CEM-T4 Cells Do Not Lack an APOBEC3G Cofactor

Human APOBEC3G inhibits the replication of Vif-deficient HIV-1 by hypermutating nascent viral cDNA (reviewed in [1– 3]). Zheng and colleagues recently reported that the HIV-1 restriction activity of APOBEC3G requires a cellular co-factor [4]. Their conclusion depended on three critical observations: i) CEM-T4 cells support the replication of Vif-deficient HIV-1, ii) CEM-T4 cells express restrictive levels of APOBEC3G, and iii) CEM-T4 cells engineered to express more APOBEC3G still permitted Vif-deficient HIV-1 replication [4]. These observations suggested that APOBEC3G alone is insufficient for restriction, and together with subsequent cell fusion experiments, that a recessive cellular co-factor is required. However, the fact that APOBEC3G is capable of restricting a broad number of retroelements, including yeast Ty elements, strongly suggests that other human cellular proteins are not absolutely required for restriction (e.g., [5,6] and reviewed in [1–3]). An alternative explanation that could account for the observed permissive phenotype of the CEM-T4 line is that it is mixed, composed of a population of cells expressing low and/or variable levels of APOBEC3G. If this were the case, then the permissive phenotype could simply be due to virus replication in the subset of cells expressing low levels of APOBEC3G. To address this hypothesis, we generated subclones of the CEM-T4 line by serial dilution and determined the level of APOBEC3G expression by immunoblotting (Figure 1A). First, we observed that CEM-T4 cells expressed levels of APOBEC3G that were considerably lower than those in the non-permissive line CEM, regardless of whether they were obtained from the AIDS Research and Reference Reagent Program (CEM-T4-A) or directly from the Zheng laboratory (CEM-T4-Z). Second, we found that representative CEM-T4 subclones, regardless of source, expressed both low and variable APOBEC3G levels. It is further notable that none of the subclones expressed fully nonpermissive, CEM-like APOBEC3G levels. This heterogeneity is reflected by the kinetics of Vif-deficient virus replication, with some subclones being fully permissive and others being semi-permissive (Figure 1B Citation: Haché G, Harris RS (2009) CEM-T4 Cells Do Not Lack an APOBEC3G Cofactor. PLoS Pathog 5(7): e1000528. doi:10.1371/journal.ppat.1000528