Inactivation of the Celf1 gene that encodes an RNA-binding protein delays the first wave of spermatogenesis in mice

Abstract

Background: The first wave of spermatogenesis in mammals is characterized by a sequential and synchronous appearance of germ cells in the prepubertal testis. Post-transcriptional controls of gene expression play important roles in this process but the molecular actors that underlie them are poorly known.

Methodology/principal findings: We evaluated the requirement for the RNA-binding protein CELF1 during the first wave of spermatogenesis in mice. Mice inactivated for Celf1 gene were not viable on pure genetic backgrounds. On a mixed background, we observed by histology and gene profiling by RT-qPCR that the testes of inactivated prepubertal mice were characterized by several features. (i) Spermiogenesis (differentiation of post-meiotic cells) was blocked in a subset of prepubertal inactivated mice. (ii) The appearance of the different stages of germ cell development was delayed by several days. (iii) The expression of markers of Leydig cells functions was similarly delayed.

Conclusions/significance: Celf1 disruption is responsible for a blockage of spermiogenesis both in adults and in prepubertal males. Hence, the spermiogenesis defects found in Celf1-inactivated adults appear from the first wave of spermiogenesis. The disruption of Celf1 gene is also responsible for a fully penetrant delayed first wave of spermatogenesis, and a delay of steroidogenesis may be the cause for the delay of germ cells differentiation.

Figure 1. Incompletely penetrant blockage of spermiogenesis in Celf1^−/− mice during the first wave of spermatogenesis.

Representative histological sections of testes of +/+ (left panels) and −/− (two right panels) mice at 42 days post-partum (dpp). A–C, seminiferous tubules. D–F, epididimes of the same respective mice. Depending on the presence of elongated spermatids in seminiferous tubules, KO mice were classified as “non affected” (B, E) or “affected” (C, F). Sg, Spermatogonia; PS, Pachytene Spermatocyte; RS, Round Spermatid; ES, Elongated Spermatid; Spz, Spermatozoa; *, multinuclear giant cell. Scale bars, 100 µm (A–C), 80 µm (D–F).

Figure 2. The first wave of spermatogenesis is delayed in Celf1^−/− mice.

Representative histological sections of testes of homozygous Celf1 ^+/+ mice (left panels) and Celf1 ^−/− mice (right panels) at different prepubertal ages. A, 7 dpp; B, 15 dpp; C, 24 dpp; D, 35 dpp. G, Gonocyte; Sg, Spermatogonia; St, Sertoli cell; ZS, Zygotene Spermatocyte; PS, Pachytene Spermatocyte; RS, Round Spermatid; ES, Elongated Spermatid. Scale bars, 50 µm.

Figure 3. Quantification of the delay of the first wave of spermatogenesis. A

, For each mouse, we analysed 300 seminiferous tubules in 15 testis sections. We classified each tubule according to the most differentiated germ cells that it contained (ZS, Zygotene Spermatocyte; PS, Pachytene Spermatocyte; RS, Round Spermatid; ES, Elongated Spermatid; Spz, Spermatozoa). Next, we calculated for each mouse the percentage of seminiferous tubules of each class, and we plotted the percentage of seminiferous tubules of the most advanced class against the age. Blue diamonds and orange squares correspond to individual Celf1 ^+/+ and Celf1 ^−/− mice respectively. B, We quantified by real-time RT-PCR the relative amounts of the indicated mRNAs (Mouse Vasa Homolog, Ddx4; Proacrosine, Acr; Heat Shock Protein 70.2, Hspa2; Lactate Dehydrogenase 3, Ldhc; Miwi, Piwil1) for +/+ and −/− testes of similar stages of spermatogenesis but different ages: lower panels and inserts, 17 dpp +/+ (blue bars) and 24 dpp −/− (orange bars) testes with 20% of seminiferous tubules of the PS class; upper panels, 24 dpp +/+ (blue bars) and 42 dpp −/− (orange bars) testes with 60% of seminiferous tubules of the RS class. Results are expressed as the means of 3–5 animals for each genotype. Error bars are standard deviations. We used a Student’s t-test to statistically compare the +/+ and the −/− genotypes, and the p-values below 0.1 are given on the right of the corresponding bars.

Figure 4. The expression of Leydig cells markers is delayed in prepubertal Celf1 ^−/− mice.

We quantified by real-time RT-PCR the relative amounts of the indicated mRNAs in +/+ (blue diamonds) and −/− (orange squares) testes at the indicated prepubertal ages. NA (not affected) and A (affected) refer to 42 dpp −/− mice with and without elongated spermatids respectively based on histological analyses. A, Lhr. B, The main pathway of steroidogenesis in rodents and the corresponding enzymes . Hsd3b1 and Hsd3b6 have different expression patterns but encode enzymes with similar activities . Results are expressed as the means of 3–5 animals for each age and genotype. Error bars are standard deviations. We used a Student’s t-test to statistically compare the different genotypes of identical ages, and we show the p-values below 0.1 above the corresponding symbols.