Pumilio 1 suppresses multiple activators of p53 to safeguard spermatogenesis

Abstract

During spermatogenesis, germ cells initially expand exponentially through mitoses. A majority of these cells are then eliminated through p53-mediated apoptosis to maintain germline homeostasis. However, the activity of p53 must be precisely modulated, especially suppressed in postmitotic spermatogenic cells, to guarantee robustness of spermatogenesis. Currently, how the suppression is achieved is not understood. Here, we show that Pumilio 1, a posttranscriptional regulator, binds to mRNAs representing 1,527 genes, with significant enrichment for mRNAs involved in pathways regulating p53, cell cycle, and MAPK signaling. In particular, eight mRNAs encoding activators of p53 are repressed by Pumilio 1. Deleting Pumilio 1 results in strong activation of p53 and apoptosis mostly in spermatocytes, which disrupts sperm production and fertility. Removing p53 reduces apoptosis and rescues testicular hypotrophy in Pumilio 1 null mice. These results indicate that key components of the p53 pathway are coordinately regulated by Pumilio 1 at the posttranscriptional level, which may exemplify an RNA operon.

Figure 1. Pum1 expression and function in the testis

(A)qRT-PCR analysis of Pum1 mRNA levels in different organs and tissues, as normalized to Gapdh. ESCs: embryonic stem cells; BMW: bone marrow. Bars indicate SD (B) Immunofluorence microscopy of Pum1 in wild-type testis thin sections. Sg: spermatogonium; Sc: spermatocyte; Sd: spermatid. (C) Immunoblot analysis of Pum1 in testicular cytoplasmic and nuclear fractions. Nu: nucleus; Cy: cytoplasm. (D) Mature sperm counts in wild-type and Pum1 mutant males at different ages, with bars indicating SEM. (E) Size of litters sired by wild-type and Pum1 mutant males, with bars indicating SEM. (F and G) Testicular histology revealed by H&E staining in wild-type (F) and Pum1 null (G) mice. Arrow points to a spermatogenic cell undergoing cell death. (H) Quantification of apoptotic cells in adult wild-type and Pum1 mutant testes. Apoptosis was stained by TUNEL analysis. For each mouse analyzed, apoptotic cells were counted in 80 random seminiferous tubules sections and presented in (H) as numbers per 10 random tubules cross sections, with bars showing SD. *** indicates P<0.001 in (H). Scale bars in (B) and (F) both represent 50μm.

Figure 2. Pum1 binds to and represses mRNA targets via a defined binding motif

(A) mRNAs associated with Pum1. Three independent experiments using wild-type testicular lysates and two mock experiments using Pum1^−/− testicular lysates are shown. The Pum1 target mRNAs were co-immunoprecipitated by an anti-Pum1 antibody. mRNAs associated with Pum1 were analyzed by mouse cDNA microarrays. ENS: Ensembl; FDR: False Discovery Rate. (B) The consensus sequence among Pum1-associated mRNAs. A de novo sequence motif search was performed using the top 124 Pum1-assocated mRNAs by Multiple Expectation Maximization for Motif Elicitation (MEME). (C) Number of motifs per one kilobase of sequence in the 5′UTRs, CDS and 3′UTRs of Pum1 target mRNAs. (D and E) Expression levels of select Pum1-associated mRNAs and their protein levels in wild-type, Pum1^+/− and Pum1^−/− testes. Blue lines indicate non-Pum1 targets; red lines indicate Pum1 targets that are p53 regulators; and orange lines indicate Pum1 targets that are not p53 regulators. Bars in (D) indicate SEM.

Figure 3. Pum1 deficiency leads to activation of p38 MAPK, JNK and p53

(A) Proteins encoded by Pum1 target mRNAs activate p53 through the MAPK signaling pathway. (B) Immunoblot analysis of total and activated levels of p38, JNK and Map3k1 in total testicular lysates. While the total protein levels of p38 and JNK remain unchanged, phosphorylated p38 (Thr180/Tyr182) and JNK (Thr183/Tyr185) are increased in Pum1^−/− testes. Map3k1 is a Pum1 target that is upstream of both p38 and JNK. Both its total and phosphorylated (Thr1383) forms are elevated in Pum1^−/− testes.( C–H) Immunofluorescence microscopy of activated p38 in testicular thin sections. DAPI staining (E and F) provides nucleus morphology that enables identification of spermatogonium (Sg), spermatocyte (Sc), and spermatid (Sd). The strongest elevation of p38 activation in Pum1^−/− testes occurs in primary spermatocytes (C and D also G and F as merged images with DAPI). (I and J) Immunofluorescence microscopy of activated p53 (Serine 15) in testicular thin sections. White arrow heads indicate some of spermatocytes with activated p53. Extratubular green fluorescence in (I and J) is non-specific background from Leydig cells. Scale bars in (C), (E),(G) and (I) all represent 20 μm.

Figure 4. Inhibition of p38 reduces p53 activation and removing p53 rescues Pum1 phenotype

(A–C) Effects of p38 inhibition on p53 over-activation in Pum1^−/− testes (D–F) H&E staining of thin-sectioned testes from wild-type (D), Pum1^−/− (E) and Pum1^−/− ;Trp53^−/− testes at 15 dpp (F). (G) Removing p53 reduces apoptosis in Pum1^−/− testes at 15 dpp. Apoptotic cells were counted in 80 random seminiferous tubules sections and presented in (D) as numbers per 10 random tubules cross sections, with bars showing SEM. (H) Removing p53 rescues the testis hypotrophy in Pum1^−/− mice at 15 dpp. Bars indicate SEM. (I) A model illustrating that Pum1 represses eight mRNAs that encode activators of p53. Pum1 binds to these different mRNAs through a conserved binding motif in their 3′UTRs. Scale bars in (C) and (D) represent 5 and 50 μm for (A–C), and (D–F), respetively.