RNA-Seq in human and Drosophila testes

Total RNA from two snap-frozen testicular tissue samples of the azoospermic individual M1911 (RNF113B variant) as well as from one sample each from three unrelated individuals with normal spermatogenesis (M1544 [obstructive azoospermia], M2224 [anorgasmia] and M2234 [previous vasectomy]; Siebert-Kuss et al., 2023), were extracted using the Direct-zol RNA Microprep kit (Zymo Research), following the manufacturer’s instructions. RNA quality was estimated by electrophoresis (Agilent Technologies), with all samples having a RNA integrity number (RIN) >4.5 (range: 4.5–5.6), except one replicate of M1911 (with RIN = 2.0). rRNA depletion was performed with the NEBNext rRNA Depletion Kit v2 (New England Biolabs), and total RNA libraries were prepared with the NEBNext Ultra II Directional RNA Library Prep Kit for Illumina (New England Biolabs), using 700 ng of RNA per sample. Paired-end sequencing with 150 bp per read was performed using the NextSeq2000 system (Illumina) at the University of Münster, Germany. An average of 44 million reads was generated per sample. The expression pattern of RNF113B was determined in a control testis by retrieving the corresponding data from our recently published single-cell RNA-Seq dataset of normal human spermatogenesis (Di Persio et al., 2021).

For fruit flies, RNA was extracted from 40 pairs of adult testes per sample per condition (3–7 days post‐eclosion) using the PureLink RNA Mini Kit (Thermo Fisher Scientific) following the manufacturer’s instructions. Three conditions were analyzed: dRNF113 RNAi (meiotic arrest), Prp19 RNAi (meiotic arrest) and the mCherry RNAi negative control (normal spermatogenesis), with three independent biological replicates per condition. Extracted RNA was treated with DNAse (Thermo Fisher Scientific), with a RNA quality number of 10 for all samples, as assessed by electrophoresis (Advanced Analytical Technologies). Total RNA libraries were prepared using the Zymo-Seq RiboFree Total RNA-Seq Library Kit (Zymo Research), using 1000 ng of RNA per sample. Paired-end sequencing with 150 bp per read was performed using DNBSEQ technology (BGI Group), with an average of 46 million reads per sample. Confirmation of knockdown efficiency via quantitative RT-PCR was performed for both the dRNF113 and Prp19 RNAi. For this, cDNA was synthetized using the Transcriptor First Strand cDNA Synthesis Kit (Roche), according to the manufacturer’s instructions. Reactions were performed in a QuantStudio 6 Real-Time PCR System (Thermo Fisher Scientific), using SYBR Green chemistry (Applied Biosystems). All reactions were set up in triplicates and Act57B and RpL32 were used as internal controls. Primers are listed in Supplementary file 3.

RNA-Seq reads were aligned against the human and fruit fly genomes (GRCh38 and BDGP6.32, respectively) using the STAR aligner (Dobin et al., 2013). Gene-level counts were obtained using feature-counts, taking into account the strandedness. Counts were normalized with the TMM method (Robinson and Oshlack, 2010), and differential gene expression analysis was performed using a quasi-likelihood F-test (Lun et al., 2016), as implemented in the edgeR package. DEGs were defined using the previously listed criteria (see ‘Differential gene expression analysis of the mitosis-to-meiosis transcriptional burst’). To determine to what extent the human patient sample with lower RIN could be affecting our results, we repeated this analysis correcting for sample degradation using DegNorm (Xiong et al., 2019) This revealed that 97.8% of all DEGs, as defined by our criteria, were also listed as such after taking into account sample degradation. This fits with a largely similar number of expressed genes (at TPM >1) detected in the control and in the experimental groups (average of 53.9% and 46.8% expressed protein-coding genes, respectively).