High-dose clevudine impairs mitochondrial function and glucose-stimulated insulin secretion in INS-1E cells

Background Clevudine is a nucleoside analog reverse transcriptase inhibitor that exhibits potent antiviral activity against hepatitis B virus (HBV) without serious side effects. However, mitochondrial myopathy has been observed in patients with chronic HBV infection taking clevudine. Moreover, the development of diabetes was recently reported in patients receiving long-term treatment with clevudine. In this study, we investigated the effects of clevudine on mitochondrial function and insulin release in a rat clonal β-cell line, INS-1E. Methods The mitochondrial DNA (mtDNA) copy number and the mRNA levels were measured by using quantitative PCR. MTT analysis, ATP/lactate measurements, and insulin assay were performed. Results Both INS-1E cells and HepG2 cells, which originated from human hepatoma, showed dose-dependent decreases in mtDNA copy number and cytochrome c oxidase-1 (Cox-1) mRNA level following culture with clevudine (10 μM-1 mM) for 4 weeks. INS-1E cells treated with clevudine had reduced total mitochondrial activities, lower cytosolic ATP contents, enhanced lactate production, and more lipid accumulation. Insulin release in response to glucose application was markedly decreased in clevudine-treated INS-1E cells, which might be a consequence of mitochondrial dysfunction. Conclusions Our data suggest that high-dose treatment with clevudine induces mitochondrial defects associated with mtDNA depletion and impairs glucose-stimulated insulin secretion in insulin-releasing cells. These findings partly explain the development of diabetes in patients receiving clevudine who might have a high susceptibility to mitochondrial toxicity.


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][7][8][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][14][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.

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 Gen-Bank 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.  (5 × 10 4 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]. 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.

Results
Effects of clevudine on mtDNA copy number and mRNA levels of mtDNA encoded genes INS-1E cells were cultured with different concentrations of clevudine for 4 weeks and the in vitro effects on mtDNA replication and translation were measured. Treatment with clevudine (10 μM to 1 mM) reduced the mtDNA copy number in a dose-dependent manner (Figure 1A). Two weeks treatment with clevudine induced 39% reduction of mtDNA level (n = 3), which was smaller than four weeks treatment (51%). The mRNA levels of mtDNA-encoded Cox-1 were also dose-dependently attenuated by clevudine ( Figure 1B). Interestingly, we observed upregulation of PPAR-γ coactivator 1α (PGC-1a), mitochondrial transcription factor A (Tfam), and nuclear respiratory factor 1 (NRF1) in clevudine-treated INS-1E cells. Upregulation of these transcription factors could be a nuclear response to mitochondrial dysfunction [3]. Succinate dehydrogenase (SDH), a nuclear-encoded mitochondrial enzyme, was also upregulated by clevudine-treatment ( Figure 1C). We next examined the effects of clevudine on the levels of mtDNA and RNA in the human hepatoma cell line HepG2, major target cells of insulin action. Clevudine showed suppressive effects on mtDNA replication and transcription in HepG2 cells, similar to the effect in INS-1E cells ( Figures 1D and 1E).

Mitochondrial dysfunction induced by clevudine
The amount of formazan reaction product formed in the MTT assay reflects the total mitochondrial enzymatic activity in each well. INS-1E cells treated with or without clevudine for 4 weeks were seeded 48 hrs before the MTT assay. Exposure to clevudine decreased the MTT absorbance (71% by 100 μM and 56% by 1 mM, Figure  2A). However, we observed that there was no significant difference in protein amount between control cells (59 ± 3 μg; n = 17) and cells incubated with 100 μM (62 ± 4 μg; n = 17) or 1 mM clevudine (59 ± 3 μg; n = 17) for 48 hours after seeding (3 × 10 5 cells). This result implies that reduction of MTT signal by clevudine might be resulted from decreased mitochondrial reducing capacity. To demonstrate the functional significance of decreased Cox1 mRNA, we performed the enzymatic activity measurement of cytochrome c oxidase (Cox) based on previous reports [16,17]. We observed the reduced Cox activity of clevudine (1 mM)-treated cells compared to that of control INS-1E cells ( Figure 2B). We measured the cellular contents of ATP in control and clevudine-treated INS-1E cells using a bioluminescence method after incubation with low (2.8 mM) or high (16.7 mM) concentrations of glucose for 15 min. As shown in Figure 3A, cells cultured with 1 mM clevudine had lower cytosolic ATP level in both low and high glucose conditions than control cells. Lactate production from INS-1E cells was markedly elevated by incubation with glucose for 15 min ( Figure 3B). The glucose-induced lactate production was increased in cells treated with 1 mM clevudine compared to that in control cells ( Figure 3B).
One of the metabolic consequences of mitochondrial dysfunction is an impairment of fatty acid oxidation, thus leading to lipid accumulation [19,20]. To detect the lipid droplet in cytosol, we performed Oil-red O staining to control and clevudine-treated cells. Without exogenous fatty acid loading, there was no significant difference between two groups. When we incubated cells with a mono-unsaturated fatty acid, 0.7 mM oleate, clevudinetreated cells showed a pronounced lipid accumulation, which was much less in control cells (Figure 4).

Inhibition of glucose-stimulated insulin secretion by clevudine
To identify whether clevudine-induced mitochondrial dysfunction affects insulin secretory activity, we measured the released and cellular contents of insulin via an enzyme immunoassay. The cellular insulin contents were not significantly different between control INS-1E cells (816 ± 167 ng/well) and cells treated with clevudine (1 mM) for 4 weeks (769 ± 170 ng/well). After incubation for 30 min with low (2.8 mM) or high (16.7 mM) concentrations of glucose, the released insulin was normalized to the cellular content and expressed as a percentage of the content released. We observed that high concentration glucose stimulated the release of insulin by 5.1-fold in control cells but by only 3.1-fold and 1.9-fold in cells treated with 100 μM and 1 mM clevudine for 4 weeks, respectively (Figure 3). There was no difference in % insulin releases induced by low concentration glucose between the control and clevudinetreated groups ( Figure 5A). We observed that the glucose-stimulated insulin secretion was completely abolished by the treatment with oligomycin (0.75 μg/ml), a mitochondrial ATP synthase inhibitor ( Figure 5B). This result demonstrates the cause-effect relationship Nuclear-encoded genes related to mitochondrial biogenesis such as PPAR-γ coactivator 1α (PGC-1a), mitochondrial transcription factor A (Tfam), and nuclear respiratory factor 1 (NRF1) were upregulated by clevudine treatment (C). Succinate dehydrogenase (SDH) was also upregulated by clevudine (C). Data presented are means ± SEM. The number of experiments is shown in parenthesis. *, **, and *** denote p < 0.05, p < 0.01, and p < 0.001, respectively. between mitochondrial dysfunction and impaired insulin secretion.

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, clevudineinduced 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 NRTItreated 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 clevudineinduced 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 Figure 5 Clevudine inhibited glucose-stimulated insulin secretion. Released insulin and cellular insulin content were measured using an enzymatic immunoassay after incubation with Krebs buffer containing 2.8 mM or 16.7 mM glucose for 15 min. Glucose-stimulated insulin secretion was impaired by clevudine incubation (A), as well as by inhibiting mitochondrial ATP synthase with oligomycin (B). Insulin release was normalized to the insulin content. Data presented are mean ± SEM. ** denotes p < 0.01.
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.

Conclusions
In summary, clevudine, used as an antiviral agent against chronic hepatitis B, significantly decreased the mtDNA copy number at higher doses compared to therapeutic concentration. Mitochondrial dysfunction due to depleted mtDNA and defective ATP synthesis in insulin-releasing cells, consequently led to the impairment of glucose-stimulated insulin secretion. Clevudine-induced mitochondrial dysfunction may contribute to diabetogenesis among clevudine-treated patients who might be more susceptible to mitochondrial toxicity.