Extended Data Fig. 4: RPF signals on piRNA precursors respond to translation inhibitors.

a, Examples of pachytene piRNA genes with single uORF (two on the left), and pachytene piRNA genes with multiple uORF (one on the right). Browser views of piRNA precursor loci with normalized reads of RPFs and piRNAs. Grey shades highlight the uORF regions. Ppm, parts per million. b, A₂₅₄ absorbance profile of testis lysates from adult mice with (blue) and without (black) harringtonine treatment following separation in 15% to 60% sucrose density gradients. Harr, harringtonine. c, Aggregated data for RPF abundance on pachytene expressed mRNAs from untreated testis (top), and from harringtonine treated testis (bottom) across 5´UTRs, ORFs, and 3´-UTRs of pachytene-expressed mRNAs. The x-axis shows the median length of these regions, and the y-axis represents the 10% trimmed mean of relative abundance. d, Cumulative distribution of Z scores comparing lengths of the longest ORFs from mRNAs (black line), lncRNAs (grey line), and pachytene piRNA precursors (pre-piRNAs) (purple line) with the length of the predicted longest ORF from the random shuffled background. Given a sequence X, its nucleotides were shuffled to generate a sequence Y. The length N of the longest ORF in Y was counted. After repeating this procedure 10,000 times, the Empirical Null Distribution of ORF lengths conditioned on the nucleotide distribution of X was obtained. The Z score of the ORF length of X was calculated based on this empirical null distribution. The dashed line denotes a Z-score of 2.58, which corresponds to p = 0.01 according to a two-side Z-test. The 29 ORFs from piRNA precursor transcripts with a Z score > 2.58 were from transposons embedded in the transcripts. e, Length histograms of total testis small RNAs from Control (Mov10l1^CKO/∆, left) and Mov10l1 CKO (Mov10l1^CKO/∆ Neurog3-cre, right). Statistical Source Data are provided in Source Data Extended Data Fig. 4.

Source data