Although children born to HIV-infected (HIV+) women receiving antiretroviral therapy during pregnancy show virtually no adverse clinical effects at birth, the antiretroviral nucleoside analog drugs are known to damage nuclear and mitochondrial DNA. In this study, biomarkers of mitochondrial toxicity and genotoxicity have been examined in a well-characterized sample set consisting of infants born to HIV-uninfected (HIV-) mothers (n = 30), and HIV- infants (n = 20) born to HIV-infected (HIV+) mothers who received either no antiretroviral therapy (n = 10) or zidovudine (3`-azido-3`-deoxythymidine [AZT]) during pregnancy (n = 10). DNA from cord blood leukocytes and peripheral blood leukocytes taken at 1 and 2 years of age was examined for loss of mitochondrial DNA (mtDNA) and telomere integrity. Telomere length, a measure of nuclear DNA damage, was the same in all infants at birth and at age 1 year. The quantity of mtDNA was assessed relative to nuclear DNA using a polymerase chain reaction-based chemiluminescence detection (PCR-CID) method that determined mitochondrial D Loop gene copies relative to nuclear 18S RNA gene copies by comparison with a standard curve. MtDNA quantity was expressed as a ratio of gene copy numbers. In infants of uninfected mothers (AZT-/HIV-) at the three time points, the ratios were 442 to 515, whereas in infants of untreated AZT-/HIV+ mothers the ratios were 261 to 297, and in infants of AZT-treated (AZT+/HIV+) mothers the ratios were 146 to 203. At all three time points, differences between the AZT-/HIV- group and the two HIV+ groups were statistically significant (p < .05), and differences between the AZT-/HIV+ and AZT+/HIV+ groups were also statistically significant (p < .05), demonstrating that AZT exposure causes a persistent depletion of mtDNA. The study shows that children of HIV+ mothers are at risk for mitochondrial damage that is further increased in infants of mothers receiving AZT during pregnancy.
The current spread of HIV to the heterosexual population in the United States has resulted in approximately 7,000 children born yearly to HIV-infected mothers. When given during pregnancy, the antiretroviral nucleoside analog zidovudine (3`-azido-3`-deoxythymidine [AZT]) significantly reduces maternal-fetal viral transmission and thereby protects more than 1,000 children per year in the United States (1-3). The Centers for Disease Control's recommended treatment (from Protocol ACTG076 [2] and http://www.hivatis. org/guideline/perinatal) is AZT given for the last 6 months of pregnancy, intravenous AZT dosing during labor and delivery, and 6 weeks of AZT administration to the infant after birth. In current clinical practice, the widely used drug combivir plus lamivudine, which is a second nucleoside analog (3`-dideoxy-3`-thiacytidine [3TC]), is more effective than AZT alone in inhibiting maternal-fetal HIV transmission. A number of studies have concluded that there are no persistent clinical changes in nucleoside analog-exposed infants compared with unexposed infants (4-8). The studies presented here were performed because mitochondrial compromise has been observed in some children exposed transplacentally to AZT (9) and telomere length shortening was observed in newborn mice transplacentally exposed to tumorigenic levels of AZT (10).
Long-term AZT use in adults has been associated with mitochondrial toxicity that is manifested in skeletal muscle and cardiac wasting (11-14), elevated serum lactic acid (15,16), abnormal oxidative phosphorylation (OXPHOS) enzyme-specific activities (17,18), depletion in the quantity of mitochondrial DNA (mtDNA) (12,14), and morphologically damaged mitochondria with disrupted cristae and vacuolization (17,19-21). Until recently, similar nucleoside analog-induced mitochondrial toxicity was thought not to occur in HIV-uninfected infants born to HIV-infected mothers. However, Blanche et al. (9) reported several nucleoside analog-exposed children between the ages of 7 months and 4.4 years with clinical mitochondrial dysfunction. The same study (9) reported the deaths of two children, at 11 and 13 months of age, with severe persistent mitochondrial dysfunction. In contrast, a retrospective analysis of infant mortality in several clinical trials (22) and a study of echocardiograms in exposed HIV-uninfected infants (23) did not provide evidence of clinical mitochondrial toxicology after in utero exposure to AZT and/or 3TC. Further investigation will be required to resolve the apparent controversy.
Several studies have demonstrated that AZT targets telomeric DNA. Preferential incorporation of AZT and subsequent shortening of telomeric DNA occur in AZT-exposed cultured cells (24-26). The tumorigenic properties of AZT in a transplacental mouse model have been described (10,27), and newborn mice in the same study had AZT incorporated into DNA and shortened telomeres in many of the same organs from which tumors arose (10). If telomere length is altered in DNA of human infants exposed to AZT, then there may be long-term consequences related to lifespan or other functions.
The Women and Infants Transmission Study (WITS), an ongoing, multi-center, prospective, cohort investigation of HIV-infected women and their infants, provides a well-characterized set of biological samples (28). The current analysis used DNA from infant cord blood leukocytes and peripheral blood leukocytes taken at ages 1 year and 2 years from HIV-uninfected (HIV-) infants of HIV-infected (HIV+) women, all of which were supplied by the WITS. The samples were obtained between 1990 and 1995, spanning a time when many HIV-infected women did not receive AZT or other antiretroviral therapy. Using the WITS leukocyte samples and age-matched leukocytes from HIV- pregnancies, we have selected a panel of biomarkers to investigate toxicities known to be associated with antiretroviral nucleoside analog use. The study included three groups: infants born to uninfected mothers (AZT-/HIV-), infants born to HIV+ mothers receiving no antiretroviral drugs (AZT-/HIV+), and infants born to HIV+ women receiving AZT (AZT+/HIV+). All samples were examined for mtDNA quantity and telomere length.