Disruption of CHTF18 causes defective meiotic recombination in male mice

Abstract

CHTF18 (chromosome transmission fidelity factor 18) is an evolutionarily conserved subunit of the Replication Factor C-like complex, CTF18-RLC. CHTF18 is necessary for the faithful passage of chromosomes from one daughter cell to the next during mitosis in yeast, and it is crucial for germline development in the fruitfly. Previously, we showed that mouse Chtf18 is expressed throughout the germline, suggesting a role for CHTF18 in mammalian gametogenesis. To determine the role of CHTF18 in mammalian germ cell development, we derived mice carrying null and conditional mutations in the Chtf18 gene. Chtf18-null males exhibit 5-fold decreased sperm concentrations compared to wild-type controls, resulting in subfertility. Loss of Chtf18 results in impaired spermatogenesis; spermatogenic cells display abnormal morphology, and the stereotypical arrangement of cells within seminiferous tubules is perturbed. Meiotic recombination is defective and homologous chromosomes separate prematurely during prophase I. Repair of DNA double-strand breaks is delayed and incomplete; both RAD51 and γH2AX persist in prophase I. In addition, MLH1 foci are decreased in pachynema. These findings demonstrate essential roles for CHTF18 in mammalian spermatogenesis and meiosis, and suggest that CHTF18 may function during the double-strand break repair pathway to promote the formation of crossovers.

Figure 1. Derivation of Chtf18 ^−/− and Chtf18^flox/flox mice.

(A) Gene targeting strategy (exons are not drawn to scale). LoxP sites are represented by triangles. Diamonds represent FRT sites. (B) Southern blot analysis of targeted ES cells (lanes 1, 2, and 3) and wild-type ES cells (lane 4). The probes used for Southern blot analysis are depicted as SP3 and neo. (C) Western blot analysis of testis protein extracts from Chtf18-null mice with CHTF18 antibody. A ß-actin specific antibody was used as a control.

Figure 2. Chtf18 ^−/− mice are born at submendelian ratios.

Data collected from Chtf18^+/− intercross matings. (A and B) Numbers of expected and observed Chtf18 ^+/+, Chtf18 ^+/−, and Chtf18 ^−/− offspring and embryos (14.5–18.5 dpc), respectively. (C) Histograms showing observed/expected ratios of Chtf18 ^+/+, Chtf18 ^+/−, and Chtf18 ^−/− embryos at 14.5–18.5 dpc. Values are the mean ratio ± SEM; p≤0.007 using ANOVA.

Figure 3. Chtf18 ^−/− testes are significantly smaller than those of wild-type mice.

(A) Dramatic size reduction in Chtf18 ^+/+ and Chtf18 ^−/− adult testes. Scale bar represents 2 mm. (B) Histograms showing testis weight per body weight for Chtf18 ^+/+ and Chtf18 ^−/− mice (mg/g, means ± SEM) Chtf18 ^+/+ 6.98±0.30, n = 9 mice; Chtf18 ^−/− testes 3.03±0.32, n = 8 mice, P<0.0001 using the Student's t-test.

Figure 4. Impaired spermatogenesis in Chtf18 ^−/− mice.

Seminiferous tubules stained with hematoxylin and eosin from 6–12 week old wild-type (A) and Chtf18 ^−/− testes (B–F) are shown. Chtf18 ^−/− tubules show a range of abnormalities. A paucity of spermatids and spermatozoa (asterisks in C) or almost complete absence of spermatogenic cells and large vacuoles (B) are seen, although a few Chtf18 ^−/− tubules demonstrate almost normal morphology (D). Disorganized testes morphology and aberrant spermatogenic cells (E and F) are also seen. Spermatogenic cells appear to be in pachynema, P, are located at the lumen, and arrows in E and F indicate aberrant cells in Chtf18 ^−/− seminiferous tubules. Scale bar is 25 µm.

Figure 5. Oligospermia and decreased fertility in Chtf18-null mice.

(A) Histograms showing caudal epididymal sperm concentrations from 10–12 week old Chtf18 ^−/− mice and Chtf18 ^+/+ controls. Values are the number of sperm ×10⁶ per ml, means ± SEM, Chtf18 ^+/+ 39.60±7.71, n = 9 mice; Chtf18 ^−/− 6.90±3.04, n = 8 mice, P≤0.002 using Student's t-test (B) Histograms showing the number of offspring from wild-type females bred with Chtf18 ^−/− and Chtf18 ^+/+ male mice over a 5 month period. Each male was paired with two wild-type females (number of offspring from 12 pairs of wild-type females mated with Chtf18 ^+/+ and Chtf18 ^−/− males, means ± SEM, Chtf18 ^+/+ 77.33±7.36, n = 3 males; Chtf18 ^−/− 21.00±10.97, n = 3 males, P≤0.0130 using Student's t-test) (C) TUNEL staining of seminiferous tubule sections from three Chtf18 ^−/− and three wild-type 7 week old mice demonstrates that apoptotic cells are increased in Chtf18 ^−/− tubules. Scale bar is 25 µm. (D) Histograms showing mean apoptotic cells per tubule ± SEM; wild-type (+/+) 3.083±0.197; N = 145 vs. Chtf18 ^−/− (−/−) 7.658±0.477; N = 155; p<0.0001 (E and F) Tubules from 13–14 week old Chtf18^flox/−; TNAP Cre (cKO) mice stained with hematoxylin and eosin demonstrate a morphological phenotype that is indistinguishable from Chtf18 ^−/− tubules (Figure 4). Scale bars are 25 µm and 50 µm for E and F, respectively.

Figure 6. The number of prospermatogonia is decreased in Chtf18 ^−/− neonatal testes.

(A and B) Testis sections from postnatal day 3 Chtf18-null and wild-type males were stained with anti-MVH (green) and DAPI (blue). Scale bars are 50 µm. (C) Histogram showing the mean number of MVH-positive cells per seminiferous tubule ± SEM; wildtype (+/+) 2.399±0.097; N = 278 tubules and Chtf18 ^−/− (−/−) 1.750±0.090; N = 280 tubules; p<0.0001 using Student's t-test (D) Histogram showing the number of seminiferous tubules completely lacking prospermatgonia in Chtf18 ^−/− and wild-type testes; wild-type (+/+) 30, N = 278 and Chtf18 ^−/− 61, N = 280, p = 0.0005 using Fisher's exact test.

Figure 7. Formation of univalent chromosomes in Chtf18 ^−/− spermatocytes.

Meiotic chromosome spreads from Chtf18 ^−/− juvenile males at postnatal day 21 were stained by immunofluorescence with anti-SYCP1 serum (green), anti-SYCP2 serum (red), and DAPI (blue) (A–D). SYCP1 localizes to central elements of the synaptonemal complex and SYCP2 localizes to the axial/lateral elements of the synaptonemal complex. The leptone, zygotene, pachytene, and diplotene stages of prophase I Chtf18 ^−/− spermatocyte (−/−) are shown (A, B, C, and D respectively). Arrows in D reveal the presence of univalent chromosomes in the diplotene stage of prophase I. Wild-type and Chtf18 ^−/− spermatocytes from juvenile males at postnatal day 21 (E and F, respectively) were stained with CREST autoimmune serum, which stains centromeres, and anti-SYCP3, which stains the axial/lateral elements of the synaptonemal complex. Arrows in F indicate univalent chromosomes in a Chtf18 ^−/− diplotene spermatocyte. Metaphase I spread analysis of spermatocyte nuclei from juvenile males at postnatal day 21 stained with Giemsa (G and H). Twenty bivalent chromosomes are present in the wild-type metaphase I spermatocyte (G). A Chtf18 ^−/− metaphase I spermatocyte (H) demonstrates the presence of four univalent chromosomes (arrows in H).

Figure 8. Meiotic recombination is defective in Chtf18 ^−/− spermatocytes.

Meiotic chromosome spreads from wild-type (A–D) and Chtf18 ^−/− (E–H) juvenile males at postnatal day 14 or 21 were stained with anti-SYCP3 (red) and γH2AX (green). A & E represent leptonema; B & F represent zygonema; C & G represent pachynema; D & H represent diplonema. Arrows in G and H show persistence of γH2AX in Chtf18 ^−/− spermatocytes. Meiotic chromosome spreads from wild-type (I and L) and Chtf18 ^−/− (J and M) males were stained with anti-SYCP3 (green) and anti-RAD51 (red). I and J represent zygonema; L and M represent pachynema. RAD51 focus counts per nucleus in zygonema (K and O) and pachynema (N and O) are shown. Meiotic chromosome spreads from wild-type and Chtf18 ^−/− spermatocytes were stained with anti-SYCP3 (red) and anti-MLH1 (green) (P and Q). Arrows indicate homologues without MLH1 foci. MLH1 focus counts per nucleus are shown in R.