Fig 1.
L1 bodies (LBs) are large cytoplasmic compartments enriched in ORF1p, L1 mRNA, and ribosomes.
(A–E) Electron micrographs exemplifying the formation of LBs in Mael-/- spermatocytes. Boxed areas in A and B identify small cytoplasmic aggregates magnified in A’ and B’, respectively. Boxed area in C identifies an intermediate size LB magnified in C’. D and E show more prominent LBs, partially surrounded by a bilayer membrane (red arrows). Nuc: nucleus. Cyto: cytoplasm. NE: nuclear envelope. (F) Immunofluorescence staining of L1 ORF1p on Mael-/- spermatocytes shows ORF1p accumulation in LBs (left panel, magenta). 3D reconstruction of LBs (magenta) and nuclei (blue) (middle and right panels); numbers [1–5] are used to label corresponding cells. Scale bars: 5 μm. (G) HCR RNA-FISH of L1 RNA (green) and immunofluorescence staining of L1 ORF1p (magenta) on Mael+/- and Mael-/- testis sections. Scale bars: 5 μm. (H) Double immunofluorescence staining of ribosomal protein S6 (RPS6, top two rows, green) or L28 (RPL28, bottom two rows, green) and L1 ORF1p (magenta) showing their colocalization in LBs. Dotted white lines mark seminiferous tubule boundaries. Scale bars: 15 μm. See also S1 Fig.
Fig 2.
L1-polysome and L1 RNP complexes in Mael-/- testes.
(A) Western blot analysis of L1 ORF1p and RPS6 in Mael-/- testicular extracts fractionated through 10–50% sucrose gradients in the presence of cycloheximide (CHX). Fractionated Mael+/- extract served as a negative control. The absorbance profile of the gradient is shown in red. (B) qRT-PCR analysis of L1 RNA (L1 ORF1 amplicon) and beta-actin mRNA (Actb) in Mael-/- testicular extracts fractionated through 10–50% sucrose gradients in the presence of CHX. RNA amounts per fraction are reported as percentages of the total RNA detected across the gradient; barplot shows mean percentage values + SEM of four biological replicates. (C) Western blot analysis of L1 ORF1p and RPS6 in Mael-/- testicular extracts fractionated through 10–50% sucrose gradients in the presence of EDTA (+ EDTA); the absorbance profile of the gradient is shown in blue. Upon addition of RNases A and T1 (+ EDTA + RNase) ribosomes are degraded (profile in green) and L1 RNPs disaggregated. (D) qRT-PCR analysis of L1 RNA (L1 ORF1 amplicon) and beta-actin mRNA (Actb) in Mael-/- testicular extracts fractionated through 10–50% sucrose gradients in the presence of EDTA. Barplot shows mean percentage values + SEM of three biological replicates. See also S2 Fig.
Fig 3.
L1 ORF1p binds to L1, other TEs and abundant genic mRNAs.
(A) RNA levels of selected genomic repeats across anti-ORF1p co-immunoprecipitation samples (TOTAL, INPUT, Beads Only (BO), IP; in triplicate). Heatmap reports log2 normalized collapsed read counts; color indicates RNA levels. (B) Identification of mRNAs bound by L1 ORF1p in IP samples (n = 2081). For every comparison (IP vs. BO, IP vs. TOTAL or IP vs. INPUT), mRNAs with log2FoldChange > 1 and padj < 0.05 were considered enriched. (C) Linear regression analysis of n = 2081 ORF1p-bound mRNAs showing the relationship between their expression level in Mael-/- testes (x axis, TOTAL) and their enrichment in L1 RNPs (y axis, IP). Scatterplot shows log2 mean normalized counts. (D) Linear regression analysis of n = 2081 ORF1p-bound mRNAs showing the relationship between their length (x axis, log2 length in nucleotides) and their enrichment in L1 RNPs (y axis, IP). (E) Multiplexed HCR RNA-FISH of MMERVK10c-int and Setx with L1 RNA on Mael+/- and Mael-/- testis sections. Images of spermatocytes (SPC) and round spermatids (RS) are reported. Scale bars: 5 μm. See also S3 and S4 Figs.
Fig 4.
L1 ORF1p associates with cytoplasmic RNA granules components, the CCR4-NOT complex and ribosomal proteins.
Network of physical and functional interactions among the 80 high-confidence L1 ORF1p interactors identified in this study. The network was generated using the STRING database; thickness of connection lines indicates the strength of data support. Proteins identified for the first time in this work (new) are marked in blue, while previously known ORF1p interactors are marked in green. ORF1p interactors resistant to RNase treatment are enclosed in dashed grey areas. *: unspecific interactor. See also S5 Fig.
Fig 5.
LBs are enriched with DDX6 and CNOT7 proteins.
(A) Double immunofluorescence staining of RNA helicase DDX6 (green) and L1 ORF1p (magenta) in Mael+/- and Mael-/- testes. DDX6 is detectable in cytoplasmic foci in both Mael+/- and Mael-/- spermatogonia and spermatocytes (yellow arrows); in Mael-/- both ORF1p and DDX6 concentrate in LBs (white arrowheads). Dotted white lines mark seminiferous tubule boundaries. Scale bars: 10 μm. (B) Double immunofluorescence staining of CCR4-NOT complex subunit CNOT7 (green) and L1 ORF1p (magenta) in Mael+/+ and Mael-/- testes. CNOT7 is detectable in cytoplasmic foci in Mael+/+ spermatocytes (yellow arrows); in Mael-/- CNOT7 colocalizes with ORF1p in LBs (white arrowheads). Scale bars: 10 μm. (C) Double immunofluorescence staining of DCP1A (green) and L1 ORF1p (magenta) in Mael+/- and Mael-/- testes. DCP1A shows a characteristic granular pattern in both samples (yellow arrows), although PBs appear overall less abundant in Mael-/- seminiferous tubules; weak diffuse DCP1A signal is also detected in LBs (white arrowheads). Boxed areas in Mael-/- panels are magnified in insets. Scale bars: 10 μm. See also S6 Fig.
Fig 6.
L1 ORF1p binding to endogenous mRNAs affects neither their stability nor their translation efficiency.
(A) RNA-seq coverage of RNAs detected in anti-L1 ORF1p co-immunoprecipitation samples (IP and TOTAL; shown in triplicate). Each of the segments on the x axis (labeled 1 to 10) corresponds to 10% of transcript lengths; the cumulative number of reads across each segment is plotted (y axis, read count x106). (B) Immunofluorescence staining of L1 ORF1p (magenta) in P16 Mael+/- and Mael-/- testes. Scale bars: 100 μm. (C) Analysis of the poly(A) tail of Sycp1 and Setx mRNAs in P16 testes of Mael+/- and Mael-/- littermates. A schematic of the assay is shown on the left; numbers mark the assay steps: 1) poly(U) tailing; 2) cDNA synthesis using an RT adapter; and 3) PCR spanning the poly(A) tail. Representative gels are shown on the right; the assay was performed in biological triplicate. (D) Scatterplot showing the relationship between mRNA level changes (x axis, RNA-seq log2FoldChange) and ribo-footprint changes (y axis, Ribo-seq log2FoldChange) in Mael-/- over Mael+/- testes. A subset of 1313 mRNAs found associated with ORF1p previously in this study are labeled in red. ΔTE n.s.: difference in Translation Efficiency between Mael-/- and Mael+/- non significant. (E) Western blot analysis of phosphorylated eIF2α levels (P-eIF2α) in P16 testes of Mael+/- and Mael-/- littermates. Blot #1 and #2 were run and processed in parallel. Blot #1 was first probed with an anti-P-eIF2α antibody, then stripped and re-probed with an anti-eIF2α antibody (pan); blot #2 was probed sequentially with the monoclonal anti-ORF1p (Abcam) and an anti-DDX4/MVH antibody.
Fig 7.
RNase-resistant ORF1p-interactor PRKRA localizes to LBs in vivo and promotes mouse L1 retrotransposition in cultured human cells.
(A) Western blot analysis of PRKRA in anti-L1 ORF1p co-immunoprecipitation samples from Mael-/- testes. Immunoprecipitation with an isotype IgG was used as a negative control; note the higher yield of ORF1p in the presence of RNases. Super: supernatant. (B) Double immunofluorescence staining of PRKRA (green) and L1 ORF1p (magenta) in Mael+/- and Mael-/- testes shows PRKRA enrichment in LBs (yellow arrows). Scale bars: 10 μm. (C) Representative L1 retrotransposition assay results using the retrotransposition-competent or control plasmids listed on the left, in combination with pEGFP or pPRKRA in HeLa cells. neoR: neomycin-resistant colonies. (D) Overexpression of mouse PRKRA stimulates retrotransposition of a mouse wild type L1 in HeLa cells. Box-and-whisker plot shows corrected retrotransposition efficiencies from five biological replicates. (E) Double immunofluorescence staining of 3XFLAG-PRKRA (anti-FLAG antibody, magenta) and ORF1p (yellow) in HeLa cells co-transfected with pTN201 together with pEGFP or pPRKRA. Boxed areas in pPRKRA images are magnified in the corresponding insets and identify a cell with a moderate expression level of pPRKRA plasmid; exogenous 3XFLAG-PRKRA shows a cytoplasmic distribution that partially overlaps with L1 ORF1p granules (yellow arrows). Scale bars: 10 μm. (F) qRT-PCR analysis of mouse L1 RNA (ORF1 amplicon) derived from pTN201 co-transfected with pEGFP or pPRKRA in HeLa cells. Human GAPDH was used as the reference gene. Barplot shows relative normalized expression ± SEM of three biological replicates. Two-tailed unpaired Student’s t-test: *** p ≤ 0.001. (G) Western blot analysis of mouse ORF1p derived from pTN201 co-transfected with pEGFP or pPRKRA in HeLa cells. Alpha-tubulin (α-tub) serves as a loading control. (H) Barplot showing mean fold change of mouse ORF1p derived from pTN201 co-transfected with pPRKRA in HeLa cells. ORF1p band intensities are normalized to alpha-tubulin (α-tub) and further shown as fold change relative to control (pTN201 + pEGFP). Data are expressed as mean ± SEM of three biological replicates. Two-tailed unpaired Student’s t-test: ** p ≤ 0.01. (I) Western blot analysis of human phosphorylated eIF2α (P-eIF2α) in the presence of pTN201 co-transfected with pEGFP or pPRKRA in HeLa cells. eIF2α (pan) serves as a loading control. (J) Barplot showing mean fold change of human P-eIF2α in the presence of pTN201 co-transfected with pPRKRA in HeLa cells. P-eIF2α band intensities are normalized to eIF2α and further shown as fold change relative to control (pTN201 + pEGFP). Data are expressed as mean ± SEM of three biological replicates. Two-tailed unpaired Student’s t-test: * p ≤ 0.05. See also S7 Fig.