The antiretroviral 2',3'-dideoxycytidine causes mitochondrial dysfunction in proliferating and differentiated HepaRG human cell cultures

Abstract

Nucleoside reverse transcriptase inhibitors (NRTIs) were the first drugs used to treat human immunodeficiency virus infection, and their use can cause mitochondrial toxicity, including mitochondrial DNA (mtDNA) depletion in several cases. The first-generation NRTIs, including 2',3'-dideoxycytidine (ddC), were originally and are still pursued as anticancer agents. NRTI-sensitive DNA polymerases localizing to mitochondria allow for the opportunity to poison proliferating cancer cell mtDNA replication as certain cancers rely heavily on mitochondrial functions. However, mtDNA replication is independent of the cell cycle creating a significant concern that toxicants such as ddC impair mtDNA maintenance in both proliferating and nonproliferating cells. To examine this possibility, we tested the utility of the HepaRG cell line to study ddC-induced toxicity in isogenic proliferating (undifferentiated) and nonproliferating (differentiated) cells. Following ddC exposures, we measured cell viability, mtDNA copy number, and mitochondrial bioenergetics utilizing trypan blue, Southern blotting, and extracellular flux analysis, respectively. After 13 days of 1 μM ddC exposure, proliferating and differentiated HepaRG harbored mtDNA levels of 0.9% and 17.9% compared with control cells, respectively. Cells exposed to 12 μM ddC contained even less mtDNA. By day 13, differentiated cell viability was maintained but declined for proliferating cells. Proliferating HepaRG bioenergetic parameters were severely impaired by day 8, with 1 and 12 μM ddC, whereas differentiated cells displayed defects of spare and maximal respiratory capacities (day 8) and proton-leak linked respiration (day 14) with 12 μM ddC. These results indicate HepaRG is a useful model to study proliferating and differentiated cell mitochondrial toxicant exposures.

Keywords: 2′-3′-dideoxycytidine (ddC, zalcitabine); HepaRG; bioenergetics; cell biology; human immunodeficiency virus (HIV); mitochondrial DNA (mtDNA) maintenance; mitochondrial toxicity; nucleoside reverse-transcriptase inhibitor (NRTI).

Figure 1

HepaRG differentiation timeline and mitochondrial DNA (mtDNA) content.A, timeline of HepaRG differentiation. CM, Combination Medium; WDM, Working Differentiation Medium; WGM, Working Growth Medium (see Experimental procedures). B, a representative Southern blot of relative proliferating and differentiated HepaRG mtDNA content. Triplicate 1 μg reactions of whole-cell extracted (WCE) DNA were digested with BamHI, then loaded and electrophoresed on an agarose gel before blotting. The Southern blot was probed simultaneously with mtDNA (MT) and nDNA (N) probes. Proliferating HepaRG WCE DNA samples were prepared from cells obtained 7 days post-seeding at 2 × 10⁴ cells/cm². Differentiated HepaRG DNA samples were prepared from cells obtained 7 days post-differentiation. BamHI-digested human genomic DNA samples generate a 16.6-kb mtDNA genome-length band and a 2.2-kb nDNA band as described (79). Using the relative comparison of the MT to the N probe, differentiated HepaRG cells contain ∼2-fold more mtDNA than proliferating cells. Data are presented as mean ± standard deviation (SD); n = 12 from n ≥ 3 whole-cell DNA extract preparations derived from independent passages. Four Southern blots containing triplicate sample replicates were used for the analysis. For each lane of the dual-probed blots, the mtDNA peak 'Percent' value was normalized to the 18S nDNA peak 'Percent' value. A peak percent value is defined as the area of a peak measured as a percent of the total size of all measured peak areas on a blot. The mean normalized band intensity values of the proliferating HepaRG samples were set to 100% (∗∗∗∗p < 0.0001). C, quantitation of mtDNA molecules per cell. Representative blots are shown for each of proliferating (Prolif.) and differentiated (Diff.) HepaRG. Triplicate 1 μg reactions of WCE DNA were digested with BamHI then loaded and electrophoresed on an agarose gel. In parallel, 7.0, 3.5, and 1.7 ng of HindIII linearized pCR2.1-TOPO-mtDNA plasmid (4.4-kb) were loaded. The blots were detected with the MT probe. Three-point standard curves based on the pCR2.1-TOPO-mtDNA peak area percent values were used to estimate the mtDNA copy number and the 7-ng plasmid bands were set to 1. Two curves from two blots are plotted below the representative blots. The average mtDNA peak area percent values for 1 μg proliferating or differentiated HepaRG whole-cell DNA extracts are shown in gray with errors reported as SD (n = 6; triplicate measurements from two independent experiments using different passages; P10 and P14, differentiated; P8 and P9, proliferating HepaRG). On average, 1 μg of proliferating HepaRG WCE DNA was prepared from 1.4 × 10⁵ ± 4.4 × 10⁴ cells and 1 μg of differentiated WCE DNA was prepared from 9.8 × 10⁴ ± 1.7 × 10⁴ cells (n = 6 for each; triplicate cell counts from two independent experiments using different passages). The calculated copies of mtDNA are 3.7 × 10⁸ mtDNA/μg or 2600 mtDNA genomes per proliferating cell and 7.2 × 10⁸/μg or 7400 mtDNA genomes per differentiated cell.

Figure 2

HepaRG growth rate and survival following 2′,3′-dideoxycytidine (ddC) treatment.A, proliferating HepaRG growth rate was determined by growing cells in Working Growth Medium (WGM) at 5% CO₂, 37 °C. Cells were seeded at 2 × 10⁴ cells/cm². The mean doubling time (DT) based on three independent experiments utilizing HepaRG at passages P8, P9, and P16 is reported in hours (h) with error as SD. Mean viable cell count values (n = 4) and SDs for a representative experiment (P8) are shown in the graph. DT values were calculated using the least-squares fit of the exponential growth equation in Graph Pad Prism. B, a representative proliferating HepaRG survival curve is shown following exposure to ddC. The mean IC₅₀ value based on two independent experiments utilizing HepaRG at passages P15 and P16 is reported in micromolar (μM) with error as SD. Cells were seeded at 2 × 10⁴ cells/cm² in WGM the day before the addition of WGM treatment media. C, a representative differentiated HepaRG survival curve is shown following exposure to Working Differentiation Medium ddC treatment media. The mean IC₅₀ value based on two independent experiments utilizing differentiated cells at passages P8 and P10 is reported in micromolar with error as SD. In B, proliferating and C, differentiated experiments, cells were exposed to 128, 8, 4, 2, 1, 0.5, 0.25, and 0 μM ddC for 15 days, and mean survival values (n ≥ 3) and SDs are reported for each graph. IC₅₀ values were calculated using the least-squares fit of inhibitor concentration versus normalized response in Graph Pad Prism.

Figure 3

A greater amount of mitochondrial DNA (mtDNA) depletion is observed in proliferating HepaRG as compared with differentiated HepaRG after 2′,3′-dideoxycytidine (ddC) exposure.A, workflow for seeding, culturing, and exposing HepaRG cells to ddC. Twenty-four tissue culture dishes were seeded with 2 × 10⁴ proliferating HepaRG cells/cm², then either subjected to the differentiation protocol (thin black arrows) or incubated overnight (O/N), then exposed to ddC on the next day in Working Growth Medium (WGM) (thick gray arrows). After the 31-day differentiation process, cells were exposed to ddC in Working Differentiation Medium (WDM), Day 0. For each experiment, one-quarter of the tissue culture dishes were harvested on day 6 post-ddC exposure for Southern blotting (SB) and one-quarter were harvested on day 7 for Seahorse (SH). The remaining half of the tissue culture dishes were harvested on day 13. Experiments were repeated three times with three different cell passages (see Experimental procedures for details). Southern blots of proliferating and differentiated HepaRG mtDNA content following B, 6 and C, 13 days of ddC exposure. A representative blot is shown for each experiment. Triplicate 1 μg reactions of whole-cell DNA extracts were digested with BamHI then loaded and electrophoresed on an agarose gel before blotting and nonradioactive probe hybridization. Except for blots containing samples obtained from proliferating HepaRG at day 13, mtDNA content was measured on blots that were simultaneously probed with the 18S nDNA probe (N) and the mtDNA-specific probe (MT) as described in the legend for Figure 1, and the mean normalized band intensity values of the vehicle control samples (0 μM ddC) were set to 100%. Dual-probed blots containing day 13 samples from proliferating HepaRG treated with 1 μM could not be quantitated owing to low mtDNA signal; however, when the blots were stripped and probed individually with the mtDNA probe, stripped again, and then probed with the 18S probe, 0 and 1 μM mtDNA bands could be quantitated in Fiji as described (79). To confirm equal loading of whole-cell extracted DNA into each lane, the %CV for the nDNA bands on the single-probed 18S blots were at most 11% (100% ± 11% for 0 μM; 102% ± 10% for 1 μM; 101% ± 8% for 12 μM, errors are SDs), and the 0, 1, and 12 μM ddC–treated 18S sample band intensities are not significantly different from one another as judged by a one-way ANOVA. Data presented in the graphs are mean ± SD, n ≥ 9 (≥3 blots from three independent experiments using different preparations/passages of cells); statistically significant differences in mtDNA content were determined by ANOVAs except for proliferating HepaRG day 13, which was done by a Welch’s t test as the 12 μM ddC–treated sample had no detectable mtDNA. ∗∗∗∗p ≤ 0.0001 between vehicle control (0 μM) and 1 or 12 μM. D. An overexposed image of proliferating HepaRG day 13 samples with the mtDNA-specific probe (same image as in C). BCA, bicinchoninic acid.

Figure 4

HepaRG viable cell counts following 2′,3′-dideoxycytidine (ddC) exposure. HepaRG cells were exposed to vehicle, 1, or 12 μM ddC for 7 and 13 days, then cells were counted. The number of viable cells remaining on the cell culture dishes was determined utilizing the trypan blue exclusion method. In each set of experiments, the vehicle control–treated (0 μM ddC) cell counts were set to 100 (%). Data are presented as mean viable cell counts ±SD, n ≥ 12 (≥quadruplicate from three independent experiments using different cell passages). The percentage of the viable (clear) cells of the total remaining cell population (blue and clear) is reported under the graphs as percent viability. The statistical significance was determined by one-way ANOVA for parametric data and by Kruskal–Wallis tests for nonparametric data. ∗∗∗∗p ≤ 0.0001 between vehicle control (0 μM) and 1 or 12 μM; ∗∗∗p ≤ 0.001 between 0 and 1 μM; ##p ≤ 0.01 between 1 and 12 μM.

Figure 5

Exposure to 2′,3′-dideoxycytidine (ddC) causes aberrant bioenergetics in proliferating and differentiated HepaRG cells.A, proliferating (Pro.) and differentiated (Dif.) HepaRG Mito Stress test oxygen consumption rates (pmol O₂/min/μg cellular protein) and extracellular acidification rates (mpH/min/μg cellular protein) following 8 days of treatment with 0, 1, or 12 μM ddC. Metabolic stressors were injected sequentially from Ports a (2 μM oligomycin final well concentration), b (1 μM FCCP final well concentration), and c (0.5 μM antimycin A + 0.5 μM rotenone, final well concentrations). A.A., antimycin A; Olig., oligomycin; Rot., rotenone. Data are mean values ± SD; n ≥ 12, ≥quadruplicate from three independent experiments using different preparations/passages of cells. B, scatter dot plots of proliferating and differentiated HepaRG mitochondrial bioenergetic parameters following 8 and 14 days of ddC exposure. Bioenergetic parameters: Proton Leak, and nonmitochondrial respiration (Non-Mito Resp.), ATP-linked respiration (ATP-linked Resp.), basal respiration (Basal Resp.), spare respiratory capacity (Spare Resp. Cap.), and maximal respiratory capacity (Max. Resp. Cap.). Data are presented with mean values and SD errors on the plots, n ≥ 11 (greater than triplicate from three independent experiments using different passages). The fold-changes (Δ) of 0 μM ddC (control)-treated bioenergetic parameters relative to 12 and 1 μM ddC–treated cells are listed under each plot. The statistical significance was determined by three-way ANOVA. ∗∗∗∗p ≤ 0.0001, 0 versus 1 or 12 μM; ∗∗∗p ≤ 0.001, 0 versus 1 μM; ∗p ≤ 0.05, 0 versus 12 μM; $$$$p ≤ 0.0001, D8 versus D14 at 1 μM; $$p ≤ 0.01, D8 versus D14 at 1 μM; ####p ≤ 0.0001, D8 versus D14 at 12 μM; ##p ≤ 0.01, D8 versus D14 at 12 μM; #p ≤ 0.05, D8 versus D14 at 12 μM. The p-values for differentiated HepaRG spare respiratory and maximal respiratory capacities on day 14 post-treatment with 12 μM ddC were 0.07 and 0.09, respectively. Bioenergetic parameters were calculated as previously described (31).

Figure 6

Mitochondrial DNA (mtDNA) replication occurs in proliferating and differentiated cells. Nucleoside reverse-transcriptase inhibitors (NRTIs) are metabolically activated by intracellular kinases to nucleotide reverse transcriptase inhibitors (NtRTIs). The insets highlight the mitochondrial matrix, the location of the DNA polymerases, and the mtDNA genome encoding 13 polypeptides of the oxidative phosphorylation (OXPHOS) machinery. The mtDNA genes are colored by OXPHOS complex, orange, purple, green, and red. Orange, OXPHOS complex I (NADH dehydrogenase); black, OXPHOS complex II (succinate dehydrogenase); purple, OXPHOS complex III (cytochrome bc1 complex); green, OXPHOS complex IV (cytochrome c oxidase); red, OXPHOS complex V (ATP synthase). The OXPHOS machinery is localized within the mitochondrial inner membrane (MIM). Within the mitochondrion, NtRTIs compete with native nucleotides at DNA polymerase active sites to inhibit mtDNA replication through chain termination and persistence in the mtDNA genome. H+, proton; O₂, oxygen; H₂O, water; ADP, adenosine diphosphate; ATP, adenosine triphosphate; TS, mitochondrial transcription; TL, mitochondrial translation; DNA polymerases gamma (Polγ), beta (Polβ), theta (Polθ), zeta (Polζ), and PrimPol. See the text for details.