Knockout of cyclin-dependent kinases 8 and 19 leads to depletion of cyclin C and suppresses spermatogenesis and male fertility in mice

Abstract

CDK8 and CDK19 paralogs are regulatory kinases associated with the transcriptional Mediator complex. We have generated mice with the systemic inducible Cdk8 knockout on the background of Cdk19 constitutive knockout. Cdk8/19 double knockout (iDKO) males, but not single Cdk8 or Cdk19 KO, had an atrophic reproductive system and were infertile. The iDKO males lacked postmeiotic spermatids and spermatocytes after meiosis I pachytene. Testosterone levels were decreased whereas the amounts of the luteinizing hormone were unchanged. Single-cell RNA sequencing showed marked differences in the expression of steroidogenic genes (such as Cyp17a1, Star, and Fads) in Leydig cells concomitant with alterations in Sertoli cells and spermatocytes, and were likely associated with an impaired synthesis of steroids. Star and Fads were also downregulated in cultured Leydig cells after iDKO. The treatment of primary Leydig cell culture with a CDK8/19 inhibitor did not induce the same changes in gene expression as iDKO, and a prolonged treatment of mice with a CDK8/19 inhibitor did not affect the size of testes. iDKO, in contrast to the single knockouts or treatment with a CDK8/19 kinase inhibitor, led to depletion of cyclin C (CCNC), the binding partner of CDK8/19 that has been implicated in CDK8/19-independent functions. This suggests that the observed phenotype was likely mediated through kinase-independent activities of CDK8/19, such as CCNC stabilization.

Keywords: CDK19; CDK8; Spermatogenesis; developmental biology; mouse.

Figure 1.. Changes in male urogenital system in Cdk8/19 knockout mice.

(A) Crossing of Cdk8^fl/fl, Cdk19^-/- and Cre/ER^T2 mice and formation of experimental (Cdk8^fl/fl/Cdk19^-/-/ ROSA26^CreERT2+tamoxifen, Cdk8^fl/fl/ROSA26^CreERT2+tamoxifen, and Cdk19^-/-) and control (Cdk8^fl/fl/Cdk19^-/-/ROSA26^CreERT2 without tamoxifen and wild-type +tamoxifen) groups. (B) Confirmation of tamoxifen-induced CDK8 iKO in testes by Western blot. (C) After 2 mo of KO induction, iDKO mice had significantly lower body weight [nonparametric t-test, ***p<0.001, n=6]. (D) Male urogenital system atrophy in iDKO mice. (E) Sexual behavior and fertility of tamoxifen-treated control, single KOs, and iDKO male mice.

Figure 1—figure supplement 1.. Genotype and phenotype confirmation.

(A) Crossing of Cdk8^fl/fl, Cdk19^-/-, and Cre/ER^T2 mice. The red square indicates mice that were maintained in a homozygous state. (B) Proof of Cdk8 2^nd exon excision by PCR and agarose gel electrophoresis as described in Ilchuk et al., 2022 Cre and Cdk8^fl/wt probes on first and third lanes for comparison, mQ as negative control. (C) Confirmation of CDK8 and Cyclin C absence in the intestine 2 mo after inducible double knockout (iDKO) induction. (D) Periodic acid–Schiff stained sections of intestine of tamoxifen-treated R26/Cre/ER^T2 (Cre +Tam) and iDKO mice. Red arrows indicate goblet cells, black arrows indicate Paneth cells. Both cell types numbers are significantly reduced in iDKOs. Magnification 100 X [nonparametric t-test, n=173 and 178 crypt and n=240 and 261 villus for control and iDKO respectivelly, ****p<0.0001].

Figure 2.. Cdk8/19 knockout blocks spermatogenesis in mice.

(A) H&E staining of prostate, epididymis, and testes of inducible double knockout (iDKO) mice and tamoxifen-treated control. 100 X magnification (B) H&E staining of wild-type (WT) and iDKO seminiferous tubules. 400 X magnification. (С) Time course of experiments. CDK8 iKO was activated by tamoxifen administration in 8–10 wk old males. Urogenital abnormalities became visible in 2 wk. Spermatogenesis was analyzed by flow cytometry and immunofluorescence (IF) after 2, 8, and 20 wk since activation. Single-cell RNA sequencing was performed at 7 wk after KO. (D) Western blot analysis of Cre/ERT2 (Cre+Tam), single (CDK8 iKO and CDK19 KO), and double (iDKO) knockout testes, 2 mo after tamoxifen injections. CCNC protein is absent in iDKO, but not in the single KO in the testes. pSTAT1 727 is independent of CDK8/19 KO. Stars mark nonspecific staining by CDK19 antibodies. (E) CDK8 and CDK19 IHC staining of testes sections, 200 x (upper row) and 630 x magnification (bottom row) showing staining in various types of testicular cells. (F–K). Flow cytometry analysis of CDK8/CCNC expression in different testicular cell types. Figures F and I show major CDK8 (50.68%) and CCNC populations (44.52%), figures G and J show that 1 n (round, but not elongated spermatids), 2 n and 4 n cells can be CDK8 and CCNC positive, figures H and K indicate, not only cKit⁺ cells among 2 n-4n can be CDK8 and CCNC positive.

Figure 2—figure supplement 1.. CDK8 and CDK19 are mutually compensatory for cyclin C stabilization and male urogenital system phenotype.

(A) H&E staining of testes and prostate of wild-type (WT) mice and CDK8 iKO mice. Despite the lack of sexual behavior CDK8 iKO mice have normal urogenital system. (B) Western blot of cultured mouse embryonic fibroblasts (MEFs) with knockout induced in vitro by 4-hydroxytamoxifen. Single KOs have compensatory elevated levels of paralogous kinase.

Figure 2—figure supplement 2.. CDK19 antibody specificity.

Figure 3.. Absence of postmeiotic 1 n cells in inducible double knockout (iDKO) 2 mo after KO induction.

(A) Distinctive histograms of wild-type (left) and iDKO (right) mice. Violet - 4 n population, red - 2 n population, green - round spermatids, blue - elongated spermatids, orange - apoptotic subG1 cells. (B) Quantitative distribution of testicular cells between these groups. Wild-type, with and without tamoxifen, as control groups, have similar distribution to that of CDK8 and CDK19 single KO. iDKO testes have a greatly reduced number of round spermatids and no elongated spermatids [repeated measurements two-way ANOVA, mean ± SD, n=5]. (C) Overall cellularity is significantly reduced only in the iDKO testes [nonparametric t-test, n > 3, ***p<0.001, ****p<0.0001]. (D) IF staining of the control and iDKO testes frozen sections. Nuclei are stained by DAPI (blue pseudocolor), SYCP3 is depicted as green, γH2A.X - as red. All stages of spermatogenesis are visible in control testes, while pachytene is the last detected stage in iDKO. Confocal microscopy, magnification 600 X. M - meiotic entry spermatocytes; L - leptotene; Ph - pachytene; PM - post-meiotic stages.

Figure 4.. Single-cell RNA sequencing reveals loss of spermatids due to steroidogenesis failure.

Raw data single-cell RNA sequencing (scRNA) sequencing are available in the SRA (SRP470231). (A) UMAP projection and relative cell numbers for all testicular cell types in control and iDKO samples. Number of secondary spermatocytes is significantly decreased and spermatids are almost absent in iDKO samples. (B) UMAP projection and relative cell numbers for spermatogonia and primary spermatocytes. Post-pachytene spermatocytes are severely depleted in iDKO samples. (C) GO Biological Processes pathways enriched among the Leydig cells differentially expressed genes (DEGs). Lipid metabolism and steroid biosynthesis are severely perturbed. (D) Violin plots for key Leydig cells genes.

Figure 4—figure supplement 1.. Flow cytometry analysis of propidium iodide stained cell fractions from testes of 2 wild-type (WT) and 2 double knockout (DKO) testes later subjected to single-cell RNA sequencing (scRNAseq).

DKO cells do not contain elongated spermatids and only small fraction of round spermatids.

Figure 5.. Inducible double knockout (iDKO) Sertoli cells re-enter the cell cycle and lose characteristic cytoskeleton organization.

(A) Violin plots for Reactome cell cycle gene sets indicate that Sertoli cells in iDKO lose terminal differentiation and re-enter cell cycle. Percentage of cells in G1-S and G2-M transitions are increased in iDKOs. (B) Violin plots for key cytoskeleton and intercellular contacts related to differentially expressed genes (DEGs). (C) Immunofluorescence (IF) staining for vimentin demonstrates the blood-testis barrier (BTB) integrity disruption and a loss of characteristic striation cytoskeleton patterns in iDKOs. Magnification 600 X. (D) Enrichment of GO stress pathways in Sertoli cells indicates their dysfunction in iDKOs.

Figure 5—figure supplement 1.. Venn diagram of SCARCO, SCARIBO и iDKO mice differentially expressed genes (DEGs).

Figure 5—figure supplement 2.. Volcano plots of differentially expressed genes (DEGs) of spermatogonia and spermatocytes clusters (А), Sertoli (B), and Leydig cells (C).

Graphs were made by GraphPad Prism.

Figure 5—figure supplement 3.. GO terms for spermatogonia depicted as a treemap.

Area of each square is proportional to the number of the genes in the term. The graph is built with R Treemap library.

Figure 6.. Confirmation of single-cell RNA sequencing (scRNA) sequencing data by other methods.

(A) Immunofluorescence (IF) staining for CYP17A1 of testes frozen sections, magnification 600 X. CYP17A1 is visualized in extratubular space in Leydig cells in control mice and is completely absent in inducible double knockout (iDKOs). (B) Western blot for CYP17A1 confirms disappearance of the protein in iDKOs, but not in other genotypes [nonparametric t-test, n=5, **p<0.01, ***p<0.001]. (C) Testosterone blood level is decreased only in iDKO mice. (D) Luteinizing hormone production is not impaired by CDK8 iKO and CDK19 KO or iDKO [nonparametric t-test, n=5].

Figure 6—figure supplement 1.. OilRed-stained frozen sections of testes of Cre+Tam and inducible double knockout (iDKO) mice 2 mo of tamoxifen injection.

Magnification 200 X (A) and 400 X (B). There increased lipid accumulation is observed in the interstitial space of testes in iDKO mice. To further prove the steroid hypothesis, we performed CYP17A1 immunostaining of testicular cross sections and western blotting, as well as direct testosterone measurement in the blood via LC-MS and cell culture experiments with ex vivo Leydig cells.

Figure 7.. Limited recovery of spermatogenesis, 5 mo after inducible double knockout (iDKO) induction.

(A) Round and elongated spermatids become detectable by flow cytometry 5 mo after iDKO [repeated measurements two-way ANOVA, mean ± SD, n=3]. (B) The overall testes cellularity is only slightly increased [nonlinear regression, mean ± SD, n=3]. (C) The postmeiotic cells become visible with H&E staining of the tubules, however, epididymal ducts remain empty. (D) Post-pachytene and post-meiotic (PM) cells became visible on the SYCP3 + γH2A.X-stained frozen sections, magnification 600 X. (E) CYP17A1 level remains at the background level, 5 mo after KO induction.

Figure 8.. The effects of CDK8/19 inhibitor on spermatogenesis in mice.

(A–C) Male C57BL/6 J mice were treated with SNX631-6 medicated chow (500 ppm, 40–60 mg/kg/d dosage, on average) for 3 wk. (A) Representative images of H&E histology analysis of the testicular tissues collected from animals in control or treated groups. (B) Organ weights of testes (left and right) at endpoint [unpaired t-test, ns – non significant, n=10]. (C) SNX631-6 concentrations in plasma and testicular tissues at endpoint [unpaired t-test, ns – non significant, n=10]. (D) qPCR analysis of steroidogenic Star and Fads genes in ex vivo cultured Leydig cells in response to the CDK8/19 inhibitor Senexin B (1 μM) or hydroxytamoxifen-induced CDK8/19 inducible double knockout (iDKO) [nonparametric t-test, n=3 for Star and n=2 for Fads].

Author response image 1.