Cellular factors leading to AZT resistance in HIV

Based on the structural analogy between AZT and some anticancer agents, we recently asked whether HIV-1 resistance to AZT can depend on cellular mechanisms of acquired drug-resistance in addition to those based on viral mutations. We recently reported that a lymphoblastoid cell line, CEMVBL100 which constitutively expresses high levels of the multi-drug resistance protein (P-glycoprotein), is less sensitive to the antigrowth and antiviral effects of AZT, suggesting that this protein can also interfere with the activity of this antiviral agent. Subsequently, we addressed the question whether AZT itself is able to induce cellular resistance to the drug. Indeed, CEM cells propagated in the presence of increasing concentrations of AZT (from 1 μM to 2000 μM) become resistant to the antigrowth and antiviral activity of AZT, but do not express detectable levels of P-glycoprotein. Sensitivity of these cells to other compounds, such as vinblastine, vincristine, ddI and ddC remained unchanged, indicating that, in contrast to P-glycoprotein-positive cells, AZT-induced resistance is specific for AZT. Interestingly, in AZT-induced resistant cells the intracellular accumulation of AZT, and of exogenous deoxythymidine (dT), are significantly reduced when compared with the parental cell line. Furthermore, the level of incorporation of ³H-dT into DNA of CEMAZT cells, determined by TCA precipitation, as well as thymidine kinase (TK) activity are lower than the values observed in CEM cells. To further characterize this phenomenon the kinetics parameters of TK activity have been defined in CEMAZT in comparison with the parental AZT sensitive CEM cells. The result revealed that the value of the maximum velocity (Vmax) of TK activity for dT phosphorylation is decreased in CEMAZT as compared to the wild type cell line (Vmax: CEM = 105.3 ± 17.6 nmol/hr/mg of protein; CEMAZT = 0.3 ± 0.02 nmol/hr/mg of protein; p <0.001). Furthermore, the enzyme affinity versus dT is lower in CEMAZT as compared to CEM (Km: CEM = 0.9 ± 0.2 μM; CEMAZT = 1.6 ± 0.2 μM; p <001). Consequently phosphorylation efficiency, expressed as the ratio between Vmax and Km, is also reduced in CEMAZT. To evaluate whether such a phenomenon may occur also in patients, ex vivo experiments were carried out by using PBMC from HIV-infected patients, treated or not treated with AZT. The results indicate that a prolonged treatment with AZT may modify the enzymatic kinetics of TK, leading to a significant reduction in the phosphorylation efficiency of the enzyme (4.07 ± 1.7 in treated patients versus 13.5 ± 1.7 in untreated patients; p <0.001) The results indicate that AZT treatment can induce a defect of the TK activity also in patients.