Rapid Analysis of Mitochondrial DNA Depletion by Fluorescence In Situ Hybridization and Immunocytochemistry: Potential Strategies for HIV Therapeutic Monitoring

Abstract

Nucleoside reverse transcriptase inhibitors (NRTIs) have been a mainstay in the treatment of human immunodeficiency virus since the introduction of azidothymidine (AZT) in 1987. However, none of the current therapies can completely eradicate the virus, necessitating long-term use of anti-retroviral drugs to prevent viral re-growth. One of the side effects associated with long-term use of NRTIs is mitochondrial toxicity stemming from inhibition of the mitochondrial DNA (mtDNA) polymerase γ, which leads to mtDNA depletion and consequently to mitochondrial dysfunction. Here we report the use of fluorescence in situ hybridization (FISH) and immunocytochemistry (ICC) to monitor mtDNA depletion in cultured fibroblasts treated with the NRTI 2',3'-dideoxycytidine (ddC). These techniques are amenable to both microscopy and flow cytometry, allowing analysis of populations of cells on a single-cell basis. We show that, as mtDNA depletion progresses, a mosaic population develops, with some cells being depleted of and others retaining mtDNA. These techniques could be useful as potential therapeutic monitors to indicate when NRTI therapy should be interrupted to prevent mitochondrial toxicity and could aid in the development of less toxic NRTIs by providing an assay suitable for pharmacodynamic evaluation of candidate molecules.

Keywords

mitochondria oxidative phosphorylation mitochondrial DNA HIV AIDS mitochondrial disease nucleoside reverse transcriptase inhibitor cytochrome c oxidase immunocytochemistry FISH

Mitochondria play a critical role in many cellular processes. They provide more than 95% of the cellular ATP through the process of oxidative phosphorylation, function in several other metabolic processes, have a role in ion homeostasis, and are a key player in apoptosis or programmed cell death (Scheffler 1999).

There is increasingly strong evidence that mitochondria originated as symbiotic bacteria. They retain a prokaryote-like DNA, which in all species encodes a fraction of the proteins of the organelle. In mammals, mtDNA is a 16-kb double-stranded circular DNA that encodes 13 different polypeptides, all involved in oxidative phosphorylation, along with two rRNAs and 22 tRNAs (Anderson et al. 1981). Unlike the nuclear genome, mtDNA is present in thousands of copies in mammalian cells, all of which are used in translation of gene products made within the organelle on bacterium-like ribosomes (Taanman 1999). mtDNA lacks protective histones and has minimal repair mechanisms and, as a result, has a much higher spontaneous mutation rate than nuclear DNA (nDNA), which is significantly enhanced because the mitochondrion is a free radical-rich environment (Sastre et al. 2003). Accumulation of mutations and deletions occurs throughout life and becomes physiologically relevant when there are sufficient copy numbers to alter oxidative phosphorylation (Schon et al. 1997; Schapira 2002). Protein synthesis in mitochondria involves several transcription factors, replication enzymes, DNA polymerases, and regulatory factors, all encoded by the nuclear genome and imported into the organelle (Taanman 1999). Mutations in mtDNA occur that decrease the amount of mtDNA and result in concomitant mitochondrial dysfunction, leading to a set of diseases of variable phenotypes called mtDNA depletion syndromes (Taanman et al. 1997; Elpeleg et al. 2002; Carrozzo et al. 2003).

The current strategy for managing mitochondrial toxicity associated with NRTI therapy requires temporarily removing patients from the treatment to recover mitochondrial function before the viral load becomes life threatening (Lewis 2003; Walker 2003). This treatment regimen would be greatly aided by the availability of efficient and reliable methods for monitoring mtDNA depletion and by the development of less toxic NRTIs. Here we describe simple approaches for monitoring the levels of mtDNA that should be suitable for such clinical applications. Recent reports have demonstrated the utility of monitoring mtDNA abundance by real-time PCR in patients treated with NRTIs (Cote et al. 2002; Petit et al. 2003). A key observation in one of these reports is that measurable loss of mtDNA precedes measurable metabolic changes, such as hyperlactemia, associated with drug toxicity (Cote et al. 2002). This means, in effect, that changes in mtDNA content may predict the onset of some side effects. However, cells contain thousands of copies of mtDNA and, because of variable segregation of the mtDNA, often only a proportion of the mitochondrial genomes within a cell carry a mutation. In fact, both heteroplasmy (intracellular variation) and mosaicism (intercellular variation) are characteristic of mtDNA mutations. Therefore, techniques that allow analysis of populations of cells on a single-cell basis are necessary for accurate determination of mitochondrial changes.

Here we report the use of mtDNA FISH and ICC to monitor mtDNA depletion in cultured fibroblasts treated with the NRTI ddC and compare the results obtained with these methods to those obtained by the population-based technique of conventional realtime PCR.

Materials and Methods

Cell Culture

MRC5 fibroblasts were obtained from the American Type Culture Collection (Manassas, VA). Fibroblasts were grown in high- glucose DMEM containing 110 μg/ml sodium pyruvate supplemented with 2 mM l-glutamine, 10% FBS, 50 μg/ml uridine (Sigma-Aldrich; St Louis, MO), and 10 mM HEPES buffer to maximize growth rates, as modified from a previously described method of cell culture (Marusich et al. 1997). mtDNA depletion was induced by addition of 4 μM ddC (Sigma-Aldrich) to the culture medium and cells were analyzed at progressive population doublings (PDs) after NRTI treatment. Unless specified, cell culture reagents were purchased from Invitrogen (Carlsbad, CA).

Real-time PCR

Total DNA was extracted from cells using the Qiagen DNeasy kit per manufacturer's recommendations (Qiagen; Valencia, CA). The primers used to amplify the mtDNA were forward 5'-dTATTAGTGGCATCGCCTGCC-3’ (1951–1970) and reverse 5'-dGACCCTCGTTTAGCCATTCATTC-3’ (2071–2049). The carboxytetramethylrhodamine (TAMRA)- and carboxyfluorescein (FAM)-conjugated probe used was FAM 5'-dAAGGTAGGATAATCACTTGTTC-3’ TAMRA (2012–2033). The total DNA quantity was corrected by simultaneous measurement of 28SrRNA using 5'-dAGAAATTCAATGAAGCGCGG-3’ and 5'-dAGGGACAGTGGGAATCTCGTT-3’ as primers and FAM 5'-dACGGCGGGAGTAAC-3’ TAMRA as the probe (Yamagata et al. 2002). An initial 10-min denaturation at 95C was followed by 50 cycles of denaturation at 95C for 15 sec and annealing at 60C for 30 sec. A standard curve for real-time PCR was generated using serial dilutions of DNA obtained from human fibroblasts. The amount of mtDNA is expressed as relative quantity to nDNA using Ct values calculated by the sequence detector (Perkin Elmer; Boston, MA).

Fluorescence In Situ Hybridization

Six overlapping PCR products, averaging 3.3 kb in length, representing the entire human mtDNA were prepared. All primers were custom synthesized (Operon; Alameda, CA) and PCR reactions were performed using Amplitaq Gold PCR Master Mix (Applied Biosystems; Foster City, CA). PCR primers used for mtDNA FISH probe preparation were as follows: MtF1 5'-dTCCATGCATTTGGTATTTTCGTC-3'; MtR1 5'-dCGAAGGGTTGTAGTAGCCCGTAG-3'; MtF2 5'-dACCTTCAAATTCCTCCCTGTACG-3'; MtR2 5'-dTGATGGCCCCTAAGATAGAGGAG-3'; MtF3 5'-dGGCAACCTTCTAGGTAACGACCA-3'; MtR3 5'-dGAGGAGCGTTATGGAGTGGAAGT-3'; MtF4 5'-dGTAAGCCTCTACCTGCACGACAA-3'; MtR4 5'-dGTGGATGCGACAATGGATTTTAC-3'; MtF5 5'-dGCTCACTCACCCACCACATTAAC-3'; MtR5 5'-dTGAAGGGCAAGATGAAGTGAAAG-3'; MtF6 5'-dTGAGGGGCCACAGTAATTACAAA3'; and MtR6 5'-dGAGTGGGAGGGGAAAATAATGTG-3'. Each PCR reaction was confirmed by agarose gel electrophoresis and the product was purified by QIAquick PCR Purification Kit (Qiagen). An equal amount of PCR product from each reaction was combined to represent the whole mtDNA. Ten Arabidopsis cDNA PCR products from the SpotReport cDNA Array Validation System (Stratagene; La Jolla, CA) were pooled and used as the template for labeled negative control probe generation. Pooled DNA for the mtDNA and the negative control probes was nick-translated and labeled with the Alexa Fluor 488 dye as instructed in the ARES Alexa Fluor 488 DNA Labeling Kit (Molecular Probes; Eugene, OR).

For microscopic analysis, MRC5 fibroblasts were grown to 50–70% confluency on glass coverslips. In some instances, cells were incubated for 20 min at 37C in 300 nM MitoTracker Red dye (Molecular Probes) in culture medium to counterstain the mitochondria. Cells were then washed twice with PBS before fixation in 4% paraformaldehyde in complete medium for 20 min at 37C. After fixation, cells were washed in PBS and permeabilized in 0.1% Triton X-100 for 10 min at room temperature. Cells were again washed twice in PBS before treatment with 0.1 mg/ml DNase-free RNase (Roche Diagnostics; Indianapolis, IN) for 90 min at 37C. Meanwhile, microscope slides were pre-warmed to 74C on a slide warmer. The cells were then washed twice with PBS and excess liquid was blotted from the coverslips. To the pre-warmed slide, 15 μl of the mtDNA or negative control probe was added [0.5 ng/μl in 50% formamide, 10% dextran sulfate, 2 × saline sodium citrate (SSC) buffer]. Coverslips were mounted on the pre-warmed slides and incubated on the slide warmer for 5 min at 74C, allowing simultaneous denaturation of targets and probes. Cover-slips were sealed to the slides with rubber cement and the probes were left to hybridize overnight at 37C in a dark humidified chamber. Coverslips were removed from the slides and washed once in 2 × SSC buffer at RT, followed by one wash in 2 × SSC and one in 0.5 × SSC each for 5 min at 40C. Samples were then rinsed in PBS, counterstained with 0.2 ng/ml 4',6-diamidino-2-phenylindole dihydrochloride (DAPI) (Molecular Probes) for 2 min, and mounted in Pro-Long antifade reagent (Molecular Probes). All samples prepared for microscopic evaluation were visualized with a Nikon Eclipse 800 fluorescence microscope (Nikon Instruments; Melville, NY) equipped with a Micromax 1300 YHS camera (Roper Scientific; Trenton, NJ). Images were acquired with MetaMorph software v4.6.9 (Universal Imaging; Downingtown, PA) and appropriate filter sets (Omega Optical; Brattleboro, VT).

For flow cytometric analysis, fibroblasts were labeled in suspension. The FISH protocol is the same as above, with the following exceptions. Cells were harvested by removing the culture medium from a nearly confluent culture dish, washing the cells once with pre-warmed (37C) PBS, releasing the cells by incubation in a pre-warmed solution of trypsin and EDTA (0.25% and 1 mM, respectively) for ≃5 min, and adding an equal volume of pre-warmed culture medium containing 10% FBS to quench the trypsin. Fibroblasts were collected by low-speed centrifugation, resuspended in complete medium containing 4% paraformaldehyde, and fixed for 20 min at 37C before a single PBS wash and cell count. Approximately two million cells per sample were re-suspended, and both detergent and RNase treatments were carried out as above. Each sample for flow cytometry was hybridized with 50 μl of mtDNA or negative control probe. Simultaneous denaturation of the probes and targets took place in a water bath at 74C and hybridization took place in a 37C water bath protected from light. After labeling and posthybridization washing as described above, cells were re-suspended in ≃0.3 ml of PBS for flow cytometric analysis. Samples prepared for flow cytometric analysis were not counterstained with MitoTracker Red dye or with DAPI. All samples prepared for flow cytometry were analyzed using an Elite flow cytometer equipped with EXPO 32 software v. 1.2, a 488-nm argon ion laser, a 633-nm HeNe laser, and bandpass filters at 525 ± 10 nm and 675 ± 10 nm with a 640-nm dichroic mirror (Beckman Coulter; Miami, FL).

Immunocytochemistry

For microscopic analysis, fibroblasts were grown to ≃50–70% confluency on glass coverslips. The coverslips were removed from culture medium and washed in PBS before fixation in complete medium containing 4% paraformaldehyde for 20 min at 37C. Coverslips were then washed with PBS and incubated in antigen retrieval buffer (0.1 M Tris-HCl at pH 9.5 containing 5% urea) for 20 min at ≃95C before permeabilization in 0.1% Triton X-100 for 10 min at RT. After a 30-min block in PBS containing 10% normal goat serum (blocking buffer), cells were incubated for 1 hr at RT with primary antibody cocktails consisting of mouse monoclonal 31HL anti-porin IgG_2b at 5 μg/ml (Calbiochem; San Diego, CA) and mouse monoclonal 1D6 anti-OxPhos Complex IV subunit I (COX I) IgG_2a at 2 μg/ml (Molecular Probes) diluted in blocking buffer. Coverslips were washed three times in blocking buffer and then incubated for 30 min at RT with secondary antibodies diluted in blocking buffer. Alexa Fluor 647 dye-conjugated goat anti-mouse IgG_2b-specific IgG and Alexa Fluor 488 dye-conjugated goat anti-mouse IgG_2a-specific IgG were used at 5 μg/ml to detect anti-porin and anti-COX-I primary antibodies, respectively. Isotype-specific secondary antibodies (Southern Biotech; Birmingham, AL) were conjugated to either Alexa Fluor 488 dye or Alexa Fluor 647 dye as described previously (Haugland 2000). Negative ICC controls were perfomed by replacement of described specific primary antibodies with nonspecific mouse IgG_2a and IgG_2b mouse antibodies, exclusion of primary antibodies in the labeling scheme, or incubation of isotype secondary antibodies with mismatched isotypes of specific primary antibodies. Cover-slips were washed three times in PBS before counterstaining with 0.2 ng/ml DAPI for 2 min at RT and mounting in Pro-Long antifade reagent.

For flow cytometric analysis, cells were harvested and fixed as described above for FISH methods. Approximately two million fixed cells per sample underwent antigen retrieval as described above, with vortexing every few minutes. Samples were washed once with PBS before being permeabilized with 0.1% Triton X-100 and blocked with PBS containing 30 μg goat IgG for 10 and 30 min at RT, respectively. Primary and secondary antibody incubations were for 60 and 30 min at RT, respectively, with a wash in between and after incubations using PBS containing 1% BSA. All antibodies were used at 2 μg per sample and ICC controls were performed as described above.

Results

Real-time PCR

Depletion of mtDNA has been quantitated previously by real-time PCR, a method considered highly accurate for monitoring changes in DNA levels (Cote et al. 2002; Gourlain et al. 2003). Therefore, the effect of ddC on the levels of mtDNA in cells was monitored with this method at the same time that the samples were analyzed by FISH and by ICC. As shown in Figure 1, the levels of mtDNA measured as the amount relative to the levels of nDNA decreased with time and PD in ddC. The rate of this disappearance was then compared with rates obtained by the FISH and ICC methods.

Figure 1.

Real-time PCR. The ratios of mitochondrial to nuclear DNA (mtDNA:nDNA) in cells were assayed by real-time PCR over 7 PDs in the presence of ddC (•). Maximum (▪) and minimum (△) relative contents are also shown to demonstrate the range of DNA content observed. The levels of mtDNA decreased relative to nuclear DNA levels over time in the presence of ddC.

Fluorescence In Situ Hybridization

Microscopic analysis of untreated MRC5 fibroblasts hybridized with a full-length mtDNA probe produced punctate spots throughout the cytoplasm. Co-localization studies with MitoTracker Red dye revealed that these spots lay within the mitochondrial reticulum (Figure 2A). FISH analysis was also carried out on MRC5 fibroblasts grown for 1, 3, 5, or 7 PDs in 4 μM ddC. This analysis revealed that cells were progressively depleted of mtDNA as a function of time in culture with ddC (PD 7 shown in Figure 2B). Negative control samples hybridized with a probe to Arabidopsis thaliana DNA showed no specific staining in the mitochondria (Figure 2C).

FISH samples prepared for more quantitative analysis by flow cytometry revealed a pattern of mtDNA depletion consistent with microscopic observations (Figure 3). Untreated populations of fibroblasts labeled with the mtDNA probe produced signals exceeding but similar to cells hybridized after one PD in ddC (Figure 3). Longer exposures to ddC resulted in mtDNA probe signals approximately threefold lower than untreated cells, levels corresponding to background as indicated by the negative control A. thaliana DNA probe (Figure 3).

Immunocytochemistry

MRC5 fibroblasts grown for 0, 1, 3, 5, or 7 PDs in 4 μM ddC were double-labeled with a control antibody against porin and an antibody against COX I. Porin is an nDNA-encoded mitochondrial protein whose levels remain relatively constant even in cells with oxidative phosphorylation defects, and was used to control for mitochondrial mass. Microscopic analysis revealed that whereas levels of porin remain relatively constant with increased PD in ddC, cell populations exposed to the drug became depleted of the mtDNA-encoded COX I (Figure 4). Cells exposed to ddC quickly became mosaic with respect to COX I as mtDNA depletion proceeded (Figure 4). These samples were uniformly depleted of COX I by 7 PDs in ddC (Figure 4).

Figure 2.

mtDNA FISH analysis by microscopy. MitoTracker Red dye-labeled MRC-5 human lung fibroblasts hybridized with Alexa Fluor 488 dye-labeled mtDNA probe (green) and counterstained with DAPI (blue). Untreated cells showed extensive incorporation of the mtDNA probe, which co-localized with MitoTracker Red dye labeling (A). Cells exposed to ddC for 7 PDs showed virtually no hybridization signals of mtDNA probe (B). Fibroblasts hybridized with a negative control Arabidopsis thaliana DNA probe produced no hybridization signals (C). Bar = 10 μm.

Figure 3.

mtDNA FISH analysis by flow cytometry. MRC-5 human lung fibroblasts in suspension hybridized with Alexa Fluor 488 dye-conjugated mtDNA probe after exposure to ddC for various PDs. Samples were excited with a 488-nm argon ion laser and emission at 525 ± 10 nm was collected. Control cells hybridized with a negative control A. thaliana DNA probe are also shown. Levels of mtDNA probe incorporation decreased with exposure to ddC and reached background levels, as defined by the negative control probe, by PD 3 in NRTI. Values are the mean fluorescence ± SD of three samples.

MRC5 fibroblasts grown for 0, 1, 3, 5, or 7 PDs in 4 μM ddC were also double-labeled as above for flow cytometric analysis. With increasing PD in ddC, flow cytometry revealed a decrease in COX I while levels of porin were maintained (Figure 5). These results demonstrated not only the potential utility of these antibodies in therapeutic monitoring but also their amenability to rapid analysis of cell populations by flow cytometry. All ICC controls were negative, and signals reported in all relevant figures were considered specific.

Discussion

In this study, three detection methods were used to follow mtDNA depletion. The first two methods, FISH and real-time PCR, were used to monitor the amounts of mtDNA in cells directly. The third method, ICC, was used to detect the mtDNA-encoded COX I protein and thus indirectly measure the loss of mitochondrial DNA. As levels of the mitochondrial genome decreased, the progressive loss of COX I was observed and, as expected, trailed the loss of mtDNA as measured directly. Levels of the nDNA-encoded porin, as measured immunocytochemically, were maintained throughout the exposure of cells to NRTI.

mtDNA depletion can occur through several mechanisms and, when extensive, is detrimental to cellular respiration and ATP synthesis. Mutations in enzymes responsible for mtDNA transcription have been described that lead to mtDNA depletion syndromes (Taanman 1999). Most often, mtDNA depletion and subsequent mitochondrial dysfunction lead to undesirable side effects during drug therapies (Brinkman and Kakuda 2000). For example, there is evidence of a myopathy induced by statin therapy for hypocholesterolemia that could be caused by loss of mtDNA (Fosslien 2001; Phillips et al. 2002). The use of NRTI treatment for HIV infection also leads to mtDNA depletion, with grave consequences for the patient. NRTIs have been the mainstay for treatment of HIV infections since the introduction of AZT in 1987. There are now 15 different anti-retroviral drugs that are FDA-approved for HIV treatment, six of which are NRTIs, three of which are non-nucleoside reverse transcriptase inhibitors, and six of which are protease inhibitors (Temesgen 2001). Although mortality rates have decreased since the introduction of combination NRTI drug therapy, none of the current therapies can completely eradicate the virus and so long-term use of anti-viral drugs is needed to prevent viral regrowth.

Recent studies have indicated that mtDNA depletion can lead to many of the severe side effects that occur during drug therapy. The best-documented and most important example is the now widely described NRTI-induced mtDNA depletion present in a significant proportion of patients being treated with these drugs for HIV infection (Lewis 2003; Walker 2003). Mitochondrial toxicity associated with long-term use of NRTIs may manifest itself in a variety of forms, including peripheral neuropathy, myopathy, lipidopathy, lactic acidosis, and hepatic dysfunction (Chariot et al. 1999; Brinkman and Kakuda 2000; Carr and Cooper 2000). NRTIs competitively inhibit HIV reverse transcriptase as they are incorporated into the newly synthesized DNA strand, but because the NRTIs lack the 3'-hydroxyl group needed for further chain elongation, the growing viral DNA chain is terminated. Although most cellular replicative DNA polymerases are not significantly inhibited by NRTIs, mtDNA polymerase γ, the sole polymerase responsible for replicating the mitochondrial genome, is an exception. mtDNA polymerase γ shows relatively high incorporation rates of NRTIs into mtDNA as well as low excision rates (Kukhanova et al. 1998; Feng et al. 2001; Lim and Copeland 2001). The resulting inhibition of the mtDNA polymerase γ leads to mtDNA depletion and its pathological consequences (Brinkman and Kakuda 2000; Walker 2003).

Figure 4.

Immunocytochemical analysis by microscopy. MRC-5 human lung fibroblasts labeled with antibodies against porin (red) and COX I (green). Untreated cells show fairly equivalent labeling of both targets, as is apparent by merged red and green signals that appear yellow. Cells at PD 1, 3, 5, and 7 after ddC treatment show a progressive loss of COX I and maintenance of porin, as can be seen in panels with merged images. Cells were counterstained with DAPI (blue) to visualize nuclei. Bar = 20 μm.

The current strategy for managing side effects arising from mitochondrial toxicity associated with NRTI treatment is the temporary removal of HIV-affected patients from therapy to allow for new mtDNA replication before a deleterious increase in viral load ensues (Moyle 2000). Deciding when to remove a patient from the therapy and when to reinstate NRTI treatment is subjective until measurable criteria can be established. In this study we have demonstrated two approaches, mtDNA FISH and ICC, for measuring mtDNA levels that should be useful in this regard. Both methods are advantageous in that they require very little live cell/tissue manipulation and are amenable to analysis of populations of cells on a single-cell basis by flow cytometry. In contrast to real-time PCR, both FISH and ICC do not require DNA isolation, and the antibody-based approach is particularly expeditious relative to FISH and real-time PCR.

Figure 5.

Immunocytochemical analysis by flow cytometry. MRC-5 human lung fibroblasts labeled in suspension with antibodies against COX I and porin and isotype-specific secondary antibodies conjugated to either the Alexa Fluor 488 dye or the Alexa Fluor 647 dye. Samples were excited with 488-nm argon ion and 633-nm HeNe lasers. Emission at 525 ± 10 nm and 675 ± 10 nm was collected for control cells and cells exposed to ddC for various periods of time. Values are expressed as the percent of mean fluorescence ± SD of three samples from untreated cells associated with either nDNA-encoded porin or mtDNA-encoded COX I. For direct comparison of methods, the inset graph shows data from real-time PCR (•), FISH (△), and ICC (♦) samples expressed as percentages of untreated cells as an indication of mitochondrial health. As reflected in Figure 3, FISH signals from samples exposed to ddC for 5 and 7 population doublings were variable and were comparable to background levels obtained with the negative control probe.

We have described microscopic observations of samples prepared for FISH and ICC that provide a simple qualitative assessment of mtDNA depletion suitable for even the most basic clinical situations. Furthermore, microscopy revealed clear mosaicism by 3 PDs in ddC, a clinically significant observation not evident from real-time PCR analysis in which subpopulations are lost in an average of the whole population. Inasmuch as this additional information reflects compromised oxidative phosphorylation and subsequent side effects, the use of monoclonal antibodies should prove quite valuable in therapeutic monitoring during management of pathologies such as HIV infection. A variety of different cell and tissue types, including leukocytes and adipose tissue, are affected by NRTI therapy and represent potential targets for immunological detection of mtDNA depletion in the context of HIV therapeutic monitoring (Cote et al. 2002; Cossarizza 2003; Hellerstein 2003).

The FISH method follows the loss of mtDNA copies directly, but the protocol requires 24 hr and multiple steps to perform. The loss of COX I is an indirect measure of mtDNA depletion because the levels of subunit observed are determined by the transcription and translation rates from available mtDNA copies and the rate of polypeptide loss as previously assembled cytochrome c oxidase complexes are degraded. Not surprisingly, the loss of COX I trails the loss of mtDNA copies but, in effect, both the ICC and FISH methods can be used to establish a standard curve for mtDNA depletion (Figure 5, inset graph). As demonstrated here, one advantage of the ICC approach is that it can be completed and the data analyzed within 4 hr. Moreover, this method is readily adaptable to a 96-well plate format for simultaneously screening multiple patient samples. Such a high-throughput format should aid in the development of less toxic NRTIs and in drug evaluation where mitochondrial toxicity is of concern. Although all three methods were effective in monitoring changes in mtDNA content, the ICC approach was considerably faster, less expensive, and is easily amenable to high-throughput screening. Importantly, ICC data were confirmed by direct detection of mtDNA content by real-time PCR and FISH analysis. This illustrates the utility of antibodies directed against mtDNA-encoded proteins as surrogate markers for mtDNA in the context of NRTI screening and as a front-line method to monitor mtDNA toxicity.

Footnotes

Acknowledgements

We are grateful to Michael J. Ignatius for review of this manuscript, to James D. Hirsch for conjugation of secondary antibodies, and to Marci Cardon for assistance with manuscript preparation. Alexa Fluor, MitoTracker, and ProLong are registered trademarks of Molecular Probes, Inc.

References

Anderson

Bankier

Barrell

de Bruijn

Coulson

Drouin

Eperon

, et al. (1981) Sequence and organization of the human mitochondrial genome. Nature 290:457–465

Brinkman

Kakuda

(2000) Mitochondrial toxicity of nucleoside analogue reverse transcriptase inhibitors: a looming obstacle for long-term antiretroviral therapy? Curr Opin Infect Dis 13:5–11

Carr

Cooper

(2000) Adverse effects of antiretroviral therapy. Lancet 356:1423–1430

Carrozzo

Bornstein

Lucioli

Campos

de la Pena

Petit

Dionisi-Vici

, et al. (2003) Mutation analysis in 16 patients with mtDNA depletion. Hum Mutat 21:453–454

Chariot

Drogou

de Lacroix-Szmania

Eliezer-Vanerot

Chazaud

Lombes

Schaeffer

, et al. (1999) Zidovudine-induced mitochondrial disorder with massive liver steatosis, myopathy, lactic acidosis, and mitochondrial DNA depletion. J Hepatol 30:156–160

Cossarizza

(2003) Tests for mitochondrial function and DNA: potentials and pitfalls. Curr Opin Infect Dis 16:5–10

Cote

Brumme

Craib

Alexander

Wynhoven

Ting

Wong

, et al. (2002) Changes in mitochondrial DNA as a marker of nucleoside toxicity in HIV-infected patients. N Engl J Med 346:811–820

Elpeleg

Mandel

Saada

(2002) Depletion of the other genome-mitochondrial DNA depletion syndromes in humans. J Mol Med 80:389–396

Feng

Johnson

Anderson

(2001) Insights into the molecular mechanism of mitochondrial toxicity by AIDS drugs. J Biol Chem 276:23832–23837

10.

Fosslien

(2001) Mitochondrial medicine—molecular pathology of defective oxidative phosphorylation. Ann Clin Lab Sci 31:25–67

11.

Gourlain

Amellal

Ait Arkoub

Dupin

Katlama

Calvez

(2003) Quantitative analysis of human mitochondrial DNA using a real-time PCR assay. HIV Med 4:287–292

12.

Haugland

(2000) Antibody conjugates for cell biology. In Bonifacino

Dasso

Harford

Lippincott-Schwartz

Yamada

, eds. Current Protocols in Cell Biology. New York, John Wiley & Son, 16.5.1–16.5.22

13.

Hellerstein

(2003) Turnover of adipose componentsand mitochondrial DNA in humans: kinetic biomarkers for human immunodeficiency virus-associated lipodystrophy and mitochondrial toxicity. Clin Infect Dis 37(suppl 2):S52–61

14.

Kukhanova

Chen

King

Doyle

Prusoff

Cheng

(1998) Interaction of beta-L-2',3'-dideoxy-2',3'-didehydro-5-fluoro-CTP with human immunodeficiency virus-1 reverse transcriptase and human DNA polymerases: implications for human immunodeficiency virus drug design. Mol Pharmacol 53:801–807

15.

Lewis

(2003) Mitochondrial DNA replication, nucleoside reverse-transcriptase inhibitors, and AIDS cardiomyopathy. Prog Cardiovasc Dis 45:305–318

16.

Lim

Copeland

(2001) Differential incorporation and removal of antiviral deoxynucleotides by human DNA polymerase gamma. J Biol Chem 276:23616–23623

17.

Marusich

Robinson

Taanman

Kim

Schillace

Smith

Capaldi

(1997) Expression of mtDNA and nDNA encoded respiratory chain proteins in chemically and genetically-derived Rho⁰ human fibroblasts: a comparison of subunit proteins in normal fibroblasts treated with ethidium bromide and fibroblasts from a patient with mtDNA depletion syndrome. Biochim Biophys Acta 1362:145–159

18.

Moyle

(2000) Clinical manifestations and management of antiret-roviral nucleoside analog-related mitochondrial toxicity. Clin Ther 22:911–936

19.

Petit

Mathez

Barthelemy

Leste-Lasserre

Naviaux

Sonigo

Leibowitch

(2003) Quantitation of blood lymphocyte mitochondrial DNA for the monitoring of antiretroviral drug-induced mitochondrial DNA depletion. J Acquir Immune Defic Syndr 33:461–469

20.

Phillips

Haas

Bannykh

Hathaway

Gray

Kimura

Vladutiu

, et al. (2002) Statin-associated myopathy with normal creatine kinase levels. Ann Intern Med 137:581–585

21.

Sastre

Pallardo

Vina

(2003) The role of mitochondrial oxidative stress in aging. Free Radic Biol Med 35:1–8

22.

Schapira

(2002) Primary and secondary defects of the mitochondrial respiratory chain. J Inherit Metab Dis 3:207–214

23.

Scheffler

(1999) Mitochondria. New York, John Wiley & Sons

24.

Schon

Bonilla

DiMauro

(1997) Mitochondrial DNA mutations and pathogenesis. Bioenerg Biomembr 29:131–149

25.

Taanman

(1999) The mitochondrial genome: structure, transcription, translation and replication. Biochim Biophys Acta 1410:103–123

26.

Taanman

Bodnar

Cooper

Morris

Clayton

Leonard

Schapira

(1997) Molecular mechanisms in mitochondrial DNA depletion syndrome. Hum Mol Genet 6:935–942

27.

Temesgen

(2001) Current status of antiretroviral therapies. Expert Opin Pharmacother 2:1239–1246

28.

Walker

(2003) Update on mitochondrial toxicity: where are we now? J HIV Ther 8:32–35

29.

Yamagata

Muro

Usui

Hagiwara

Kai

Arakawa

Shimizu

, et al. (2002) Mitochondrial DNA mutations in focal segmental glomerulosclerosis lesions. J Am Soc Nephrol 13:1816–1823