Unraveling transcriptome dynamics in human spermatogenesis

Abstract

Spermatogenesis is a dynamic developmental process that includes stem cell proliferation and differentiation, meiotic cell divisions and extreme chromatin condensation. Although studied in mice, the molecular control of human spermatogenesis is largely unknown. Here, we developed a protocol that enables next-generation sequencing of RNA obtained from pools of 500 individually laser-capture microdissected cells of specific germ cell subtypes from fixed human testis samples. Transcriptomic analyses of these successive germ cell subtypes reveals dynamic transcription of over 4000 genes during human spermatogenesis. At the same time, many of the genes encoding for well-established meiotic and post-meiotic proteins are already present in the pre-meiotic phase. Furthermore, we found significant cell type-specific expression of post-transcriptional regulators, including expression of 110 RNA-binding proteins and 137 long non-coding RNAs, most of them previously not linked to spermatogenesis. Together, these data suggest that the transcriptome of precursor cells already contains the genes necessary for cellular differentiation and that timely translation controlled by post-transcriptional regulators is crucial for normal development. These established transcriptomes provide a reference catalog for further detailed studies on human spermatogenesis and spermatogenic failure.

Keywords: Gamete development; Human; RNA-binding proteins; RNA-sequencing; Spermatogenesis.

Fig. 1.

Scheme of spermatogenesis and stages of the seminiferous epithelium in humans. (A) A timeline showing the approximate relative time required for each developmental step of spermatogenesis. (B) Scheme of the various germ cell subtypes found within the testis. The germ cell subtypes collected for this study are boxed in blue. (C) An adapted scheme (Muciaccia et al., 2013) showing the stages of the seminiferous epithelium in man. Germ cell subtypes collected based on cellular association are depicted by the following colors: light purple (A_dark spermatogonia), dark purple (A_pale spermatogonia), pale green (leptotene/zygotene spermatocytes), bright green (early pachytene spermatocytes) and dark green (late pachytene spermatocytes).

Fig. 2.

Germ cell subtype identification. Microscopic images showing the similarities in morphology of the germ cell subtypes in this study in testis tissue prepared using an alcohol-based fixative (modified methacarn) compared with tissue prepared using the standard fixative diluted Bouin's.

Fig. 3.

Transcriptomic dynamics during spermatogenesis. (A-D) Spermatogenic phase analysis of in silico pooled data representing the pre-meiotic (average of A_dark and A_pale spermatogonia, n=11), meiotic (average of leptotene/zygotene, early and late pachytene spermatocytes, n=17) and post-meiotic (round spermatids, n=6) phases of spermatogenesis from six patients. (A) Mean number of transcripts expressed based on counts per million (mean±s.d.) in the pre-meiotic, meiotic and post-meiotic phases. *P<0.05. (B) An MDS plot showing the transcriptomic variability between pre-meiotic, meiotic and post-meiotic samples. (C) The number of differentially expressed genes (DEGs) and their direction of change (red arrow, upregulated; blue arrow, downregulated) found between the different phases of spermatogenesis (adjusted P-value≤0.05). (D) K-means clustering of 6717 DEGs across spermatogenesis. (E-H) Germ cell subtype analysis of data (mean±s.d.) obtained from the germ cell subtypes: A_dark (n=6), A_pale (n=5) spermatogonia, leptotene/zygotene (n=5), early pachytene (n=6), late pachytene (n=6) spermatocytes, round spermatids (n=6). (E) Mean number of transcripts expressed in the germ cell subtypes. *P≤0.05. (F) An MDS plot showing the transcriptomic variability between the different germ cell subtypes: spermatogonia (purple; circles), spermatocytes (green; squares), spermatids (red; triangles). (G) Differentially expressed genes (red arrow, upregulated; blue arrow, downregulated) between the germ cell subtypes (adjusted P-value≤0.05). (H) K-means clustering of 4622 DEGs across spermatogenesis.

Fig. 4.

Expression levels of selected pre-meiotic, meiotic and post-meiotic genes in man and mouse. (A,B) Heatmaps displaying the expression levels of genes involved in spermatogenic phases. (C) Expression levels of selected genes from previously described spermatogenic phases analyses of human spermatogonia, spermatocytes and spermatids (Zhu et al., 2016) compared with our spermatogenic phase (SP) and germ cell subtype (GC subtypes) data. (D,E) Heatmaps comparing our spermatogenic phase and germ cell subtype data with reported mouse data on spermatogonia (spg), spermatocytes (spc) and spermatids (spt) (Namekawa et al., 2006) (D) and leptotene/zygotene spermatocytes (LZ), pachytene spermatocytes (PS) and round spermatids (RS) (da Cruz et al., 2016) (E).

Fig. 5.

Gene expression profiles of meiotic, post-meiotic and PIWIL genes. (A,B) Bar charts showing expression levels represented by counts per million (log scale) (mean±s.d.) for meiotic (A) and post-meiotic (B) associated genes. (C) Expression measured by RT-PCR of four meiotic genes (SYCP3, DMC1, MSH4 and DMRTC2), four post-meiotic genes (ODF2, HOOK1, TCP11, PHF7) and a reference gene (EPN3) in RNA isolated from spermatogonia collected from a spermatogonial arrest patient (URO0074; lane 1), and spermatogonia collected from vasectomy reversal men used for this study (Stl141 and Stl142, respectively; lanes 2 and 3). Note that owing to a limited amount of RNA from Stl142 we were unable to analyze SYCP3, ODF2 and TCP11. (D) Bar chart showing expression level of genes (mean±s.d.) encoding the PIWIL proteins in each germ cell subtype during spermatogenesis.

Fig. 6.

Expression patterns of differentially expressed RNA-binding protein genes. (A) K-means clusters showing the expression pattern of 110 differentially expressed RNA-binding proteins. The predicted functional phase is indicated by a gray box and ‘PFP’. (B) Immunohistochemistry of five selected RNA-binding proteins on testicular tissue of a prostate cancer patient (URO0368), repeated on three biopsies. Scale bars: 50 µm. Colored lines in the k-means clusters indicate the expression pattern of the mentioned corresponding gene.