Both ATM and DNA-PK Are the Main Regulators of HIV-1 Post-Integrational DNA Repair

Abstract

The integration of a DNA copy of an HIV-1 RNA genome into the host genome, carried out by the viral enzyme integrase, results in the formation of single-stranded gaps in cellular DNA that must be repaired. Here, we have analyzed the involvement of the PI3K kinases, ATM, ATR, and DNA-PKcs, which are important players in the DNA damage response (DDR) in HIV-1 post-integrational DNA repair (PIR). The participation of the DNA-PK complex in HIV-1 PIR has been previously shown, and the formation of a complex between the viral integrase and the DNA-PK subunit, Ku70, has been found to be crucial for efficient PIR. Now, we have shown that the inhibition of both DNA-PKcs and ATM, but not ATR, significantly reduces PIR efficiency. The activation of both kinases is a sequential process, where one kinase, being activated, activates the other, and it occurs simultaneously with the integration of viral DNA. This fact suggests that the activation of both kinases triggers PIR. Most interestingly, the activation of not only DNA-PKcs, but also ATM depends on the complex formation between integrase and Ku70. The elucidation of the interactions between viruses and DDR is important both for understanding the modulation of host cell functions by these pathogens and for developing new approaches to combat viral infections.

Keywords: ATM; DNA-PK; HIV-1; Ku70; integrase; post-integrational repair.

Figure 1

The effect of inhibitors of ATM, ATR, and DNA-PKcs on HIV-1 post-integrational repair. (A–C) Analysis of the relative luciferase expression in 293T cells transduced by HIV_wt in presence of increasing concentrations of Ku-55933 (iATM), AZ20 (iATR), and Nu7441 (iDNA-PK). (D) The effect of azidothymidine (AZT), raltegravir (Ral), iATM, and iDNA-PK on luciferase expression in 293T cells transduced by HIV_wt. (E,F) The level of the total viral DNA (E) and integrated viral DNA (F) in 293T cells transduced by HIV_wt in presence of AZT, Ral, iATM, and iDNA-PK. (G) Efficiency of post-integrational DNA repair in 293T cells transduced by HIV_wt in presence of iATM or iDNA-PK. Mean values ± SD of three (D–G) or four (A–C) independent experiments are presented. Significance was determined by one-way ANOVA with Dunnett’s correction for multiple comparisons, ns—not significant, * = adj. p-value < 0.05, ** = adj. p-value < 0.01, **** = adj. p-value < 0.0001.

Figure 2

The effect of ATM and DNA-PKcs inhibitors on luciferase activity in 293T cells transduced by HIV_wt and HIV_mut. (A) HIV_mut transduced 293T cells five times worse than HIV_wt. Significance was determined by two-tailed Student’s t-test, *** = p < 0.001. (B,C) Analysis of the relative luciferase expression in 293T cells transduced by HIV_wt or HIV_mut in presence of increasing concentrations of Nu7441 (iDNA-PK) or Ku-55933 (iATM). Data for HIV_wt or HIV_mut were normalized to 0 μM points for each vector independently. Mean values ± SD of three (A) or four (B,C) independent experiments are presented. Significance was determined by two-way ANOVA with Sidak’s correction for multiple comparisons, * = adj. p-value < 0.05, ** = adj. p-value < 0.01, *** = adj. p-value < 0.001, **** = adj. p-value < 0.0001.

Figure 3

Dependence of the activity of inhibitors on the time of their addition to HIV_wt transduced cells. 293T cells were transduced by HIV_wt for 1 h, then the excess of the vector was removed; inhibitors (AZT, Ral, iATM or iDNA-PK) were added to the cells 1–24 h.p.i. Luciferase expression was tested 48 h.p.i. The mean values of three independent experiments are presented.

Figure 4

Models of ATM and DNA-PKcs activation during HIV-1 PIR. (A) The parallel activation model suggests that activation of either of the two kinases at the site of integration can lead to PIR activation and successful completion of this process. (B) The sequential activation model suggests that activation of both kinases is required for successful PIR to occur. For both models, activation of ATM and DNA-PKcs depends on the formation of the IN-Ku70 complex.

Figure 5

The inhibitory effect of iATM, iDNA-PK or its combinations on luciferase activity of HIV_wt in 293T cells normalized to cell viability (MTT-test). The expected inhibition efficiency was calculated for the parallel model, in which the inhibitory effect of a combination of two compounds is the sum of inhibitory effects of each individual compound. Mean values ± SD from four independent experiments are presented. Significance was determined by two-way ANOVA with Sidak’s correction for multiple comparisons, **** = adj. p-value < 0.0001.

Figure 6

γH2AX as a marker of HIV-1 post-integrational DNA repair. (A) Kinetics of γH2AX formation in 293T cells transduced by HIV_wt, HIV_mut, or HIV_E152A analyzed using Western blot (left) and its quantitative analysis (right). HIV_mut contains IN with E212A/L213A substitutions defective in Ku70 binding and PIR initiation. HIV_E152A contains IN with E152A substitution defective in catalytic activity (impaired integration). The sample, designated cntr, was not transduced by any pseudovirus. Mean values ± SD of three independent experiments are presented. (B) γH2AX formation in 293T cells transduced by HIV_wt and treated with Nu7441 (iDNA-PK, 1 μM) or Ku-55933 (iATM, 5 μM) analyzed using Western blot (left) and its quantitative analysis (right). Samples designated as DMSO were transduced with pseudovirus but without any inhibitor, and DMSO was added to them in the same amount as to the samples treated with inhibitors. The sample denoted as cntr was treated with DMCO but not transduced by pseudovirus. Mean values ± SD of three independent experiments are presented. Significance was determined by two-way ANOVA with Sidak’s correction for multiple comparisons, ns—not significant, *** = adj. p-value < 0.001, **** = adj. p-value < 0.0001.