Disruption of mitochondrial energetics and DNA synthesis by the anti-AIDS drug dideoxyinosine

Tenofovir renal toxicity targets mitochondria of renal proximal tubules

Tenofovir disoproxil fumarate (TDF) is an analog of adenosine monophosphate that inhibits HIV reverse transcriptase in HIV/AIDS. Despite its therapeutic success, renal tubular side effects are reported. The mechanisms and targets of tenofovir toxicity were determined using '2 x 2' factorial protocols, and HIV transgenic (TG) and wild-type (WT) littermate mice with or without TDF (5 weeks). A parallel study used didanosine (ddI) instead of TDF. At termination, heart, kidney, and liver samples were retrieved. Mitochondrial DNA (mtDNA) abundance, and histo- and ultrastructural pathology were analyzed. Laser-capture microdissection (LCM) was used to isolate renal proximal tubules for molecular analyses. Tenofovir increased mtDNA abundance in TG whole kidneys, but not in their hearts or livers. In contrast, ddI decreased mtDNA abundance in the livers of WTs and TGs, but had no effect on their hearts or kidneys. Histological analyses of kidneys showed no disruption of glomeruli or proximal tubules with TDF or ddI treatments. Ultrastructural changes in renal proximal tubules from TDF-treated TGs included an increased number and irregular shape of mitochondria with sparse fragmented cristae. LCM-captured renal proximal tubules from TGs showed decreased mtDNA abundance with tenofovir. The results indicate that tenofovir targets mitochondrial toxicity on the renal proximal tubule in an AIDS model.

Brain mitochondrial injury in human immunodeficiency virus-seropositive (HIV+) individuals taking nucleoside reverse transcriptase inhibitors

Nucleoside reverse transcriptase inhibitors (NRTIs) suppress human immunodeficiency virus (HIV) replication, but are often associated with mitochondrial toxicity. Although well studied outside of the central nervous system, no investigation has examined the effects of these drugs on brain mitochondria of individuals living with HIV. The authors used proton magnetic resonance spectroscopy to evaluate NRTI-related changes in brain mitochondria. N-acetylaspartate (NAA; sensitive to alterations in mitochondrial integrity) was measured in frontal lobe white and gray matter of 18 HIV+ individuals taking didanosine and/or stavudine (two NRTIs likely to cause mitochondrial toxicity), 14 HIV+ individuals taking zidovudine and lamivudine, 16 HIV+ individuals not currently taking antiretrovirals, and 17 HIV- controls. The HIV+ groups were comparable on demographic measures, estimates of illness severity, and estimated length of HIV infection. Those taking didanosine and/or stavudine had a significant 11.4% decrease in concentrations of frontal white matter NAA compared to HIV- controls, whereas NAA levels of the other HIV+ groups were intermediate. Group differences in metabolites were not found in frontal gray matter. Lower levels of frontal white matter NAA were associated with longer periods of didanosine and/or stavudine treatment (r = -.41, P = .06). Levels of NAA were not related to length of zidovudine/lamivudine treatment (r = -.04, P = .44). Furthermore, taking more than one of stavudine, didanosine, and abacavir increased the likelihood of having reduced NAA. The results are consistent with previous studies finding HIV-related changes in neuronal integrity. However, because NRTIs can injure mitochondria, we propose that the observed reductions in NAA in individuals taking didanosine and/or stavudine may be the result of depleted brain mitochondria and/or alterations in cellular respiration. Measurement of brain metabolites sensitive to impairments in energy metabolism, including NAA, may aid in early detection of subclinical NRTI-mediated mitochondrial toxicity.

Increased long-term mitochondrial toxicity in combinations of nucleoside analogue reverse-transcriptase inhibitors

BACKGROUND

Some nucleoside analogue reverse transcriptase inhibitors (NRTI) may cause depletion of mitochondrial (mt) DNA in liver by inhibiting polymerase-gamma. mtDNA depletion may contribute to lactic acidosis, steatohepatitis and liver failure.

OBJECTIVE

To evaluate the long-term mitochondrial toxicity of NRTI combinations.

METHODS

The HepG2 human hepatoma cell line was cultivated in the presence of zalcitabine (ddC), didanosine (ddI), stavudine (d4T), lamivudine (3TC), zidovudine (ZDV) and efavirenz at concentrations equivalent to steady-state peak plasma levels (C ), and also in one-third and 10 times C. The NRTI were added to the medium alone or in combination. Control cells were incubated without any NRTI or with efavirenz. Cell growth, lactate production, intracellular lipid droplets, mtDNA and the mtDNA-encoded respiratory chain subunit COX II were monitored over a period of up to 30 days.

RESULTS

Time- and dose-dependent mtDNA depletion was observed with ddC > ddI > d4T and mtDNA depletion preceded or coincided with a decline in COX II expression, a decrease in cell growth, increased lactate production and increased intracellular lipids. 3TC and efavirenz did not affect any measurement. ZDV increased lactate moderately and cell growth was inhibited, despite normal mtDNA and COX II levels. The negative effects on some measurements were more pronounced in the 3TC-ZDV and ddC-d4T combinations, than in the single-NRTI incubations. The combination of ddI-d4T was not more toxic than ddI alone. Mitochondrial damage by ZDV, d4T, ddI, and ddC did not reach steady-state by day 25. Using a Southern blot technique, mtDNA deletions were never observed.

CONCLUSION

The data indicate additive or synergistic long-term mitochondrial toxicity in some NRTI combinations.

Pharmacology of nucleoside and nucleotide reverse transcriptase inhibitor-induced mitochondrial toxicity

OBJECTIVE

This paper reviews the function of the mitochondria and the mechanisms by which nucleoside and nucleotide reverse transcriptase inhibitors (NRTIs) cause mitochondrial toxicity.

BACKGROUND

Highly active antiretroviral therapy (HAART) reduces rates of morbidity and mortality due to HIV disease. However, long-term treatment with these drugs may be associated with adverse effects. Nucleoside and nucleotide analogues are potent inhibitors of HIV reverse transcriptase and have become the cornerstone of HAART. Unfortunately, these drugs have also been shown to inhibit cellular polymerases, most notably mitochondrial DNA polymerase gamma.

RESULTS

Studies of the NRTIs in enzyme assays and cell cultures demonstrate the following hierarchy of mitochondrial DNA polymerase gamma inhibition: zalcitabine > didanosine > stavudine > lamivudine > zidovudine > abacavir. In vitro investigations have also documented impairment of the mitochondrial enzymes adenylate kinase and the adenosine diphosphate/adenosine triphosphate translocator. Inhibition of DNA polymerase gamma and other mitochondrial enzymes can gradually lead to mitochondrial dysfunction and cellular toxicity. The clinical manifestations of NRTI-induced mitochondrial toxicity resemble those of inherited mitochondrial diseases (ie, hepatic steatosis, lactic acidosis, myopathy, nephrotoxicity, peripheral neuropathy, and pancreatitis). Fat redistribution syndrome, or HIV-associated lipodystrophy, is another side effect attributed in part to NRTI therapy. The morphologic and metabolic complications of this syndrome are similar to those of the mitochondrial disorder known as multiple symmetric lipomatosis: suggesting that this too may be related to mitochondrial toxicity. The pathophysiology of less common adverse effects of nucleoside analogue therapy, such as diabetes, ototoxicity, and retinal lesions, may be related to mitochondrial dysfunction but have not been adequately studied.

CONCLUSION

NRTls can block both HIV reverse transcriptase and mitochondrial DNA polymerase gamma. Inhibition of the latter enzyme is the most likely cause of the adverse effects associated with these drugs.

Effects of zidovudine (AZT) and dideoxyinosine (ddI) on human trophoblast cells

The anti-HIV agents AZT (zidovudine) and ddl (dideoxyinosine) are being used clinically during pregnancy. The toxicity of these agents to the fetus and placenta remains a concern because few human pregnancy exposure data are available, and pregnant rodent studies with AZT indicate increased embryonic resorptions and developmental arrest. The current study used a human choriocarcinoma cell line (JAr), which exhibits many characteristics of the early placenta, to assess the effects of a single 24 h exposure of 7.6 or 0.076 mM AZT, and the effects of a single 24 h exposure of 7.6 or 0.076 mM ddI upon cell proliferation and hormone production of human chorionic gonadotropin (hCG), estradiol (E2), and progesterone (P4). The higher concentration of AZT and ddI produced significant (P < 0.025) reductions in cell numbers and growth rate while producing significant increases in hormone production (hCG, E2, and P4). The lower concentration of AZT and ddI produced significant increases in E2 production, but no changes in cell numbers, hCG, or P4. Because placental cells require androgen precursor for E2 synthesis, exogenous androstenedione was added to confirm observations of increased estradiol synthesis after AZT or ddl exposure. These results demonstrate that single 24 h high dose exposures of AZT or ddI produce significant inhibition of cell proliferation and alterations in hormone production in this paradigm of human placental cells.

Mitochondrial, but not peroxisomal, beta-oxidation of fatty acids is conserved in coenzyme A-deficient rat liver

Hepatic coenzyme A (CoA) plays an important role in cellular lipid metabolism. Because mitochondria and peroxisomes represent the two major subcellular sites of lipid metabolism, the present study was designed to investigate the specific impact of hepatic CoA deficiency on peroxisomal as well as mitochondrial beta-oxidation of fatty acids. CoA deficiency (47% decrease in free CoA and 23% decrease in total CoA) was produced by maintaining weanling male Sprague-Dawley rats on a semipurified diet deficient in pantothenic acid (the precursor of CoA) for 5 weeks. Hepatic mitochondrial fatty acid oxidation of short-chain and long-chain fatty acids were not significantly different between control and CoA-deficient rats. Conversely, peroxisomal beta-oxidation was significantly diminished (38% inhibition) in livers of CoA-deficient rats compared to control animals. Peroxisomal beta-oxidation was restored to normal levels when hepatic CoA was replenished. It is postulated that since the role of hepatic mitochondrial beta-oxidation is energy production while peroxisomal beta-oxidation acts mainly as a detoxification system, the mitochondrial pathway of beta-oxidation is spared at the expense of the peroxisomal pathway when liver CoA plummets. The present study may offer an animal model to investigate mechanisms involved in peroxisomal diseases.

Immunotoxicity of 2',3'-dideoxyinosine in female B6C3F1 mice

2',3'-dideoxyinosine (ddI) is one of several purine analogues used for the treatment of HIV and the acquired immunodeficiency syndrome (AIDS). These nucleoside analogues are promising in their inhibition of viral reverse transcriptase and termination of DNA synthesis. However, each of these drugs has toxicity associated with its use. A previous immunotoxicological evaluation of 2',3'-dideoxyadenosine (ddA), the parent compound of ddI, showed that ddA suppresses humoral immunity. These studies were undertaken to determine the potential for immunotoxicity due to treatment with ddI. This evaluation included an assessment of innate and acquired immunity after exposure to ddI (100, 250, 500, and 1000 mg/kg/day) for 14, 28 or 180 days. There were no overt signs of toxicity related to treatment with ddI except for a decrease in body weight in the group treated with the highest dose of ddI for 180 days. Overall, 6 months of treatment with ddI showed minimal effects on specific organs with the exception of the spleen and thymus. ddI selectively targets the immune system, with assays that challenge humoral immunity being more affected than those testing cell-mediated immunity. Innate immunity was unaffected by ddI treatment. Cell-mediated immunity, as measured by proliferative response to allogeneic cells (MLR) and the T cell mitogen (Concanavalin A), was moderately suppressed. There were no ddI associated effects on NK function or macrophage function as measured by the vascular clearance rate and phagocytic uptake of the tissue macrophages. The most sensitive indicator of ddI-induced immunotoxicity is suppression of the response to the T-dependent antigen, sheep red blood cells (sRBC). The No Observable Adverse Effect Level (NOAEL) for toxicity to the immune system following 14 days of exposure to ddI is 250 mg/kg. A suppression of the humoral immune response was seen at the lowest dose tested after treatment for 28 and 180 days. Thus, the NOAEL for both of these treatment periods is below 100 mg/kg/day.

2'3'-Dideoxyinosine inhibits the humoral immune response in female B6C3F1 mice by targeting the B lymphocyte

2',3'-Dideoxyinosine (ddI) is a purine nucleoside analog currently being used for the treatment of HIV-positive individuals and patients with AIDS. Preliminary immunotoxicity studies have shown that a consequence of ddI treatment in female B6C3F1 mice is the inhibition of the humoral immune response. This effect was dose dependent in a range of 100 to 1000 mg/kg with a no observed adverse effect level of less than 100 mg/kg for a 28-day treatment period. These studies were undertaken to investigate the immune cell target of ddI and to determine the mechanism of this toxicity. B6C3F1 mice were treated with 1000 mg/kg/day by oral gavage for 28 days. The B lymphocyte was identified as the cellular target of ddI through separation-reconstitution experiments of the adherent and nonadherent cell populations and of the T and B lymphocyte populations. These studies revealed a deficit in the ability of the nonadherent cells from ddI-treated mice to mount a normal antibody response to sRBC. A further separation of the nonadherent cells into T and B cells revealed a decreased ability of ddI-treated B cells to develop specific humoral immunity. Additional studies were undertaken to determine the mechanism by which ddI is affecting the B cell. Surface marker analysis of splenocytes revealed no difference in the cell populations between vehicle- and ddI-treated mice. B cell proliferation was also unaffected as shown by incubation with either a polyclonal stimulator, lipopolysaccharide, or anti-IgM plus IL-4. These results indicate that the primary cellular target of ddI is the B lymphocyte.

Mechanism of phthalate-induced inhibition of hepatic mitochondrial beta-oxidation

Studies show that peroxisome proliferators inhibit mitochondrial beta-oxidation of fatty acids. However, mechanism(s) of this inhibitory effect has not been identified. This study was undertaken to delineate such mechanism(s). Ketogenesis was significantly diminished in perfused livers from rats pre-treated with diethylhexyl phthalate (DEHP) compared with livers from control rats. Monethylhexyl phthalate (MEHP; 200 microM), a primary metabolite of DEHP and a known peroxisome proliferator, inhibited the oxidation of palmitic acid as well as its acyl-CoA and acylcarnitine derivatives in isolated mitochondria by about 50-60%. Similar concentrations of MEHP also inhibited mitochondrial respiration of succinate and malate plus glutamate. However, respiration of ascorbate was not influenced by MEHP. Considering the assembly of the mitochondrial respiratory chain, these data indicate that phthalates inhibit fatty acid metabolism as a result of inhibiting the respiratory chain at the level of the cytochrome c reductase. This effect may represent an early step in the mechanism by which phthalates cause hepatic peroxisome proliferation.

Randomized study of two doses of didanosine in children infected with human immunodeficiency virus

2'3'-Dideoxyinosine (didanosine) is a nucleoside analog active in vitro against human immunodeficiency virus. Few data are available regarding its use for the treatment of children. In a single-center, randomized, open-label trial, we compared two dosages of didanosine (120 vs 270 mg/m2 per day) for at least 6 months in 34 children infected with human immunodeficiency virus who had become resistant to or were intolerant of zidovudine. Serum levels of didanosine 1 hour after administration were significantly different in the two groups and remained stable with time. There was a significant reduction in human immunodeficiency virus-p24 antigenemia and quantitative cellular viremia with time but no difference between the two groups. The intensity of the biologic response, however, was significantly higher in the patients who had more than 50 CD4+ cells 10(6)/L at inclusion. No pancreatic or neurologic toxic effects were observed. In five children, liver function abnormalities developed that are unusual in this setting, and the death of one child from unexplained hepatocellular failure suggests that didanosine may be hepatotoxic. Three of these five children had preexisting liver disease. Although no definite conclusion can be made as to the optimal dose, there were no major differences between the two administration schedules in terms of biologic effects and tolerability.