INTRODUCTION
Argonautic proteins associate with small regulatory RNAs — such as microRNAs (miRNAs) or small interfering RNAs (siRNAs) — and coordinate silencing of complementary RNAs
1. Unlike miRNA-mediated translational repression of the target RNAs
2,3, siRNAs direct Argonautes to bind and cleave complementary RNAs via the slicing activity of Argonautes and thus rapidly destroy targets
4,5. In plants and animals, exogenous siRNAs function as a potent defense mechanism against invading nucleic acids, such as viruses and transgenes
4. Meanwhile, studies have uncovered a rich repertoire of endogenous siRNAs (endo-siRNAs) that are generated from cellular transcripts
6. Particularly, endo-siRNAs are highly enriched in animal oocytes across species
7 and may participate in the regulation of maternal materials required for embryonic development.
In
Caenorhabditis elegans, Argonaute protein CSR-1 is predominantly found in germ cells, where it specifically associates with endogenous 22G siRNAs to slice complementary target mRNAs
8. The production of these 22G-siRNAs is orchestrated by the RNA-dependent RNA polymerase EGO-1, which employs mRNA as a template to synthesize antisense 22G siRNAs
9. Moreover, among the 27 Argonaute proteins identified in
C. elegans, CSR-1 plays an essential role during embryonic cell divisions and is the only Argonaute that is indispensable for fertility and embryo viability
10,11. Previous studies reported that CSR-1 regulates spindle assembly and chromosome organization by targeting the microtubule depolymerase KLP-7
10. Depletion of CSR-1 in oocytes causes defects in spindle assembly in ~20% of early embryos but results in complete embryonic lethality
9,10,12. However, correcting these spindle defects by knocking down KLP-7 levels did not rescue the lethality of CSR-1-depleted embryos
10, suggesting that CSR-1/endo-siRNAs may have other, more critical physiological functions yet to be recognized.
In mouse oocytes, AGO2 is also found to be highly expressed and largely loaded with endo-siRNAs during oogenesis
5,13. The biogenesis of these endo-siRNAs is controlled by an oocyte-specific isoform of ribonuclease III protein DICER, DICER
O14. DICER
O processes long double-stranded RNA (dsRNA) precursors into endo-siRNAs
14. These endo-siRNAs are primarily sourced from transposons, which undergo active transcription during oogenesis
7. They are present from the germinal vesicle stage of oocyte maturation and persist until the 4-cell stage of the mouse embryo. This persistence is likely a response to the activation of transposons, which is mediated by DNA hypomethylation and histone acetylation. These epigenetic modifications occur within the larger context of DNA methylation reprogramming and the activation of oocyte-specific enhancers
15-17. The transposon-derived endo-siRNAs are thought to repress transposon activity and safeguard genome stability in oocytes
7,18. Notably, blocking the biogenesis of endo-siRNAs by
DICERO knockout or inactivating the endo-siRNA pathway by
AGO2 knockout in oocytes also leads to early embryo arrest and infertility of female mice
19,20, similar to
csr-1 knockout in
C. elegans. Together, these requirements of Argonaute/endo-siRNAs for embryogenesis suggest a conserved and undefined physiological function of endo-siRNAs across species.
In this study, we report a novel physiological function of endo-siRNAs. We show that Argonaute/endo-siRNAs dampen proteasome activity to finely control ribosome function and activation of the zygotic genome during the transition from oocyte to early embryo. We show that Argonaute-endo-siRNA complexes in worms and mice suppress the expression of proteasome components, thereby limiting 26S proteasome assembly. Remarkably, proteasome inhibition rescues the early embryonic arrest phenotypes caused by loss of CSR-1 in worm oocytes or AGO2 in mouse oocytes. Increased proteasome activity in csr-1-mutant oocytes in worms or Ago2-mutant oocytes in mice dramatically reduces the translation of transcription factors vital for early embryonic development. This study uncovers, for the first time, a defined physiological function of endo-siRNAs in mammalian oocytes — regulating a proteasome-ribosome axis — a role that is evolutionarily conserved in C. elegans. This finding also provides the first mechanistic explanation for the long-standing mystery of why CSR-1 deficiency leads to complete embryonic lethality. Therefore, our findings expand the functional repertoire of the Argonaute/siRNA pathway beyond antiviral defense and transposon silencing to include a physiological role in regulating proteasome activity to ensure maternal quality — fundamental for generating competent oocytes capable of initiating life. Moreover, our work provides a novel perspective on mammalian transposons, not merely as "junk" DNA, but as functional regulatory elements in proteasome control.
RESULTS
CSR-1/22G siRNAs affects proteasome assembly via regulating proteasome subunit levels
The catalytic activity of CSR-1 is essential for embryogenesis
8,10, suggesting that CSR-1 slices target RNAs that are crucial for early embryo development. To directly identify targets of CSR-1, we generated a
C. elegans strain using CRISPR/Cas9 genome editing to introduce a
3×flag multiplex tag into the endogenous
csr-1 locus. We utilized our previously established cross-linking ligation and sequencing of hybrids (CLASH)
21, combined with deep sequencing to identify RNAs that purify with the CSR-1–small RNA complexes (Fig. 1a). To maximize the identification of small RNA–target RNA hybrids, we split each read at different positions (17–27 nt) and mapped small RNA and target RNA candidates to the antisense and sense strands of mRNA sequences, respectively (Supplementary Fig. S1a; Materials and Methods). This analysis successfully identified CSR-1/small RNA-mediated mRNA cleavage candidates, including 674 unique small RNA reads and 536 unique mRNA reads (Supplementary Fig. S1a).
Analysis of the small RNAs showed that CSR-1 predominantly binds 22-nt small RNAs with a 5′ G, referred to as 22G-RNAs (Supplementary Fig. S1b)
8,9. CSR-1 was previously shown to cleave mRNAs
in vitro and to account for the slicing activity in
C. elegans extracts
22. To refine the identification of CSR-1-associated small RNA-target chimeras, we calculated the hybridization energy and structural conformation of the RNA duplexes. We then extended the target sequences at the 3′ end and reassessed the hybridization structure and energy of the small RNAs with the extended targets. Candidate sites with the lowest energy were considered the most probable source of target RNA (Supplementary Table S1; Materials and Methods). Our CLASH results, likely for the first time, identified cleavage events mediated by CSR-1/small RNAs
in vivo. The chimeric sequences revealed that CSR-1/small RNAs were complementary to the target site and sliced the target between guide positions g10 and g11 from the 5′ end of 22G RNAs (Fig. 1b; Supplementary Fig. S1c). We identified 560 cleavage events on a total of 438 target mRNAs, including a range of proteasome mRNAs (Fig. 1c). More than 100 genes of 438 CLASH-identified CSR-1 targets are important for embryonic development and viability
23-25 (Supplementary Fig. S1d). To investigate the mechanism of CSR-1/small RNAs in embryo development, we focused on CSR-1 targets related to early embryonic lethality. Protein–Protein Interaction analysis of these targets revealed a significant enrichment for proteasome, followed by enrichments for microtubule and centromere (Supplementary Fig. S1e). Given the marked enrichment of transcripts from proteasome components as CSR-1 targets (Fig. 1c; Supplementary Fig. S1e), we sought to test whether CSR-1 regulates maternal protein levels by tuning proteasome activity.
The 26S proteasome comprises a 20S core particle that mediates proteolysis and a 19S regulatory particle that caps either end of the 20S. In
C. elegans, the 20S core is made up of 14 subunits, while the 19S regulatory particle consists of ~18 subunits
26 (Fig. 1d). RT-qPCR analyses revealed that a majority of the mRNAs of 19S regulatory particle subunits (11 of 18) were upregulated in the germlines of
csr-1 RNA interference (RNAi)-treated or
csr-1ADH catalytic mutant worms (Fig. 1d). Notably, the upregulated genes are predominantly RPN subunits, which function as regulatory and structural components
27, suggesting a selective mechanism to fine-tune proteasome assembly and homeostasis in
C. elegans.
csr-1ADH encodes a catalytically inactive CSR-1 in which the D606 active-site residue (in CSR-1)
28 was mutated to alanine (Supplementary Fig. S2a), the depletion of wild-type
csr-1 and expression of
csr-1ADH were validated by qPCR with specific primers (Supplementary Fig. S2a, b). We validated the upregulation of a number of proteasome subunits in
csr-1 RNAi germ cells by western blot (Fig. 1e; Supplementary Fig. S1f). Importantly, we further confirmed the upregulation of 19S subunits
rpn-2 and
rpn-12 by fluorescence
in situ hybridization (FISH) and immunofluorescence (IF), particularly in oocytes (Fig. 1f, g; Supplementary Fig. S1g, h). CSR-1 primarily regulates 19S regulatory subunits, which activate the 20S proteasome
29. We therefore posited that the upregulation of 19S regulatory subunits might enhance 26S proteasome assembly and activity. Indeed, native PAGE analysis revealed that 19S and 26S proteasome levels were increased in
csr-1 RNAi germ cells compared to wild-type control germ cells (Fig. 1h; Supplementary Fig. S1i). By contrast, 20S proteasome levels remained constant (Fig. 1h; Supplementary Fig. S1i, j). These results suggest that CSR-1 suppresses 19S proteasome levels, thus regulating 26S proteasome assembly in
C. elegans oocytes.
CSR-1 regulates embryo development through restricting proteasome activity in C. elegans
To measure proteasome activity, we monitored the degradation of specific fluorogenic peptide substrates in
C. elegans lysates
30. The turnover of fluorogenic peptide substrates in wild-type
C. elegans lysates was enhanced by treating lysates with the proteasome activator PA28 and suppressed by treating lysates with the proteasome inhibitor MG132 (Supplementary Fig. S2c). While the proteasome activity was not increased in somatic lysates from
csr-1 RNAi worms (Supplementary Fig. S2d), we observed a significant increase in proteasome activities (chymotrypsin-like, caspase-like, and trypsin-like activities) in germ cell lysates from
csr-1 RNAi or
csr-1ADH worms (Fig. 2a). The increased proteasome activity in
csr-1 RNAi germlines appears to be restricted to oocytes since proteasome activity was not increased in
csr-1 lysates from L4 stage germlines (Supplementary Fig. S2e), which lack oocytes. Consistent with the increased proteasome activity, polyubiquitin levels in germ cells were reduced in
csr-1 RNAi worms compared to wild-type worms (Fig. 2b). Similarly, a
ub::G76V::gfp reporter (Supplementary Fig. S2f) that is constitutively targeted for proteasome-mediated degradation was detected in wild-type oocytes, but not in
csr-1 RNAi oocytes (Fig. 2c, d). Inhibiting the proteasome via RNAi of
pas-5, an essential subunit of the proteasome core, dramatically increased the Ub::G76V::GFP intensity in the germline (Supplementary Fig. S2g). Expression of a non-hydrolyzable Ub::M::GFP reporter was comparable in the wild-type and
csr-1 RNAi germlines (Supplementary Fig. S2h). To investigate the proteasome activity during germ cell development, spatiotemporal transcriptome data
31 were analyzed, and the results revealed that proteasome transcript levels reached the highest at the pachytene region followed by a slight decrease during oogenesis, and the highly expressed transcripts correspond to a high count of small RNAs (Supplementary Fig. S3a–c). Consistently, proteasome activity measured by the fluorescent probe Me4BodipyFL
32 showed decreased proteasome activity during oogenesis, but remained elevated in CSR-1-deficient oocytes (Supplementary Fig. S3d, e). Together, these results suggest that CSR-1/small RNAs play a key role in restraining proteasome activity during oogenesis in
C. elegans.
To determine whether the embryonic lethal phenotype of CSR-1-depleted worms results from activation of the proteasome, we chemically or genetically inactivated the proteasome in CSR-1-depleted worms. We treated
csr-1 RNAi worms with a range of concentrations of the proteasome inhibitor MG132. We also modulated the expression level of
rpn-6.1, a key subunit of the 19S proteasome complex, by titrating the level of
rpn-6.1 RNAi
33. Remarkably, inhibiting proteasome activity via MG132 or
rpn-6.1 RNAi in CSR-1-depleted worms improved the hatching rate to 25–30% (Fig. 2e, f). Control experiments indicate that the proteasome is not required for RNAi (Supplementary Fig. S2i, j). Notably, the highest rescue rate was achieved when proteasome activity was restored to levels comparable to the wild type, whereas excessive inhibition of proteasome activity reduced rescue efficiency (Fig. 2e, f; Supplementary Fig. S2k), suggesting proper level of proteasome activity is required for embryonic development. Additionally, proteasome activity in the
csr-1ADH strain also showed a significant rescue of hatching rates (Supplementary Fig. S2l), indicating that the regulation of proteasome for embryo development depends on the catalytic activity of CSR-1. Taken together, these findings suggest that CSR-1/22G regulation of precise proteasome assembly levels is vital for early embryonic development.
The proteasome affects translation of transcription factors (TFs) in C. elegans germ cells
To delve deeper into how CSR-1/22G regulates the proteasome during gametogenesis and embryogenesis, we employed tandem ubiquitin-binding entities (TUBEs) and mass spectrometry to isolate and identify ubiquitinated proteins, potentially proteasome target proteins, in
C. elegans germ cells (Fig. 3a). TUBEs enrichment assay showed ubiquitinated proteins were significantly enriched in the TUBEs immunoprecipitation fraction (Supplementary Fig. S4a). We identified 569 ubiquitinated proteins, potentially targeted for degradation by the 26S proteasome (Supplementary Fig. S4b and Table S2). To identify proteins that are dysregulated in CSR-1-depleted germ cells, we isolated control or
csr-1 RNAi gonads, and utilized signal intensity-based label-free quantification with mass spectrometry. We identified 219 proteins that are decreased in
csr-1 RNAi germ cells compared to the control (Supplementary Fig. S4c and Table S3). Ninety-nine of these proteins were also polyubiquitinated in
csr-1 RNAi germ cells (Fig. 3b). An analysis of previous mRNA-seq data
34 showed that the mRNA levels of these proteins remained similar in CSR-1-depleted and control germ cells (Supplementary Fig. S4d), indicating that their decreased protein levels resulted from increased proteasome activity. It’s worth noting that many of these ubiquitin–proteasome system-degraded proteins are ribosomal proteins and translation initiation and elongation proteins, which are involved in the translational process (Fig. 3b; Supplementary Fig. S4e). Co-Immunoprecipitation (co-IP) experiments using FLAG-tagged alleles of endogenous
rpl-11 and
iff-1 revealed that both RPL-11 and IFF-1 were polyubiquitinated, and their expression was reduced by CSR-1 depletion and restored by MG132 treatment (Supplementary Fig. S4f). Thus, the degradation of these translational proteins in the CSR-1-depleted germ cells was proteasome-mediated.
Using the O-propargyl-puromycin (OPP) assay, we observed no significant differences in protein synthesis activity between the gonads of wild-type control and CSR-1-depleted worms (Fig. 3c, d). However, protein synthesis was significantly reduced in early embryos (1–8 cells) from CSR-1-depleted worms compared to control early embryos (Fig. 3c, d), suggesting that the depletion of translation proteins in CSR-1-depleted oocytes impairs translation in early embryos. Translation control is important for the maternal-to-embryonic transition
35 and zygotic genome activation (ZGA)
36. To further address the embryo arrest induced by CSR-1 depletion, we attempted to rescue the phenotype by directly injecting mRNAs of translation-related genes into the gonads (Supplementary Fig. S4g). The results showed that injection of different combinations or concentrations of the mRNA mix successfully rescued the full embryonic lethality in the CSR-1-depleted worms (Fig. 3e; Supplementary Fig. S4h). To determine whether this rescue restored translational activity in early embryos, we performed OPP incorporation assays and revealed that protein synthesis in
csr-1-deficient early embryos was also rescued by mRNA injection (Supplementary Fig. S4i, j). Taken together, these results underscore the critical role of the translation process in CSR-1/22G function during embryogenesis.
Maternal mRNAs accumulated during oocyte maturation must be precisely regulated, either degraded or translated, to ensure successful embryonic development
37. To elucidate the impact of Argonaute depletion on these processes, we conducted Ribo-lite sequencing
38 to map the ribosome landscape and translation patterns in control and CSR-1-depleted germ cells of
C. elegans (Supplementary Fig. S4k). The results showed an overall decrease in ribosome occupancy (1,411 decreased and 81 increased) in CSR-1-depleted germ cells (Fig. 3f). These genes were enriched for TFs or factors related to transcriptional processes (Supplementary Fig. S4l). Consistently, proteomic analyses confirmed that reduced ribosome occupancy on these mRNAs was accompanied by an overall decreased protein abundance, indicating that the widespread translational downregulation drives proteome remodeling in germ cells (Supplementary Fig. S4m). Notably, more than half of the TFs with decreased translation, whose mRNA levels remained unchanged, belonged to the nuclear hormone receptor (NHR) and
C. elegans homeobox (CEH) families (Fig. 3g; Supplementary Fig. S4n). For example, CEH-43, CEH-53, and NHR-192 protein levels were all reduced in embryos of
csr-1 RNAi worms compared to that in embryos from wild-type control-treated worms (Fig. 3h). Depletion of CEH-43, CEH-53, and NHR-192 via RNAi reduced embryo viability by 50–70%, and combined RNAi of all three genes reduced embryo viability by 90% (Fig. 3i, j; Supplementary Fig. S4o). These findings suggest that CSR-1/22G RNA-regulated proteasome activity is crucial for the translation of key transcription factors, such as those from the NHR and CEH families, in early embryos, thereby facilitating embryogenesis.
AGO2/endo-siRNAs targeting the proteasome in mouse oocytes are involved in embryonic development
In mice, AGO2 is the only Argonaute essential for female fertility
7. In mouse oocytes, AGO2 is guided by endo-siRNAs to silence mRNAs, and the loss of AGO2 or AGO2 catalytic activity leads to early embryonic arrest
39. To investigate whether AGO2 regulates the proteasome in mouse oocytes, we bred
Ago2flox/flox mice to mice expressing Cre recombinase driven by the oocyte-specific
Zp3 promoter to produce
Ago2flox/flox;
ZP3-Cre female, resulting in an oocyte-specific
Ago2 knockout, referred to as
AGO2-KO.
In vitro fertilization (IVF) of
AGO2-KO oocytes produced zygotes that arrest development after the first division (Supplementary Fig. S5a)
7. To determine whether AGO2 directly regulates the expression of proteasome mRNAs in mouse oocytes, we immunoprecipitated AGO2 from growing germ vesicle (GV) oocytes and used Cas9-assisted small RNA sequencing (CAS-seq)
40 to identify the small RNAs loaded onto AGO2 in oocytes. More than half of the AGO2-associated small RNAs corresponded to endo-siRNAs derived from repeat sequences (Supplementary Fig. S5b); retrotransposons (LINE, LTR, and SINE) comprised 79% of the AGO2-associated endo-siRNAs (Supplementary Fig. S5c). We then performed degradome sequencing of RNA isolated from growing GV oocytes to identify targets cleaved by endo-siRNAs. Using rules for target RNA binding and cleavage by mouse AGO2
in vitro41, we integrated our AGO2-associated small RNA data and degradome-seq data to identify endo-siRNA cleavage products specific to mouse growing GV oocytes (Supplementary Fig. S5d; Materials and Methods). We successfully identified 9,405 cleavage events on 2,704 coding gene transcripts in mouse GV oocytes (Fig. 4a, b; Supplementary Table S4). The results showed that a substantial proportion of cleavage events on protein-coding mRNAs were executed by the repeat-derived endo-siRNAs (Fig. 4a; Supplementary Fig. S5e). We further compared our AGO2-sliced targets data with previously published AGO2 LACE-seq data, which identifies AGO2 binding sites, and identified 1,493 shared AGO2 target genes between the two datasets (Supplementary Fig. S5f). This overlap suggests a correlation between the two independent datasets generated by different approaches.
Similar to that in C. elegans, proteasome genes were also found to be cleaved by AGO2/endo-siRNAs in mouse oocytes and encoded components of both the 19S and 20S complexes (Fig. 4b). Notably, we observed that endo-siRNAs cleave imperfectly matched target sites (Fig. 4c; Supplementary Fig. S6a, b). To verify the cleavage of candidate proteasome mRNA cleavage sites by AGO2/endo-siRNA complexes, we performed in vitro cleavage assays using AGO2/endo-siRNAs with Psmd11, Psmc3, and Psmb2 target sequences, with several mismatches distributed in central and 3’ regions (Supplementary Fig. S6c–e). The results showed that the partially complementary substrates are indeed cleaved by AGO2/endo-siRNAs (Fig. 4d; Supplementary Fig. S6f). This tolerance of AGO2/endo-siRNAs for partially complementary target sites confirmed that proteasome transcripts were cleaved by AGO2/endo-siRNAs, and highlights an intriguing role for retrotransposon-derived endo-siRNAs in slicing coding transcripts in mouse oocytes.
RT-qPCR analysis of each subunit of the 19S and 20S proteasome particles identified a number of subunits (from both particles) whose mRNAs were upregulated in
AGO2-KO oocytes (Fig. 4e). Immunoblot analyses confirmed the upregulation of several proteasome subunits (e.g., PSMD11, PSMC3, and PSMB2; Fig. 4f), and native PAGE followed by immunoblot verified an increase in the level of assembled 26S proteasome in
AGO2-KO oocytes compared to wild-type (Fig. 4g). Using the fluorogenic proteasome assay, we measured a ~2-fold increase in proteasome activity in
AGO2-KO metaphase II (MII) oocytes compared to wild-type MII oocytes (Fig. 4h; Supplementary Fig. S6g).
AGO2-KO oocytes also showed increased proteasome activity using the cell-permeable fluorescent proteasome activity probe Me4BodipyFL
42 (Fig. 4i; Supplementary Fig. S6h). We also transcribed
ub::G76V::gfp and
ub::M::gfp reporter mRNAs
in vitro and injected the reporter mRNAs into oocytes, the decreased GFP fluorescence intensity further confirmed enhanced proteasome activities in
AGO2-KO oocytes (Supplementary Fig. S6i, j). To rule out the possibility that the increase in proteasome activity we observed was due to a nonspecific overall increase in proteolytic activity, we measured lysosomal activity in oocytes using LysoTracker. The results showed that
AGO2-KO did not lead to an increase in lysosomal activity (Supplementary Fig. S6k). Thus, AGO2 regulates proteasome activity in mouse oocytes.
To examine the contribution of the heightened proteasome activity to the lethality of AGO2-KO embryos, we inhibited the proteasome activity with different concentrations of MG132 in AGO2-KO oocytes (Materials and Methods). MG132 was added to a final concentration of 0.05 μM at 6 h post-GV breakdown (GVBD) and then intracytoplasmic sperm injection (ICSI) was performed at the MII stage (Fig. 4j). Whereas fertilized AGO2-KO oocytes rarely divided, greater than 50% of the AGO2-KO oocytes treated with MG132 developed into 2-cell (2C) embryos and ~30% progressed to the morula and blastocyst stage (Fig. 4j, k). Taken together, our findings indicate that AGO2/endo-siRNAs down-regulate the expression of proteasome subunits in mouse oocytes by cleaving their mRNAs, which is crucial for early embryonic development and parallels the role of CSR-1 in C. elegans.
Given that AGO2-associated endo-siRNAs in mouse oocytes also target a substantial number of transposon-derived transcripts, we next asked whether the transposon upregulation in
AGO2-KO oocytes could be restored by proteasome inhibition. We therefore examined the expression levels of the AGO2-targeted retrotransposons
RLTR10 and
MT in oocytes
7. Results showed that MG132 treatment failed to restore their elevated expression levels in
AGO2-KO oocytes (Supplementary Fig. S6l). These results indicate that proteasome regulation and transposon control act as parallel downstream pathways of the AGO2/endo-siRNAs, while potential contributions of these two pathways to embryonic development remain to be explored in future studies.
Proteasome-mediated regulation of translation controls the ZGA of the mouse embryo
We then identified 220 proteins potentially degraded by enhanced proteasome activity with the TUBE pull-down assay and mass spectrum (Fig. 5a; Supplementary Fig. S7a–c and Tables S5 and S6). Analysis of published mRNA-seq data
20 indicates that the most mRNAs (177) of these 220 proteins are present at normal levels in
AGO2-KO oocytes (Supplementary Fig. S7d). Notably, ribosomal proteins were predominantly degraded by the proteasomes (Fig. 5a; Supplementary Fig. S7e). The preferential degradation of ribosomal proteins prompted us to investigate the ubiquitination landscape in oocytes. We identified 2,389 proteins with at least one ubiquitin site, including ribosomal proteins such as RPL5 and RPS7 (Fig. 5b, c; Supplementary Fig. S7f). Western blot validated that the levels of RPL5 and RPS7 were significantly reduced in
AGO2-KO MII oocytes but were restored with MG132 treatment (Fig. 5d). To further confirm whether
AGO2-KO leads to reduced ribosome levels, we performed transmission electron microscopy (TEM), which revealed a significant decrease in ribosome abundance in
AGO2-KO oocytes (Fig. 5e, f). These results indicate that AGO2 plays a critical role in restraining proteasome activity to preserve ribosomal proteins for early embryonic development.
To further investigate the effects of dysregulated ribosomes on translational landscapes, we performed Ribo-lite analysis and the results showed an overall decrease in translation efficiency in
AGO2-KO oocytes, with ribosome occupancy increased on only 10 genes and reduced on 437 genes, including those known to be essential for early embryo development and ZGA (e.g.,
Faf1,
Yy1, and
Igf2bp143-45; Fig. 5g). The protein levels were decreased in most of the 437 genes, consistent with their reduced translational activity (Supplementary Fig. S7g). To assess whether
AGO2-KO impaired ZGA in zygotes, we performed RNA-seq and found that the mRNA levels of 495 well-known ZGA genes were dramatically reduced in maternal
AGO2-KO (m
AGO2-KO) zygotes (Fig. 5h), indicating a near-complete blockade of ZGA progression. To further assess the impact of these dysregulated TFs on early embryo development, we knocked down three TFs (
Gabpb1,
Yy1, and
Sp1) that showed decreased translation in AGO2-KO oocytes (Supplementary Fig. S7h). Knockdown of
Gabpb1,
Yy1, or
Sp1 in wild-type oocytes (by ~80%) followed by ICSI significantly delayed or blocked embryo development, yet some embryos still reached the blastocyst stage (Supplementary Fig. S7i, j). Simultaneous knockdown of
Gabpb1,
Yy1, and
Sp1 by ~50% in wild-type oocytes — comparable to their translation levels in m
AGO2-KO zygotes — caused embryos to arrest at the 1–2 cell stages (Fig. 5i, j; Supplementary Fig. S7h) Furthermore, OPP incorporation assays showed that MG132 treatment markedly restored translational activity in m
AGO2-KO zygotes, accompanied by a significant reactivation of representative ZGA genes
46 (Fig. 5k, l), indicating successful ZGA initiation. Taken together, our results suggest that Argonaute/small RNAs restrict proteasome activity to support translation of transcription factors essential for ZGA and early development.
DISCUSSION
How animals generate competent oocytes to launch the beginning of life remains poorly understood. In this study, we identified Argonaute/endo-siRNAs as critical regulators of competent oocyte generation by restricting the activity of the ubiquitin–proteasome system during oogenesis. Our findings uncover that endo-siRNAs suppress the ubiquitin–proteasome system in ribosomal catabolism within oocytes. This regulation enables the translation of transcription factors that are essential for early embryonic development. Our work illuminated the molecular framework underlying the fundamental question of how competent oocyte is generated by animals to begin life (Fig. 6). In this discussion, we explore the implications of our findings for understanding the functions of Argonaute/small RNAs, the ubiquitin–proteasome system, and ribosome and their relationship during the transition from oogenesis to early embryonic development.
Ubiquitin–proteasome system controlled by Argonaute/small RNAs pathway
Although Argonaute/small RNAs are well-recognized in previous studies for their requirements in early embryonic development
40,47, the potential reasons why Argonaute mutant embryos are not viable remain unclear. We found that Argonaute/small RNAs in worms and mice regulate the proteasome complex, a fundamental protein degradation machinery, to ensure proper oocyte maturation and zygotic gene activation. This previously unobserved Argonaute-proteasome regulatory mechanism highlights the versatility of Argonaute/small RNAs and expands the list of Argonaute functions. Of the 27 Argonautes in
C. elegans, CSR-1 is particularly noteworthy as its inhibition leads to almost complete embryonic lethality
8,10,28. Likewise, AGO2 deficiency in mouse oocytes causes early embryonic arrest, resulting in complete female sterility
48. Previous studies reported that mutation of CSR-1 in
C. elegans or AGO2 in mice leads to spindle defects and chromosome misalignment in a subset of oocytes, yet all fertilized embryos are nonviable
8,10,39, suggesting embryonic lethality is unlikely to result from the spindle assembly defects. Moreover, rescuing spindle assembly by tuning the expression of the microtubule depolymerase KLP-7 failed to restore the viability of CSR-1-depleted embryos
10. In the mouse, deletion of AGO2 caused spindle disorganization in ∼20% of oocytes but resulted in lethality in all fertilized embryos (Supplementary Fig. S8a, b)
39. These results indicate that CSR-1 and AGO2 regulate more critical pathways required for embryonic development in
C. elegans and mice. Our data reveal that both CSR-1 and AGO2 primarily target proteasome transcripts in oocytes. Importantly, our findings show that inhibiting proteasome activity rescues CSR-1-depleted embryos to viability and
AGO2-KO embryos to the blastocyst stage, suggesting Argonaute depletion-induced aberrant proteasome activity is the major cause of embryonic lethality. Additionally, given that a subset of CSR-1/AGO2 targets identified in our study is involved in microtubule and chromosome regulation, it is possible that combined modulation of proteasome activity with spindle assembly may potentially improve embryo development and warrants further investigation.
Notably, during C. elegans gametogenesis, an uptick in proteasome transcription levels is observed from the distal region to the pachytene region, which is then followed by a slight decrease during oogenesis (Supplementary Fig. S3b). The trend of proteasome activity mirrors that of the mRNA levels in wild-type gonads, but the depletion of CSR-1 results in an increase in activity during oogenesis (Supplementary Fig. S3d). In mice, proteasome activity escalates from growing GV to mature MII oocytes during oogenesis and reaches the strongest activity at the 1-cell stage, which may result from the need for rapid maternal protein degradation, and this activity is further intensified by AGO2 knockout. Additionally, CSR-1 and AGO2 associate with distinct small RNA populations in C. elegans and mice. Nevertheless, both Argonaute/small RNA pathways regulate proteasome genes through preferential targeting of 19S subunits by CSR-1 in C. elegans and combined targeting of 19S and 20S subunits by AGO2 in mice. Together, these findings reveal a conserved regulatory principle that Argonaute/small RNA pathways control proteasome activity during oogenesis and early embryonic development, while their specific molecular mechanisms have diverged across evolution.
Previous studies report that the principal role of transposon-derived siRNAs in mouse oocytes is to silence transposon activity
7. Here, our study introduces a novel perspective, proposing that these transposon-derived siRNAs perform an additional crucial function during oogenesis: they inhibit numerous proteasome subunit transcripts, thereby modulating proteasome activity, which is integral to the production of functional oocytes. This highlights the physiological significance of transposons. This physiological tactic serves as a quintessential representation of the transposon functioning as a "controlling element", a concept initially introduced by McClintock
49. Further exploration is needed to uncover other potential biological roles of transposon-derived siRNAs and to understand the structural mechanisms through which the AGO2/siRNA complex facilitates target RNA cleavage.
Oogenic translation control and ZGA
Oogenesis is a unique cell differentiation process. As the oocyte grows and develops, it accumulates RNA and protein to support early embryonic development
37, and transcription is gradually suppressed until it finally ceases as the oocyte reaches maturity
50-52. Thus, small RNA regulation becomes more important for late stages of oogenesis. Indeed, many mRNAs continue to be translated in transcriptionally inactive oocytes and are essential for oocyte maturation
38,53. Maternal mRNAs, destined for the embryo and transcribed during mitosis or the early stages of meiosis, are stored and then translated or degraded in the early stages of embryogenesis, also crucial for early embryo development
12,54. During the oocyte-to-embryo transition, translation plays a pivotal role in regulating meiotic resumption, chromosome remodeling, ZGA, and early embryonic development
53. We have shown that in oocytes the proteasome catabolizes translation initiation factors, elongation factors, and ribosomal proteins in both
C. elegans and mice. We speculate that CSR-1 and AGO2 restrict the activity of the proteasome in order to maintain a sufficient abundance of functional ribosomes and thus enable translation for the oocyte-to-embryo transition and early embryonic development.
Beyond the regulation of maternal mRNA translation, proper control of maternal protein stability is also critical for oocyte-to-embryo transition. Our study further reveals that
AGO2 KO-induced proteasome hyperactivity leads to degradation of a subset of maternal proteins. Functional annotations of these proteins identified several well-characterized maternal factors, such as CUL1, NLRP4F, and RNF114
55-58, whose loss has been shown to cause early embryonic lethality (Supplementary Fig. S8c). These findings suggest that AGO2/endo-siRNA-mediated regulation of proteasome activity functions as a coordinated mechanism during the maternal-to-zygotic transition, balancing the degradation of maternal proteins with the regulation of zygotic translation.
MATERIALS AND METHODS
C. elegans
All wild-type, mutant, and transgenic strains used in this study were maintained on NGM agar plates supplemented with the E. coli OP50 bacteria strain. They were kept at room temperature (~20 °C). Genotypes were confirmed by PCR and sequencing. Transgenic worms were generated by CRISPR/Cas9 genome editing.
Mice
Ago2loxP/+ mice were generated by GemPharmatech, China. A mixture of Cas9 mRNA, sgRNA and loxP sequences containing pLSODN-1 donor plasmids was microinjected into C57BL/6 fertilized eggs. The injected zygotes were then transferred into the uterus of pseudopregnant ICR females. Targeted alleles were identified by PCR and Sanger sequencing. The mice were backcrossed for at least 6 generations onto C57BL/6J. Ago2loxP/+ lines were crossed with Zp3-Cre mice, and their progeny were intercrossed to produce Ago2loxP/loxP, Zp3-cre female mice. All mice were bred and kept in the SPF animal facility of Westlake University, and all animal experiments and protocols, specifically our animal care and use protocol (AP#21-042-SEZ) have been reviewed and approved by the Institutional Animal Care and Use Committee of Westlake University.
Cell line
HEK-293F cells for plasmid transfection and protein purification were cultured in SMM 293-II medium (Sino Biological), supplemented with 1% penicillin-streptomycin (Gibco) at 37 °C with 5% CO2.
Generation of C. elegans Strains
The generation of
C. elegans strains was done using previously described CRISPR-Cas9 Ribonucleoprotein (RNP) methods
59. This involved the co-injection of Cas9 RNPs, donors, and either a
rol-6 or
myo-2::gfp marker for easy identification of the transformed organisms. This was done by adding different tags or inducing mutations at endogenous loci, thereby manipulating the genome to create genetically modified organisms for further study. To ensure genetic diversity and to test for off-target effects, at least two independent alleles were generated for each strain.
Generation of the single-copy transgene
csr-1ADH strain (
Si[csr-1::D605A;unc-119(+)]II) was performed according to a Cas9-triggered homologous recombination strategy described previously
60. Briefly, the sequence of
csr-1 with RNAi targeting region reencoded to maintain coding information and encodes a catalytic inactive CSR-1 at the D605 active-site residue, which contains the unc-119(+) expressing cassette in one of its introns, was amplified and inserted into the donor plasmid, co-injected with markers Podr-1::mCherry, Pmyo-2::mCherry and Pmyo-3::mCherry into unc-119(ed4) at the young adult stage. The single-copy transgene was identified based on 100% non-unc and without any co-injection marker expression and validated via PCR-based genotyping and Sanger sequencing. Life spans and brood sizes of the strains generated in this study were confirmed to be comparable with wild-type strains.
26S proteasome fluorogenic peptidase assays
The
in vitro assay of 26S proteasome activities was performed as previously described
30. Worms, gonads, and somatic cells were collected in proteasome activity assay buffer (50 mM Tris-HCl, pH 7.5, 250 mM sucrose, 5 mM MgCl
2, 0.5 mM EDTA, 2 mM ATP, and 1 mM dithiothreitol) and lysed by FastPrep-24 (MP Biomedicals). Lysate was centrifuged at 10,000
× g for 10 min at 4 °C. 10 μg of total protein of cell lysates were transferred to a black bottom 96-well microtiter plate (BD Falcon) and incubated with fluorogenic substrate. To measure the chymotrypsin-like activity of the proteasome, we used Suc-Leu-Leu-Val-Tyr-AMC (Enzo). We used Z-Leu-Leu-Glu-AMC (Enzo) to measure the caspase-like activity of the proteasome, and Ac-Arg-Leu-Arg-AMC for the proteasome trypsin-like activity. Fluorescence (380 nm excitation, 460 nm emission) was monitored on a microplate fluorometer (Varioskan LUX, Thermo Fisher Scientific) every 5 min for 1 h at 37 °C.
Native gel immunoblotting of the proteasome.
Cell lysates were run on 3.5% native gels prepared in resolving buffer (90 mM Tris base, 90 mM boric acid, 5 mM MgCl2, 0.5 mM EDTA, and 1 mM ATP) with 5 mM ATP, 1 mM DTT, and 3.5% acrylamide from a 40% stock solution of acrylamide and bisacrylamide in a 37.5:1 ratio (Bio-Rad). These were run at 100 V for 3 h at 4 °C. Before transfer, the gels were incubated in transfer buffer (25 mM Tris base and 192 mM glycine) with 1% SDS for 10 min followed by a 10 min incubation in transfer buffer. The protein was transferred to 45 μm polyvinylidene difluoride (PVDF) membranes (Millipore) at 200 mA for 3 h in transfer buffer and immunoblotting analysis was performed. Extracts were also analyzed by SDS-PAGE to determine loading control with anti-β-Tubulin (Sigma-Aldrich).
Worm gonads dissection
Age-synchronized young adult animals were harvested in M9 buffer. Animals were transferred to a glass dish prior to dissection. Incisions with a 27-gauge syringe needle were made close to the pharynx, forcing the intestine and gonad to extrude.
Oocyte Collection and Microinjection
14-day-old female mice were sacrificed to collect growing GV oocytes without super-ovulation. To collect fully grown GV oocytes, 4-week-old female mice were super-ovulated by intraperitoneal injection with 5 U pregnant mare serum gonadotropin (PMSG) (Aibei). 5 U human chorionic gonadotropin (HCG) was intraperitoneally injected after 48 h to collect MII mature oocytes. For siRNA microinjection, fully grown GV oocytes were harvested in M2 medium (Sigma) with 2.5 μM milrinone (Sigma-Aldrich) to inhibit meiotic resumption. Approximately 10 pL of siRNA was injected at a concentration of 20 μM. After injections, oocytes were arrested at the GV stage in M2 medium containing 2.5 μM milrinone for 12 h to allow sufficient siRNA silencing, then washed in milrinone-free M16 medium, and cultured for 3 h to observe meiotic resumption (GVBD) or 14 h to detect the first polar body (Pb1) extrusion. The survival oocytes were subsequently incubated at 37 °C in a 5% CO2 environment until the performance of ICSI.
Immunoblotting
Samples of worms, gonads or mouse oocytes were collected into a 1.5 mL tube and lysed by incubation for 10 min on ice in RIPA buffer (50 mM Tris pH 7.6, 150 mM NaCl, 1% Igepal CA-630, 0.5% sodium deoxycholate, 0.1% SDS) with 1× cOmplete EDTA-free protease inhibitor cocktail (Roche), and worm samples were additionally crushed for 20 s of 6 m/s on FastPrep-24 (MP Biomedicals). Insoluble material was removed by centrifugation at > 16,000× g for 15 min at 4°C. The supernatant was quantified by BCA assay (Beyotime) and then denatured by the addition of 5× SDS sample buffer (250 mM Tris·HCl, pH 6.8, 10% SDS, 30% Glycerol, 10 mM DTT, 0.05% bromophenol blue) and heating to 95 °C for 5 min. Samples were normalized and loaded in 4–12% PAGE gels (Genscript) and run at 130 V for ∼1 h in MOPS buffer (25 mM Tris and 192 mM glycine). Gels were transferred onto PVDF membranes for 15 min with an eBlot L1 Fast Wet Transfer System (Genscript), PVDF membranes were blocked with 5% milk TBS with 0.1% Tween-20 (PBST) for 1 h at room temperature, and then incubated with primary antibody overnight at 4 °C followed by horseradish peroxidase (HRP)-conjugated secondary antibody incubation for 1 h at room temperature, with 4 washes of PBST after each antibody incubation. Antibodies used include anti-FLAG (Sigma-Aldrich), anti-β-Tubulin (Sigma-Aldrich), anti-Ubiquitin (Cell Signaling Technology), anti-AGO2 (Abcam), anti-PSMA2 (Abclonal), anti-PSMB7 (Abclonal), anti-PSMC3 (Abclonal), anti-PSMD7 (Abclonal), anti-PSMD9 (Abclonal), anti-PSMD11 (Abclonal), anti-RPL3 (Abclonal), anti-RPS12 (Abclonal), anti-EIF2B4, Goat anti-Rabbit IgG (H+L) HRP (Jackson Immunoresearch), and Goat anti-Mouse IgG (H+L) HRP (Jackson Immunoresearch). Chemiluminescence was performed with enhanced chemiluminescence (ECL) reagent (Biosharp) and imaged with a ChemiDoc Imagers (Bio-rad). Relative protein levels were quantified using ImageJ (v.1.54d).
MG132 treatments
MG132 treatment in
C. elegans was performed following the protocol of Bortezomib treatment with minor modifications
61. Briefly, MG132 (Selleck) was reconstituted in DMSO (Sigma-Aldrich), diluted to different concentrations in 1× M9 buffer, and added on top of agar plates seeded with OP50 bacteria or RNAi bacteria. L1 stage larvae were added to either drug-treated or DMSO (0.1%) treated plates and incubated at 20 °C. Fertility assays or imaging were performed using treated animals as described above. For MG132 treatments in mouse oocytes, the concentration and timing of MG132 treatment were determined through a series of preliminary experiments. A range of concentrations was tested (5, 2.5, 1, 0.5, 0.1, 0.05, 0.025 µM) at the GV, MI, and MII stages. The experimental conditions were then established as follows: Fully grown GV oocytes were cultured in M2 medium, MG132 was added to a final concentration of 0.05 μM 6 h post GVBD (MI stage), and the surviving MII oocytes were subsequently incubated at 37 °C in a 5% CO
2 environment until the performance of ICSI.
26S proteasome staining of mouse oocytes
For Me4BodipyFL (R&D) labelling, Me4BodipyFL was reconstituted in DMSO (Sigma-Aldrich), and oocytes were incubated for 1 h at 37°C with 500 nM Me4BodipyFL or 0.1% DMSO before imaging. Images were acquired with an ANDOR Dragonfly Spinning Disk Confocal System scan head mounted on an inverted microscope with a 40× or 60× oil lens. Images were processed using Imaris and ImageJ software.
mRNA in vitro transcription and injection
The cDNAs were used to amplify the coding sequence with Phanta Super-Fidelity DNA Polymerase (Vazyme) by PCR and then recombined with the eukaryotic expression vector pcDNA3.1. The plasmids were linearized with
XbaI enzyme (New England Biolabs) and then transcribed to RNAs using HiScribe T7 ARCA mRNA Kit (New England Biolabs) according to the manufacturer's standard mRNA synthesis protocols. The mRNA preparation for injection into the gonads was performed following the methods previously described
62. Injected translation-related mRNAs include ribosome genes:
rps-2, rps-3, rps-8, rps-11, rps-14, rps-15, rps-20, rps-23, rps-26, rps-2, rps-3, rpl-6, rpl-9, rpl-10, rpl-11, rpl-19, rpl-20, rpl-23, rpl-26, rpl-28, rpl-29, rpl-31, rpl-32; translation initiation and elongation genes:
iff-1, iff-2, iffb-1, eif-2.b, eif-3.c, eif-3.g, eif-3.f, eif-1g, eef-1b.1, tumf-1.
RNAi
RNAi experiments were conducted using the standard feeding RNAi method. Bacterial clones expressing the control (empty vector pL4440) construct and the dsRNA targeting different
C. elegans genes were obtained from
C. elegans ORFeome Library v1.1. All RNAi clones used in experiments were sequenced for verification before use. For RNAi experiments, RNAi bacteria with empty (pL4440 vector as control) or the dsRNA-expressing plasmid were grown overnight in liquid LB media containing ampicillin (Sangon Biotech) and then induced with IPTG (AMRESCO) for 2 h before seeding the bacteria (10×) on NGM plates supplemented with ampicillin (100 mg/mL) and IPTG (1 mM). For different concentrations of
rpn-6.1 RNAi,
rpn-6.1 dsRNA-containing bacteria were concentrated by 10,000×
g centrifugation and then resuspended in 1/10 of the original volume of LB medium, as a 100% concentration of
rpn-6.1 RNAi bacteria, and then diluted to different concentrations. The mix volume ratio of
csr-1 RNAi and different concentrations of
rpn-6.1 RNAi is 9:1. For
csr-1 RNAi, an equal volume of control RNAi (PL4440 bacteria) is added to the
csr-1 RNAi bacteria. Bleach-hatched L1 worms were dispersed onto RNAi plates. To eliminate the potential impact of somatic cells on the experimental results, all RNAi experiments in this study were performed using a
C. elegans strain with germline-specific RNAi (
mkcSi13 II; rde-1(mkc36) V)
63.
IF and mRNA FISH
The gonads were dissected on poly-lysine slides in 1× M9, fixed in 4% formaldehyde, and washed with PBT (PBS, 0.1% BSA, and 0.1% Triton X-100). For IF, samples were incubated with primary antibody anti-FLAG in PBT overnight at 4 °C, washed with PBT, and incubated with secondary antibody in PBT for 1 h at room temperature. Alexa fluor 594 conjugated anti-mouse IgG (Thermo Fisher Scientific) secondary antibodies in 1:100 dilution were used for visualization. Samples were washed and mounted with Vectashield containing 1 mg/mL DAPI (Beyotime). For mRNA FISH, samples were hybridized with FISH probes and processed for imaging according to the manufacturer’s recommendations for C. elegans samples of Stellaris RNA FISH protocols (Biosearch Technologies). Images were acquired with an ANDOR Dragonfly Spinning Disk Confocal System scan head mounted on an inverted microscope with a 40× or 60× oil lens. Images were processed using Imaris and ImageJ software.
CSR-1 CLASH and data processing
The CLASH was performed, and a library was created as previously described
21. In this study, CSR-1-associated RNAs are well characterized in length and nucleotide bias, but lack a fixed reference sequence set, to comprehensively infer each potential chimeric read (small RNA-target sequence pair). To achieve this, we iteratively processed each raw read under two guiding hypotheses: (1) the length of CSR-1-associated small RNAs varies between 17 to 27 nucleotides, potentially initiating with A, U, C, or G, and (2) the target sequence is ligated at the 3’ end of each small RNA within the chimeras.
As a case study, we employed the 22G small RNA subtype to illustrate this iterative process for preprocessing, filtering, and isolating the most probable chimeric reads. Adaptor sequences were removed using Cutadapt (version 1.1.0). Reads lacking adaptor ligation were excluded, ensuring that the length of the remaining reads was no shorter than 22 nucleotides. The trimming process was expedited via multithreading, tailored to the computational capacity available.
Our custom script, pi_tar_splict_T.py, initiated the division of each trimmed read from the 5’ end. Setting the RNA length to 22 nt and the initial base to G enabled the selective extraction of 22G RNAs. The rationale for focusing on chimeric reads is grounded in the hypothesis that target sequences are likely derived from the transcriptome, and that small RNAs enriched through immunoprecipitation exhibit complementarity to transcriptomic sequences. Bowtie (version 1.1.0) facilitated the mapping of segmented small RNA sequences onto the forward strand, and the corresponding target sequences onto the reverse strand. The scripts sRNA_expr.R and sRNA_filter.R were utilized to determine the enrichment of small RNAs in comparison to negative controls. Reads demonstrating simultaneous successful mapping and enrichment were classified as probable chimeric reads. This process was analogously applied to other small RNA types. Subsequently, sub-sequences of small RNAs were consolidated into a unique pool using the Integrate_bases.R script.
To refine the identification of CSR-1-associated small RNA-target chimeras, we employed a BEB (Blast, Extension, and Blast) approach, complemented by analyses of the Minimal Free Energy (MFE) of interaction. The compatibility between small RNAs and their targets was initially assessed using Blastn (version 1.1.0). RNAduplex (version 2.6.4) was then used to calculate the hybridization energy and structural conformation of the RNA duplexes. Target RNAs were subsequently mapped back to the transcriptome using Blastn, and target sequences were extended at the 3’ end based on the mapping coordinates. RNAduplex was again utilized to reassess the hybridization structure and energy of small RNAs with the extended targets. The mapping site with the lowest energy was considered the most probable source of the target RNA. The ratio of energies pre- and post-extension was computed. Finally, the definitive inference of authentic chimeric reads, base pairing, and comparative information post-extension was compiled into a comprehensive table.
OPP translation assay
The OPP assay was performed following standardized procedures previously described with minor modifications
64. Briefly, OP50 bacteria were cultured to a density of ∼2 × 10
9 CFU/mL and then incubated with 1% paraformaldehyde (PFA) for 1 h with gentle shaking. The bacteria were subsequently pelleted by centrifugation and washed with M9 buffer to remove residual PFA before further use. Each assay started with washing approximately 100 worms from NGM plates using liquid NGM. The worm pellets were then washed with liquid NGM to remove OP50. The OPP assay was based on the Click-iT Plus OPP Alexa Fluor 488 kit (Thermo Fisher Scientific), manufacturer’s instructions with slight alterations for use in
C. elegans. Worms were incubated for 3 h at 20 °C with gentle shaking in 10 μM OPP diluted in liquid NGM containing 2 × 10
8 CFU/mL of PFA killed OP50 in a total volume of 1 mL in a 1.5 mL microcentrifuge tube. The gonads were dissected for fixation, 100 μL of freshly prepared 4% PFA in PBS was added for 1 h at 10°C with shaking. After fixation, the pellet was washed three times with PBS. The fixed worms were incubated in 877.5 μL of Click-iT reaction buffer containing 2.5 μL of 500 μM Alexa Fluor picolyl azide, 20 μL 100 mM copper protectant, and 100 μL of the reaction buffer additive, and incubated overnight (16 h) at 10 °C in shaking incubator at 900 rpm. For early embryo OPP incorporation, young adult worms were bleached and the early embryos (mostly 1–4 cell embryos), the embryos were then incubated with OPP and proceeded with the protocols described above. The pellets of gonads or embryos were then washed 3 times in PBS with shaking for 30 min each wash to remove unconjugated Alexa Fluor picolylazide before images were acquired with an ANDOR Dragonfly Spinning Disk Confocal.
qRT-PCR
For mRNA analysis by qRT-PCR, total RNA was extracted using Trizol (Invitrogen). cDNA was generated using HiScript II 1st Strand cDNA Synthesis Kit (Vazyme). Primers designed to generate ~200 bp fragments were used to perform the qPCR reactions with an AceQ qPCR SYBR Green Master Mix (Vazyme) on a LightCycler® 96 Instrument (Roche). For qPCR of pre-mRNA, forward primers were positioned in the intron, and the reverse primer was positioned in the exon. In each experiment comparing mRNA levels of selected genes, measurements with three biological replicates were performed. Relative changes in mRNA expression levels of the genes tested were normalized to that of
pie-1 in
C. elegans and
act-1 in mice. Quantification of relative mRNA expression was determined using 2^−(ΔΔcq)
65.
Embryo hatching statistics
Self-brood counts of C. elegans were assessed by placing single age synchronized worms on individual blank control, RNAi-seeded, or proteasome inhibitor-treated plates at 20 °C, passaging parent worms to new plates every 24 h for 72 h and then scoring the total brood and hatched larvae from 3 independent trials (n ≥ 25 each trial). All fertility assays were conducted at 20 °C. The total brood consists of embryos and hatched larvae.
IVF and ICSI
Oocyte isolation and IVF were conducted following standardized procedures as previously described
66. Briefly, epididymal spermatozoa were obtained from the cauda epididymis of 8 to 10-week-old C57BL6 mice and subsequently incubated in HTF medium (Aibei) for 1 h at 37°C in an environment containing 5% CO
2, female oocytes were euthanized the same day at 4 weeks of age, after superovulation induced with 5 U of PMSG and 5 U of HCG before being killed for oocyte collection. The sperm and the oocytes were co-cultured for 4–6 h. Subsequently, the fertilized oocytes were transferred and incubated in KSOM medium at 37 °C and 5% CO
2. Rates of different development stages were microscopically checked and measured to assess embryonic development.
For ICSI, epididymal spermatozoa were obtained from the cauda epididymis of 8 to 10-week-old C57BL6 mice. Afterwards, 1 µL of the spermatozoa suspension was combined with 10 µL of 3% polyvinyl pyrrolidone (PVP)-HEPES-buffered CZB medium in an ICSI manipulation chamber. The ICSI procedure was carried out following previously described methods
67 with a minor modification. Briefly, the sperm head was isolated from the tail using piezo pulses targeting the neck region. The separated head was promptly injected into an oocyte. Following a 10-min recovery period at room temperature, the oocytes were thoroughly washed at least three times before being transferred to KSOM medium (Aibei). Fertilization success was determined by the presence of two pronuclei (2 PN) in the zygote, observed 5 h after the ICSI procedure. The developmental progression of the embryos to the blastocyst stage was assessed every 24 h post-insemination.
TUBE pulldown
The TUBE pulldown assay for enrichment of ubiquitinated proteins was performed according to the manufacturer’s instruction (LifeSensors). Briefly, TUBE magnetic beads were washed with TBST (20mM Tris-HCl, pH 8.0, 0.15M NaCl, 0.1% Tween-20) twice on a magnetic stand before use. Gonads from C. elegans or mouse oocytes were incubated in 100 μL TUBE lysis buffer (50mM Tris-HCl, pH 7.5, 0.15M NaCl, 1mM EDTA, 1% NP-40, 10% glycerol) and incubated on ice for 10 min. Lysates were clarified by centrifugation at 14,000× g for 15 min at 4 °C. The supernatant was collected, and 10 μL was reserved as the input sample. The remaining lysate was incubated with pre-equilibrated magnetic TUBE beads for 3 h at 4 °C on a rotating platform. Beads were then collected using a magnetic stand, and the supernatant was retained as the unbound fraction. The beads were washed three times with 1 mL TBST, each wash consisting of resuspension followed by magnetic separation and removal of the supernatant. Both input and TUBE-enriched samples were subsequently subjected to mass spectrometry analysis.
Liquid chromatography with tandem MS (LC–MS/MS) analysis
Two hundred gonads of each group of C. elegans or 150 MII stage oocytes of wild-type or AGO2-KO mice were lysed with 100 μL RIPA cleavage buffer on ice for 30 min and then centrifuged at 16,000× g for 30 min. For LC-MS/MS, lyophilized peptides were resuspended in 20 µL of 0.1% formic acid and 3 µL aliquots were injected using the nanoElute UHPLC System (Bruker) onto a 25 cm × 75 µm ID, 1.6 µm C18 column (Aurora series, IonOpticks). Peptides were eluted with a gradient of water/0.1% formic acid (A) and acetonitrile/0.1% formic acid (B) over 60 min at a flow rate of 300 nL/min, with the linear gradients starting from 2% B and increasing to 22% in 45 min, followed by an increase to 35% B in 50 min, 80% B in 55 min, 80% B in 55–60 min. Eluted peptides were analyzed with a TIMS quadrupole time-of-flight timsTOF Pro 2 instrument (Bruker Daltonics) using a CaptiveSpray nano-electrospray source. The dda-PASEF mode was employed with a m/z range of 300 to 1500 and a 1/K0 range of 0.75–1.3, with a Ramp Time of 166 ms. LC-MS/MS data files were processed using the Parallel Search Engine in Real-time (PaSER version 2023) against the Mus musculus proteome from Uniprot (https://www.uniprot.org/). Search parameters included a peptide mass tolerance of 20 ppm for 1 isotopic peak, precursor mass range of 600–6,000 Da, and tryptic digestion specificity. Variable modifications included oxidation of methionine and carbamidomethylation of cysteine. The False Discovery Rate was set to 1% for both protein and peptide levels. For the analysis of ubiquitination, GG(K) (+114.042927 Da) and GGRL(K) (+383.228103 Da) were set as additional variable modifications. A minimum Andromeda score of 40 was demanded for identifications of modified peptides.
Ribo-lite-seq library preparation and data processing
The Ribo-lite library was created and analyzed as previously described with minor modifications
38. Briefly, 200 gonads of
C. elegans or 100 oocytes of mice were lysed in 20 μL of ice-cold Ribo-lite lysis buffer (20 mM Tris-HCl, pH7.4, 150 mM NaCl, 5 mM MgCl
2, 1mM DTT, 100 μg/mL CHX, 1% Triton; 25 U/mL Turbo DNase) on ice for 10 min. The
C. elegans samples were mixed with 50
S. pombe cells lysis as spike-in and the mouse oocyte samples were mixed with 2 μL from one single
C .elegans 100 μL lysis sample as spike-in. 300 μL ice-cold Ribo-lite lysis buffer was supplemented and the sample was incubated for another 10 min on ice. The lysate was clarified for 10 min at 20,000×
g at 4 °C, then treated with 1 μL RNase I 100 U/mL (Ambion) and incubated at room temperature for 45 min with gentle mixing. The digested extracts were stopped by the addition of 10 μL SUPERase•In (Ambion) and were overlaid onto a 4 mL sucrose cushion of 1M sucrose in 20 mM total Tris-HCl, pH7.4, 150 mM NaCl, 5 mM MgCl
2, 1 mM DTT, 100 μg/mL CHX and 20 U/mL SUPERase•In (Ambion) for 4 h at 76,400 rpm in SW-60Ti rotor in Beckman Optima XPN-100 ultracentrifuge. The supernatant was removed, and the pellet was resuspended in 50 μL pellet buffer (10 mM Tris, pH 7.5, 1% SDS). Ribosome-protected fragments (RPF) were extracted from pellet buffer by using 1 mL TRIzol and 200 μL chloroform. The aqueous phase was precipitated by adding 750 μL isopropanol and 20 μg glycogen. Precipitation was carried out at –20 °C overnight and RPFs were then pelleted by centrifugation for 40 min at 12,000×
g, 4 °C. The RPF sample was then subjected to sequencing library preparation by CATS Small RNA-seq Kit (Diagenode). CATS Small RNA-seq Kit utilizes a single-tube strategy based on polyadenylation, reverse transcription, template switch, and PCR amplification. Two replicates of each library were prepared.
All Ribo-lite data were trimmed with cutadapt v1.18, and the trimmed reads were sequentially mapped to C. elegans/mouse rRNA sequences (ws276/mm9) using STAR v2.5.3a. Those aligned to rRNA were discarded, and the remaining reads were mapped to the transcriptome of ws276/mm9. The reads from C. elegans and mouse were also mapped to the transcriptome of S. pombe and C. elegans (ASM294v2/ws276), respectively. The gene expression level was then calculated by normalizing to the reads that mapped to the S. pombe and C. elegans, respectively.
AGO2-IP-enriched small RNA-seq library preparation and data processing
The AGO2-IP-enriched small RNA library was created and analyzed as previously described
40. Adaptors in raw sequencing data were trimmed using Cutadapt (version 4.4), with reads containing ambiguous base N discarded and only those ranging in length from 19 to 25 nt retained. Reads in the sense orientation for ribosomal RNAs, deemed degradation products, were identified by downloading rRNA sequences from the RNAcentral (https://rnacentral.org/) database and mapping reads using Bowtie2 (version 2.4.2, with parameters -L 20 -N 1 -k 1). Remaining clean reads were mapped to the genome (UCSC mm10) and transcriptome (GENCODE v23) using Bowtie (version 1.2.3) in end-to-end mode, allowing only one mismatch. AGO2-associated small RNAs were identified, requiring at least a twofold increase in associated raw counts and a nonzero count in immunoprecipitation.
Annotation of identified AGO2-associated small RNAs was conducted primarily following the methods of Hiroyuki et al
7. To ascertain the exact genomic positions of the small RNAs, we aligned the sequences to the genome using Bowtie (1.2.3), allowing one mismatch. To identify small RNAs corresponding to various repeat elements (e.g., rRNA, tRNA, retrotransposon, DNA transposon), genomic positions of these elements were redefined by integrating data from RepeatMasker (https://www.repeatmasker.org/) and retrotransposon coordinates from Choi et al. A small RNA overlapping any repeat by at least 15 nucleotides was classified as repeat-derived.
To identify small RNAs corresponding to tRNAs, rRNAs, snRNAs, snoRNAs, miRNAs, piRNAs, and mRNAs based on sequence similarity, we extracted non-coding RNA sequences from the RNAcentral database and mRNA sequences from GENCODE v23. Alignments were performed using the Bowtie (1.2.3) programme, employing identified AGO2-associated small RNA sequences as queries against the aforementioned database sequences, allowing for one mismatch considering that our reference sequences may not encompass all RNA species present in cells.
Finally, annotations based on genomic positions and sequence similarities were integrated. If a small RNA received multiple annotations, they were prioritized as follows: rRNA, tRNA, snoRNA, snoRNA, miRNA, piRNA, pseudogene transcripts, lncRNA, and mRNA. Non-annotated sequences were categorized as unknown.
Degradome sequencing library preparation
The degradome RNA library was prepared as previously described with minor modification
68. Four hundred wild-type and
AGO2-KO growing GV oocytes were collected and total RNAs were extracted using TRIzol (Invitrogen) followed by Turbo DNase (Invitrogen) treatment. Next, RNA bearing 5′ monophosphate was enriched by ligating a 5′ adaptor with T4 RNA ligase1 (NEB), followed RNA purification by VAHTS RNA clean beads (Vazyme). Reverse transcription was performed with SuperScript III (Thermo Fisher Scientific) and a primer containing a degenerate sequence at its 3′ end, which also introduced the 3′ adaptor. PCR was done with Q5 High-Fidelity 2× Master Mix (NEB) to amplify cDNA and to introduce sequencing primer and index sequences. Two replicates of the wild-type and
AGO2-KO library were prepared.
AGO2-mediated slicing of transcripts identification pipeline
We developed GuidewithTarget 2.0, a computational pipeline designed to analyze the complementarity of “guide” pairs in the context of AGO2-mediated slicing of transcripts directed by AGO2-associated small RNAs in mouse oocytes. The initial version was tailored for investigating Ago3-mediated slicing of transcripts by piRNAs in
Drosophila melanogaster69 (available at
https://github.com/CMACH508/GuidewithTarget).
The analysis is structured into three segments: identification of AGO2-associated small RNAs from RNA immunoprecipitation sequencing libraries; detection of cleavage products from degradome sequencing libraries; and pinpointing authentic cleavage products guided by AGO2-associated small RNAs.
A comprehensive description of the first segment is provided in the corresponding section. Adaptors from degradome sequencing reads were removed using Cutadapt (version 4.4) and aligned solely to the forward strand of the mouse transcriptome using Bowtie (version 1.2.3). Alignments sharing identical starting positions were considered equivalent cleavage events. Following a comparative analysis of product abundance, only those exclusively detected in the AGO2 wild-type group were designated as definitive cleavage products. To identify AGO2-associated small RNA-guided cleavage events, a sequence window centered on the starting position and extending L (defaulting 20) nucleotides both downstream and upstream was extracted. Blastn (version 2.5.0) was employed to align queries (AGO2-associated small RNAs) against subjects (extended target sequences) using makeblastdb for database construction and Blastn for sequence alignment. Authentic guide pairs were distinguished based on alignment results, specifically if the starting position of the cleavage product aligned with the 10th nucleotide from the 5' end of the small RNA.
The pipeline is powered by custom scripts incorporating several widely used bioinformatics tools. These scripts were developed in R (version 4.2.0), utilizing the ggplot2 package (version 3.4.0) for data visualization. The updated pipeline has been made publicly available on GitHub at
https://github.com/CMACH508/GuidewithTarget2.0. New version enables the pipeline to cope with duplicates in the cleavage products identification of degradome sequencing analysis with more replicates instead of just one trial.
Protein expression and purification
The full-length mouse AGO2 coding sequence was codon-optimized for expression in HEK 293F cells (Invitrogen) and cloned into the pcDNA3.1(+) vector with an N-terminal 3× FLAG tag followed by a TEV cleavage site. The plasmid was transfected transiently into HEK 293F cells. Mouse AGO2 and AGO2–small RNA complex purification were performed as described previously
69. Purified AGO2-small RNA complex was concentrated, flash-frozen in liquid nitrogen, and stored at –80 °C.
Cleavage assays
Purified AGO2-small RNA complexes were incubated with equimolar amounts of 5′-FAM-labeled target RNAs in 50 mM HEPES (pH 8.0), 150 mM NaCl, 0.5 mM TCEP, 2 mM magnesium acetate, 2 mM MnCl2, and 40 U RNase inhibitor for 1 h at 37 °C. Cleavage reactions were stopped by adding proteinase K (NEB), and incubating for 30 min at 55 °C. Aliquots of each reaction were then mixed with an equal volume of RNA loading dye (47.5% formamide, 0.01% SDS, 0.01% bromophenol blue, 0.005% xylene cyanol, 0.5 mM EDTA, and 0.1 M formaldehyde) and incubated for 10 min at 70 °C. Samples were then resolved on 18.5% denaturing urea-PAGE gels and visualized by a ChemiDoc MP imager (Bio-Rad) for FAM fluorescence.
TEM imaging
The ampullas of the wild-type and AGO2-KO mouse oviduct were dissected for fixation at 4 °C with 2% glutaraldehyde in 0.1 M sodium cacodylate buffer overnight and then washed in ice-cold 0.1 M sodium cacodylate buffer three times. Next, the cells were postfixed in reduced 1% osmium tetroxide in cacodylate buffer for 1 h and rinsed with 0.1 M sodium cacodylate buffer three times. After washing with ddH2O three times, the cells were stained with 1% uranyl acetate in ddH2O for 1 h. Next, the cells were rinsed with ddH2O, dehydrated with a graded ethanol series, and infiltrated with Durcupan ACM resin. Finally, ultrathin sections were prepared, and TEM images were acquired with a transmission electron microscope (Thermo Fisher Scientific, Talos L120C G2) at 80 kV.
Functional enrichment analysis
We use Wormbase analysis tools (
https://wormbase.org/tools/enrichment/tea/tea.cgi) for the phenotype enrichment analysis of CSR-1 targets. And we searched for statistically enriched Gene Ontology terms using the DAVID database (https://david.ncifcrf.gov/), and the background was set to the entire
C. elegans or mouse genome data set. The threshold level for all functional enrichment analyses was set for
P values < 0.05. The terms were ranked according to their enrichment
P values.
Statistical analysis
Statistical parameters, including the exact value of ‘n’, and descriptive statistics (mean ± SD) are stated in the Figure legends. Each replicate experiment was performed with independent populations of animals. For quantification of fluorescence, ‘n’ indicates the number of individual animals analyzed. Individuals were selected for imaging at random from populations of animals of the indicated genotype. All statistical analyses were performed using GraphPad Prism software. Statistical significance was defined as P < 0.05, as determined by one-way or two-way ANOVA and subsequent correction for multiple testing.
DATA AND MATERIALS AVAILABILITY
All data needed to evaluate the conclusions in the paper are present in the paper and/or the Supplementary Materials. Sequencing data generated in this study were deposited in the Sequence Read Archive (SRA) under the accession number PRJNA1149515 (
https://www.ncbi.nlm.nih.gov/bioproject/PRJNA1149515/). Additional data and materials in this study are available upon request. Several key strains are also deposited in the
Caenorhabditis Genetics Center (CGC).
The Author(s) 2026. Published by Higher Education Press. This is an Open Access article distributed under the terms of the CC BY license (https://creativecommons.org/licenses/by/4.0/).
