The base excision repair pathway is required for efficient lentivirus integration

Abstract

An siRNA screen has identified several proteins throughout the base excision repair (BER) pathway of oxidative DNA damage as important for efficient HIV infection. The proteins identified included early repair factors such as the base damage recognition glycosylases OGG1 and MYH and the late repair factor POLß, implicating the entire BER pathway. Murine cells with deletions of the genes Ogg1, Myh, Neil1 and Polß recapitulate the defect of HIV infection in the absence of BER. Defective infection in the absence of BER proteins was also seen with the lentivirus FIV, but not the gammaretrovirus MMLV. BER proteins do not affect HIV infection through its accessory genes nor the central polypurine tract. HIV reverse transcription and nuclear entry appear unaffected by the absence of BER proteins. However, HIV integration to the host chromosome is reduced in the absence of BER proteins. Pre-integration complexes from BER deficient cell lines show reduced integration activity in vitro. Integration activity is restored by addition of recombinant BER protein POLß. Lentiviral infection and integration efficiency appears to depend on the presence of BER proteins.

Figure 1. Viability of BER glycosylase deletion cell lines treated with DNA damaging agent H₂O₂.

BER glycosylase deletion cells and matched wild type littermate cells were treated with increasing concentrations of H₂O₂. Cells were stained with trypan blue and viable cells counted. The percentage of viable cells remaining is shown. (A) Matched wild type, Myh null and Ogg1 null MEFs treated with H₂O₂. (B) PCR analysis of Myh and Ogg1 genotypes. Primer sets detected wild type Myh (Myh+/+), the Myh deletion construct (Myh−/−), wild type Ogg1 (Ogg1+/+) and the Ogg1 deletion construct (Ogg1−/−). PCR targets included water (No target) and genomic DNA from wild type (WT), Myh−/−, and Ogg1−/− cell lines. (C) Matched wild type and Neil1 null MEFs treated with H₂O₂. (D) PCR analysis of Neil1 genotypes. Primer sets detected wild type Neil1 (Neil1+/+) and the Neil1 deletion construct (Neil1−/−). PCR targets included water (No target) and genomic DNA from wild type (Neil1+/+) and Neil1−/− cell lines.

Figure 2. Viability of BER polymerase b cell lines treated with DNA damaging agents.

(A) Polß deletion cells and matched wild type littermate cells were treated with increasing concentrations of the oxidative base damage inducing agent H₂O₂. (B) Polß cell lines were treated with increasing concentrations of the methylation base damage inducing agent MMS. Cells were stained with trypan blue and viable cells counted. The percentage of viable cells remaining is shown. (C) Western analysis of wild type (PolB+/+) and Polß null (PolB−/−) cell lines for Polß expression. Blots were stripped and re-probed for PCNA as a loading control.

Figure 3. Infection of BER cell lines with HIV, MMLV, and FIV.

Cells were infected with HIV, MMLV, and FIV retroviral vectors expressing GFP following integration. Cells were analyzed at 72 hpi by flow cytometry for GFP expression indicating successful infection. Wild type MEFs were from littermates. (A) Wild type (WT), Myh−/−, and Ogg1−/− cell lines, (B) Wild type (Neil1+/+) and Neil1−/− cell lines, (C) Wild type (Polß+/+) and Polß−/− cell lines. Infections were performed at two MOI in duplicate at least three times. Error bars indicate the standard deviation after normalization.

Figure 4. Infection of BER cell lines with HIV with and without accessory genes or the cPPT.

Wild type (WT), Myh−/−, and Ogg1−/− cell lines, wild type (Neil1+/+) and Neil1−/− cell lines, or wild type (Polß+/+) and Polß−/− cell lines were infected with HIV retroviral vectors. Wild type murine embryonic fibroblasts were from littermates. Cells were analyzed at 72 hpi by flow cytometry for GFP expression. (A) Cells were infected with an HIV vector with accessory genes or without accessory genes (ΔvifΔvprΔvpuΔnef). (B) Total RNA was isolated from wild type HIV and HIV(ΔvifΔvprΔvpuΔnef) vector producer cells, treated with DNaseI to digest producer plasmids, and re-isolated. RNA fractions were amplified with the same conditions by RT-PCR or PCR, to confirm the absence of contaminating producer plasmids. RT-PCR and PCR targets were Lane 1 water negative control, Lane 2 HIV(ΔvifΔvprΔvpuΔnef) RNA, Lane 3 wild type HIV RNA, and Lane 4 wild type HIV producer plasmid DNA, positive control for PCR. Primers amplified the accessory genes vpu, vpr, vif, and nef as well as the gag gene. (C) Cells were infected with an HIV vector with or without the cPPT. Infections were performed at two MOI in duplicate at least three times. Error bars indicate the standard deviation after normalization.

Figure 5. Time course of HIV cDNA accumulation in BER cells.

BER wild type and deletion cell lines were infected with HIV. DNA was purified at multiple time points and analyzed by qPCR for HIV late reverse transcripts and the 18S gene. The 18S gene is a measure of the number of genomes present. The late reverse transcript primer set amplifies all complete HIV cDNAs including linear, 1LTR and 2LTR circles, and integrated provirus. The number of late reverse transcripts was divided by the number of genomes to yield HIV cDNA per cell (HIV cDNA/cell). Infections were performed in duplicate three times. Error bars indicate the standard deviation.

Figure 6. HIV autointegration, 2LTR circles, and integrated provirus in BER cells.

BER wild type and deletion cell lines were infected with HIV. DNA was purified at 24 hpi or 72 hpi and analyzed by qPCR for the indicated HIV cDNA and the 18S gene. The number of HIV cDNA was divided by the number of genomes, measured by the 18S gene, to yield the HIV cDNA products per cell. Each HIV cDNA per cell is expressed relative to matched wild type cells. (A) HIV autointegration products at 24 hpi. (B) HIV 2LTR circles at 24 hpi. (C) HIV integrated provirus at 72 hpi. Infections were performed in duplicate three times. Error bars indicate the standard deviation after normalization.

Figure 7. Integration activity of HIV PICs from BER cell lines.

HIV PIC extracts were generated at 6 hpi. Purified human genomic DNA was added to HIV PICs. Integration products and total HIV cDNA were measured by qPCR. The number of integration products divided by the total HIV cDNA indicates the integration activity. Integration activity of PICs from mutant cells is expressed relative to PICs from matched wild type cells. (A) HIV PICs from matched wild type and Ogg1, Neil1, or Polß null cells. (B) Increasing concentrations of purified recombinant human POLß protein were added to PICs from wild type and Polß−/− cells. PICs were generated from three infections. Integration reactions for each PIC preparation were performed in duplicate. Error bars indicate the standard deviation. (C) Silver stained gel of recombinant human POLß protein. Lanes 1 (left) and 7 (right) show the molecular weight markers. Lanes 2–6 correspond to 20, 40, 80, 120 and 200 ng purified POLß, respectively.