SIMR-1 and ENRI-2 colocalize at somatic granules in embryos.

A. Summary of IP-mass spectrometry interactions detected between NRDE-3, ENRI-2, ENRI-1, and SIMR-1 from previously published studies (Chen and Phillips, 2024, Lewis et al., 2021). The number of replicates from which the interaction was detected relative to the total number of replicates performed is indicated.

B. Live imaging of GFP::3xFLAG::NRDE-3 and SIMR-1::mCherry::2xHA; ENRI-2::2xTy1::GFP embryos at different stages (4-cell, 8-cell, 28-cell, 100-cell, 200-cell, and comma). Boxes identify the location of Z2 and Z3 primordial germ cells, showing that SIMR-1 is present in germ granules while ENRI-2 is not. At least five individual embryos were imaged for each genotype and stage. Scale bars, 5 μm.

C. Box plot of SIMR-1::mCherry::2xHA granule number quantification at different embryonic stages (4-cell, 8-cell, 28-cell, 100-cell, and 200-cell). At least ten individual embryos at each stage were used for quantification. Each dot represents an individual embryo, and all data points are shown. Bolded midline indicates median value, box indicates the first and third quartiles, and whiskers represent the most extreme data points within 1.5 times the interquartile range. Lines connect the mean granule number for each stage, illustrating the change in number of SIMR granules across the developmental stages of the embryo. Two-tailed t-tests were performed to determine statistical significance and p-values were adjusted for multiple comparisons. See Materials and Methods for a detailed description of quantification methods.

Summary of NRDE-3 and CSR-1 small RNA pathway and components.

A. Illustration of NRDE-3 and CSR-1 pathways in C. elegans. In the NRDE-3 pathway, primary ERGO-class 26G-RNAs are synthesized by the ERI complex, including ERI-1. The primary Argonaute protein, ERGO-1, is loaded with the ERGO-class 26G-RNAs. The secondary ERGO-dependent 22G-RNAs, a subset of the WAGO-class 22G-RNAs, are synthesized from ERGO-1-targeted mRNAs. Their synthesis depends on the Mutator complex, which includes the RNA-dependent RNA polymerase, RRF-1, and the poly(UG) polymerase, RDE-3. They are subsequently loaded into NRDE-3 and other cytoplasmic WAGO-family Argonaute proteins to promote gene silencing. In the CSR-1 pathway, the CSR-class 22G-RNAs are synthesized by the RNA-dependent RNA polymerase, EGO-1. CSR-1 is loaded with CSR-class 22G-RNAs to promote gene licensing in the germline and clearance of maternal transcripts in early embryos.

B. A table summarizing the function of some key proteins in NRDE-3 and CSR-1 pathways.

Expression of NRDE-3, SIMR-1, and ENRI-1.

A. Live imaging of the germline from SIMR-1::mCherry::2xHA; GFP::3xFLAG::NRDE-3 day-one-old adult animals. Late pachytene (left) and oocyte (right) regions are shown, revealing that SIMR-1 and NRDE-3 do not colocalize in the germline. At least seven individual germlines were imaged. Scale bars, 5 μm.

B. Live imaging of SIMR-1::GFP(R159C)::GFP::3xFLAG embryo at 100-cell stage, showing that SIMR-1(R159C) fails to not localize to cytoplasmic granules. At least five individual embryos were imaged. Scale bars, 5 μm.

C. Live imaging of 2xTy1::GFP::ENRI-1 (top two rows) and ENRI-1::mCherry::2xHA (bottom two rows) embryos at different stages (4-cell, 8-cell, 28-cell, 100-cell, 200-cell, and comma). Fluorescence and brightfield channels are shown, demonstrating that ENRI-1 does not show prominent localization in embryos. At least seven individual embryos were imaged for each genotype and stage. Scale bars, 5 μm.

D. Western blot of ENRI-1::mCherry::2xHA (left lane) and wild-type (right lane) gravid adult animals (72 hours at 20°C after L1 arrest), showing that the ENRI-1::mCherry::2xHA is expressed in this strain. Anti-HA-HRP and anti-actin antibodies were used for the western blot. Source data of the original blot is provided.

Unloaded NRDE-3 localizes to cytoplasmic granules with SIMR-1.

A. Live imaging of GFP::3xFLAG::NRDE-3 embryos in eri-1, rde-3, and nrde-3(HK-AA) mutants at 8-cell, 100-cell, and comma stage embryos. At least five individual embryos were imaged for each genotype and stage. Arrows point to granule localization of NRDE-3 in the 100-cell stage. Asterisks highlight the nuclear localization of NRDE-3 in somatic cells of the 100-cell stage embryos and in the primordial germ cells of the comma stage embryos. Scale bars, 5 μm.

B. Box plot of GFP::3xFLAG::NRDE-3 granule number quantification in different mutants.

C. Box plot of GFP::3xFLAG::NRDE-3(HK-AA) granule number quantification at different embryonic stages. Lines connect the mean granule number (red dots) for each stage, illustrating the change in change in number of NRDE-3 granules across embryonic development.

D. Live imaging of SIMR-1::mCherry::2xHA; GFP::3xFLAG::NRDE-3(HK-AA) at 100-cell stage. Arrows and insets show examples of colocalization between SIMR-1 and NRDE-3(HK-AA). At least ten individual embryos were imaged. Scale bars, 5 μm.

For box plots in B and C, at least twelve individual embryos in each mutant were used for quantification. Each dot represents an individual embryo, and all data points are shown. Bolded midline indicates median value, box indicates the first and third quartiles, and whiskers represent the most extreme data points within 1.5 times the interquartile range. Two-tailed t-tests were performed to determine statistical significance and p-values were adjusted for multiple comparisons. See Materials and Methods for a detailed description of quantification methods.

NRDE-3 and SIMR-1 localization and expression in various mutants.

A. Live imaging of GFP::3xFLAG::NRDE-3 embryos in the simr-1 mutant, enri-1 mutant, enri-2 mutant, and enri-1; enri-2 double mutant in 8-cell, 100-cell, and comma stage embryos. At least five individual embryos were imaged for each genotype and stage.

B. Western blot of wild-type N2, GFP::NRDE-3, and GFP::NRDE-3(HK-AA) mixed staged embryos, showing that the level of NRDE-3 protein is not significantly changed in the nrde-3(HK-AA) mutant compared to GFP::NRDE-3. Anti-FLAG and anti-ACTIN antibodies to detect NRDE-3 and ACTIN, respectively.

C. Live imaging of SIMR-1::mCherry::2xHA; GFP::3xFLAG::NRDE-3 embryos in an eri-1 mutant at the 100-cell stage. Arrows point to examples of colocalization between SIMR-1 and NRDE-3. At least five individual embryos were imaged. Dotted white line indicates the outline of the embryo.

D. Box plot quantifying the number of SIMR-1::mCherry::2xHA granules in a nrde-3(HK-AA) mutant at different embryonic stages. At least twelve individual embryos for each genotype were used for quantification. Each dot represents an individual embryo, and all data points are shown. Bolded midline indicates median value, box indicates the first and third quartiles, and whiskers represent the most extreme data points within 1.5 times the interquartile range. Lines connect the mean granule number (red dots) at each stage, illustrating the change in number of SIMR-1 granules across embryonic development. Two-tailed t-tests were performed to determine statistical significance and p-values were adjusted for multiple comparisons. See Materials and Methods for a detailed description of quantification methods.

E. Live imaging of germline regions from SIMR-1::mCherry::2xHA; GFP::3xFLAG::NRDE-3(HK-AA) day-one-old adult animals shows that SIMR-1 and NRDE-3(HK-AA) do not colocalize and NRDE-3(HK-AA) does not form foci in the germline (late pachytene (top) and oocyte (bottom)). At least five individual germlines were imaged. All scale bars, 5 μm. Source data are provided as a Source Data file.

SIMR-1 recruits ENRI-2 and then NRDE-3 to cytoplasmic granules.

A. Live imaging of GFP::3xFLAG::NRDE-3(HK-AA) embryos in simr-1 and enri-2 mutants, and GFP::3xFLAG::NRDE-3 embryos in a simr-1; eri-1 double mutant at 8-cell, 100-cell, and comma stages. At least five individual embryos were imaged for each genotype and stage. Asterisk marks the nuclear localization of GFP::NRDE-3 in the simr-1; eri-1 mutant, visible in somatic cells of 8- and 100-cell stage embryos and in a primordial germ cell of the comma stage embryo. Scale bars, 5 μm.

B. Live imaging of ENRI-2::2xTy1::GFP embryos in a simr-1 mutant and SIMR-1::GFP::3xFLAG embryos in an enri-2 mutant. At least five individual embryos were imaged for each genotype and stage. Arrows point to examples of cytoplasmic SIMR granules still visible in the enri-2 mutant. Box surrounds a primordial germ cell displaying germ granule localization of SIMR-1. Scale bars, 5 μm.

NRDE-3(HK-AA) and ENRI-1 localization in various mutants.

A. Western blot of wild-type (N2), GFP::NRDE-3, GFP::NRDE-3; simr-1, GFP::NRDE-3(HK-AA), GFP::NRDE-3(HK-AA); simr-1 mixed staged embryos, showing that the GFP::NRDE-3 and GFP::NRDE-3(HK-AA) does not exhibit higher expression level in the simr-1 mutant. Anti-FLAG and anti-ACTIN antibodies were used to detect NRDE-3 and ACTIN, respectively.

B. Live imaging of GFP::3xFLAG::NRDE-3(HK-AA) embryos in an enri-1 mutant and an enri-1; enri-2 double mutant. At least five individual embryos were imaged for each genotype and stage. Arrows point to cytoplasmic NRDE-3(HK-AA) granules visible in an enri-1 mutant. Scale bars, 5 μm.

C. Live imaging of ENRI-1::mCherry::2xHA embryo in an enri-2 mutant at 100-cell stage. Both fluorescence (left) and brightfield (right) channels are shown to demonstrate that no specific localization can be observed for ENRI-1 in the enri-2 mutant. At least five individual embryos were imaged. Scale bars, 5 μm.

D. Live imaging of ENRI-1::mCherry::2xHA; GFP::3xFLAG::NRDE-3(HK-AA) embryo at 100-cell stage, showing that no specific localization can be observed for ENRI-1 in the nrde-3(HK-AA) mutant. Arrows point to cytoplasmic NRDE-3(HK-AA) granules. At least five individual embryos were imaged. Scale bars, 5 μm.

CSR and WAGO pathway proteins localize to distinct cytoplasmic granules.

A. Live imaging of SIMR-1::mCherry::2xHA; RDE-3::GFP embryo at 100-cell stage, showing that RDE-3 colocalizes with SIMR-1. At least five individual embryos were imaged for each genotype and stage. Arrowheads and insets show examples of colocalization between SIMR-1 and RDE-3 at cytoplasmic granules. Scale bars, 5 μm.

B. Box plot of Pearson’s R value quantifications among different pairs of proteins at 100-cell embryonic stage. At least 20 granules from at least 4 individual embryos were used for quantification. Each dot represents an individual quantification, and all data points are shown. Box indicates the first and third quartiles, and whiskers represent the most extreme data points within 1.5 times the interquartile range. See Materials and Methods for a detailed description of quantification methods.

C. Live imaging of SIMR-1::GFP::3xFLAG; HA::EGO-1::mCherry::RRF-1 at 100-cell stage embryo, showing that RRF-1 colocalizes with SIMR-1. At least five individual embryos were imaged for each genotype and stage. Arrowheads and insets show examples of colocalization between SIMR-1 and RRF-1 at cytoplasmic granules. Scale bars, 5 μm.

D. Live imaging of mCherry::EGO-1::GFP::RRF-1 in a simr-1 mutant, showing that RRF-1 no longer associates with cytoplasmic granules, while EGO-1 remains associated with granules in the simr-1 mutant. At least five individual embryos were imaged. Arrowheads point to examples of cytoplasmic EGO-1 granules in a simr-1 mutant. Insets show examples of cytoplasmic localization of RRF-1 and granule localization of EGO-1 in a simr-1 mutant. Scale bars, 5 μm.

E. Live imaging of SIMR-1::mCherry::2xHA embryos in a mut-16 mutant at 8-cell, 100-cell, and comma stages. At least five individual embryos were imaged. Asterisks indicate spindle localization of SIMR-1 in a mut-16 mutant. Box highlights germ granule localization of SIMR-1 in a comma-stage, mut-16 mutant embryo. Scale bars, 5 μm.

F. Live imaging of GFP::3xFLAG::CSR-1 embryos at different stages (4-cell, 28-cell, 100-cell, 200-cell, and comma), shows that CSR-1 localizes to cytoplasmic granules in early embryos and is restricted to germ granules in late embryos. At least three individual embryos were imaged for each stage. Dotted white line marks perimeter of the embryo. Box marks germ granule localization of CSR-1. Scale bars, 5 μM.

G. Box plot quantifying GFP::3xFLAG::CSR-1 granules at different embryonic stages. At least ten embryos at each stage were used for quantification. Each dot represents an individual embryo, and all data points are shown. Bolded midline indicates median value, box indicates the first and third quartiles, and whiskers represent the most extreme data points within 1.5 times the interquartile range. Lines connect the mean granule number (red dots) at each stage, illustrating the change in number of CSR granules across embryonic development. Two-tailed t-tests were performed to determine statistical significance and p-values were adjusted for multiple comparisons. See Materials and Methods for a detailed description of quantification methods.

H. Live imaging of mCherry::EGO-1; GFP::3xFLAG::::CSR-1 embryo at 28-cell stage, showing CSR-1 colocalization with EGO-1. At least ten individual embryos were imaged. Arrowheads and insets show examples of CSR-1 and EGO-1 colocalization at cytoplasmic granules. Scale bars, 5 μm.

I. Live imaging of SIMR-1::mCherry::2xHA; GFP::3xFLAG::CSR-1 embryo at 28-cell stage, showing the absence of colocalization between SIMR-1 and CSR-1 with occasional adjacent localization. At least ten individual embryos were imaged. Arrowheads point to examples of SIMR and CSR granules that do not colocalize. Insets show examples of SIMR and CSR granules that are found adjacent to each other or fail to colocalize. Dotted white line marks perimeter of embryo. Scale bars, 5 μm.

J. Model of CSR and SIMR granules in the somatic cells of C. elegans embryos. The RdRP EGO-1, which synthesizes CSR-class 22G-RNAs, localizes to CSR granules, where CSR-1 loading may take place. The RdRP RRF-1, along with RDE-3, ENRI-2, and unloaded NRDE-3 localize to SIMR granules. SIMR-1 and ENRI-2 recruits unloaded NRDE-3 to granule where RRF-1 may synthesize ERGO-dependent 22G-RNAs for loading into NRDE-3. After loading, NRDE-3 translocates to the nucleus and silences genes co-transcriptionally.

SIMR-1 does not colocalize with any previously identified embryonic granules.

A. Live imaging of SIMR-1::GFP::3xFLAG; mCherry::CGH-1 embryo at 100-cell stage, showing that SIMR-1 does not colocalize with CGH-1. Arrowheads and insets show examples of SIMR-1 and CGH-1 granules that do not colocalize.

B. Live imaging of mCherry::CGH-1; eri-1; GFP::3xFLAG::NRDE-3 at 100-cell stage embryo, showing that NRDE-3 in the eri-1 mutant does not colocalize with CGH-1. Arrowheads and insets show examples of NRDE-3 and CGH-1 granules that do not colocalize.

C. Live imaging of SIMR-1::GFP::3xFLAG; RSD-2::mCherry embryo at 100-cell stage, showing that RSD-2 does not localize to cytoplasmic granules in the embryo. Arrowheads and insets show examples of SIMR granules.

D. Live imaging of SIMR-1::mCherry::2xHA; HRDE-2::2xTy1::GFP embryo at 100-cell stage, showing that HRDE-2 does not localize to cytoplasmic granules in the embryo. Arrowheads and insets show examples of SIMR granules.

E. Live imaging of SIMR-1::GFP::3xFLAG; hrde-2 embryo at 100-cell stage, showing that SIMR-1 still localizes to cytoplasmic granules in the hrde-2 mutant. Arrowheads and inset show examples of SIMR granules.

F. Live imaging of SIMR-1::mCherry::2xHA; GFP::3xFLAG::RDE-12 embryo at 100-cell stage, showing that SIMR-1 does not colocalize with RDE-12. Arrowheads and insets show examples of RDE-12 and SIMR granules that do not colocalize.

G. Live imaging of SIMR-1::mCherry::2xHA RSD-6::GFP::3xFLAG embryo at 100-cell stage, showing that SIMR-1 does not colocalize with RSD-6. Arrowheads and insets show examples of RSD-6 and SIMR granules that do not colocalize.

H. Live imaging of SIMR-1::GFP::3xFLAG; HA::tagRFP::ZNFX-1 embryo at 100-cell stage, showing that SIMR-1 does not colocalize with ZNFX-1. Arrowheads and insets show examples of SIMR and ZNFX-1 granules that do not colocalize.

I. Live imaging of SIMR-1::mCherry::2xHA; tubulin::GFP embryo at 100-cell stage, showing that SIMR-1 does not colocalize with tubulin. Arrowheads and insets show examples of tubulin and SIMR granules that do not colocalize.

J. Live imaging of MUT-16::GFP at 100-cell stage, showing that MUT-16 localizes to somatic granules. Arrowheads and inset show examples of MUT-16 somatic granules.

K. Live imaging of RDE-3::GFP in a mut-16 mutant embryo at 100-cell stage, showing that RDE-3 no longer associates with granules in the mut-16 mutant.

L. Live imaging of SIMR-1::GFP::3xFLAG; mCherry::EGO-1 embryo at 100-cell stage, showing that SIMR-1 does not colocalize with EGO-1. Arrowheads and insets show examples of SIMR-1 and EGO-1 granules that do not colocalize. At least five individual embryos were imaged for all experiments. All scale bars, 5 μm.

Autophagy regulates the removal of SIMR granules and other embryonic granules.

A. Live imaging of SIMR-1::mCherry and PGL-1::BFP at 8-cell, 100-cell, and comma stage embryos following treatment of parental animals with control (L4440) and lgg-1 RNAi. At least ten individual embryos were imaged for each genotype and stage. Dotted white line marks perimeter of embryo. Scale bars, 5 μm.

B. Box plot quantifying the number of SIMR granules at the 100-cell stage following treatment of parental animals with control (L4440) and lgg-1 RNAi. Fourteen embryos at each stage were used for quantification. Each dot represents an individual embryo, and all data points are shown. Bolded midline indicates median value, box indicates the first and third quartiles, and whiskers represent the most extreme data points within 1.5 times the interquartile range. Two-tailed t-test was performed to determine statistical significance.

C. Live imaging of GFP::CSR-1 and RFP::ZNFX-1 at 8-cell, 100-cell, and comma stage embryos following treatment of parental animals with control (L4440) and lgg-1 RNAi. At least ten individual embryos were imaged for each genotype and stage. Dotted white line marks perimeter of embryo. Scale bars, 5 μm.

D. Live imaging of SIMR-1::GFP; RFP::ZNFX-1; PGL-1::BFP embryo and SIMR-1::mCherry; GFP::CSR-1 embryo at 100-cell stage embryos following treatment of parental animals with lgg-1 RNAi. At least ten individual embryos were imaged for each genotype. Dotted white line marks perimeter of embryo. Insets show examples of granule adjacency. Scale bars for embryos, 5 μm. Scale bars for insets, 0.2 μm. Images in A, C-D are maximum projections of deconvolved 12.5 μm z-stacks (about two-third of embryo depth).

Components of the SIMR granule are regulated by autophagy.

A. Live imaging of GFP::NRDE-3; eri-1 at 8-cell, 100-cell, and comma stage embryos following treatment of parental animals with control (L4440) and lgg-1 RNAi. At least ten individual embryos were imaged for each genotype and stage. Scale bars, 5 μm. Images are maximum projections of deconvolved 12.5 μm z-stacks (about two-third of embryo depth).

B. Box plot quantifying Pearson’s R values of SIMR-1 with ENRI-2, RDE-3/MUT-2, RRF-1, and NRDE-3; eri-1 at comma stage on lgg-1 RNAi condition. At least twenty embryos at each stage were used for quantification. Each dot represents an individual embryo, and all data points are shown. Bolded midline indicates median value, box indicates the first and third quartiles, and whiskers represent the most extreme data points within 1.5 times the interquartile range.

NRDE-3 switches small RNA targets during development.

A. Diagram of IP-sRNA seq on NRDE-3 early embryos (<=100-cell stage) and late embryos (>=300-cell). GFP::FLAG::NRDE-3 was immunoprecipitated from embryo lysate and its associated small RNAs were isolated for sequencing.

B. Box plots depicting log2(fold change small RNA abundance) in NRDE-3 IP compared to input for at least two biological replicates.

C. Normalized NRDE-3-bound small RNA read distribution across a CSR-target gene (ztf-27) and an ERGO-target gene (Y37E11B.2) in early embryos, late embryos, and young adults. One representative replicate is shown.

D. Normalized NRDE-3 IP compared to input small RNA reads in early embryos, late embryos, and young adults. CSR-target and ERGO-target genes are indicated in blue and red, respectively. One representative replicate is shown. Insets are pie charts describing numbers of CSR targets, ERGO targets, and other targets that are significantly enriched in the NRDE-3 IP. The enriched targets were defined as small RNAs with at least 2-fold enrichment in IP compared to input, average RPM >10, and p-values ≤0.05.

E. Box plot depicting log2(fold change small RNA abundance) in mutants compared to wild-type in late embryos for two or three biological replicates.

F. Box plots depicting log2(fold change of ERGO-class small RNA abundance) in NRDE-3 IP compared to input in wild-type and mutants in late embryos for two or three biological replicates.

G. Box plots depicting log2(fold change of CSR-class small RNA abundance) in NRDE-3 IP compared to input in wild-type and mutants in late embryos for two or three biological replicates.

For box plots in B,E-G, bolded midline indicates median value, box indicates the first and third quartiles, and whiskers represent the most extreme data points within 1.5 times the interquartile range, excluding outliers. Two-tailed t-tests were performed to determine statistical significance and p-values were adjusted for multiple comparisons.

Defining NRDE-3-bound ERGO-target genes.

A. Normalized NRDE-3 IP compared to input small RNA reads in young adults (left), nrde-3 mutant compared to wild-type small RNA reads in mixed-stage embryos (middle), nrde-3 mutant compared to wild-type mRNA reads in mixed-stage embryos (right). ERGO-target genes from Manage et al (2020) are indicated in yellow. One representative replicate is shown.

B. Normalized NRDE-3 IP compared to input small RNA reads in young adults (left), nrde-3 mutant compared to wild-type small RNA reads in mixed-stage embryos (middle), nrde-3 mutant compared to wild-type mRNA reads in mixed-stage embryos (right). ERGO-target genes from Fischer et al (2011) are indicated in green. One representative replicate is shown.

C. Normalized NRDE-3 IP compared to input small RNA reads in young adults (left), nrde-3 mutant compared to wild-type small RNA reads in mixed-stage embryos (middle), nrde-3 mutant compared to wild-type mRNA reads in mixed-stage embryos (right). NRDE-3-bound, ERGO-target genes are defined as genes with at least four-fold enrichment and an average of 100 RPM from the young adult NRDE-3 IP-sRNA seq libraries, and are indicated in red. One representative replicate is shown.

D. Venn diagrams indicate overlap of ERGO - Manage and ERGO - Fischer targets gene lists with the newly defined NRDE-3-bound, ERGO-target genes.

E. Box plots depicting log2(fold change of small RNA abundance) in a nrde-3 mutant compared to wild-type in mixed-stage embryos for three biological replicates.

F. Box plots depicting log2(fold change of mRNA abundance) in a nrde-3 mutant compared to wild-type in mixed-stage embryos for three biological replicates.

For box plots in E,F, bolded midline indicates median value, box indicates the first and third quartiles, and whiskers represent the most extreme data points within 1.5 times the interquartile range, excluding outliers. Two-tailed t-tests were performed to determine statistical significance and p-values were adjusted for multiple comparisons. Adjusted p-values are a comparison of the indicated gene list to all genes in the mixed-stage embryos.

NRDE-3-bound small RNA in different mutants.

A. Normalized NRDE-3 IP compared to input small RNA reads in eri-1 mutant early (left) and late (right) embryos. CSR-target and ERGO-target genes are indicated in blue and red, respectively. One representative replicate is shown.

B. Normalized NRDE-3 IP compared to input small RNA reads in simr-1 mutant (left) and enri-2 mutant (right) late embryos. CSR-target and ERGO-target genes are indicated in blue and red, respectively. One representative replicate is shown.

NRDE-3 associates with CSR-class 22G-RNAs in early embryos.

A. Venn diagrams indicate overlap of NRDE-3 IP enriched targets in early embryos (this work), CSR-1 IP enriched targets in young adults (Nguyen et al., 2021), and CSR-1 IP enriched targets in embryos (Quarato et al., 2021).

B. Normalized NRDE-3 IP compared to input small RNA reads in early embryos. CSR-target genes with 1-50 RPM, with 50-100 RPM, and with more than 150 RPM are indicated in light blue, medium blue, and dark blue, respectively. One representative replicate is shown.

C. Density plot of small RNA enrichment on CSR targets in CSR-1 IP (dark blue), NRDE-3 IP (light blue) in embryos (left) and adults (right). Transcription start site (TSS) to transcription end site (TES) were plotted using normalized small RNA reads. All replicates are shown as individual lines.

D. Box plot quantifying the number of embryos laid per adult csr-1::degron or csr::degron, gfp::nrde-3(HK-AA) animal on 4mM auxin plate. At least 65 individuals from each strain were scored. Each dot represents an individual animal, and all data points are shown.

E. Box plot depicting log2(fold change of H3K9me3 level in IP vs input) in wild-type (grey) and nrde-3 mutant (green) mixed-stage embryos, indicating that the H3K9me3 level of NRDE-3 targets in early embryos are not affected in nrde-3 mutant. Anti-H3K9me3 ChIP-seq data was obtained from Padeken et al. (2021).

For box plots in D-E, bolded midline indicates median value, box indicates the first and third quartiles, and whiskers represent the most extreme data points within 1.5 times the interquartile range, excluding outliers. Two-tailed t-tests were performed to determine statistical significance and p-values were adjusted for multiple comparisons.

NRDE-3 associates different classes of small RNAs during development.

A. Normalized NRDE-3 IP compared to input small RNA reads in late embryos (left) and young adults (right). CSR-target genes with 1-50 RPM, with 50-100 RPM, and with more than 150 RPM are indicated in light blue, medium blue, and dark blue, respectively. One representative replicate is shown.

B. Density plot of small RNA enrichment on ERGO targets in CSR-1 IP (dark red), NRDE-3 IP (light red) in the embryos (left) and adults (right). Transcription start site (TSS) to transcription end site (TES) were plotted using normalized small RNA reads. All replicates are shown as individual lines.

C. Box plots depicting normalized log2(fold change of small RNA abundance in IP vs input) in a NRDE-3 IP in early embryos and CSR IP in embryos for two or three biological replicates. All genes list includes all genes expressed in IP or input. Early degraded mRNAs are maternal mRNAs that show at least twofold reduction in mRNA levels in early embryos (4 to 20 cell-stage) compared to 1-cell embryos (Quarato et al., 2021). Late degraded mRNAs are maternal mRNAs that show stable levels of mRNAs in early embryos and at least twofold reduction in late embryos (more than 20-cell stage) (Quarato et al., 2021). Zygotic mRNAs are mRNAs that are not detectable in 1-cell embryos but accumulate in early and late embryos (Quarato et al., 2021).

D. Box plot depicting mRNA RPM in one-cell embryo (dark blue), early embryo (medium blue), and late embryo (light blue) on NRDE-3 targets in early embryos (this study) and CSR-1 targets in embryos (Quarato et al., 2021) (left), and mRNA RPM in one-cell embryo (dark pink), early embryo (medium pink), and late embryo (light pink) on ERGO targets (this study) (right). All biological replicates (two or three) are included. mRNA-seq data was obtained from Quarato et al. (2021).

E. Quantification of the fertility for csr-1::degron and csr-1::degron, gfp::nrde-3(HK-AA) animals on ethanol and 4mM auxin plate. The fertility of each animal was categorized as: hatched (some embryos laid by the animal hatched), >=50 dead embryos (animal laid at least 50 unhatched embryos and no hatched embryos), <50 dead embryos (animal laid less than 50 unhatched embryos and no hatched embryos), no embryos (animal did not lay any embryos). At least 65 individual adult animals of each strain and each condition were used for quantification.

NRDE-3 associates with CSR-class 22G-RNAs to represses RNA Polymerase II in oocytes.

A. Live imaging of GFP::3xFLAG::NRDE-3 in one-day-adult germlines for wild-type, eri-1, rde-3, and nrde-3(HK-AA) mutants, showing that NRDE-3 localizes to the nuclei of oocytes in wild-type, eri-1 mutant, and rde-3 mutants, and to the cytoplasm in the nrde-3(HK-AA) mutant. At least five individual gonads were imaged for each genotype. Dotted white line traces the proximal portion of the C. elegans gonad and outlines the individual oocytes. Scale bars, 25 μm.

B. Live imaging of one-day-adult germlines and 8-cell embryos for EGO-1::degron; GFP::3xFLAG::NRDE-3 (top) and EGO-1::degron; GFP::3xFLAG::NRDE-3 in a rde-3 mutant (bottom) with ethanol (control) and 4mM auxin treatment, showing that loss of both WAGO-class and CSR-class 22G-RNAs (rde-3 mutant and degron-mediated EGO-1 depletion) is necessary to result in cytoplasmic localization of NRDE-3 in both oocytes and early embryos. At least five individual gonads and embryos were imaged for each treatment condition. Dotted white line traces the proximal portion of the C. elegans gonad and outlines the individual oocytes. Arrowheads indicate granule localized NRDE-3 in 8-cell stage embryos. Scale bars, 25 μm in adults and 5 μm in embryos.

C. Box plot quantifying the RNA Pol II PSer2 signal intensity in oocytes of wild-type (GFP::NRDE-3 strain), and nrde-3(HK-AA) mutant (GFP::NRDE-3(HK-AA) strain), showing that PSer2 signal is significantly increased in all oocytes in the nrde-3(HK-AA) mutant. Each dot represents an individual oocyte, and all data points are shown. Bolded midline indicates median value, box indicates the first and third quartiles, and whiskers represent the most extreme data points within 1.5 times the interquartile range. Two-tailed t-tests were performed to determine statistical significance. See Materials and Methods for a detailed description of quantification methods.

D. Bar plot quantifying the RNA Pol II PSer2 expression pattern in wild-type (GFP::NRDE-3 strain), and nrde-3(HK-AA) mutant (GFP::NRDE-3(HK-AA) strain) oocytes, showing that the PSer2 signal is retained on DNA longer in the nrde-3(HK-AA) mutant. At least 10 oocytes were used for quantification for each strain. Examples of three patterns of PSer2 signal are shown on right. Arrows point to a region of DNA to highlight PSer2 enrichment or exclusion. Scale bars, 2 μm.

E. Immunofluorescence imaging of PSer2 signal and DAPI stained DNA in oocytes of wild-type (GFP::NRDE-3), and nrde-3(HK-AA) mutant (GFP::NRDE-3(HK-AA)), showing that the PSer2 signal appears earlier on DNA in the nrde-3(HK-AA) mutant. Images are maximum intensity projections of 12.5 μm z-stack, which allows for optimal visualization of the DNA-associated PSer2 signal, but obscures the ‘excluded from DNA’ pattern. At least five individual animals for each genotype. Arrows indicate the PSer2 signal on DNA. Scale bars, 25 μm.

Model for temporally- and developmentally-regulated NRDE-3 function.

Model of NRDE-3, SIMR-1, and CSR-1 function during C. elegans development. In early embryos, CSR-1 and EGO-1 localize to CSR granules and synthesize CSR 22G-RNAs to slice and clear maternal mRNAs. NRDE-3 binds CSR 22G-RNA in the nucleus, which are inherited from the oocytes. During mid-embryogenesis (e.g. around the 100-cell stage), unloaded NRDE-3, ENRI-2, RRF-1, and RDE-3 are localized to SIMR granules in somatic cells dependent on SIMR-1, where we propose that ERGO-dependent 22G-RNAs are produced and loaded into NRDE-3. In late embryos, NRDE-3 binds ERGO-dependent 22G-RNAs and silences ERGO-target genes in the nucleus, while autophagy controls selective degradation of SIMR and other embryonic granules. In adult C. elegans, somatic localized NRDE-3 associates with ERGO-dependent 22G-RNAs to transcriptionally silence ERGO-target genes, while germline localized NRDE-3 associates with CSR-class 22G-RNAs to globally repress transcription and promote chromatin compaction in oocytes, ultimately being deposited into early embryos.

Predicted structures of SIMR-1, nuclear Argonaute proteins, and interactors.

A. Graph displaying disorder tendency for the SIMR-1 protein sequence and AlphaFold3 predicted structures of SIMR-1. The disorder prediction was made using PONDR (http://www.pondr.com) with VSL2 and VSL3 parameters. Regions with PONDR scores of greater than 0.5 indicate disorder and regions with PONDR score less than 0.5 indicate order. The AlphaFold3 predicted confidence score of SIMR-1 is shown. The extended Tudor domain (aa. 89-259) of SIMR-1 is highlighted in orange.

B. AlphaFold3 predicted structures of HRDE-1 - HRDE-2, and NRDE-3 - ENRI-2 complexes, showing that the C-terminus of HRDE-2 and ENRI-2 are inserted into small RNA binding pockets of HRDE-1 and NRDE-3 respectively. Domains of HRDE-1 and NRDE-3 are indicated and AlphaFold3 predicted confidence scores are shown.