The identities of host or viral factors that mediate repair of this integration intermediate are unknown. Three siRNA library screens of host factors that affect HIV infection efficiency failed to conclusively identify DNA repair pathways that might complete repair of the integration intermediate. Studies of repair with recombinant proteins in vitro indicated that any polymerase, endonuclease and ligase could repair the integration intermediate, suggesting that multiple DNA repair pathways may mediate this process in vivo. A recent study described an siRNA screen targeted to host DNA repair proteins. This study 5α-Androstan-3β-ol identified multiple host genes throughout the oxidative BER pathway that were required for efficient HIV infection. Using a panel of deletion cell lines, we have found that several BER proteins affect lentiviral infection but not infection by a gamma retrovirus. The role of the BER pathway appears to be at the integration step of the viral life cycle. One obvious mechanism for BER proteins during lentiviral integration is that these proteins complete repair of the integration intermediate. It is possible that lentiviruses rely largely on BER while retroviruses are less restricted. It is not yet clear how glycosylases might be involved in repair of gapped DNA. It is possible that glycosylases target downstream BER proteins to the integration intermediate. Other host factors have been identified that play a role during lentiviral but not retroviral infection. Significantly, LEDGF has been shown to enhance lentiviral integration by directly binding to lentiviral integrase and chromatin. Mouse cells with a deletion of the Ledgf gene have been engineered and show a pronounced defect in lentiviral infection and no effect on retroviral infection. While LEDGF is known to affect HIV integration to chromatin DNA targets, HIV PICs generated in Ledgf null cells have no integration defect with a naked DNA target. Results with HIV PICs from BER deficient cells Amsacrine hydrochloride indicate that BER affects integration to naked DNA. The ability of BER to direct integration to chromatin targets remains to be tested. BAF and HMGA1 proteins were also shown to stimulate HIV PIC integration activity, but reduced expression of these genes showed no effect on HIV infection efficiency. This is the first example of putative HIV integration co-factors that show a difference in the integration efficiency of PICs in vitro and infection efficiency in vivo. Retroviral integration sites display a subtle sequence preference unique to each virus.