Figure 3. Features of the spermatocyte transcription program.

(A, B) UMAPs of snRNA-seq data showing: (A) number of genes detected as expressed and (B) Number of unique molecular identifiers (UMIs) detected per nucleus. (C) Plot of average number of genes expressed (dot size) and UMIs detected per nucleus per germline-annotated Leiden 6.0 cluster, ordered by estimated progression of germ line differentiation. (D, E) UMAPs of snRNA-seq data showing average expression of: (D) Y chromosome tor (E) X chromosome encoded transcripts relative to an expression-matched control set (gene sets with binned expression matching transcript lists). Note the dramatically different relative expression scales in (D) vs (E) UMAPs taken directly from ASAP (Li et al., 2022; Gardeux et al., 2017). (F, G) UMAP plots of snRNA-seq raw counts (log-transformed) showing expression of: (F) Hemolectin (Hml) in hemocytes and late spermatocytes and (G) Myosin heavy chain (Mhc) in muscle and late spermatocytes. Yellow: relative expression high. (H) Testis hybridized in situ with biotinylated antisense RNA probe to Hml, showing expression (blue) in spermatocytes. Scale bar, 200 µm. (I) Dot Plot showing expression of pairs of tissue specific markers across cell types in the testis plus seminal vesicle snRNA-seq sample. Average expression of each gene in a given cell type (here, a cluster at Leiden 0.4 resolution) denoted by color intensity. Percent of nuclei of the given cell type scored as expressing each gene denoted by dot size. Colors along axes (see Figure 1B) indicate the relevant cell type for tissue-specific marker pairs. (see also Figure 3—figure supplements 1 and 2 and Figure 3—source data 1).

Figure 3—figure supplement 1. Magnitude of the mid-to-late spermatocytes transcription program.

(A) Average UMI and numbers of genes expressed per cluster (ordered by cluster number In Li et al., 2022) inFCA snRNA-seq data from testis, malpighian tubules, heart, and male reproductive tract. Dot size proportional to the number of genes detected as expressed (aveGenes). Note difference between panels in dot size scale bars. Across these tissues, Mid-to-late spermatocytes had the highest average UMI and number of genes expressed per cluster. The highest fat body signal was found to be due to contaminating spermatocytes, and is therefore not relevant here. In the FCA dataset from Li et al., 2022, Mid-to-late spermatocytes are represented by clusters 18, 29, 30, 31, 32; in our dataset, these correspond to clusters 51, 35, 33, 48, and some of 105 (Figure 2A and Figure 2—figure supplement 1A). (B) Germ line portion of UMAP from single cell sequencing (scRNAseq) of adult testes showing number of genes detected as expressed per cell. (C) Violin plot of average number of UMIs per cell type for scRNA-seq data.

Figure 3—figure supplement 2. Markers of other tissues also expressed in mid-to-late spermatocytes.

(A–D) UMAPs of snRNA-seq raw counts (log-transformed) showing expression of: (A) Synaptotagmin 1 (Syt1); (B) grainy head (grh); (C) eyes absent (eya); (D) unpaired 1 (upd1). Yellow: relative expression high. (E) Dot Plots showing expression of Mhc, Hml, Syt1, and grh in spermatocytes detected in single cell RNA-seq (scRNA-seq) analysis of adult testes. (F) Testis from a male mutant for the tMAC subunit comr hybridized in situ with biotinylated antisense RNA probe to Hml, showing expression in spermatocytes. Scale bar, 200 µm. (G) UMAP of snRNA-seq raw counts (log-transformed) showing expression of geko. (H) Testis showing Visual system homeobox 1 (Vsx1) mRNA (FISH, green), Coracle (Cora) protein (red), and DNA (blue). White arrows: somatic cyst cells, white arrowheads: spermatocytes. Scale bar, 100 µm. (I) UMAP of snRNA-seq raw counts (log-transformed) showing expression of Vsx1.