HIVresistanceWeb - Perspectives and Opinions

Cellular resistance to AZT

written by Charles Boucher, M.D.
published on HIVresistanceWeb: March 1, 1998

Several reports have been published over the last couple of years which show that in the test tube it is possible to makes cell cultures resistant to AZT. This cellular resistance is caused by changes in the cellular enzymes that convert thymidine—the natural substrate and one of the components of DNA—into its active form. The structure of AZT is almost identical to that of deoxythymidine (a deoxythymidine analogue). As a consequence, the same cellular enzymes that activate thymidine also activate AZT. The modification in the structure of AZT has the result that once AZT is incorporated into the DNA chain no further nucleotides can be incorporated. But only activated AZT can be used by the enzyme that catalyzes DNA production. This activation consists of the addition of three phosphate groups to AZT (or thymidine), and each phosphorylation step is performed by a different enzyme.

In vitro studies by Dianzani and colleagues from the Institute of Virology, University La Sapienza (Rome) show that growing cells in the presence of very high AZT concentrations can lead to the selection of a AZT resistant cell line. These cells were found to exhibit a reduction in the activity of thymidine kinase, the enzyme responsible for the addition of the first phosphate group. The activity of this enzyme in the resistant cell line was dramatically reduced, and as a consequence, AZT could not be activated and incorporated in the growing DNA chain.

To evaluate whether this phenomenon also occurs in patients, ex vivo experiments were carried out using peripheral blood mononuclear cells (PBMC) from HIV-infected patients in different disease stages, who were treated or not treated with AZT. The function of thymidine kinase in PBMC from ten AZT-treated (500 mg/day, mean duration of therapy 19.9 months) and ten AZT naïve patients was then compared. The kinetic parameters of thymidine kinase acticity were studied using deoxythymidine as a substrate (according to the authors, the overall level of AZT phosphorylation in PBMC is low, preventing the evaluation of thymidine kinase function using AZT ).

Analyzing cell extracts, the authors found that both the mean binding affinity and the mean maximum velocity of thymidine kinase activity differed significantly between the AZT exposed and AZT naïve patients. However, the magnitude of these differences was not very pronounced.

This paper is the first published to demonstrate that AZT treatment of HIV-infected patients may affect the efficiency of AZT phosporylation by thymidine kinase, and Dianzani et als' findings support the hypothesis that cellular metabolism of AZT wanes over time. If this does occur in vivo, then clinical AZT resistance may result in part from the lowered availablity of activated AZT caused by changes in the cellular enzymes which normally convert the drug to its active form.

A limitation of this study is that the authors used cell extracts to evaluate thymidine kinase kinetics. It therefore remains difficult to exclude the possibility that the findings are mainly due to proportional differences in the cell populations analyzed in the AZT-treated and -untreated groups. If this was the case, differences in the number of cells expressing thymidine kinase could account for the differences observed in the study. The authors imply that their findings may also be explained by a reduction in the amount of thymidine kinase expression per cell. More experiments are required to confirm the concept that prolonged AZT exposure can lead to a reduction in the capacity of PBMC to convert AZT to it active, triphosphate form.

Dianzani et als' data are in contrast to findings reported by Back et al in a late breaker session at the 5th Conference on Retroviruses and Opportunistic Infections in Chicago. Back et al used a different and more direct approach to evaluate the effect of AZT treatment on the phosphorylation of AZT over time and in vivo. Instead of evaluating the kinetics of the phosphorylating enzyme(s) in vitro, they directly measured intracellular concentrations of AZT and its mono-, di-, and triphosphate forms in PBMC from patients. Ten patients (who were exposed to AZT for at least twelve months in combination with either ddI or 3TC) were evaluated after two weeks and 1,2,3,6,9 and 12 months. Intracellular concentrations of total AZT phosphate levels or AZT triphosphate levels were not different when analyzed with a HPLC-RIA method. Back and colleagues concluded from their data that AZT phosphorylation does not systematically decrease over time.

In another presentation at the Retroviruses meeting, however, Sommadossi and colleagues reported in vivo data regarding the phosphorylation of two other nucleoside analogues—3TC and d4T—by PBMC of patients not responding to combination d4T/3TC therapy after AZT pretreatment. The amount of d4T and 3TCtriphosphate in PBMC were com pared in three groups of patients. d4T and 3TC triphosphate levels were significantly lower in patients who were preexposed to AZT and not responding to combination d4T/3TC than in AZT naïve patients responding to this combination. Sommadossi et al claim that the absence of a treatment response to d4T/3TC therapy can be accounted for by a defect(s) in nucleoside analogue phosphorylation resulting from pretreatment with AZT.

Again, similar to the study by Dianzani et al, the numbers of patients in the study were small and the differences in intracellular concentrations of triphosphate drug levels were modest. In conclusion, there is growing interest in the relationship between the cellular metabolism of nucleoside analogues and the antiviral activity of these compounds in vivo. Whereas some studies find differences in the efficiency of phosphorylation of AZT or other nucleoside analogues in patients exposed to AZT, another study could not confirm this. It is clear that additional studies will be required to clarify this area of investigation. Until the discrepancies cited above are resolved, one should be cautious in applying the conclusions of any of the above studies in clinical practice.

References

Long term exposure to zidovudine affects in vitro and in vivo the efficiency of phosphorylation of thymidine kinase. Antonelli G; Turriziani O; Verri A; Narciso P; Ferri F; D'Offizi G; Dianzani F. AIDS Res Hum Retroviruses.1996 Feb 10;12(3):223-8.
Zidovudine phosphorylation detemined at intervals over 12 months in naïve and antiretroviral experienced HIV+patients Back DJ, Hoggard PG, Gibbons SE, Lloyd J, Khoo SH, Wilkins EG, Merry C, Barry MG. 5th Conference on Retroviruses and Opportunistic Infections, February 1-5, 1998, Chicago, Illinois. LB 10 page 224.
Intracellular phosphorylation of stavudine (d4T) and 3TC correlates with their antiviral activity in naïve and zidovudine experienced patients Sommadossi J-P , Valantin MA, Zhou X-J, Xie M-Y, Moore J, Calvez V, Desa M, Katlama C. 5th Conference on Retroviruses and Opportunistic Infections, February 1-5, 1998, Chicago, Illinois. Abstract 362 page 146.

(click titles to read abstracts)

back to the top of this page