Histone H2AX is integral to hypoxia-driven neovascularization

Abstract

H2A histone family member X (H2AX, encoded by H2AFX) and its C-terminal phosphorylation (gamma-H2AX) participates in the DNA damage response and mediates DNA repair. Hypoxia is a physiological stress that induces a replication-associated DNA damage response. Moreover, hypoxia is the major driving force for neovascularization, as the hypoxia-mediated induction of vascular growth factors triggers endothelial cell proliferation. Here we studied the role of the hypoxia-induced DNA damage response in endothelial cell function and in hypoxia-driven neovascularization in vivo. Hypoxia induced replication-associated generation of gamma-H2AX in endothelial cells in vitro and in mice. Both in cultured cells and in mice, endothelial cell proliferation under hypoxic conditions was reduced by H2AX deficiency. Whereas developmental angiogenesis was not affected in H2afx(-/-) mice, hypoxia-induced neovascularization during pathologic proliferative retinopathy, in response to hind limb ischemia or during tumor angiogenesis was substantially lower in H2afx(-/-) mice. Moreover, endothelial-specific H2afx deletion resulted in reduced hypoxia-driven retina neovascularization and tumor neovascularization. Our findings establish that H2AX, and hence activation of the DNA repair response, is needed for endothelial cells to maintain their proliferation under hypoxic conditions and is crucial for hypoxia-driven neovascularization.

Fig. 1. Generation of γ-H2AX by hypoxia in endothelial cells and importance of H2AX for endothelial cell proliferation under hypoxia

HUVEC were kept in normoxia, irradiated with 2 Gy, treated with HU, or were exposed to hypoxia for 1 h or 6 h, as indicated. (a), Immunofluorescence staining for γ-H2AX (green) followed by confocal microscopy was performed. Nuclei are shown by DAPI staining. The percentage of cells displaying γ-H2AX foci was 3.1±0.5%, 99.1±0.4%, 19.1±1.2%, 10.9±0.7% and 17.1±1.2% for normoxic, irradiated, HU-treated, 1 h hypoxia-treated or 6 h hypoxia-treated cells, respectively (Mean ± SEM). Representative images are depicted. Bar=10 µm. (b), Left panel: Representative western blot analysis of the expression of ATR and tubulin in HUVEC transfected with siRNA against ATR or with control siRNA as indicated. Right panel: Representative western blot analysis of the expression of ATM and tubulin in HUVEC transfected with siRNA against ATM or with control siRNA as indicated. (c), Western blot analysis for γ-H2AX or actin in HUVEC transfected with siRNA against ATR, siRNA against ATM or with control siRNA as indicated, that were kept in normoxia or exposed to hypoxia for 6 h. ATR knockdown decreased the hypoxia-induced γ-H2AX generation. (d), The expression of total H2AX and actin was analysed by Western blot in HUVEC transfected with siRNA against H2AX or with control siRNA as indicated. (e), The proliferation of HUVEC transfected with control siRNA (open bars) or with siRNA targeting H2AX (filled bars) was assessed under nomoxic conditions or under hypoxic conditions, as indicated, in the presence of 10% FCS or FGF2 (50 ng/ml). Proliferation of HUVEC is expressed as percent in relation to control. Baseline proliferation under each condition in the absence of stimuli was defined as the 100% control. Data are Mean ± SEM (n=4). *: P<0.05. (f), The proliferation of WT and H2ax^−/− primary mouse lung endothelial cells was assessed under normoxic conditions or hypoxic conditions in the presence of 1% FBS (baseline conditions) or in the presence of 20% FBS together with 1% endothelial cell growth serum (ECGS) (stimulated conditions). Proliferation was assessed by cell counting and is expressed as absolute cell number. Data are Mean ± SEM (n=3). *: P<0.05. (g), Mice were subjected to the ROP model. Eyes were removed and retinas were extracted at day p15. Immunofluorescence analysis in retinas for lectin (red) indicating blood vessels, γ-H2AX (green) and DAPI (blue) in mice subjected to the ROP model and in mice kept in room air (non-ROP). In addition, γ-H2AX and DAPI staining were merged. Bar=10 µm.

Fig. 2. A role for H2ax in retina neovascularisation under hypoxia

Mice were subjected to the ROP model. (a), Retinal neovascularization was quantified on day p17 in WT and H2ax^−/− mice as described under Materials and Methods. Retinal neovascularization is presented as the number of epiretinal neovascular nuclei per section. Data are Mean ± SEM (n=3–5). *: P<0.05. (b) Six µm paraffin-embedded axial sections of the retina were stained with periodic acid Schiff and hematoxylin. In the ROP model pathologic neovascularizations are observed anterior to the internal limiting membrane. H2ax^−/− mice that were subjected to the ROP-model displayed less neovascular regions anterior to the internal limiting membrane as compared with WT mice. The arrows indicate neovascularizations anterior to the internal limiting membrane. Bar=100 µm. (c), Lectin staining of retina whole-mounts at p17 from a WT or a H2ax^−/− mouse that were subjected to the ROP model, demonstrating less extensive neovascular tufts in H2ax-deficient mice. For each image, multiple individual images were stitched together. Bar=500 µm. (d), Lectin (green) and BrdU (red) staining of retina whole-mounts at p17 from a WT or a H2ax^−/− mouse that were subjected to the ROP model. Arrows indicate the neovascular tufts. Bar=100 µm. (e), Endothelial cell proliferation in the retina of mice subjected to the ROP model was assessed by BrdU incorporation as described under Materials and Methods. The number of BrdU-positive endothelial cell nuclei per section was reduced in the retinas of H2ax^−/− mice as compared to WT mice. Endothelial cell proliferation is shown as % of control, i.e. endothelial cell proliferation in WT mice represents the 100% control. Data are Mean ± SEM (n=3–4). *: P<0.05. (f), The apoptosis in retina vessels was assessed by TUNEL staining, showing that endothelial cells in retinas of H2ax^−/−mice displayed increased apoptosis compared to retinas of WT mice. Endothelial cell apoptosis is shown as % of control, i.e. endothelial cell apoptosis in WT mice represents the 100% control. Data are Mean ± SEM (n=3). *: P<0.05. (g), Retinal neovascularization was quantified on day p17 in H2ax-sufficient (VEC-Cre⁻H2ax^flox/−; open bars) and endothelial-specific H2ax-deficient (VEC-Cre⁺ H2ax^flox/−; filled bars) mice. Retinal neovascularization is presented as the number of epiretinal neovascular nuclei per section. Data are Mean ± SEM (n=5–6). *: P<0.05.

Fig. 3. Reduced angiogenesis in response to hind limb ischemia due to H2ax deficiency

Recovery ability of H2ax^−/− mice and WT mice after hind limb ischemic surgery. (a), Doppler scanning monitoring results revealed significant difference between the two groups, with the WT group recovered to 96.6% and 70.7% of H2ax^−/− group recovered to only 48% and 42.9% (n=5). (b), Endothelial cell proliferation in the ischemic muscle in H2ax^−/− and WT mice: BrdU and CD31 double immunostaining revealed that the CD31⁺/BrdU⁺ ratio of H2ax^−/− mice was significantly less than the CD31⁺/BrdU⁺ ratio of WT mice. Data are Mean ± SEM. (c) Representative images of BrdU (red) and CD31 (green) double immunostaining in ischemic muscles of WT and H2ax^−/− mice. Blue: DAPI staining to identify nuclei was used. Bars=10 µm. *, P<0.05; **, P<0.01.

Fig. 4. Reduced tumor angiogenesis and tumor growth due to H2ax deficiency

(a), Tumors implanted in H2ax^−/− mice exhibited decreased vascularity. Tumors that were implanted s.c. into WT or H2ax^−/− mice, were excised, fixed and stained immunofluorescently for blood vessels (CD31). Tumor angiogenesis is shown as vessels per section at days 21 and 25 and the data are expressed as % of control, i.e. tumor angiogenesis at day 21 or day 25 in WT mice was set as the 100% control, respectively. Data are Mean ± SEM (n=6–10 tumors). *: P<0.05; **: P<0.01. (b), Representative images of CD31 staining from tumors implanted into WT or H2ax^−/− mice are shown. Bar=100 µm. (c), Endothelial cell proliferation in tumors excised from WT and H2ax^−/− mice was assessed by evaluating the number of BrdU-positive endothelial cell nuclei per field. Endothelial cell proliferation is shown as % of control, i.e. endothelial cell proliferation in WT mice represents the 100% control. Data are Mean ± SEM (n=6–9 tumors). *: P<0.05. (d), Endothelial cell apoptosis in tumors excised from WT and H2ax^−/− mice was assessed, by evaluating the number of caspase 3-positive endothelial cells per field. Endothelial cell apoptosis is shown as % of control, i.e. endothelial cell apoptosis in WT mice represents the 100% control. Data are Mean ± SEM (n=4 tumors). *: P<0.05. (e), Pericyte coverage of vessels in tumors of WT and H2ax^−/− mice was assessed by NG-2 staining. Pericyte coverage is shown as the percentage of vessels that were covered by NG-2-positive cells. Data are Mean ± SEM (n=8 tumors). (f), Tumors implanted into H2ax^−/− mice exhibited decreased tumor progression, as assessed by monitoring tumor volume after day 6. Data are Mean ± SEM (n=8–10 tumors). **: P<0.01, as compared to the tumor volume of WT mice on the same day. (g), The weight of tumors excised from WT or H2ax^−/− at day 25 is shown. Data are Mean ± SEM (n=8–10 tumors). *, P<0.05.