Gene editing of the multi-copy H2A.B gene and its importance for fertility

Abstract

Background: Altering the biochemical makeup of chromatin by the incorporation of histone variants during development represents a key mechanism in regulating gene expression. The histone variant H2A.B, H2A.B.3 in mice, appeared late in evolution and is most highly expressed in the testis. In the mouse, it is encoded by three different genes. H2A.B expression is spatially and temporally regulated during spermatogenesis being most highly expressed in the haploid round spermatid stage. Active genes gain H2A.B where it directly interacts with polymerase II and RNA processing factors within splicing speckles. However, the importance of H2A.B for gene expression and fertility are unknown.

Results: Here, we report the first mouse knockout of this histone variant and its effects on fertility, nuclear organization, and gene expression. In view of the controversy related to the generation of off-target mutations by gene editing approaches, we test the specificity of TALENs by disrupting the H2A.B multi-copy gene family using only one pair of TALENs. We show that TALENs do display a high level of specificity since no off-target mutations are detected by bioinformatics analyses of exome sequences obtained from three consecutive generations of knockout mice and by Sanger DNA sequencing. Male H2A.B.3 knockout mice are subfertile and display an increase in the proportion of abnormal sperm and clogged seminiferous tubules. Significantly, a loss of proper RNA Pol II targeting to distinct transcription-splicing territories and changes to pre-mRNA splicing are observed.

Conclusion: We have produced the first H2A.B knockout mouse using the TALEN approach.

Keywords: Chromatin; Genome editing; H2A.B; Histone variants; Pre-mRNA splicing; RNA polymerase II; Splicing speckles; TALENs.

Fig. 1

Interrogation of predicted off-target heterozygous deletions in H2A.B.3^−/y mice. a 300–400-bp regions surrounding 19 common predicted heterozygous deletions were amplified from gDNA of H2A.B.3^−/y(KO) and FVB/NJArc in-house wt mice. PCR product sizes for wt and H2A.B.3^−/y were compared by resolving the products on a 7% polyacrylamide gel, side by side. The predicted deletion is indicated with △, followed by the number of nucleotides (nt) predicted to be deleted. b Positive controls showing that a deletion as small as 5 and 10 nt can be resolved on 7% polyacrylamide gel. Lane 1, a heterozygous △5 nt, lanes 2, 3, 4 wt.; lane 5, homozygous △10 nt. c 5 out of the 19 PCR products were sequenced by Sanger sequencing to show that FVB/NJArc and H2A.B.3^−/y mice have identical sequences but they are different from the mm10 reference genome. d Table showing that the prediction tools are often inaccurate at predicting the presence or the size of a putative deletion

Fig. 2

H2A.B.3 KO males have impaired fertility. a The mRNA level of each H2A.B.3 gene was determined by real-time quantitative PCR, relative to HPRT. Primers were designed to anneal within the TALEN-induced deletions. b Confirmation that H2A.B.3 protein is not produced in H2A.B.3^−/y mice. H2A.B was immunoprecipitated from total testis lysates followed by Western blotting using anti-H2A.B.3 polyclonal antibodies. c The breeding was conducted using age-matched wt females (3–6 months) and age-matched wt (n = 5) and H2A.B.3^−/y (n = 5) males. Total number of pups produced was 201 and 256 for wt and KO males, ***p value = ≤ 0.001 (ANOVA test)

Fig. 3

H2A.B.3 KO males have a higher percentage of abnormal sperm. a wt and H2A.B.3 KO mice were scored for abnormal sperm (N = 3). *p value <0.01 (T-test).  b The major abnormalities seen in both wt and H2A.B.3 KO sperm are those with a twisted tail around the head of the sperm and a kinked tail

Fig. 4

Characterization of the incorporation of TP1 and H2A.L.2 into spermatid chromatin in the absence of H2A.B.3. a Elongating/condensing spermatids immunostained with anti-TP1 antibodies (turquoise) and counterstained with DAPI (blue). The outlines of the spermatids nuclei are shown as a white trace in TP1 panels. b The quantification of the number of wild type and H2A.B.3^−/y step 10–12 spermatids with a delayed wave of TP1 release. c HA.L.2 localization in condensing wild type and H2A.B.3^−/y spermatid. White arrows point to pericentric heterochromatin. d The quantification of the number of wild type and H2A.B.3^−/y condensing spermatids with H2A.L.2 enriched in pericentric heterochromatin (HC)

Fig. 5

H2A.B.3 KO males have a higher percentage of clogged seminiferous tubules. a Image of hematoxylin and eosin-stained seminiferous tubule from wt and H2A.B.3^−/y mice. b Histogram plot of the percentage of a seminiferous tubule with a clogged lumen in H2A.B.3^−/y mice (n = 9) and their wt siblings (n = 11). ANOVA test, **p = 0.006547 (≤ 0.01)

Fig. 6

Localization of the active form of RNA polymerase II within splicing speckles is altered in H2A.B.3^−/y round spermatids. a Immunofluorescence staining of round spermatids from wt and H2A.B.3^−/y mice testis. Cells were indirectly labeled with anti-Pol II S2P antibodies and co-stained with DAPI. Scale bar is 10 μm. White arrows show accumulation of RNA Pol II S2P signal in splicing speckles of WT round spermatid cells. The signal is diffused in H2A.B.3^−/y. Detailed quantification of RS in wt (n = 90) and H2A.B.3^−/y (n = 92) mice showing that majority of RS cells in H2A.B.3 KO mice have diffused Pol II S2 localization. b Immunofluorescence staining of round spermatids from wt and H2A.B.3^−/y mice testis. Cells were indirectly labeled with anti-Pol II S2P and Y2 (a splicing speckle marker) antibodies. Scale bar is 10 μm. White arrows show colocalization of RNA Pol II S2P and Y12 in splicing speckles of wt round spermatid cells. RNA Pol II S2P no longer co-stains with Y12 in the majority of round spermatids from H2A.B.3^−/y testes. Detailed quantification of RS in wt (n = 90) and H2A.B.3^−/y (n = 72) `