McComsey, Grace MD*; Bai, Ren-Kui PhD†§; Maa, Jen-Fue PhD‡; Seekins, Daniel MD‡; Wong, Lee-Jun PhD†
The advent of antiretroviral (ARV) therapy has led to an unsurpassed reduction in HIV-related mortality and morbidity in HIV-infected subjects, but this success has been threatened by long-term toxicities. These include complications linked to mitochondrial toxicities such as peripheral neuropathy, lactic acidosis, myopathy, and cardiomyopathy. Others, like lipoatrophy, still have an unclear etiology, although recent evidence points to the potential for mitochondrial toxicity as a pathogenic factor.1,2 The mitochondrial toxicity of nucleoside analogue reverse transcriptase inhibitor (NRTI) therapy results from its ability to inhibit DNA polymerase γ, the enzyme necessary for the replication of mitochondrial DNA (mtDNA).3 Muscle or fat seems to be a reliable tissue when studying NRTI-induced mitochondrial toxicities4-6; however sampling is invasive and impractical for routine use in clinical practice. Thus, the determination of mtDNA levels in peripheral blood, which is easily accessible, has been investigated as a marker of lipoatrophy or mitochondrial toxicities in general.7-11 Available data on blood mtDNA levels are conflicting; our pilot study7 and others8,9 showed no depletion of peripheral blood mononuclear cell (PBMC) mtDNA of patients with lipoatrophy, whereas others did.11 The discrepancies could be caused by the variability of mtDNA content among individuals, methods of study, tissues studied, type of ARV agent used, or duration of ARV therapy at the time of blood sampling. In past investigations in the HIV setting, studies looked at mtDNA deletions, depletions, and a few specific mutations but not at the entire mtDNA genome to correlate with toxicity. Our present study assessed the utility of quantitative and qualitative analysis of mtDNA in longitudinal peripheral blood samples from the same individual, with the performance of mtDNA depletion and extensive mutation analysis.
Study Design and Patient Population
The study included a cross-sectional component and a longitudinal component.
For the cross-sectional component, 3 groups of subjects were enrolled: (1) HIV-infected ARV-treated patients, (2) HIV-infected patients naive to all ARV agents, and (3) HIV-negative healthy individuals.
Within the HIV-infected treated group, subjects were further classified by the presence or absence of clinical lipoatrophy, defined as patient self-report of fat wasting in the face, buttocks, or limbs, as confirmed by the investigator, and the presence or absence of clinical mitochondrial toxicities, including symptomatic hyperlactatemia or lactic acidosis, otherwise unexplained peripheral neuropathy, and cardiomyopathy, which, in the opinion of the investigator, was the result of ARV therapy. Symptomatic hyperlactatemia was defined as a nonexercise venous lactate level of more than twice the upper limit of normal (ie, lactate >4.8 mmol/L), drawn as per the recommendations of the AIDS Clinical Trials Group (ACTG),12 in addition to consistent symptoms such as nausea, vomiting, abdominal pain, weight loss, anorexia, or unexplained liver enzyme abnormalities. Written informed consent was obtained from each patient before enrollment, and this study was approved by the Institutional Review Board at the University Hospitals of Cleveland (Cleveland, OH).
The longitudinal components of the study included follow-up evaluations for 2 groups of subjects: the ARV-treated group and HIV-positive ARV-naive group.
The evaluations were similar for the initial cross-sectional part and the later longitudinal part of the study. At each study evaluation, subjects answered a patient questionnaire with emphasis on changes in body habitus and underwent a physical examination with special emphasis on signs of fat loss and blood testing for serum electrolytes, anion gap, glucose, liver enzymes, fasting cholesterol, triglycerides, and blood lactate (collected per ACTG guidelines). In addition, the CD4 cell count and HIV-1 RNA level were extracted from the subjects' clinical chart within 2 months of the study evaluation. Blood was also collected for mtDNA depletion and mutation analysis. These tests were performed at the laboratory of one of the authors (L.-J.W.), Georgetown University (Washington, DC).
Total DNA was isolated from peripheral blood using the salting-out method.13 The concentration of DNA was determined using a DyNA Quant 200 fluorometer with Hoechst dye 33258. The DNA solution was diluted appropriately for real-time quantitative polymerase chain reaction (RT qPCR) analysis to obtain optimal results. Two microliters of the diluted solution containing 2 ng/μL of total DNA was used in 20 μL of RT qPCR assay.
Primers and TaqMan Probes
The primers and TaqMan probes used for measurement of mtDNA content in peripheral blood cells are listed in Table 1. Two mtDNA regions were used to determine mtDNA copy numbers. The mtDNA tRNAleu region is usually not deleted and the mtDNA ND4 region is usually deleted in deletion molecules. Two single-copy nuclear genes, the β2 microglobulin (β2M) gene and amplified in breast cancer 1 (AIB1) gene (also known as the nuclear receptor coactivator 3 [NCOA3] gene), were used as the control nuclear genes for normalization of mtDNA content. To minimize primer mismatch, the regions containing no reported mtDNA mutations or polymorphisms were chosen for primer design. Deletion of the tRNAleu gene region was observed in approximately 3% (16 of 525) of mtDNA deletion molecules. Conversely, deletion of the ND4 region has been observed in approximately 97% (508 of 525) of deletion mtDNA molecules (http://www.mitomap.org, accessed February 10, 2003).
The TaqMan probes containing 19 to 26 mers of oligonucleotide at the middle of the target sequences were labeled with a fluorescent reporter, 6-carboxy-fluorescein (FAM) or proprietary fluorescent dye (VAC, Applied Biosystems, Foster City, CA) at the 5′ end and a quencher 6-carboxy-tetramethyl rhodamine (TAMRA) at the 3′ end. The close proximity of the reporter and quencher prevents emission of fluorescence while the probe is intact. The probes bind to the corresponding specific target sequence. During the reaction, the 5′-to-3′ exonuclease activity of Taq DNA polymerase cleaves the probes and separates the reporter dye from the quencher if the probe is binding to the target sequence in the PCR products. This results in increased fluorescence as amplification proceeds and is recorded as the real-time of each cycle.
Real-Time Quantitative Polymerase Chain Reaction
The real-time PCR reaction was performed in triplicate for each probe. The 20-μL PCR reaction contained 1 × Platinum qPCR SuperMix-UDG Master Mix (Invitogen, catalog no. 11730-025), 300 nM of each primer, 100 nM of TaqMan probe, 0.4 μL of Rox dye (supplied by the manufacture), and 4 ng of total genomic DNA extract. Real-time PCR conditions were 2 minutes at 50°C and 10 minutes at 95°C, followed by 45 cycles of 15 seconds of denaturation at 95°C and 60 seconds of annealing and/or extension at 60°C. Fluorescent signal intensity was recorded and analyzed on an ABI-Prism 7700 Sequence Detector (Applied Biosystems, Foster City, CA) using SDS version 1.9 software (Applied Biosystems, Foster City, CA). The threshold cycle (CT) value within the linear exponential increase phase was used to construct the standard curve and to measure the original copy number of the DNA template. If a sample had a measurement greater than 10,000,000 or less than 1000 copies of mtDNA or a measurement greater than 1,000,000 or less than 100 copies of β2M or AIB1 gene, the assay was repeated at a lower or higher dilution of the DNA extract so that the measurement would fall within a linear range of the DNA copy number. Strict quality control measures were taken. First, all samples from the same patient were performed on the same run and same plate. Second, we used the same dilution of the reagents to perform real-time PCR for all the samples. Finally, 4 control DNA samples were included in each run to ensure that results from different runs were comparable.
Preparation of Standard DNA
Standard DNA for each target sequence was generated by cloning the PCR products of the corresponding gene into the pCR2.1-TOPO vector. The PCR products of the mtDNA ND4 and β2M genes are those listed in Table 1. The mtDNA-tRNAleu region was amplified with mtF3212 and mtR3471, a 260-base pair (bp) fragment. Exon 5 of the AIB1 gene was amplified using forward primer 5′ CAAGCGATCAAATGAGGGTAG 3′ and reverse primer 5′ CATTGTTTCATATCTCTGGCG 3′, giving a 439-bp PCR product. The copy number was calculated based on the molecular weight of the plasmid DNA. Serial dilutions were made, and RT qPCR reactions were performed to construct the standard curve for each gene.
Mutational Analysis of the Entire Mitochondrial Genome
Thirty-two overlapping PCR primers were used to amplify the entire mitochondrial genome,14,15 followed by temporal temperature gradient gel electrophoresis (TTGE) of the PCR products. DNA fragments that showed different TTGE patterns in longitudinal samples from the same individual were further analyzed by direct DNA sequencing using the BigDye terminator (version 3.1) cycle sequencing reagent kit and an ABI 377 (Applied Biosystems) automated sequencer.
A total of 203 venous blood specimens from 135 subjects were analyzed. For cross-sectional data, unadjusted univariate analyses for between-group comparison were performed using a Student t test with the P value evaluated by the Bonferroni adjustment for multiple comparisons. Lipoatrophy and mitochondrial toxicity were grouped separately for the ARV-treated subjects. On multivariate analysis, baseline and host factors were included for the selection of predictors for mtDNA depletion using a Proc Mixed procedure and for the development of lipoatrophy using a Proc Genmod procedure (SAS, version 8.2). The variables included in the models were age, sex, race, nadir CD4 cell count, duration of NRTI therapy, known duration of HIV infection, and NRTI types, and the laboratory markers included alanine aminotransferase (ALT), triglyceride, lactate levels, and mtDNA content. With the adjustment of the significant baseline and host factor covariates, the laboratory markers were then included in the second model for the selection of significant markers and evaluated at the 0.05 level. The mtDNA copy numbers were log transformed in the model because of nonnormality of the mtDNA levels. Summary statistics were presented as mean ± SD unless otherwise specified.
Cross-Sectional Phase of Study
The study involved a total of 203 venous blood specimens: 155 blood samples were collected from 87 HIV-infected subjects (108 specimens from 54 ARV-treated subjects and 47 from 33 ARV-naive subjects), and 48 specimens were collected from 48 age-matched healthy control subjects. The demographics and clinical and laboratory parameters of the study participants are shown in Table 2. Blood samples were drawn between January 1998 and January 2004.
Longitudinal Phase of Study
Among the 87 HIV-infected subjects enrolled in this study, 49 participated in the prospective longitudinal study, with a median of 2 (range: 2-6) assessments. Eleven of the 49 subjects remained naive to all ARV therapy throughout the study, and the median duration between samples was 13 months (range: 6-24 months). For the remaining 38 subjects in the longitudinal study, the median duration of follow-up (ie, between first and last study evaluation) was 13 months (range: 3-24 months). Characteristics of the 38 subjects who were treated with ARV therapy and had participated in the longitudinal phase of the study are shown in Table 3.
Among ARV-treated subjects, 67% (n = 36) had clinical lipoatrophy and 17% (n = 9) had clinical mitochondrial toxicity. These included 4 cases of peripheral neuropathy that, in the opinion of the investigator, were clearly related to NRTI therapy, 2 cases of lactic acidosis, and 2 cases of symptomatic hyperlactatemia. Additionally, 1 case of cardiomyopathy was included; the cardiomyopathy was documented on APV/3TC/abacavir (Trizivir) therapy and completely reversed after switching to an NRTI-sparing regimen. No other changes in medication or clinical management were performed. During the study, 5 subjects switched their ARV therapy: 2 secondary to virologic failure, 1 because of cardiomyopathy, and 2 because of lactic acidosis.
Mitochondrial DNA Depletion Data
On univariate analysis, the mtDNA/nuclear DNA (nDNA) ratio (mtDNA content) was similar in HIV-infected ARV-naive subjects and in HIV-negative controls (66 ± 27 vs. 68 ± 18, respectively; P = 0.7353). When compared with HIV-infected ARV-naive subjects, ARV-treated subjects had significantly lower average mtDNA content (53 ± 23 vs. 66 ± 27, respectively; P = 0.0021). Similarly, HIV-treated subjects had a more significantly lower average mtDNA content when compared with HIV-negative controls (53 ± 23 vs. 68 ± 18, respectively; P = 0.0005). As illustrated in Figure 1, longitudinal measurements of mtDNA in subjects initiating their first ARV regimen showed a decrease in mtDNA levels over time.
Among ARV-treated subjects, those with established lipoatrophy had similar mtDNA content as ARV-treated subjects without lipoatrophy (50 ± 24 vs. 57 ± 22, respectively; P = 0.1138). Similar results were obtained for clinical mitochondrial diseases; subjects with a clinical diagnosis of mitochondrial toxicity had similar mtDNA levels when compared with treated subjects without mitochondrial toxicity (44 ± 17 vs. 55 ± 24, respectively; P = 0.0552). Results were similar when excluding cases of peripheral neuropathy from the definition of clinical mitochondrial toxicity.
On multivariate analysis, duration of NRTI therapy was the only predictor of mtDNA level. Significant predictors of lipoatrophy were older age (>40 years), male sex, and longer duration of NRTI therapy. Elevated triglyceride level was the only laboratory marker associated with lipoatrophy. Neither lactate nor mtDNA level was found to be a marker of lipoatrophy.
Because deletions or point mutations (or polymorphisms) in the primer or probe region may affect the RT qPCR results, the mtDNA content was quantified using 2 nuclear genes, β2M and AIB1, and 2 mitochondrial genes, tRNA-leu and ND4 (see Table 1). Most (∼97%) of the mtDNA deletion molecules delete the ND4 region but not the tRNAleu region. Comparative analysis of the mtDNA copy numbers in these 2 regions can provide additional information on mtDNA deletion.
Our results showed that the ratio of the mtDNA copy number from the tRNAleu (np3212-3319) region to that from the ND4 (np12093-12170) region is between 0.85 and 1.15, suggesting that if there are deletions, the percentage is low. Two samples gave inconsistent mtDNA/nDNA ratios: 0.5 for patient 17 and 2.5 for patient 38. Multiple PCR assays were performed to detect large deletions in these 2 patients' DNA samples. No PCR-detectable large deletions were found. Sequencing analysis revealed a novel G3316A mutation in patient 17 and a novel G12164A mutation in patient 38. The novel G3316A mutation changes amino acid alanine at position 4 of ND1 to threonine. This amino acid is not located in the evolutionary conserved region, however. The novel G12164A mutation occurs in the anticodon stem region of tRNA; thus, it is highly conserved among different species (http://mamit-trna.u-strasbg.fr/table/Histidine.html). The 2 mutations are located in the reverse primer of the 3212-to-3319 region and 12093-to-12170 region, respectively. These results explained the inconsistency of copy numbers obtained from region 3212 to 3319 and region 12093 to 12170 in patients 17 and 38. Repetitive analysis using 2 single-copy nuclear genes, β2M and AIB1, as the normalizer, gave consistent results. Table 3 summarizes the results using the β2M nuclear gene.
Mitochondrial DNA Mutations
The entire mitochondrial genome of all patients with at least 2 blood specimens drawn at different times before or during treatment was analyzed by TTGE. Only 2 patients showed detectable differences in TTGE banding patterns in the longitudinal blood samples. On the first analyzed specimen, patient 31 had heteroplasmic 16441A→T but homoplasmic T after 52 months of ARV treatment. The other subject (patient 4) had at baseline T1 (before ARV therapy) heteroplasmic 228G→A, 295C→T, and 303-to-309 insC (C7/C8) but homoplasmic 228A, 295T, and 303-to-309 C8 after 14 months of ARV therapy. These mutations are in the noncoding d-loop region. We believe there are additional low levels of heteroplasmic mtDNA mutations that are undetectable by our current method. We are currently applying PCR and/or cloning sequencing methods to detect low levels of mutant mtDNA.16
This is the largest prospective longitudinal study to date that has extensively investigated mtDNA quantitatively and qualitatively in HIV-infected subjects. Using extensive rigid mitochondrial testing, we were unable to detect conclusive mtDNA depletion in the blood samples of HIV-infected subjects with lipoatrophy, regardless of lactate levels or the presence or absence of other clinical presentations thought to be linked to mitochondrial toxicity such as neuropathy and lactic acidosis. This could be attributable to the inappropriateness of the tissues examined and to the fact that the mitochondrial toxicity of ARV therapies is tissue specific. Therefore, the use of noninvasive testing such as blood mtDNA levels does not replace the need to perform tissue biopsies when investigating cases of potential mitochondrial toxicities of ARV agents. These results are consistent with the findings of a recent 24-week trial in which significant improvement of subcutaneous fat after NRTI switches was not associated with any changes in blood mtDNA.17
Other investigators have previously reported depletion in blood mtDNA of subjects with symptomatic hyperlactatemia18 and subjects with lipoatrophy.11 This was not what we observed in the current study. Approximately the same number of patients with lipoatrophy showed increased or decreased mtDNA content in blood. It is worth mentioning, however, that other studies have included small numbers of subjects7,9,10,18 or subjects on therapy but without a described clinical phenotype.19 The inhibitory effect of NRTIs on DNA polymerase-γ has been implicated in the molecular mechanism leading to mtDNA depletion. A recent study20 indicated that mtDNA replication was not affected by long-term treatment with azidothymidine (AZT), however. No difference in mtDNA copy number was observed between control and treated rats. The results suggested that AZT had a direct effect on mitochondrial targets independent of inhibition of mtDNA polymerase-γ. Mitochondrial proliferation and mtDNA overreplication have been reported to be defensive mechanisms to compensate for mitochondrial dysfunction.21-24 Thus, it is possible that mtDNA is overreplicated to compensate for mitochondrial dysfunction caused by NRTI treatment. Although mitochondrial proliferation is most common in the muscle and liver specimens of patients with mitochondrial disorders,21-25 the mtDNA content may also increase in blood. The level of compensatory amplification of mtDNA in response to mitochondrial dysfunction seemed to vary among different individuals.26 This would potentially explain the absence of correlation between mtDNA depletion and clinical symptoms.
In contrast to other studies,18,19 we did not find mtDNA depletion in HIV-infected subjects who are naive to ARV therapy. Indeed, on multivariate analysis, mtDNA levels were only associated with duration of NRTI therapy. These observations argue against a role for HIV itself in mtDNA depletion.
All patients (see Table 3A) for whom pretreatment samples were analyzed demonstrated mtDNA depletion on ARV treatment. All these patients had not developed lipoatrophy, perhaps because this toxicity takes longer to manifest. As shown in Table 3B, 16 (55%) of 29 patients with longitudinal samples during treatment displayed a decrease in mtDNA content, whereas 10 (34%) patients showed an increase in mtDNA content. Another 3 subjects had virtually no change (±5) in their mtDNA content. This could actually be attributed to compensatory mtDNA amplification or could be a reflection of noncompliance with ARV treatment. It was noted that almost all patients with increased in mtDNA content had been on ARV therapy for longer than 5 years. It is possible that the compensatory response of mtDNA amplification takes time to develop. Conversely, pretreatment mtDNA levels for the patients included in Table 3B were not available, and higher baseline mtDNA levels cannot be excluded.
In this study, the only factor associated with mtDNA depletion was duration of NRTI in general. This is consistent with prior clinical observations that linked the duration of ARV therapy to the advent of clinical mitochondrial toxicities.27,28 The use of a specific ARV drug, particularly stavudine (d4T), zidovudine (ZDV), or didanosine (ddI), was not specifically associated with such depletion. This is in contrast to fat tissue data, where mtDNA was significantly lower in subjects treated with thymidine analogues (d4T > ZDV) when compared with thymidine-sparing regimens.29,30
Our group had previously shown variations and/or polymorphisms in peripheral blood mtDNA from several patients-both ARV-naive subjects and subjects with lipoatrophy. We expand here our earlier observations to include extensive analysis of mtDNA mutations in longitudinal samples. To date, only 1 other study has documented the emergence of heteroplasmic mtDNA populations in peripheral blood of HIV-infected subjects during treatment with ARV agents.16 Data from this recent report indicated that NRTI therapy provides conditions permissive for the development of peripheral blood mtDNA mutations in vivo. Mutations occurred randomly in the whole mitochondrial genome and might be at extremely low levels that are generally not detectable by direct PCR and/or sequencing.16 A complete mutational analysis of the entire mitochondrial genome in the longitudinal samples revealed mtDNA mutations in only 2 patients. Because the acquisition of mtDNA mutations associated with NRTI therapy seems to be a dynamic process characterized by increased generation of novel heteroplasmic mutations at random sites without evidence for positive selection of preexisting non-wild type sequences, each distinct mutation could be present at a low level of heteroplasmy that was not detectable by TTGE or direct sequencing.16 To solve this problem, we are currently applying the PCR and/or cloning method to resolve each mtDNA molecule, followed by TTGE analysis of individual clones. This method should allow us to detect and estimate low-level heteroplasmic mutations.
Some studies have suggested platelet contamination of peripheral blood specimens as a potential confounding factor for mtDNA results. This was not investigated in the present study. In addition, we recognize the limitations of our study, including the relatively short duration of follow-up and the small number of subjects who developed clinical mitochondrial toxicities other than lipoatrophy. To date, available studies reporting on larger number of subjects and/or longer follow-up periods have not included assessments of the entire mtDNA genome but, instead, only measurements of mtDNA levels.17,19,29
In conclusion, blood mtDNA is not a good marker of clinical mitochondrial diseases and cannot obviate the need for tissue biopsies. In contrast to the association between fat mtDNA depletion and lipoatrophy found in other studies, we found no evidence of blood mtDNA depletion in such cases. Depletion of mtDNA was only related to a longer duration of NRTI therapy. Future studies with larger number of subjects are required to determine whether the few point mutations in the mitochondrial genome seen in this study are indeed correlated with clinical NRTI toxicities.
1. Brinkman K, ter Hofstede HJ, Burger DM, et al. Adverse effects of reverse transcriptase inhibitors: mitochondrial toxicity as common pathway. AIDS. 1998;12:1735-1744.
2. Brinkman K, Smeitink JA, Romijn JA, et al. Mitochondrial toxicity induced by nucleoside analogue reverse-transcriptase inhibitors is a key factor in the pathogenesis of antiretroviral therapy-related lipodystrophy. Lancet. 1999;354:1112-1115.
3. Martin JL, Brown CE, Matthews-Davis N, et al. Effects of antiviral nucleoside analogues on human DNA polymerases and mitochondrial DNA synthesis. Antimicrob Agents Chemother. 1994;38:2743-2749.
4. Arnaudo E, Dalakas M, Shanske S, et al. Depletion of muscle mitochondrial DNA in AIDS patients with zidovudine-induced myopathy. Lancet. 1991;337:508-510.
5. McComsey GA, Paulsen DM, Lonergan TJ, et al. Improvements in lipoatrophy, mitochondrial DNA content and adipose tissue apoptosis levels after replacement of stavudine with either abacavir or zidovudine. AIDS. 2005;19:15-23.
6. Nolan D, Hammond E, Martin A, et al. Mitochondrial DNA depletion and morphologic changes in adipocytes associated with nucleoside reverse transcriptase inhibitor therapy. AIDS. 2003;17:1329-1338.
7. McComsey G, Tan D-J, Lederman MM, et al. Analysis of the mitochondrial DNA genome in the peripheral blood leukocytes of HIV-infected patients with or without lipoatrophy. AIDS. 2002;16:513-518.
8. Miura T, Goto M, Hosoya N, et al. Depletion of mitochondrial DNA in HIV-1-infected patients and its amelioration by antiretroviral therapy. J Med Virol. 2003;70:497-505.
9. Cossarizza A, Pinti M, Moretti L, et al. Mitochondrial functionality and mitochondrial DNA content in lymphocytes of vertically infected human immunodeficiency virus-positive children with highly active antiretroviral therapy-related lipodystrophy. J Infect Dis. 2002;185:299-305.
10. Henry K, Erice A, Balfour HH, et al. Lymphocyte mitochondrial biomarkers in asymptomatic HIV-1-infected individuals treated with nucleoside reverse transcriptase inhibitors. AIDS. 2002;16:2485-2487.
11. Miro O, Lopez S, Pedrol E, et al. Mitochondrial DNA depletion and respiratory chain enzyme deficiencies are present in peripheral blood mononuclear cells of HIV-infected patients with HAART-related lipodystrophy. Antivir Ther. 2003;8:333-338.
12. ACTG recommendations for lactate blood collection. Available at: http://aactg.s-3.com
. Accessed March 30, 2005.
13. Lahiri D, Nurnberger J, Jr. A rapid non-enzymatic method for the preparation of HMW DNA from blood for RFLP studies. Nucleic Acids Res. 1991;19:5444.
14. Chen TJ, Boles RG, Wong L. Detection of mitochondrial DNA mutations by temporal temperature gradient gel electrophoresis. Clin Chem. 1999;45:1162-1167.
15. Wong LJ, Liang MH, Kwon H, et al. Comprehensive scanning of the entire mitochondrial genome for mutations. Clin Chem. 2002;48:1901-1912.
16. Martin AM, Hammond E, Nolan D, et al. Accumulation of mitochondrial DNA mutations in human immunodeficiency virus-infected patients treated with nucleoside analogue reverse transcriptase inhibitors. Am J Hum Genet. 2003;72:549-560.
17. Hoy JF, Gahan ME, Carr A, et al. Changes in mitochondrial DNA in peripheral blood mononuclear cells from HIV-infected patients with lipoatrophy randomized to receive abacavir. J Infect Dis. 2004;190:688-692.
18. Cote HC, Brumme ZL, Craib KJ, et al. Changes in mitochondrial DNA as a marker of nucleoside toxicity in HIV-infected patients. N Engl J Med. 2002;346:811-820.
19. Gallant JE, Staszewski S, Pozniak AL, et al. Efficacy and safety of tenofovir DF vs stavudine in combination therapy in antiretroviral-naive patients: a 3-year randomized trial. JAMA. 2004;292:191-201.
20. Lund KC, Wallace KB. Mitochondrial toxicity of 3′-azido-3′deoxythymidine (AZT) in the absence of mitochondrial DNA (mtDNA) depletion [abstract]. Presented at: Mitochondrial Medicine 2004 Meeting; 2004; Pittsburgh.
21. Oldfors A, Holme E, Tulinius M, et al. Tissue distribution and disease manifestations of the tRNA(Lys) A→G(8344) mitochondrial DNA mutation in a case of myoclonus epilepsy and ragged red fibres. Acta Neuropathol (Berl). 1995;90:328-333.
22. Mancuso M, Filosto M, Tsujino S, et al. Muscle glycogenosis and mitochondrial hepatopathy in an infant with mutations in both the myophosphorylase and deoxyguanosine kinase genes. Arch Neurol. 2003;60:1445-1447.
23. Vu TH, Tanji K, Pallotti F, et al. Analysis of mtDNA deletions in muscle by in situ hybridization. Muscle Nerve. 2000;23:80-85.
24. Clark KM, Taylor RW, Johnson MA, et al. An mtDNA mutation in the initiation codon of the cytochrome C oxidase subunit II gene results in lower levels of the protein and a mitochondrial encephalomyopathy. Am J Hum Genet. 1999;64:1330-1339.
25. Bai RK, Perng CL, Hsu CH, et al. Quantitative PCR analysis of mitochondrial DNA content in patients with mitochondrial disease. Ann NY Acad Sci. 2004;1011:304-309.
26. Wong LJ, Perng CL, Hsu CH, et al. Compensatory amplification of mtDNA in a patient with a novel deletion/duplication and high mutant load. J Med Genet. 2003;40:e125.
27. Lichtenstein KA, Ward DJ, Moorman AC, et al. Clinical assessment of HIV-associated lipodystrophy in an ambulatory population. AIDS. 2001;15:1389-1398.
28. McComsey G, Maa J. Host factors may be more important than the choice of antiretrovirals in the development of lipoatrophy. AIDS Read. 2003;13:539-559.
29. Cherry CL, Gahan ME, McArthur JC, et al. Exposure to dideoxynucleotides is reflected in lowered mitochondrial DNA in subcutaneous fat. J Acquir Immune Defic Syndr. 2002;30:271-277.
30. Nolan D, Hammond E, James I, et al. Contribution of nucleoside analogue reverse transcriptase inhibitor therapy to lipoatrophy from the population to the cellular level. Antivir Ther. 2003;8:617-626.
© 2005 Lippincott Williams & Wilkins, Inc.