Fig 2: The proposed molecular mechanism at a glance. During the epididymal transit, the activation of CB1 receptor in SPZ favours the recruitment of the heterotrimeric FUS-QKI-RNApol2 complex on CNOT6L-mRNA, thus inducing the backsplicing useful for circCNOT6L biogenesis. In cauda epididymis, a cargo of FUS protein and CNOT6L-mRNA is transferred to SPZ via epididymosomes to support the intrinsic ability of SPZ to do backsplicing. After fertilization, the paternal content of FUS, CNOT6L-mRNA and circCNOT6L is transferred from SPZ to oocyte to sustain zygote development

Fig 3: FUS drives CNOT6L backsplicing by interacting with RNApol2 and QKI in caput SPZ. The enrichment levels of circCNOT6L and CNOT6L-mRNA in the products of RIP assay (FUS-IP compared with IgG-IP) in WT caput SPZ alone (A) and after in vitro ACEA treatment (F) detected by qRT-PCR. Data are reported as mean ± SEM from three independent experiments. **P < 0.01. qRT-PCR detection of circCNOT6L and CNOT6L-mRNA expression levels (E) in caput SPZ from WT mice in vitro treated with vehicle (CTRL) or ACEA 1 µM (n = 3 different samples from three different animals for each experimental group in triplicate). Data are reported as mean value of nfe ± SEM, using Actin as endogenous control. **P < 0.01. Western blot analysis of FUS, QKI and RNApol2 in the products of IP in WT caput SPZ (B, C and D) using FUS, QKI and RNApol2 antibodies. Western blot analysis of RIP protein fraction immunoprecipitated with FUS Ab (FUS-IP) in WT caput SPZ after in vitro ACEA treatment (G). FUS-IP is analyzed in comparison to control IgG-IP and Input protein extracts. qRT-PCR detection of Fus expression levels in ESCs, cultured in RM medium, treated with siScr or siFus (5, 10, 25 and 50 nM) and harvested after 48 (H) and 72 h (I) from siRNA transfection. Data are reported as mean value of nfe ± SEM, using Actin as endogenous control. **P < 0.01; *P < 0.05. Western blot analysis of FUS protein (J) and qRT-PCR detection of circCNOT6L expression levels (K) in ESCs, cultured in RM medium, treated with siScr or siFus (25 nM) and harvested after 48 and 72 h from siRNA transfection. In (J) FUS amount was quantified by densitometry analysis, normalized against Actin (ACT) signals, and reported as mean value of fc of OD values ± SEM, while in (K), the data are reported as mean value of nfe ± SEM, using Actin as endogenous control. **P < 0.01; *P < 0.05

Fig 4: CircLIMA1-dependent sperm nuclear actin organization along the epididymis. qRT-PCR detection of circLIMA1 expression levels in caput and cauda SPZ (A) or epididymis (B) from WT and CB1-/- mice. (C) Immunofluorescence analysis of F-actin by phalloidin staining in caput and cauda SPZ from WT and CB1-/- mice. qRT-PCR detection of circLIMA1 expression levels (D) and Immunofluorescence analysis of F-actin by phalloidin staining (E) in caput SPZ from WT mice in vitro co-incubated with: WT Caput ELF (CTRL group), WT Caput ELF combined with Cytochalasin-D (C8273; Sigma-Aldrich, Milano, Italy) at the concentration of 10 μM (ELF WT+CYTOD); CB1-/- Caput ELF (ELF CB1-/-); CB1-/- Caput ELF combined with Cytochalasin-D 10 μM (ELF CB1-/- +CYTOD); (n=3 different samples for each experimental group from 8 different animals in triplicate). qRT-PCR data are normalized using Cyclophilin and RPS18, for SPZ and epididymis respectively, expressed as DeltaCq and reported as mean value ± S.E.M. Experimental groups with statistically significant differences (p<0.05; p<0.01) were indicated with different letters; the experimental groups without statistically significant differences were indicated with the same letter. White full arrowheads and asterisks represent F-actin localization (red) in sperm head and tail, respectively. Nuclei were labeled with DAPI (blue). Scale bar: 15 µm (F) Immunofluorescence analysis of QKI protein in caput and cauda SPZ from WT and CB1-/- mice. White full arrowheads and asterisks represent QKI localization (green) in sperm head and tail, respectively. Nuclei were labeled with DAPI (blue); F-actin was labeled with phalloidin staining (red). Scale bar: 15 µm (G-H) The enrichment levels of circLIMA1 in the products of RIP assay (QKI-IP compared with IgG-IP) in caput and cauda SRN from WT (G) and CB1-/- (H) mice detected by qRT-PCR. Data are reported as mean ± SEM from three independent experiments. **p<0.01. (J-K) Western blot analysis of RIP protein fraction immunoprecipitated with QKI Ab (QKI-IP) in caput and cauda SRN from WT and CB1-/- mice. QKI-IP was analyzed in comparison to control IgG-IP and Input protein extracts.

Fig 5: Characterization of backsplicing machinery in SPZ and epididymis of CTRL and HFD mice. (A) Immunofluorescence analysis of FUS protein in caput and cauda SPZ of CTRL and HFD mice. White arrowheads and white asterisks represent FUS localization (RED) in sperm head and tail, respectively. Nuclei were labeled with DAPI (blue). Scale bar: 5 µm. (B) Western blot analysis of FUS protein (75 KDa) in caput and cauda SPZ of CTRL and HFD mice (n = 5 different samples for each experimental group in triplicate). Signals were quantified by densitometry analysis and normalized to Ponceau Red (Pon.S). Data were expressed in OD values and reported as mean ± SEM. Experimental groups with statistically significant differences (p < 0.01) were indicated with different letters. (C) The enrichment levels of circADAM10 in the products of RIP assay (FUS-IP compared with IgG-IP) in caput and cauda SPZ of CTRL and HFD mice detected by qRT-PCR. Data are reported as mean ± SEM from three independent experiments. ** p < 0.01. (D) Immunocytochemistry of FUS in Bouin’s fixed caput and cauda epididymis sections (7 µm) of CTRL and HFD mice (n = 5 different samples for each experimental group in triplicate). The FUS protein localization in principal cells was indicated by black arrows. Scale bar: 50 µm; inset: H&E staining of Bouin’s fixed caput and cauda epididymis sections (7 µm) of CTRL and HFD mice; scale bar: 20 µm. (E) Western blot analysis of FUS protein in caput and cauda epididymis of CTRL and HFD mice. Signals were quantified by densitometry analysis and normalized to Ponceau Red (Pon.S). Data were expressed in OD values and reported as mean ± SEM. Experimental groups with statistically significant differences (p < 0.01) were indicated with different letters. (F) The enrichment levels of circPCSK6 in the products of RIP assay (FUS-IP compared with IgG-IP) in caput and cauda epididymis of CTRL and HFD mice detected by qRT-PCR. Data are reported as mean ± SEM from three independent experiments. ** p < 0.01. (G) Western blot analysis of QKI protein (40 KDa) in caput and cauda epididymis of CTRL and HFD mice. Signals were quantified by densitometry analysis and normalized to Ponceau Red (Pon.S). Data were expressed in OD values and reported as mean ± SEM. Experimental groups with statistically significant differences (p < 0.01) were indicated with different letters. (H) IP in caput and cauda epididymis of CTRL and HFD mice. Total proteins collected were immunoprecipitated using FUS antibody. Protein interaction between FUS and QKI was detected by Western blot analysis.