Introduction.

HIV-positive pregnant women can decrease the risk for in-utero vertical HIV transmission by intake of antiretroviral drugs (ARVs). Zidovudine (AZT) during pregnancy is a frequently used and WHO- recommended drug regimen [1]. However, it has been proven in human and animal studies that Nucleoside Reverse Transcriptase Inhibitors (NRTIs) like AZT can cause mitochondrial damages including depletion of mitochondrial DNA (mtDNA) [2]–[10].

One underlying mechanism of AZT-induced mitochondrial toxicity is the inhibition of human DNA polymerase gamma [11]–[12], the enzyme needed for replication of mtDNA. Other assumed mechanisms include increased mitochondrial oxidative stress, introduction of mtDNA mutations, negative effects on nucleotide phosphorylation and mitochondrial gene expression, depletion of L-carnitine and inhibition of the mitochondrial bioenergetic machinery [13]–[17].

However, also HIV-1 infection itself causes mitochondrial damage, like depletion of mtDNA and decreased activities of the mitochondrial respiratory chain complexes [18]–[21]. HIV-1 has been shown to induce mitochondrial toxicity in several ways: by loss of mitochondrial membrane potential, by increase of reactive oxygen species and through different mechanisms of the viral proteins Vpr, Tat and HIV protease [22].

In humans, the mitochondrial toxicity of antenatal NRTI-exposure was determined by measuring different mitochondrial parameters like the emergence of clinical mitochondriopathy or death [23]–[25], quantification of mtDNA [18], [26]–[29], analysis of mtDNA mutations [30] or expression of mitochondrial respiratory chain proteins [27].

Studies indicating NRTI-induced mitochondrial toxicity include a detailed analysis by Barret [23], who found a higher incidence of neuro-mitochondrial diseases in NRTI-exposed infants compared to NRTI-unexposed infants; Divi [28] found a decrease of mtDNA in umbilical cords of infants of HIV-positive mothers exposed to Combivir compared to infants of HIV-negative women. Shiramizu [26] measured lower mtDNA contents in placenta and cord blood of HIV-positive women following NRTI-exposure in comparison to HIV-negative individuals. Torres [30] detected a higher frequency of mtDNA mutations in umbilical cords of HIV-positive AZT exposed infants compared to HIV-negative infants.

In contrast, McComsey [27] identified increased mtDNA levels without changes in expression of mitochondrial respiratory chain proteins in infants of HIV-positive mothers having taken NRTIs compared to NRTI-unexposed infants of HIV-negative mothers. Williams [31] did not detect lower mental or motor functioning scores in HIV-exposed, uninfected infants who were in-utero exposed to ARVs including NRTIs compared to those unexposed to ARVs during pregnancy. Accordingly, two large cohort studies did not discover an increased risk for death or clinical manifestations suggestive of mitochondrial abnormalities in NRTI-exposed infants [24]–[25].

The only two studies comparing mtDNA levels exclusively among HIV-positive mothers and their infants came to contradictory conclusions. In blood samples of HIV-positive mothers and infants with and without prenatal AZT exposure, Poirier [29] found lower mtDNA levels in AZT-exposed infants, whereas Aldrovandi [18] identified higher mtDNA levels in women and newborns with antenatal AZT exposure.

Altogether, it has not been clarified whether the net effect of short-course AZT for drug-naive HIV-1 infected pregnant women and their infants is a positive or a negative one with regard to mitochondriopathy. In the present study, we therefore quantified the mtDNA content in placentas of HIV-1 positive women with and without antenatal AZT exposure and in umbilical cords of their AZT exposed/unexposed infants. Furthermore, we checked for the most common mitochondrial deletion in humans, the 4977 base pair deletion (dmtDNA4977) as a marker for mitochondrial stress [32].

Abstract.
Background.

Zidovudine (AZT) constitutes part of the recommended regimens for prevention and treatment of HIV-1 infection. At the same time, AZT as well as HIV-1 infection itself may induce mitochondrial damage. In this study, we analyzed the impact of prenatal AZT-exposure on mitochondrial alterations in HIV-infected women and their infants.

Methods.

Mitochondrial DNA (mtDNA) levels in placentas of HIV-1 infected Tanzanian women with and without prenatal AZT exposure, and in the umbilical cords of their AZT-exposed/unexposed infants were quantified using real-time PCR. Furthermore, we checked for the most common mitochondrial deletion in humans, the 4977 base pair deletion (dmtDNA4977) as a marker for mitochondrial stress.

Results.

83 women fulfilled the inclusion criteria. 30 women had been treated with AZT (median duration 56 days; IQR 43–70 days) while 53 women had not taken AZT during pregnancy. Baseline maternal characteristics in the two groups were similar. The median mtDNA levels in placentas and umbilical cords of women (311 copies/cell) and infants (190 copies/cell) exposed to AZT were significantly higher than in AZT-unexposed women (187 copies/cell; p = 0.021) and infants (127 copies/cell; p = 0.037). The dmtDNA4977 was found in placentas of one woman of each group and in 3 umbilical cords of AZT-unexposed infants but not in umbilical cords of AZT-exposed infants.

Conclusions.

Antenatal AZT intake did not increase the risk for the common mitochondrial deletion dmtDNA4977. Our data suggests that AZT exposure elevates mtDNA levels in placentas and umbilical cords possibly by positively influencing the course of maternal HIV-1 infection.