APOBEC3G protects the genome of human cultured cells and mice from radiation-induced damage

Abstract

Cytosine deaminases AID/APOBEC proteins act as potent nucleic acid editors, playing important roles in innate and adaptive immunity. However, the mutagenic effects of some of these proteins compromise genomic integrity and may promote tumorigenesis. Here, we demonstrate that human APOBEC3G (A3G), in addition to its role in innate immunity, promotes repair of double-strand breaks (DSBs) in vitro and in vivo. Transgenic mice expressing A3G successfully survived lethal irradiation, whereas wild-type controls quickly succumbed to radiation syndrome. Mass spectrometric analyses identified the differential upregulation of a plethora of proteins involved in DSB repair pathways in A3G-expressing cells early following irradiation to facilitate repair. Importantly, we find that A3G not only accelerates DSB repair but also promotes deamination-dependent error-free rejoining. These findings have two implications: (a) strategies aimed at inhibiting A3G may improve the efficacy of genotoxic therapies used to cure malignant tumours; and (b) enhancing A3G activity may reduce acute radiation syndrome in individuals exposed to ionizing radiation.

Keywords: APOBEC3G; DNA damage response; DNA repair; cytidine deaminase; homologous recombination; mouse; non-homologous end joining.

Figure 1.. A3G promotes the repair of DSBs induced by IR in cultured cells:

(A) From left to right: HEK293 vs HEK293-A3G cells; SupT11A3G vs SupT11EV and H9 vs SupT1 cells were exposed to 3 Gy of ionizing radiation (IR). At the indicated times post irradiation cells were fixed, stained with anti γH2AX antibody, and analyzed by flow cytometry. Results represent the percentage of γH2AX-positive cells; each value represents the average of two or three independent experiments, ±standard deviation (SD). Statistical significance by two-tailed two-sample t test for HEK293/HEK293-A3G: p=0.051, p=0.026 8h and 24h, and for SupT11EV/A3G: p=0.02, p=0.014, p=0.03 for 4h, 8h and 24h. (B): A3G expression was confirmed by western blotting using anti-A3G C-terminus rabbit polyclonal antibody followed by detection with enhanced chemiluminescence (ECL). Mouse anti-α-Tubulin antibody was used as a loading control. Flow cytometry histograms of γH2AX-stained the 4 Gy-irradiated HeLa and HeLa-A3G cells are presented. Time points after IR exposure are indicated. Red curves – irradiated cells; black curves – non-irradiated control cells. Marker boundary is set on the intersection of the control and the irradiated sample curves. The percentage of γH2AX-positive cells is indicated. Ten thousands cells were assessed in each sample; (C-D) Clonogenic assay of irradiated HeLa and HeLa-A3G cells is shown. Briefly, cells were seeded at low density in 6-well plates 3 hours prior to IR exposure (2 and 4 Gy). Seven to nine days post irradiation cells were fixed, stained by Crystal Violet (0.05%) and colonies with >20 cells were counted. Survival rates are plotted in (C) and a ratio between the survival rates of HeLa vs HeLa-A3G in each IR dose is plotted in (D). Average of two independent experiments is presented, ±SD. Statistical significance: *, P<0.05; **, P<0.01 by two-tailed two-sample t test.

Figure 2.. A3G rescues mice from IR-induced damage:

One-month old mice were whole body irradiated (8.25 Gy). Percentage of mice survival for each genotype group during 30 days post irradiation was determined. (A) Kaplan–Meier survival curves of mA3-competent mice (mA3+/+); the mA3 knocked-out (KO) mice (mA3−/−); the A3G^High and A3G^Low transgenic mice are shown. Genotype groups contained both female and male mice, as follows: mA3+/+ - 5 female and 8 male mice; mA3−/− - 10 female and 7 male mice; A3G^High – 5 female and 7 male mice; A3G^Low – 5 female and 8 male mice. Statistical significance: *, P<0.05, by one-sided two-sample t test; ns – not significant. (B) Body weight; (C) White blood cell count; (D) Red blood cell count and (E) Hemoglobin concentration are shown. (B-E) plots represent the results from one out of three independent experiments performed in the same conditions.

Figure 3.. Histological examination of livers and spleens derived from the irradiated mice:

(A) Livers (upper row) and spleens derived from 8.25 Gy-irradiated (first three columns) and non-irradiated (right column) mice are shown in photomicrographs. Samples in each column were prepared from the same mouse. Each sample represents its genotype group. Sizes of each genotype group are: n=7 for A3G^High +/+ (only mice survived the entire experiment were analyzed); n=17 for mA3−/−; n=13 for mA3+/+. Enlarged frames from mA3−/− liver and mA3+/+ spleen are shown in (B i) and (B ii), respectively. Spleen from non-irradiated C57BL mice (control) is presented (B iii) and (B iv). Organ-sections were stained with Hematoxylin and Eosin (H&E). Mice survival post irradiation is indicated. Scale-bars are indicated in (A): 200 μm (liver, upper row), 500 μm (spleen, middle and lower rows). Arrows indicate: necrosis (black triangles, liver sections), hemorrhage (black arrow, spleen) and brown-colored hemosiderin deposits (white arrows, spleen). (C) Spleen sections are stained with antibodies against CD3, B220 and c-Kit markers. Mice genotypes are indicated. Samples in each column were prepared from the same mouse. Scale-bars are indicated.

Figure 4.. A3G promotes accurate DSB repair independently of RNA binding:

(A) Scheme describes the working-flow for preparation the 131nt fragments used for sequencing analyses. (B) Plot represents fractions of accurate rejoining over total reads calculated for the whole 131 bp fragment and for 18 bp ISceI restriction site, in HEK293, HEK293-A3G, HEK293-A3G-W285A and HEK293-A3G-E259Q cells (for details see Table 1 and Figure S1). Each value represents the average of three independent experiments, ±SD. Statistical significance: *, P<0.05; ** P<0.01 by two-tailed paired t test. (C) Western blots analysis of HEK293 cells transfected with WT or deamination-deficient mutants. Blots were probed with anti-A3G C-terminus antibody and anti-α-Tubulin HRP-conjugated antibody followed by detection with ECL. (D) HEK293 cells were stably transfected with WT A3G and with the RNA-binding mutants (W94A and W127A). Cells were irradiated with IR (3Gy), followed by incubation for the indicated time intervals and then stained with anti-γH2AX antibody and analyzed by flow cytometry. Plot shows the percentage of γH2AX-positive cells, as analyzed by flow cytometry. Each value represents the average of two independent experiments, ±SD. (E) Western blots analysis of HEK293 cells transfected with WT or with RNA binding-impaired A3G mutants. Blots were probed with anti-A3G C-terminus antibody, as described in C.

Figure 5.. Proteomic analysis of the cellular proteins expressed in association with A3G:

(A) A3G protein content is shown in SupT11^A3G and SupT11^EV cells at the indicated time intervals after irradiation (4Gy), as identified by MS. (B) Principal component analysis (PCA) was performed on raw Mass Spec (MS) data from irradiated SupT11^A3G cells (blue balls) and SupT11^EV cells (red balls) at time points 0, 3 and 8 hours post-radiation. Presented is one of two independent experiments. (C) Interaction network of the 25 proteins as they appear in STRING database. Red color indicates the DNA Repair, while blue color refers to the HDR through Single-Strand Annealing (SSA) pathway, the most statistically significant enriched proteins pathways shown in Table 2 (for details see Results section). The connections between the genes are based on STRING evidence relating to active interaction sources, as co-expression, gene fusion, co-occurrence, neighborhood etc. For details see Supporting Information section (Table S1).