Background
Chronic infection with hepatitis B virus (HBV) frequently leads to serious liver disease such as cirrhosis, fulminant hepatic failure, and hepatocellular carcinoma [
1]. Several antiviral drugs have been developed and prescribed for HBV infection. Commonly used antiviral therapies are nucleoside analog reverse transcriptase inhibitors (NRTIs) including entecavir, lamivudine, and telbivudine. NRTIs undergo intracellular and intramitochondrial phosphorylation into active triphosphates that are capable of inhibiting HIV reverse transcriptase (RT) [
2]. However, these drugs have side effects such as lipodystrophy, neuropathy, myopathy, and liver steatosis, all of which are related to mitochondrial toxicity.
In vitro and
in vivo studies have shown that some NRTIs inhibit DNA polymerase-γ, a nuclear-encoded polymerase important for mitochondrial DNA (mtDNA) replication [
3,
4]. Depletion of mtDNA induced by NRTIs may attenuate mitochondrial oxidative phosphorylation, which could limit their clinical use.
Clevudine (1-(2-deoxy-2-fluoro-β-L-arabinofuranosyl) thymine) is an NRTI that exhibits potent and sustained antiviral activity against HBV with weaker effects on mitochondrial structure and function compared to those of other NRTIs [
2,
5]. However, long-term therapy for more than one year results in the development of considerable drug resistance and skeletal myopathy [
6‐
9]. Muscle biopsies from patients with myopathy as a complication of clevudine treatment revealed severe necrosis with cytochrome c oxidase (COX)-negative ragged red fibers, the typical phenotype of mitochondrial myopathy [
7,
10]. Clevudine-induced myopathy developed in approximately 4-5% of patients and was usually reversible after discontinuation of clevudine [
9].
It is well known that mitochondria play a critical role in nutrient-stimulated insulin secretion, as well as in insulin actions at target cells [
11]. Recently, a patient who developed diabetes mellitus after clevudine treatment was reported [
12]. We hypothesized that the mitochondrial dysfunction invoked by clevudine treatment could be a precipitating factor in diabetogenesis. Until now, the majority of
in vitro studies for antiviral agent toxicities have been performed in different cell types, yielding conflicting results [
13‐
15]. Insulin-secreting cells are highly specialized fuel sensors that maintain blood glucose level in the body by monitoring the ATP/ADP ratio, which is strictly regulated by mitochondrial oxidative phosphorylation. Thus, insulin-secreting cells are an appropriate model system for identification of mitochondrial toxicity and its functional consequences following antiviral therapy. In this study, we investigated the effects of clevudine exposure on mtDNA content, mitochondrial function, and metabolism-secretion coupling in insulin-releasing cells to elucidate the mechanism underlying the reversible diabetes observed in clevudine-treated patients.
Methods
Cell culture and drugs
Clevudine was purified from Revovir® tablets (Bukwang Pharm. Co., Seoul, Korea). The amount of harvested clevudine was analyzed using HPLC (Agilent G1315B UV Diode array detector, AD, Santa Clara, CA, USA). A single peak with the expected amount of clevudine was measured based on the known weight of one tablet. INS-1E cell, a clonal pancreatic β-cell line received from Prof. Claes B. Wollheim, were cultured in complete medium composed of RPMI 1640 (Invitrogen, Carlsbad, CA, USA) supplemented with 10% fetal calf serum, 1 mM sodium pyruvate, 50 μM 2-mercaptoethanol, 2 mM glutamine, 10 mM HEPES, 100 units/ml penicillin, and 100 μg/ml streptomycin. HepG2 cells, a human hepatoma cell line, were grown in DMEM medium (Invitrogen) containing 5.6 mM glucose, 4 mM L-glutamine and 1 mM sodium pyruvate. For the following experiments, cells were cultured with or without clevudine for 4 weeks.
Quantitative PCR
Total DNA or RNA was isolated and purified from INS-1E and HepG2 cells using DNeasy or RNeasy kits (Qiagen, Valencia, CA, USA), respectively. To obtain cDNA, reverse transcription (RT) was performed with oligo-dT (Applied Biosystems, Foster City, CA, USA) using reverse transcriptase (Promega, Madison, WI, USA). For PCR amplification, sequence-specific oligonucleotide primers for the genes of interest were designed (Bioneer, Daejeon, Korea) based on rat and human sequences in the GenBank database (Table
1). Quantitative real-time PCR using SYBR Green PCR Master Mix (Applied Biosystems) was performed in an ABI PRISM 7900HT Sequence Detection System (Applied Biosystems) according to the manufacturer's protocol. All amplifications were followed by melting curve analysis. The β-actin was used as the reference gene, and relative abundance of DNA or mRNA in clevudine-treated cells was normalized to that level in control cells calculated by using 2
-ΔΔCt method.
Table 1
Primers for quantitative PCR
rat β-actin |
+
| ATGGTGGGTATGGGTCAGAA | NM_031144.2 | 100 bp |
|
-
| TCCATATCGTCCCAGTTGGT | | |
rat Cox1 |
+
| GGAGCAGTATTCGCCATCAT | NC_001665 | 90 bp |
|
-
| GTGGGCTTTTGCTCATGTGT | | |
rat PGC1a |
+
| GGCACATCTGTTCTTCCACA | NM_031347.1 | 110 bp |
|
-
| TTCCTGGTCTTGGAGCTGTT | | |
rat NRF1 |
+
| GGACAGCAAGCCATTGTTCT | NM_001100708.1 | 98 bp |
|
-
| TACTTGCGCACCACATTCTC | | |
rat Tfam |
+
| GCTGAGTGGAAGGTGTACAAAG | NM_031326.1 | 85 bp |
|
-
| CTTCCTTCTCTAAGCCCATCAG | | |
rat SDH |
+
| TCTTTCCTACCCGCTCACAT | NM_130428.1 | 90 bp |
|
-
| AATGCCATCTCCAGTTGTCC | | |
human β-actin |
+
| AAGTTCACAATGTGGCCGAG | NM_001101.3 | 98 bp |
|
-
| ATGGCAAGGGACTTCCTGTA | | |
human Cox1 |
+
| CACACTCCACGGAAGCAATA | NC_012920.1 | 82 bp |
|
-
| GCCACCTACGGTGAAAAGAA | | |
MTT assay
3-(4,5-dimethylhioazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) was purchased from Sigma (St. Louis, MO, USA). INS-1E cells seeded onto a 96-well plate (5 × 104 cells/well) were incubated with MTT (50 μg/well) for 2 hrs, and then the medium was discarded and replaced with dimethylsulfoxide (100 μl/well). The absorbance of each well was measured at 570 nm using an enzyme-linked immunosorbent assay (ELISA) reader, after background subtraction at 650 nm.
Cytochrome c oxidase (COX) activity measurement
INS-1E cells were harvested and incubated with isosmotic medium [
16] containing 0.2% triton X-100 at 30°C for 2 min. Enzymatic activity of COX was measured spectrophotometrically at 550 nm based on previous reports [
16,
17].
ATP and lactate measurements
INS-1E cells seeded onto 24-well plates (3 × 10
5 cells/well) were preincubated with glucose-free medium for 2 hrs prior to incubation with KRBH solution (135 mM NaCl, 3.6 mM KCl, 2 mM NaHCO
3, 0.5 mM NaH
2PO
4, 0.5 mM MgSO
4, 1.5 mM CaCl
2, 10 mM HEPES, pH 7.4) containing 2.8 mM glucose for 30 min. The cells were then stimulated for 15 min with KRBH buffer at a low (2.8 mM) or high (16.7 mM) glucose concentration. The ATP content in the cell lysate (Roche HS-II Biolumniscence kit, Mannheim, Germany) and the lactate level in the cell supernatant (Biovision #K607-100, Mountain View, CA, USA) were measured as described previously [
18]. Measurement of the protein concentration in cell lysates was performed using the Bradford assay.
Oil red staining
INS-1E cells on coverslip were treated with bovine serum albumin (BSA) or oleate, a monounsaturated fatty acid for 24 hours. After fixation with 10% formalin, cells were washed with 60% isopropanol and dried at room temperature. Cells were incubated with Oil Red O (Sigma, St. Louis, MO, USA) for 10 min, and then counterstained with hematoxylin.
Insulin measurement
INS-1E cells were seeded and cultured as for ATP and lactate measurement. For insulin measurement, 0.1% BSA was included in the KRBH solution, and the cells were stimulated with low or high concentrations of glucose for 30 min, as described previously [
18]. Insulin levels in supernatant and cell extracts were measured using a rat insulin enzyme immunoassay kit (Shibayagi Co., Gunma, Japan).
Data analysis
All data are presented as mean ± SEM, and the statistical significance was determined using One-way ANOVA or Student's t test.
Discussion
In pancreatic β-cells, mitochondria are of particular importance in the regulation of insulin secretion because they produce ATP as well as other coupling factors which link nutrient metabolism and insulin exocytosis [
11]. mtDNA-depleted β-cell lines show complete absence of nutrient-stimulated insulin secretion [
21]. Patients with mtDNA mutations develop diabetes, accounting for up to 1% of the total number of diabetic patients [
22]. Moreover, postmortem islets from type 2 diabetes patients display functional deterioration of mitochondria [
23]. Therefore, factors that disturb the mitochondrial function in pancreatic β-cells might affect metabolism-secretion coupling and diabetogenesis.
The present study showed that the effective anti-HBV agent clevudine has a negative effect on the copy number and transcription of mtDNA in insulin-releasing cells and hepatoma cells. The reduced expressions of mtDNA-encoded proteins lead to attenuation of mitochondrial function. In insulin-releasing cells, clevudine-induced mitochondrial dysfunction can elicit defective insulin secretion in response to substrates for mitochondrial metabolism. To our knowledge, this is the first demonstration that an antiviral agent can impair nutrient-stimulated insulin secretion as a result of mitochondrial dysfunction. Because of their high dependency on mitochondrial function in metabolism-secretion coupling, insulin-secreting cells provide a useful model to investigate the functional consequences of drug-induced mitochondrial toxicity.
NRTIs are widely used to treat various viral diseases such as acquired immunodeficiency syndrome (AIDS) and hepatitis B [
24]. However,
in vitro studies showed that NRTIs can alter mtDNA content by inhibiting DNA polymerase-γ [
25]. Moreover, myopathy accompanied by mtDNA depletion has been reported in NRTI-treated patients [
4]. Clevudine treatment has also been associated with the development of mitochondrial complications. In contrast to early studies [
2], depletion of mtDNA in skeletal muscle has been observed in patients treated with clevudine [
7,
26]. Typical histological features of mitochondrial myopathy and abnormal mitochondrial morphology were displayed in tissues from patients with increased lactate dehydrogenase and lactate levels [
8,
10]. Although the incidence of clevudine-induced myopathy was reported to be low (~5%) [
9], a substantial proportion (~14.5%) of clevudine-treated patients have been found to experience symptoms, signs, and laboratory abnormalities relevant to clevudine-induced myopathy [
27].
To directly confirm the effects of clevudine on mitochondrial function, we cultured cells with medium containing different concentrations of clevudine for 4 weeks. Clevudine markedly decreased the MTT signal without significant changes in cellular protein implying the diminished enzyme activities for reduction of MTT. Since MTT assay is not specific to evaluate mitochondrial function, measurement of oxygen consumption rate or citrate synthase activity could provide more concrete evidence to prove the mitochondrial defects. Consistent with mtDNA depletion, COX activity and cellular ATP content were reduced. Decreased mitochondrial fatty acid oxidation could induce triglyceride accumulation [
19]. To avoid lipotoxic effects of palmitate in insulin-secreting cells [
28], we loaded unsaturated fatty acid oleate for 24 hours, which elicited a marked increase of lipid accretion within clevudine-treated cells. The inhibitory effect of clevudine on insulin secretion was more sensitive than the effect on ATP level. We can speculate that the treatment of 100 μM clevudine elicited significant reduction of ATP/ADP ratio which is the main signal for closing ATP-sensitive K
+ channel and insulin exocytosis.
We also observed some compensatory responses to reduced mtDNA copy number and its functional consequences. First, PGC-1α and its downstream transcriptional factors, NRF-1 and Tfam, were upregulated by clevudine. Second, nuclear DNA-encoded succinate dehydrogenase was also upregulated, which has already been observed in muscle of patients suffering from clevudine-induced myopathy [
10]. Third, lactate production was modestly increased in association with diminished ATP content. Pancreatic β-cells and clonal β-cell lines are known to have very low lactate dehydrogenase levels, which contribute to their dependency on mitochondrial function. The increase in lactate production observed in our study also demonstrates that clevudine imposes selective defects on mitochondria rather than overall cytotoxicity.
In our study, mtDNA copy number in clevudine (1 mM)-treated cells was decreased to half of that in control. It has been reported, however, that to evoke mitochondrial dysfunction mtDNA level should fall below 60% which was named as 'phenotypic threshold' [
29]. This can be explained by genetic and functional complementation at the levels of transcription, translation, enzyme activity and cell activity. Several investigators showed that NRTI such as zidovudine and stavudine can also induce mitochondrial dysfunction independent from lack of mtDNA [
20,
30]. Thus, we cannot exclude the possibility that clevudine could be involved in multiple site of inhibition of mitochondrial function in addition to the effects of mtDNA depletion.
Niu et al. [
31] suggested that the intracellular level of the triphosphate form of clevudine in cells exposed to 1 μM extracellular clevudine approximates the plasma level in patients receiving a 30 mg dose. Our results indicated that impairments in mitochondrial function and insulin secretion are elicited only by high concentrations of clevudine (> 100 μM). This means that clevudine would minimally affect mitochondrial function within the therapeutic concentration range. It is noteworthy, however, that mutations or polymorphisms of DNA polymerase-γ were identified in NRTI-treated patients with mitochondrial complications [
32]. This suggests that genetic alterations in DNA polymerase-γ are not normally deleterious, but that certain conditions such as NRTI treatment may push mitochondrial activity below the clinical threshold, causing pathogenic dysfunction [
33]. Differences in genetic susceptibility to mitochondrial toxicity could be one explanation for why a limited proportion of patients receiving clevudine have complications including myopathy.
Clevudine-induced depletion of mtDNA is not restricted to insulin-secreting cells but is also observed in cultured hepatoma cells or muscle tissue from patients [
7,
26]. Mitochondrial dysfunction in insulin target tissues such as liver and muscle could result in insulin resistance and diabetes [
34]. In addition to defects in insulin secretion, decreased sensitivity in insulin target cells can also participate in diabetogenesis in patients receiving clevudine who might have a high susceptibility to mitochondrial toxicity. Interestingly, several reports have shown that NRTI induces intramitochondrial pyrimidine deficiency which may aggravate mtDNA depletion and mitochondrial dysfunction [
35,
36]. They also discovered that uridine supplementation attenuates steatohepatitis or mitochondrial myopathy induced by NRTI. Further studies concerning the effects of NRTIs on mitochondrial function in different cell types may help us understanding these intractable complications and develop novel antiviral agents.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
KSP and MYK designed the project. YOJ, XQ, RD, and SX performed the experiments. YOJ, KSP, and MYK wrote the manuscript. CMA prepared a purified clevudine. CHC, CMA, SKB, and IDK contributed to the discussion of the data and the revision of the manuscript. All readers read and approved the final manuscript.