Secondary Logo

Journal Logo

Basic Science

13C-Methionine Breath Test Detects Distinct Hepatic Mitochondrial Dysfunction in HIV-Infected Patients With Normal Serum Lactate

Banasch, Matthias MD*; Goetze, Oliver MD*; Hollborn, Ivo MD*; Hochdorfer, Bettina MD; Bulut, Kerem MD*; Schlottmann, Renate MD*; Hagemann, Dirk MD*; Brockmeyer, Norbert H MD; Schmidt, Wolfgang E MD*; Schmitz, Frank MD*

Author Information
JAIDS Journal of Acquired Immune Deficiency Syndromes: October 1st, 2005 - Volume 40 - Issue 2 - p 149-154
doi: 10.1097/01.qai.0000179465.48571.d5


Mitochondrial toxicity related to HIV infection and antiretroviral therapy (ART) is associated with a broad range of metabolic disorders such as hepatic steatosis, lactic acidosis, and polyneuropathy and may be at least partially responsible for development of lipodystrophy syndrome.1-7 Several studies have pointed to the central role of nucleoside analogs (nucleoside reverse transcriptase inhibitors [NRTIs]), because these drugs also inhibit human DNA polymerase-γ, a key enzyme of mitochondrial DNA (mtDNA) replication.8,9 mtDNA depletion has been demonstrated in peripheral blood cells as well as in liver tissue and adipocytes of HIV-infected patients,10-12 especially in those treated with the so-called “d-drugs” stavudine (d4T) and didanosine (ddI).13,14 Although new insights into the pathogenesis of mitochondrial toxicity have been gained in recent years, in vivo diagnosis has not been standardized.

The use of 13C stable isotope-labeled compounds may allow a more pathophysiologic approach to different diseases by means of a battery of breath tests that examine the integrity of metabolic pathways.18 Because the amino acid methionine is metabolized by hepatic mitochondrial decarboxylation, it seems to be a suitable substrate for monitoring this functional compartment.19-21 Recently the 13C-methionine breath test (MeBT) was proposed for assessment of NRTI-related mitochondrial toxicity in HIV-infected patients with symptomatic hyperlactatemia.22 Because symptomatic hyperlactatemia is a rare complication of HIV infection and not a precondition for development of the more frequent chronic complications such as lipoatrophy, the aim of our study was to assess mitochondrial function for the first time in patients with normal serum lactate. We investigated therapy-naive and treated patients with different durations of ART under which a subpopulation had long-term treatment experience and clinical signs of lipoatrophy. In addition, mtDNA content in peripheral blood mononuclear cells (PBMCs) was determined by real-time polymerase chain reaction (PCR) quantification to assess the relation between hepatic mitochondrial impairment and DNA depletion.


Study Design

This prospective cross-sectional study was approved by the Research-Ethics Board of the University of Bochum. All patients in the study provided written informed consent.

Ten healthy controls (8 men, age range: 37.3 ± 5.6 years, body mass index [BMI]: 22.6 ± 1.4 kg/m2 [mean ± standard deviation (SD)]) and 45 patients with chronic HIV infection [40 men, age range: 40.5 ± 8.9 years, BMI: 22.4 ± 2.5 kg/m2 [mean ± SD]) participated in the study. Subject inclusion in the protocol was solely determined by subject availability with no a priori selection bias introduced.

Detailed patient characteristics are shown in Table 1. Patients with concurrent liver diseases such as viral hepatitis B and C, evidence of liver cirrhosis, or excessive alcohol consumption (>50 g/d of ethanol) were excluded, as were patients with diabetes mellitus, severe respiratory dysfunction, anemia (hemoglobin [Hb] <7.4 mmol/L) and malignancies. Pregnant and breast-feeding women were also excluded. Serum lactate had to be within normal levels (reference range: 0.5-2.2 mmol/L) at the time of the study.

Demographic and Biochemical Characteristics of the Study Groups

The patient group was subdivided in 3 subgroups depending on treatment experience and clinical evidence of (chronic) mitochondrial toxicity: patients with no treatment experience (referred to as therapy naive [n = 15]), patients with asymptomatic disease receiving ART longer than 1 year (n = 15) without clinical signs of lipoatrophy, and lipoatrophic patients with long-term (>5 years) treatment experience (n = 15). All patients rated as lipoatrophic showed progressive fat redistribution, especially facial lipoatrophy. Highly active antiretroviral therapy (HAART) regimens of all treated patients contained an NRTI backbone; 8 patients received d-drugs at the time of the study, of whom 3 had asymptomatic HIV infection and 5 showed clinical evidence of lipoatrophy.

13C-Methionine Breath Test

After an overnight fast, each control and patient received [methyl-13C]-labeled methionine (L-methionine-13C, 99% atom isotopic enrichment; Cambridge Isotope, Andover, MA) at a rate of 2 mg/kg of body weight dissolved in 100 mL of water. Breath samples (on expiration, patients breathed into closed aluminized plastic breath bags with a content of 50 mL) were obtained before substrate administration at baseline and then at 10-minute intervals for 2 hours. During the MeBT, the subjects were kept in a relaxed sitting position. Physical activity was restricted during the test. Additionally, the intraindividual reproducibility of MeBT was examined in 10 randomly selected patients (4 lipoatrophic patients and 2 patients in each of the other groups) by obtaining MeBT parameters on 2 study days (separated by 4 weeks) under the same conditions. The 13C/12C isotope ratio of the breath samples was analyzed by nondispersive isotope selective infrared spectroscopy (IRIS; Wagner Analysen Technik, Bremen, Germany). The results were expressed as the delta (δ) value per mil (‰).

Venous Lactate and Other Biochemical Measurements

Venous blood was drawn with the use of a normal tourniquet. Patients were instructed to avoid fist clenching or hand pumping. Samples were collected in sodium fluoride/potassium oxalate tubes and kept on ice. Lactate was immediately tested enzymatically in an automated analyser (Roche/Hitachi 917) according to the manufacturer's instructions (laboratory reference range: 0.5-2.2 mmol/L). Further evaluation included quantification of HIV viral loads (COBAS AMPLICOR HIV-Monitor; Roche Diagnostics, Basel, Switzerland), CD4 lymphocyte counts (Coulter counter; Beckmann Coulter, Krefeld, Germany; reference range: 300-1400 cells/μL), alanine aminotransferase (ALT; <34 U/L), cholesterol (<5.69 mmol/L), and triglycerides (<2.26 mmol/L) (Roche/Hitachi 917).

Real-Time Polymerase Chain Reaction Quantification of Mitochondrial DNA Content

The real-time PCR quantification of mtDNA content was performed according to the protocol previously described by Cote et al.10 Venous blood samples from all tested persons were collected in EDTA tubes. PBMCs were isolated by means of Ficoll density-gradient centrifugation. Thrombocyte contamination was reduced to a ratio less than 5 to 6 thrombocytes per cell by repeated washing (2 or 3 times) with phosphate-buffered saline (PBS; according to a protocol provided by Primagen, The Netherlands). The samples were stored at −70°C until used. Total DNA was extracted from 200 μL of each sample with the QIAamp DNA Blood Mini kit (Qiagen, Hilden, Germany) according to the manufacturer's instructions. For each DNA extract, the mitochondrial gene human cytochrome-C oxidase subunit I (CCOI) and the nuclear gene human polymerase-γ accessory subunit (ASPOLG) were quantified separately by real-time quantitative PCR with use of the Roche Light Cycler (Roche Applied Science, Mannheim, Germany). The PCR reactions were performed in duplicate for each gene with use of the Light Cycler Fast Start DNA Master SYBR Green I kit (Roche Applied Science). Primer sequences and PCR conditions are described elsewhere.10 Because of the use of the nonspecific double-stranded DNA (dsDNA)-binding SYBR green dye, exclusive amplification of the ASPOLG and CCOI genes was verified by melting curve analysis. Standard curves of both genes were generated with serial dilutions of pooled HIV-negative DNA (30 healthy volunteers [15 male]). The nuclear gene was quantified with the kit's control DNA as a known concentration standard. Because of the fact that both standard curves were generated with the same serial dilutions, the mtDNA/nuclear DNA (nDNA) ratio was arbitrarily set at 1.0 in the pooled DNA, which served as the concentration standard for further analyses.

For measuring the proportion of metabolized 13C-methionine, the results were expressed as well as the percentage dose of 13C recovered (PDR) over time for each time interval from which the cumulative PDR (cPDR), obtained by numeric integration from PDR values, was calculated. These calculations are based on the formula proposed by Ravussin et al.23

Statistical analysis was first carried out as a descriptive evaluation of PDR (% per hour), cPDR (%), mtDNA/nDNA ratio, and characteristics of patients and volunteers. All data are presented as mean ± standard error of the mean (SEM) unless otherwise specified. Normality of distribution was shown by a 1-sample Kolmogorov-Smirnov test and frequency distribution histograms for all subgroups. Significance between 2 groups was tested by independent sample t tests, and for comparison of the 4 groups, analysis of variance (ANOVA) was used. When significant differences were detected, subgroups were compared by the Bonferroni post hoc test. Reproducibility of the MeBT was determined with the interindividual (CVinter) and intraindividual (CVintra) coefficients of variation of cPDR after 2 hours of test time (cPDR2h) and with the Pearson correlation coefficient of intraindividual measurements (expressed as r). Linear regression analysis was used to assess the relation between cPDR2h and the mtDNA/nDNA ratio. A forward-stepwise multiple regression analysis, based on individual cPDR2h or mtDNA/nDNA ratio as the dependent univariate parameter, was used to determine the association between impaired mitochondrial function and demographic characteristics and laboratory variables such as age and BMI as well as ALT, cholesterol, and triglycerides as continuous variables and gender and the presence of hepatic steatosis as categoric variables. The results were regarded as significant when the error probability was less than 0.05. Calculations for linear regression analysis and statistical analysis were done using commercial software programs (SPSS 11.5.1; SPSS, Chicago, IL and Graph PAD Prism, version 4.01, San Diego, CA).


13C-Methionine Breath Test

Figure 1 illustrates 13CO2 excretion data expressed as percentage doses of 13C recovered in controls and patient groups. A significant decay of mitochondrial decarboxylation function (see Fig. 1; Fig. 2) was observed between controls and therapy-naive patients and patients with clinical signs of lipoatrophy for cPDR2h (7.26% ± 0.42% vs. 4.28% ± 0.31% vs. 2.96% ± 0.37%; P < 0.001). These differences were also found for the maximal percentage dose of 13C recovered (PDRmax: 7.94% ± 0.64% per hour vs. 4.18% ± 0.33 % per hour vs. 3.20% ± 0.37% per hour; P < 0.001). Controls and ART-treated patients with asymptomatic HIV disease showed no difference in 13CO2 exhalation (cPDR2h: 7.26% ± 0.42% vs. 7.02% ± 0.29%; P > 0.05).

Mean13CO2 excretion as a function of time (PDR) in 10 healthy controls (▪) and 45 HIV-infected patients: 15 therapy-naive patients (•), 15 ART-treated patients with asymptomatic disease (▴), and 15 patients with evidence of lipoatrophy (♦). All patients had normal serum lactate. Mitochondrial dysfunction was determined by measurement of the 13CO2/12CO2 isotopic ratio in breath samples collected after ingestion of 100 mL of water labeled with 13C-methionine using isotope selective infrared spectroscopy. Data are given as mean ± SEM.
cPDRs after 2-hour test time in 10 healthy controls and 45 HIV-infected patients with normal serum lactate. Data are given as mean ± SEM.P < 0.001 vs. HIV-negative control group (A), P < 0.001 vs. HIV-positive asymptomatic therapy-naive group (B), and P < 0.05 vs. HIV-positive therapy-naive group (C) by ANOVA and the Bonferroni post hoc test.

The CVintra concerning cPDR2h was distinctly lower than the CVinter (5% vs. 36%). Intraindividual measurements showed an excellent correlation (P < 0.001, r = 0.98).

Real-Time Polymerase Chain Reaction Quantification of Mitochondrial DNA Content

The mtDNA/nDNA ratios are presented in Figure 3. Although there was a slight decline in mtDNA content from controls to asymptomatic and therapy-naive HIV-infected patients (1.08 ± 0.14 vs. 1.03 ± 0.17 vs. 0.93 ± 0.15, each in mean ± standard deviation [SD]), there was no statistical significance compared with the control group.

mtDNA/nDNA ratios in PBMCs obtained by real-time PCR quantification of the mitochondrial CCOI gene and the nuclear ASPOLG gene. Data are given as mean ± SD.P < 0.001 vs. HIV-positive lipoatrophy by ANOVA and the Bonferroni post hoc test.

The lowest mtDNA/nDNA ratios were presented by the group of lipoatrophic patients, with a mean ratio of 0.57 ± 0.14, which was decreased compared with each of the other study groups (P < 0.001).

Biochemical Measurements

Detailed data are presented in Table 1. As expected, treatment-naive patients showed a higher viral load compared with patients on treatment (P < 0.05). The reverse observation was made regarding the lipid profile: triglyceride and cholesterol values were significantly higher in ART-treated subjects (P < 0.05), who also presented the highest ALT values, with statistical significance in lipoatrophic patients compared with all other study groups (P < 0.05). Multiple regression analysis was used to determine predictors for the presence of impaired mitochondrial function (ie, cPDR2h and mtDNA content). The ALT value was identified as the only independent predictor of cPDR and mtDNA/nDNA ratio (R2 = 0.344, P = 0.007).

Linear Regression Analysis and Correlation of 13C-Methionine Breath Test Results With Mitochondrial DNA Content in Peripheral Blood Mononuclear Cells

Regression analysis showed a linear relation between the cPDR2h and mtDNA/nDNA ratio in healthy controls and all patients. The Pearson correlation analysis showed a significant correlation (P < 0.001), with a correlation coefficient (r) of 0.77 (Fig. 4).

Association between mtDNA content in PBMCs and hepatic mitochondrial decarboxylation capacity as assessed by the MeBT in 10 healthy controls (▪) and 45 HIV-infected patients: 15 therapy-naive patients (•), 15 ART-treated patients with asymptomatic disease (▴), and 15 patients with evidence of lipoatrophy (♦).


In this cross-sectional study, we aimed to establish the MeBT as a noninvasive diagnostic tool for detection of hepatic mitochondrial toxicity in HIV-infected patients with different durations of infection and treatment experience.

We found, as expected, that among patients with clinical evidence of lipoatrophy, hepatic 13C-methionine metabolism was distinctly reduced, reflecting chronic mitochondrial toxicity related to long-term ART and HIV infection. In parallel, mtDNA content in peripheral blood cells was significantly decreased in this study group, suggesting NRTI-related inhibition of DNA polymerase-γ as a possible pathomechanism of mitochondrial damage. The ALT level was the only explanatory variable showing a significant correlation with MeBT results (R2 = 0.34). This association can be primarily assigned to the lipoatrophic patient group, in which elevated ALT values were more prevalent.

Interestingly, no changes in mitochondrial decarboxylation capacity were observed in asymptomatic ART-treated patients. These findings are reflected by mtDNA content in peripheral blood cells similar to those of healthy volunteers. In contrast to previous studies reporting a significant decline in mtDNA in ART-treated patients, we could not confirm this observation in our study group. A reasonable explanation is the low proportion of asymptomatic patients with d-drug-containing HAART regimens (3 [20%] of 15 patients) and even those with a history of d-drug experience (5 [33%] of 15 patients). Because of the limited sample number, there also might be selection biases in terms of rating patients as “asymptomatic” who were less sensitive to NRTI drug toxicity or had higher baseline values in mtDNA content.

The reduced hepatic mitochondrial decarboxylation capacity in treatment-naive patients was initially unexpected, because these patients had never received antiretroviral drugs. Hence, NRTI-induced mtDNA depletion has to be ruled out as a pathomechanism of mitochondrial toxicity in this study group. This assumption is consistent with our findings in real-time PCR measurements, because mtDNA content in PBMCs of therapy-naive patients was slightly declined but without statistical significance compared with that of the control group. The issue of mtDNA depletion in treatment-naive patients is still a point of controversial debate, because only a few recent studies have focused on this topic, with contrary results.24,25 The mechanisms of viral-mediated mitochondrial toxicity are not clearly understood. There is evidence that HIV-encoded proteins are directly and indirectly involved in mitochondrial damage, leading to disturbances of mitochondrial membrane potential and apoptotic cell death in HIV-infected cells.26-29 A decline in HIV viral load therefore should lead to an (at least temporary) improvement in mitochondrial function. A small subset of previous treatment-naive patients in our study started ART, yet all of them showed significant improvement of hepatic methionine metabolism as reflected in the MeBT after 3 month of HAART, despite the potential drug-related toxic effects.

A general concern of functional metabolic tests is their reproducibility. Although limited by the small number of patients recruited for a second MeBT, our data suggest excellent intraindividual reproducibility of this new method. A limitation of the present study, aside from its cross-sectional design, is the comparison of the different functional compartments in the liver and PBMCs. This approach was chosen because of the pilot character of this study and the difficulties and ethical concerns of obtaining liver tissue samples. Although the functional impact of mtDNA depletion is controversially discussed, the observed correlation with hepatic methionine metabolism (see Fig. 4) suggests at least a partial relation between mtDNA content and hepatic mitochondrial function in patients receiving ART.

Further studies with a focus on morphologic changes in liver tissue and longitudinal observational trials in larger cohorts are required to define the specificity and sensitivity of this new method and eventually to derive recommendations for clinical practice.


1. Carr A, Cooper DA. Adverse effects of antiretroviral therapy. Lancet. 2000;356:1423-1430.
2. Fortgang IS, Belitsos PC, Chaisson RE, et al. Hepatomegaly and steatosis in HIV-infected patients receiving nucleoside analog antiretroviral therapy. Am J Gastroenterol. 1995;90:1433-1436.
3. Chariot P, Drogou I, de Lacroix-Szmania I, et al. Zidovudine-induced mitochondrial disorder with massive liver steatosis, myopathy, lactic acidosis, and mitochondrial DNA depletion. J Hepatol. 1999;30:156-160.
4. Dalakas MC, Illa I, Pezeshkpour GH, et al. Mitochondrial myopathy caused by long-term zidovudine therapy. N Engl J Med. 1990;322:1098-1105.
5. Gerard Y, Maulin L, Yazdanpanah Y, et al. Symptomatic hyperlactataemia: an emerging complication of antiretroviral therapy. AIDS. 2000;14:2723-2730.
6. 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.
7. Kakuda TN, Brundage RC, Anderson PL, et al. Nucleoside reverse transcriptase inhibitor-induced mitochondrial toxicity as an etiology for lipodystrophy. AIDS. 1999;13:2311-2312.
8. Martin JL, Brown CE, Matthews-Davis N, et al. Effects of antiviral nucleoside analogs on human DNA polymerases and mitochondrial DNA synthesis. Antimicrob Agents Chemother. 1994;38:2743-2749.
9. Lim SE, Copeland WC. Differential incorporation and removal of antiviral deoxynucleotides by human DNA polymerase gamma. J Biol Chem. 2001;276:23616-23623.
10. 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.
11. Walker UA, Bauerle J, Laguno M, et al. Depletion of mitochondrial DNA in liver under antiretroviral therapy with didanosine, stavudine, or zalcitabine. Hepatology. 2004;39:311-317.
12. Pace CS, Martin AM, Hammond EL, et al. Mitochondrial proliferation, DNA depletion and adipocyte differentiation in subcutaneous adipose tissue of HIV-positive HAART recipients. Antivir Ther. 2003;8:323-331.
13. Kakuda TN. Pharmacology of nucleoside and nucleotide reverse transcriptase inhibitor-induced mitochondrial toxicity. Clin Ther. 2000;22:685-708.
14. Walker UA, Setzer B, Venhoff N. Increased long-term mitochondrial toxicity in combinations of nucleoside analogue reverse-transcriptase inhibitors. AIDS. 2002;16:2165-2173.
15. Deleted in proof.
    16. Deleted in proof.
      17. Deleted in proof.
        18. Schoeller DA, Schneider JF, Solomons NW, et al. Clinical diagnosis with the stable isotope 13C in CO2 breath tests: methodology and fundamental considerations. J Lab Clin Med. 1977;90:412-421.
        19. Finkelstein JD, Martin JJ, Harris BJ. Methionine metabolism in mammals. The methionine-sparing effect of cystine. J Biol Chem. 1988;263:11750-11754.
        20. Armuzzi A, Marcoccia S, Zocco MA, et al. Non-invasive assessment of human hepatic mitochondrial function through the 13C-methionine breath test. Scand J Gastroenterol. 2000;35:650-653.
        21. Spahr L, Negro F, Leandro G, et al. Impaired hepatic mitochondrial oxidation using the 13C-methionine breath test in patients with macrovesicular steatosis and patients with cirrhosis. Med Sci Monit. 2003;9:CR6-CR11.
        22. Milazzo L, Riva A, Sangaletti O, et al. 13C-methionine breath test detects liver mitochondrial impairment in HIV-infected patients with antiretroviral drug-related hyperlactatemia. J Acquir Immune Defic Syndr. 2004;35:429-432.
        23. Ravussin E, Pahud P, Thelin-Doerner A, et al. Substrate utilization during prolonged exercise after ingestion of 13C-glucose in obese and control subjects. Int J Obes. 1980;4:235-242.
        24. Miro O, Lopez S, Martinez E, et al. Mitochondrial effects of HIV infection on the peripheral blood mononuclear cells of HIV-infected patients who were never treated with antiretrovirals. Clin Infect Dis. 2004;39:710-716.
        25. McComsey G, Bai RK, Maa JF, et al. Extensive investigations of mitochondrial DNA genome in treated HIV-infected subjects: beyond mitochondrial DNA depletion. J Acquir Immune Defic Syndr. 2005;39:181-188.
        26. Jacotot E, Ravagnan L, Loeffler M, et al. The HIV-1 viral protein R induces apoptosis via a direct effect on the mitochondrial permeability transition pore. J Exp Med. 2000;191:33-46.
        27. Ferri KF, Jacotot E, Blanco J, et al. Mitochondrial control of cell death induced by HIV-1-encoded proteins. Ann NY Acad Sci. 2000;926:149-164.
        28. Roggero R, Robert-Hebmann V, Harrington S, et al. Binding of human immunodeficiency virus type 1 gp120 to CXCR4 induces mitochondrial transmembrane depolarization and cytochrome C-mediated apoptosis independently of Fas signaling. J Virol. 2001;75:7637-7650.
        29. Muthumani K, Hwang DS, Desai BM, et al. HIV-1 Vpr induces apoptosis through caspase 9 in T cells and peripheral blood mononuclear cells. J Biol Chem. 2002;277:37820-37831.

        HIV; mitochondrial toxicity; 13C-methionine breath test

        © 2005 Lippincott Williams & Wilkins, Inc.