 J/m2 of UV-C light (254 nm) on OP50-seeded NGM plates, post-cultured on the same plates at 25 °C for 16 h, then transferred to a drop of M9 buffer on a printed glass-slide well. The mother hermaphrodites in M9 buffer were further incubated at 25 °C for 24 h in a wet chamber, and finally their laid progeny, which were grown at 25 °C and synchronized at the L1 larval stage in the absence of food, were examined for the number of persistent cell corpses in their head region.

RNAi

RNAi analysis was performed using the “RNAi-by-soaking” method as previously described39. dsRNA for lin-9, lin-35, lin-37, lin-54, pgl-1, pgl-3, glh-1, glh-4, cgh-1, car-1, ced-3, ced-4, and sir-2.1 genes was synthesized by T7 RNA polymerase-mediated in vitro transcription from respective cDNA templates flanked by T7 promoter sequences, which were PCR-amplified from respective yk cDNA clones in lambda ZAPII vectors (gifts from Yuji Kohara) using T7 primer (5′-GTAATACGACTCACTATAGGGC-3′) and CMo422 primer (5′-GCGTAATACGACTCACTATAGGGAACAAAAGCTGGAGCT-3′). L4 larvae of appropriate genotype were soaked with each dsRNA solution for 24 h, recovered to OP50-seeded NGM plates, irradiated with UV when required, further incubated on the plates at 25 °C for 24 h, and resulting RNAi phenotypes of themselves or their progeny were analyzed.

Immunofluorescence

Embryos obtained from dissected gravid adult hermaphrodites or synchronized L1-stage larvae were placed on poly-lysine-coated glass slides, freeze-cracked with liquid nitrogen, fixed with cold methanol and cold acetone, and immunostained with primary and secondary antibodies as previously described5,6,8. The following primary and secondary antibodies were used: rabbit anti-PGL-1 antibody5 (1:4,000), rabbit anti-GFP antibody (1:400; Molecular Probes), mouse monoclonal OIC1D4 antibody that specifically recognizes PGL-1 (undiluted; Developmental Studies Hybridoma Bank), rat anti-PGL-3 antibody6 (1:1,000), rabbit anti-SIR-2.1 antibody (1:500; Novus), Alexa Fluor 488 goat anti-rabbit IgG (1:500; Molecular Probes), Alexa Fluor 546 goat anti-mouse IgG (1:500; Molecular Probes), and Alexa Fluor 546 goat anti-rat IgG (1:500; Molecular Probes). The specimens were further counter stained with 1 μM TO-PRO-3 (Molecular Probes) to stain DNA. Immunofluorescence images were acquired using a confocal microscope (Olympus, FV1000 Spectral).

Image analysis

Specimens on glass slides were observed under a fluorescence microscope (Zeiss, Axioskop 2 MOT), and their images were acquired using a cooled CCD digital camera (Hamamatsu, ORCA-ER) with operation software (Nikon, NIS-Elements). Images were acquired from multiple embryos or animals under the same exposure conditions, and mean pixel intensities of fluorescence images were calculated using Image J software (http://rsb.info).

qRT-PCR

Wild-type N2 and hpl-2(tm1489) mutant mothers were shifted from 20 °C to 25 °C, and their progeny were grown on OP50-seeded NGM agar plates until the adult stage at 25 °C. Then, ca. 400 each of those gravid adult hermaphrodite progeny were harvested for RNA extraction. Total RNA was extracted with Trizol reagent (Sigma), purified, and reverse transcribed with M-MLV reverse transcriptase (Gibco BRL, USA) using an oligo-dT primer (Promega, USA) to synthesize the first-strand cDNA. ced-4 cDNA product was PCR-amplified using ced-4 (C35D10.9) F primer; 5′-AAA CTA TCG CCA ATG GAA TCT C-3′, and ced-4 (C35D10.9) R primer; 5′-TTT GGA TAC ATC TCA CTG GCT-3′. act-1 cDNA product, which was used as an internal control, was PCR-amplified using act-1 F primer; 5′-CCAGGAATTGCTGATCGTATGCAGAA-3′, and act-1 R primer; 5′-TGGAGAGGGAAGCGAGGATAGA-3′. PCR reactions were performed in a 25 μl reaction volume using Power SYBR Green PCR Master Mix (Applied Biosystems, USA). mRNA levels of ced-4 in N2 and hpl-2 mutants were quantified by averaging triplicate qRT-PCR measurements, and normalized against that of act-1 for comparison.

Data presentation and statistical analysis

All experiments were repeated more than three times for statistical analysis. Data in graphs were presented as mean ± s.d. Statistical comparison of groups was carried out using Student’s t-test. We tested at least 50 samples for each independent experiment. Differences were considered as significant when P < 0.05.

Additional Information

How to cite this article: Al-Amin, M. et al. Somatically expressed germ-granule components, PGL-1 and PGL-3, repress programmed cell death in C. elegans. Sci. Rep. 6, 33884; doi: 10.1038/srep33884 (2016).