PBAF chromatin remodeler complexes that mediate meiotic transitions in mouse

Abstract

Gametogenesis in mammals encompasses highly regulated developmental transitions. These are associated with changes in transcription that cause characteristic patterns of gene expression observed during distinct stages of gamete development, which include specific activities with critical meiotic functions. SWI/SNF chromatin remodelers are recognized regulators of gene transcription and DNA repair, but their composition and functions in meiosis are poorly understood. We have generated gamete-specific conditional knockout mice for ARID2, a specific regulatory subunit of PBAF, and have compared its phenotype with BRG1 knockouts, the catalytic subunit of PBAF/BAF complexes. While Brg1Δ/Δ knockout acts at an early stage of meiosis and causes cell arrest at pachynema, ARID2 activity is apparently required at the end of prophase I. Striking defects in spindle assembly and chromosome-spindle attachment observed in Arid2Δ/Δ knockouts are attributed to an increase in aurora B kinase, a master regulator of chromosome segregation, at centromeres. Further genetic and biochemical analyses suggest the formation of a canonical PBAF and a BRG1-independent complex containing ARID2 and PBRM1 as core components. The data support a model in which different PBAF complexes regulate different stages of meiosis and gametogenesis.

Keywords: Meiosis; Mouse gametogenesis; PBAF; SWI/SNF; Testis development.

Fig. 1.

Expression and function of BRG1 and ARID2 in mouse spermatogenesis. (A) Images show BRG1 and ARID2 immunostaining in paraffin wax-embedded sections of wild-type mice. P, pachytene; D, diplotene; R, rounded spermatid. (B) Immunosignal quantitation of BRG1 in spermatogonia (Sg) (arbitrary units, mean±s.d.) 317±76, n=24 cells; leptotene (L) 76±19, n=66; zygotene (Z) 90±36, n=106 total cell counted; pachytene (P) 223±59, n=168; diplotene (D) 113±52, n=101; rounded spermatids (R) 50±26, n=101) and of ARID2 (spermatogonia 86±29, n=29; leptotene 77±12, n=95; zygotene 62±11, n=74; pachytene 97±22, n=100; diplotene 124±44, n=77; rounded spermatid 128±31, n= 124) from images in A. Data were obtained in three different experiments from three different mice. (C) Arid2 gene targeting design and subsequent deletion of exon 4. Gene location of F1, R1 and R2 oligonucleotides for PCR genotyping strategy used for detection of wild-type, floxed and delta mice are represented by blue arrows. (D) Diagram showing stages in gamete development and timing of expression for the Ddx4 and Spo11 promoters driving expression of the alternative Cre recombinases used in this study. (E) Western blots of 2-month-old testes of Arid2 knockout mice show no signal corresponding to the ARID2 protein compared with wild-type mice. Expression of PBRM1 and BRG1 in wild-type and Arid2^Δ/Δ testis is also shown. Tubulin was used as loading standard. Results are shown for four mice: Ddx4-Arid2 control and knockout (left); and Spo11-Arid2 control and knockout (right). Quantitation of immunosignal can be found in Fig. S3C. (F) Arid2 knockout mice (2-month old) have reduced testis size and testis weight compared with wild-type mice. Brg1 knockout mice (2-month old) have reduced testis size and weight compared with Arid2 knockout and wild-type testes. Compared with wild-type (mean±s.d.) (0.108±0.014 g, n=8; n=26 testes, 13 mice) littermate testes in Arid2^Δ/Δ adult mice are significantly smaller. Ddx4-Arid2^Δ/Δ males (0.044±0.004 g, n=6 testes, 3 mice, P≤0.0001, t-test), Spo11-Arid2^Δ/Δ males (0.066 g±0.003, n=8 testes, 4 mice, P≤0.0001, two-tailed unpaired Student's t-test). Ddx4-Brg1^Δ/Δ knockout mice (0.035 g±0.004, n=6 testes, 5 mice) showed reduced testis size compared with wild type (P≤0.0001, two-tailed unpaired Student's t-test).

Fig. 2.

Gamete developmental impairment in Arid2 knockout mice originates from spindle assembly and microtubule-chromosome attachment defects. (A) Meiosis is arrested at the end of prophase I and II in 2-month-old Arid2^Δ/Δ spermatocytes. Hematoxylin and Eosin staining corresponding to (a-d) wild type; (e-i) Ddx4-Arid2^Δ/Δ; and (j-m) Spo11-Arid2^Δ/Δ. (g) Cells at metaphase I show lagging chromosomes (some examples are highlighted with arrows). (h) Cells arrested at the second meiotic division showing abnormal chromosome figures. (i) Gigantic polynucleated cells (some examples are indicated with arrowheads) apparently formed after the second meiotic division. MI, metaphase I; MII, metaphase II; E, elongated spermatids; R, rounded spermatids. (B) Quantitation of lagging chromosomes assessed in Hematoxylin and Eosin stained 2-month-old mouse testis sections. Wild type (mean±s.d.) 2.5±0.024% of cells with lagging chromosomes, n=122 total cells counted; Ddx4-Arid2^Δ/Δ 31.3±12.8% of cells with lagging chromosomes, n=265 cells, P=0.0049, two-tailed unpaired Student's t-test; Spo11-Arid2^Δ/Δ 22.5±2.68% of cells with lagging chromosomes, n=115 cells, P=0.0054, two-tailed unpaired Student's t-test. Data were obtained from three different mice per strain. (C) Representative images of 2-month-old wild-type and Ddx4-Arid2^Δ/Δ squashed seminiferous tubules immunostained using antibodies against tubulin and DAPI. There are multiple spindles and unaligned chromosomes in knockout samples. (D) Quantitation of chromosome conformations in meiosis I and meiosis II cells at metaphase. N, normal metaphase. NA, metaphases showing abnormal metaphase with non-aligned chromosomes (e.g. chromosomes prematurely at the poles). MS, cells with multiple spindles; ND, non-determined. The data were collected from three different mice of each genotype. Wild-type meiosis I (% of cells, n=31 total number of cells counted), n=94, NA=6; wild-type meiosis II (n=30 total cells counted), n=100. Ddx4-Arid2^Δ/Δ meiosis I (% of cells, n=84 total cells counted), n=29, MS=5, NA=66; Ddx4-Arid2^Δ/Δ meiosis II (% of cells, n=34 total cells counted), n=35, MS=24, NA=41. Spo11-Arid2^Δ/Δ meiosis I (% of cells, n=35 total cells counted), n=57, MS=12, NA=31.

Fig. 3.

Characterization of Arid2 knockout mice gametogenic arrest. (A) Images show SYCP3, SYCP1 and γH2AX immunostaining in paraffin wax-embedded sections of wild-type, Ddx4-Arid2^Δ/Δ and Spo11-Arid2^Δ/Δ mice. L-Z, leptotene-zygotene; D, diplotene; P, pachytene. The images in the far-right panel correspond to the area highlighted by the white square. (B) PNA and SYCP3 immunostaining in wild-type, and Ddx4-Arid2 and Spo11-Arid2 knockout mice. Note the reduction in the number of S1 spermatids and the accumulation of cells at diplotene in tubules type I and XI in Arid2 knockouts when compared with wild type. S1-S13 indicate different types of spermatids. (C) Quantification of diplotene cells and S1 spermatids (number of cells per 0.1 mm² of tubule). Diplotene (mean±s.d.) 3.8±0.65, n=17 tubules, 1081 cells; Ddx4-Cre-Arid2 5.9±1.18, n= 26 tubules, 1733 cells, two-tailed unpaired Student's t-test P<0.0001; Spo11-Cre-Arid2 6.9±1.38, n= 25 tubules, 1812 cells, two-tailed unpaired Student's t-test P<0.0001). S1 rounded spermatids, wild type (mean±s.d.) 9.9±2.4, n=27 tubules, 4793 cells; Ddx4-Cre-Arid2 4.2±1.7, n= 33 tubules, 1904 cells, two-tailed unpaired Student's t-test P<0.0001; Spo11-Cre-Arid2 3.6±1.9, n=32 tubules, 1544 cells, two-tailed unpaired Student's t-test P<0.0001 (data were obtained from three different 2-month-old mice per strain).

Fig. 4.

Arid2^Δ/Δ mice show a higher number of apoptotic cells. (A) Increased number of apoptotic cells positive for TUNEL assay is observed in 2-month-old Ddx4-Arid2 and Spo11-Arid2 knockouts compared with wild type. (B) Higher magnification of apoptotic cells is shown. White arrows indicate examples of TUNEL-positive cells and blue arrows indicates examples of cells at metaphase I with possible lagging chromosomes. (C) Number of apoptotic cells per seminiferous tubule in 2-month-old Ddx4-Arid2 and wild-type mice. Wild-type (2.01±1.15, n=95 seminiferous tubules) and Arid2^Δ/Δ (Ddx4-Cre) mice (21.73±10.62, n=49; two-tailed unpaired Student's t-test P>0.0001). Different symbols in the graph correspond to data collected from three mice of each phenotype.

Fig. 5.

Defective spindle formation, abnormal chromosome-spindle attachment and increased aurora B kinase at centromeres in Arid2^Δ/Δ testis. (A) Western blots show an increase in aurora B kinase in 2-month-old Arid2^Δ/Δ testis compared with wild type. The diagram shows distribution of PLK1, aurora kinase B and survivin at the centromeric chromosome area. (B) Mouse spermatocyte chromosome spreads from Arid2^Δ/Δ show increased H3pS10 intensity at centromeres in diplotene cells. Antibodies against SYCP3 were used as a marker for the chromosome axis. Quantitation of H3pS10 fluorescent intensity is also shown [wild type (arbitrary units, mean±s.d.) 19.26±6.04, n=30 total cell counted, Ddx4-Arid2^Δ/Δ 68.70±31.57, n=30 total cell counted; P<0.0001, two-tailed unpaired Student's t-test (data were obtained from three mice of each phenotype)]. (C) Representative images of paraffin wax-embedded testis sections immunostained for SYCP3 and aurora B kinase in wild-type and Ddx4-Arid2 knockout mice (also see Spo11-Arid2 knockout in Fig. S8A). Cells transitioning metaphase-anaphase I are immunostained for SYCP3. (D) Quantitation corresponding to aurora B kinase immunostaining in metaphase spermatocyte nuclei of wild type, Ddx4-Arid2 and Spo11-Arid2 knockout mice shown in C. Ddx4-Arid2^Δ/Δ (arbitrary units, mean±s.d.) 142.5±63.7, n=421 total cells counted obtained from three mice; P<0.0001, two tailed two-tailed unpaired Student's t-test. Spo11-Arid2^Δ/Δ 126.0±57.9, n=493 obtained from three different mice; P<0.0001, two-tailed unpaired Student's t-test compared with wild-type controls (78.8±15.5 relative fluorescence intensity, n=373 cells obtained from three different 2-month-old mice).

Fig. 6.

Comparative analysis of Brg1^Δ/Δ and Arid2^Δ/Δ testes. (A) Hematoxylin and Eosin stained testes sections of wild-type and Ddx4-Brg1 knockout. P, pachytene; E, elongated spermatids. There is arrest at pachynema in Ddx4-Brg1^Δ/Δ. (B) Representative chromosome spreads obtained from 2-month-old wild-type, Ddx4-Brg1^Δ/Δ and Ddx4-Arid2^Δ/Δ mice. The extent of γH2AX staining in Brg1^Δ/Δ is indicative of deficient DSB repair. (C) Representative images of chromosomes and quantitation corresponding to normal and abnormal synapsis in wild-type, Ddx4-Brg1^Δ/Δ and Ddx4-Arid2^Δ/Δ mice (number of cells and corresponding percentage are represented in the table). Data were obtained from three mice of each genotype. Wild type, n=68 total counted cells; Brg1^Δ/Δ, n=58 total counted cells; Arid2^Δ/Δ, n=45 total counted cells. Data were obtained from three different 2-month-old mice.

Fig. 7.

BRG1-independent PBAF complexes formed by ARID2 and PBRM1. (A) Immunoprecipitation followed by western blotting analysis of synchronized enriched fractions of zygotene spermatocytes. (B) Western blot showing depletion of BRG1 in zygotene- and diplotene-enriched spermatocyte fractions from Spo11-Cre mice. (C) Quantitation of testis weight from 2-month-old mice is also shown. Wild type (mean±s.d.) 0.100±0.0049 g, n=6 testes from 3 mice) and Spo11-Brg1^Δ/Δ males (0.097±0.0065 g, n=6 testes from 3 mice; P=0.331, two-tailed unpaired Student's t-test). (D) Hematoxylin and Eosin-stained testes sections of Spo11-Cre Brg1^Δ/Δ and wild type. (E) Schematic of the proposed function of BRG1-dependent and -independent PBAF complexes.