Fig 2: Hbs1l is required for embryogenesis.(A) Domain structure of HBS1L and isoform II and the exons encoding the two splice variants. (B) Design of Hbs1l loss-of-function alleles that target Hbs1l and isoform II. Hbs1ltm1a allele, Hbs1l specific gene trap (hypomorph); Hbs1ldelta allele, Hbs1l specific deletion of exon 5 (Hbs1l-, knock out); and Hbs1lGTC allele, Hbs1l gene trap to target Hbs1l and isoform II (knock out). (C) Cresyl violet-stained cross-sections of testis from P45 control (Hbs1ltm1a/+) and Hbs1ltm1a/tm1a mice. Immunofluorescence was performed with antibodies to proliferating cell nuclear antigen (PCNA, red) and sections were counterstained with DAPI. (D) Cresyl violet-stained sagittal brain section from 4-month-old control (Hbs1l+/+) and Hbs1ltm1a/tm1a mice. (E) Splicing analysis of correctly spliced Hbs1l and trapped Hbs1l transcripts in various tissues from 4-month-old control (Hbs1l+/+) and Hbs1ltm1a/tm1a mice. ß-actin was used as an input control. (F) Quantitative RT-PCR analysis of Hbs1l and isoform II using cDNA from E8.5 embryos. Data were normalized to Gapdh and the fold change in gene expression is relative to that of controls (Hbs1l+/+) from each cross. Data represent mean + SEM. (G) Bright field images of control (Hbs1l+/+ and Pelo+/+), Hbs1l-/- and Pelo-/- embryos at E7.5, E8.5, and E10.5. Scale bars: 500 µm and 200 µm (higher magnification), 50 µm (immunofluorescence image) (C); 500 µm (D); 100 µm (E7.5), 200 µm (E8.5), and 2 mm (E10.5) (G). PTC, premature termination codon; Frt, flippase-mediated recombination site; loxP, Cre recombinase-mediated recombination site; En2(SA), splice acceptor of mouse Engrailed-2 exon 2; SC, spermatocytes; SG, spermatogonia; St, spermatids; LC, Leydig cells. t-tests were corrected for multiple comparisons using Holm-Sidak method (F). ns, not significant; **p=0.01; ***p=0.001.Figure 1—source data 1.Hbs1l is required for embryogenesis.

Fig 3: Loss of Upf2 and Pelo cause similar cerebellar developmental defects.(A) Parasagittal (E13.5) and sagittal (P0) cerebellar sections of control (Pelofl/fl), En1Cre; Upf2 cKO and En1Cre; Pelo cKO mice stained with cresyl violet. (B) Sagittal cerebellar sections of P21 control (Pelofl/fl), Tg(Atoh1-Cre); Upf2 cKO and Tg(Atoh1-Cre); Pelo cKO mice stained with cresyl violet. Higher magnification images of lobule VI are shown below for each genotype. Cerebellar lobules are indicated by Roman numerals. (C) Immunofluorescence with antibodies to BrdU (green) and Ki67 (red) on sections of P6 control (Tg(Atoh1-Cre); Upf2fl/+), Tg(Atoh1-Cre); Upf2 cKO and Tg(Atoh1-Cre); Pelo cKO cerebellum. Mice were injected with BrdU 24 hr prior to harvest to determine the fraction of granule cell precursors in the EGL that exited the cell cycle. Images are shown for anterior (IV/V) and posterior (IX) lobules. (D) Quantification of the fraction of granule cell precursors in the EGL (lobules IV/V and IX) that exited the cell cycle (BrdU+, Ki67- cells) of control (Tg(Atoh1-Cre); Upf2fl/+), Tg(Atoh1-Cre); Upf2 cKO, Tg(Atoh1-Cre); Pelo cKO and Tg(Atoh1-Cre); Hbs1l cKO mice. Data represent mean + SD. (E) Quantification of terminally differentiated granule cells (NeuN+ cells) in the IGL in lobules IV/V and IX of control (Tg(Atoh1-Cre); Upf2fl/+), Tg(Atoh1-Cre); Upf2 cKO, Tg(Atoh1-Cre); Pelo cKO and Tg(Atoh1-Cre); Hbs1l cKO mice. Data represent mean + SD. Scale bars: 200 µm (E13.5) and 500 µm (P0) (A); 500 µm and 20 µm (higher magnification) (B); 50 µm (C). VZ, ventricular zone, EGL, external granule cell layer; NTZ, nuclear transitory zone; SC, superior colliculus; IC, inferior colliculus; ML, molecular cell layer; PL, Purkinje cell layer; GCL, granule cell layer; oEGL, outer external granule cell layer; iEGL, inner external granule cell layer; IGL, internal granule cell layer. Two-way ANOVA was corrected for multiple comparisons using Tukey method (D, E). ns, not significant; ***p=0.001; ****p=0.0001.Figure 7—source data 1.Loss of Upf2 and Pelo cause similar cerebellar developmental defects.

Fig 4: Loss of Pelo induces greater defects in translation elongation then loss of Hbs1l.(A) Experimental strategy for in vitro studies using tamoxifen (4-OHT) treatment of primary mouse embryonic fibroblasts (MEFs). (B) Western blot analysis of HBS1L and PELO using MEF lysates from tamoxifen-treated control (Hbs1lfl/+ or Pelofl/fl), Hbs1l-/- and Pelo-/- cells. Vinculin was used as a loading control. (C) Analysis of significantly increased local ribosome occupancy (z-score =10, pause site detected in all three replicates) from tamoxifen-treated control (TgCAG-CreER) and Pelo-/- (left), or control and Hbs1l-/- cells (right). The number of genes that pause sites map to is shown below for each genotype. (D) Comparison of the z-scores (‘pause score’) for pauses observed in Pelo-/- (blue) and Hbs1l-/- (red) cells. Downward direction of the arrow indicates significant lower pause scores of Hbs1l-/-- compared to Pelo-/--specific pauses. (E) Examples of mapped footprints (27–34 nucleotides) on genes from tamoxifen-treated control (TgCAG-CreER, gray) and Pelo-/- (blue), or control and Hbs1l-/- (red) cells. Upward direction of arrows indicates significant increase in local ribosome occupancy and the dashed line indicates the pause site. 4-OHT, 4-hydroxytamoxifin; nts, nucleotides; rpm, reads per million. Wilcoxon rank-sum test was used to determine statistical significance (D).

Fig 5: UV- and 4NQO-stalled ribosomes are resistant to the ASCC(A) The effect of ASCC3 or ZNF598 depletion on the loss of collided ribosomes 1 h after treatment with emetine, anisomycin, 4NQO, or UVB. ASCC3 (A) and ZNF598 (Z) were efficiently depleted compared with the control (C).(B) The effect of ASCC3, ASCC2, ASCC1, or ASC-1 depletion on the loss of collided ribosomes 1 h after treatment with anisomycin. ASCC1 (1), ASCC2 (2), ASCC3 (3), or ASC-1 (A) were efficiently depleted compared with the control (C).(C and D) ASCC3 depletion in (C) or Pelo depletion in (D) has no effect on the loss of 4NQO-stalled ribosomes.See also Figure S3.