RNA Binding Protein Ptbp2 Is Essential for Male Germ Cell Development

Abstract

RNA binding proteins (RBPs) are increasingly recognized as essential factors in tissue development and homeostasis. The polypyrimidine tract binding (PTB) protein family of RBPs are important posttranscriptional regulators of gene expression. In the nervous system, the function and importance of PTB protein 2 (Ptbp2) as a key alternative splicing regulator is well established. Ptbp2 is also abundantly expressed during spermatogenesis, but its role in this developmental program has not been explored. Additionally, the importance of alternative splicing regulation in spermatogenesis is unclear. Here, we demonstrate that Ptbp2 is essential for spermatogenesis. We also describe an improved dual fluorescence flow cytometry strategy to discriminate, quantify, and collect germ cells in different stages of development. Using this approach, in combination with traditional histological methods, we show that Ptbp2 ablation results in germ cell loss due to increased apoptosis of meiotic spermatocytes and postmeiotic arrest of spermatid differentiation. Furthermore, we show that Ptbp2 is required for alternative splicing regulation in the testis, as in brain. Strikingly, not all of the alternatively spliced RNAs examined were sensitive to Ptbp2 loss in both tissues. Collectively, the data provide evidence for an important role for alternative splicing regulation in germ cell development and a central role for Ptbp2 in this process.

FIG 1

Conditional inactivation of Ptbp2 expression in mouse testis. (A) Ptbp2 levels during postnatal testis development. Western blot analysis of WT testis extract prepared from testes collected on the indicated day postpartum (dpp). (B) Early postnatal decline in Ptbp2 levels in cKO testes. Analysis of Ptbp2 levels in WT and cKO testis extract, with the germ cell-specific RBP Dazl serving as a loading control. (C) A low level of Ptbp2 persists in cKO testes. Analysis of Ptbp2 levels in two WT and two cKO P25 testes, compared to a serial dilution of WT testis. Hsp90 serves as a loading control. (D) Comparison of ratios of average testis weight to whole body weight in WT and cKO mice, with photographs of 42-dpp testes displayed at right (bars in centimeters).

FIG 2

Spermatogenic arrest and germ cell loss in cKO testes. (A to D) Hematoxylin and eosin staining of WT testis at 42 dpp (A) and cKO testes at 42 dpp (B), 3 months (C), and 6 months (D). Black arrowheads indicate MNCs, red arrowheads indicate MNCs with chromatin condensation, green arrowheads represent MNCs with pyknotic nuclei, and arrows indicate vacuoles. (E and F) Hematoxylin and eosin staining of cauda epididymis from 24-week-old WT (E) and cKO (F) mice. (G) Higher-magnification image of the box in panel F.

FIG 3

Spermatid arrest and multinucleate cell formation occur in the early steps of differentiation. (A) Quantification of spermatids in seminiferous tubules from 12 different animals collected on the indicated days (three animals per time point). The graph at left indicates the percentages of tubule cross sections containing round, multinucleate, or elongating spermatids. Pie charts (right) indicate proportions of spermatid-containing tubules with spermatids in the four stages of acrosome biogenesis: (1) Golgi stage, (2) cap, (3) acrosome, and (4) maturation. (B) Detection of spermatids in different stages of acrosome development in 33-dpp WT testis by PAS and hematoxylin staining. (i) Proacrosomal granules; (ii) acrosomal granules that have attached to spermatid nuclei; (iii) acrosomal granules that have attached to and flattened against spermatid nuclei; (iv) elongating spermatids with acrosomes extending over the nuclei. (C) PASH staining of 33-dpp cKO testis shows punctate PAS-positive acrosomal granules associated with spermatid nuclei and accumulation of PAS-positive material in the cytoplasm of multinucleated cells. Also evident are spermatid nuclei with crescent-shaped clearings indicative of chromatin condensation (red arrowheads) and MNCs that have undergone cell shrinkage and contain nuclei with more advanced condensation (blue arrowheads). Early-stage spermatids that are PAS negative and have not yet become MNCs are indicated with white arrowheads, while PAS-positive material accumulating in the center of MNCs is indicated with arrows.

FIG 4

Separation of Ho-stained GFP^pos germ cells by flow cytometry. (A) Schematic of the IRG transgene before and after Cre-lox recombination. (B) Western blot analysis of IRG transgene expression in brain, liver, and testes from mice containing the Stra8-iCre and/or IRG transgene. As a positive control for GFP expression, brain lysate from a Ptbp2^+/pDLTV1 mouse was included. The Ptbp2^pDLTV1 allele was generated by homologous recombination to replace essential coding sequence from exon 1 with coding sequence for enhanced GFP (13). Consistent with the work of De Gasperi et al. (46), RFP expression was detected in brain and absent in liver. Importantly, RFP expression was confirmed in testis of IRG⁺ (lane 9) and Stra8-iCre⁺;IRG⁺ (lane 7) mice, while GFP expression was detected in testes of Stra8-iCre⁺;IRG⁺ mice (lane 7). Hsp90 serves as a loading control. (C) Distribution of Hoechst 33342-stained cells from IRG⁺ WT testis. 1C, 2C, and 4C cells segregate as three distinct bands with increasing blue and red fluorescence. Panels correspond to all cells (i), GFP^pos cells (ii), and GFP^neg cells (iii). GFP^pos cells cluster into 5 subpopulations, previously demonstrated to correspond to (1) spermatogonia, (2) early-prophase spermatocytes, (3) late-prophase I spermatocytes, (4) secondary spermatocytes, and (5) spermatids. Population 6 (iii) corresponds to diploid GFP^neg cells with blue and red fluorescence values that overlap those of GFP^pos population 1. (D) Quantitative RT-PCR analysis of germ cell transcripts (Rhox13, Stra8, Dazl, and Mvh relative to Actb transcripts) and the somatic cell transcript Gata4 (relative to Actb transcripts) in GFP^pos population 1 cells versus GFP^neg population 6 cells.

FIG 5

Flow cytometry of Ho-stained cells from IRG⁺ WT and cKO testes. (A) Average percentages of GFP^pos and GFP^neg cells in testes from two WT and two cKO 9-week-old mice. (B and C) Western blot analyses of Ptbp2 levels in two IRG⁺ cKO animals at 21 dpp (B, lanes 7 and 8) and two IRG⁺ cKO animals at 25 dpp (C, lanes 7 and 8). Lanes 1 to 6 represent serial dilutions of WT testis lysate collected from 21- and 25-dpp mice (B and C, respectively). (D) Distribution of all cells (5,000 events) (i and iv), GFP^pos cells (ii and v), and GFP^neg cells (iii and vi) from IRG⁺ WT (i to iii) and IRG⁺ cKO (iv to vi) testes. (E) Distribution of GFP^pos cells in different stages of spermatogenesis based on blue fluorescence values from data presented in panel D. (F) Quantification of WT and cKO germ cells in different GFP^pos populations, relative to population 1 (mitotic) cells.

FIG 6

Increased apoptosis of spermatocytes in cKO testes. TUNEL staining of WT (A) and cKO (B and C) testes at 33 dpp. Arrows denote TUNEL-negative MNCs.

FIG 7

Ptbp2 is required for AS regulation in the testis. (A) RT-PCR analysis of AS RNAs from the Wdr7 (i), Stam2 (ii), and Actn1 (iii) genes in testes from 21-dpp WT and cKO mice without (lanes 1 and 2) and with (lanes 3 and 4) the IRG transgene. Lanes 5 and 6 correspond to embryonic brain from embryonic day 18.5 (E18.5) WT and Ptbp2-deficient mice generated by timed matings of Ptbp2^+/pDLTV1 animals (13). (B) RT-PCR analysis of AS RNA from the Dzip1 (i) and Pum2 (ii) genes in testes from two biological replicates of 25-dpp WT and cKO testes (lanes 1 and 2 and lanes 3 and 4, respectively), with embryonic brain included in the top panel (Dzip1), as described above. (C) Quantitative RT-PCR analysis of Pgk2 RNA levels (relative to Actb) in early- and late-prophase-I spermatocytes (GFP^pos populations 2 and 3, respectively) purified by flow cytometry of Ho-stained IRG⁺ WT and IRG⁺ cKO testes. FACS, fluorescence-activated cell sorting; MI, meiosis I.