Differential effects of cisplatin on cybrid cells with varying mitochondrial DNA haplogroups

Creation of cybrid cells

Cytoplasmic hybrids (cybrids), cell lines with identical nuclear genomes but different mitochondrial genomes, were created and used in the fifth passage for all experiments. All experiments were carried out in accordance with the Institutional Review Board at the University of California, Irvine, (IRB #2003-3131) and were consistent with Federal guidelines. All subjects in this study read and signed the written informed consent prior to beginning their participation. Cybrid cells were created using platelets isolated from peripheral blood that was fused with Rho0 (mtDNA free) ARPE-19 cells (Fig. 1) (Patel et al., 2019). The mtDNA haplogroups for each subject and cybrid cell line were identified using polymerase chain reaction (PCR) along with restriction enzyme digestion and mtDNA sequencing as described previously in the study conducted by Patel et al. (2019) on H and J mtDNA haplogroup cybrid cells.

Sequencing of mtDNA from L, A, B, and D Cybrids

DNA was extracted from the individual cybrids (n = 7 for L cybrids, n = 4 for [A+B] cybrids, and n = 3 for D cybrids) using a kit (DNeasy Blood and Tissue Kit, Qiagen, Germantown, MD). Sequencing techniques similar to those previously described in the study between H and J mtDNA haplogroup cybrids were conducted (Patel et al., 2019). Next Generation Sequencing (NGS) technology was used to sequence both strands of mtDNA independently in both directions. This was done to quantitate the haplogroup-defining single nucleotide polymorphisms (SNPs), private SNPs (not defining haplogroups), and low-frequency heteroplasmy SNPs across the entire mitochondrial genome. NGS technology is capable of deep sequencing (average sequencing depth of 30,000; range 1,000 to 100,000) and accurately differentiates low-frequency mtDNA heteroplasmy SNPs from DNA modification artifacts. The mtDNA sequences were analyzed using HaploGrep (https://haplogrep.i-med.ac.at/) to identify the mtDNA haplogroups (Figs. 2–4, Table 1). Defined SNP changes between the L, [A+B], and D cybrids were obtained using http://www.Phylotree.org. and amino acid changes and any associated pathogenesis resulting from differences in SNP variants were verified using http://www.MitoMap.org and/or http://www.hmtvar.uniba.it. The rs numbers were identified using http://www.ncbi.nlm.nih.gov/snp. All of the SNPs identified had a Quality Score of 100 (A Phred-scaled quality score assigned by the variant caller) and Passed all of the Filters (Patel et al., 2019) (Table 2).

Cybrid cell lines culture

Cybrid cell lines containing mitochondria with either L, [A+B], or D haplogroups were cultured in flasks until confluent. They were then plated on a 96-well or 24-well plates depending on the assay. The MTT, JC1 and LDH experiments were performed on 7 different L cybrids, each with mitochondria from different individuals. The [A+B] cybrids represented 4 different individuals. The D cybrids represented 3 different subjects. Each treatment condition was run in quadruplicate and the experiments repeated twice. The mean values of all cybrids were calculated for both experiments individually. Then these individual means were combined by haplogroup (L versus [A+B] versus D) for the final analyses.

Cells were plated in 96-well plates with each well containing 50,000 cells. After 24 h, media were removed, and cells were treated with 100 µL of 40 µM concentration of cisplatin (given as a gift from Professor Bota, UCI Medical Director, Neuro-Oncology Program). Generally, the concentration of cisplatin used to treat cancer cells for 48 h is 10 µM. However, a previous study done by Patel et al. on ARPE-19 cybrids containing either H or J mtDNA haplogroups found that the concentration best fit to inhibit these cells was 40 µM. This was concluded via IC-50 analysis which demonstrates the concentration of cisplatin necessary to inhibit cell viability by 50%. As a result, the 40 µM concentration was selected for treatment in the current study (Patel et al., 2019). Plates were then incubated at 37 °C for another 48 h.

Cellular metabolic activity measured by tetrazolium dye (MTT) assay

The MTT assay was conducted to assess cellular metabolic activity, which was in turn correlated to ATP production and overall cell viability. After the 48-hour incubation period, 10 µL of MTT solution (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide, (Biotium, Hayward, CA) was added to the 100 µL of medium in each well. Plates were then incubated at 37 °C for 2 h. 100 µL of DMSO was directly added into the medium in each well and pipetted up and down until thoroughly mixed. Absorbance was measured on an absorbance reader (BioTek, ELX 808, Winooski, VT) at 570 nm (measured) and 630 nm (reference). The reference absorbance value was subtracted from the measured value in order to attain true absorbance values. All values were normalized to the average metabolic activity of the untreated-L cybrids. Average values for the cybrids were then compared to the respective untreated cybrids using a two-tailed t-test (GraphPad Prism Software, Inc, San Diego, CA). P ≤ 0.05 was deemed statistically significant. Each treatment condition was run in quadruplicate, and the entire experiment was repeated twice.

Mitochondrial membrane potential (ΔΨm) assay (JC-1 assay)

L, [A+B], and D cybrids were plated in 24-well plates (100,000 cells/well), incubated 24 h and, treated with 0 or 40 µM of cisplatin for another 48 h. JC-1 reagent (5, 5′, 6, 6′-tetrachloro-1,1′, 3, 3′- tetraethylbenzimidazolylcarbocyanine iodide) (Biotium, Hayward, CA) was added to cultures for 15 min. Similar to the study conducted by Patel et al. a Gemini XPS Microplate Reader (Molecular Devices) was used to measure fluorescence for red (excitation 550 nm and emission 600 nm) and green (excitation 485 nm and emission 545 mm) wavelengths. Intact mitochondria with normal ΔΨm fluoresced red, while cells with decreased ΔΨm fluoresced green. Experiments were then analyzed in quadruplicate, and the entire experiment was repeated twice. All values were normalized to the average of the untreated-L cybrids, and cisplatin-treated cybrids were compared to untreated cybrids using a two-tailed t-test to assess for statistical significance (GraphPad Prism Software, Inc.) (Patel et al., 2019).

Mitochondrial membrane potential (ΔΨm) assay (IncuCyte^® live cell analyzer with ARPE-19 cells)

After cisplatin treatment, the cybrids showed decreased metabolic activity but the mitochondrial membrane potentials remained unchanged, which was unexpected. Therefore, we conducted additional experiments using the ARPE-19 cell line evaluated with the IncuCyte^® Live Cell Analyzer (Sartorius, City, State) and the Nuclight Rapid Red probe (A549 NucLight Red, cat # 4491) that allowed for comparison of growth/proliferation over 40 h in cultures treated with 20 µM or 40 µM cisplatin (n = 7 per treatment condition). The IncuCyte^® Analyzer captures changes in probes fluorescence and morphologic images in real time. The mitochondrial membrane potential was measured with the JC-1 reagent (5,5′,6,6′-tetrachloro-1,1′,3,3′- tetraethylbenzimidazolyl-carbocyanine iodide) (Biotium, Hayward, CA) that was added to cultures for 15 min. Fluorescence was measured for red (excitation 550 nm and emission 600 nm) and green (excitation 485 nm and emission 545 mm) wavelengths. Intact mitochondria with normal ΔΨm appeared red, while cells with decreased ΔΨm were in a green fluorescent state (n = 7 wells per treatment condition).

The ARPE-19 cells (ATCC, Manassas, VA); (Invitrogen-Gibco, Carlsbad, CA), which are the wildtype of the Rho0 receipt cells of the cybrids, were cultured in Dulbecco’s modified Eagle’s and Ham’s nutrient mixture F-12 (1:1 mixture, vol/vol), 0.37% sodium bicarbonate, 0.058% L-glutamine, 10% fetal bovine serum, antibiotics (100 U/ml penicillin G, 0.1 mg/ml streptomycin sulfate, 10 µg/mL gentamicin), 10 mM non-essential amino acids, and an anti-fungal agent (amphotericin-B 2.5 µg/mL). ARPE-19 cells are a homogeneous, retinal-derived cell line with functional and structural properties similar to human retinal pigment epithelial (RPE) cells (Dunn et al., 1996). All cells within the ARPE-19 culture express RPE-specific markers such as CRALBP, BEST1, and RPE-65 (Moustafa et al., 2017).

Cytotoxicity assay - lactate dehydrogenase assay (LDH assay)

L, [A+B], and D cybrids were plated in 96-well plates (10,000 cells/ well) for 24 h and treated with 0 or 40 µM of cisplatin for another 48 h. 50 µM of supernatant containing the released LDH from each corresponding well was transferred to a new 96-well plate, and 50 µM of reaction mixture was subsequently added and gently mixed. The plates were incubated at room temperature for 30 min, and the resulting reactions were stopped by adding 50 µM of stop solution to each well. Absorbance readings were taken at both 490 nm and 680 nm using an absorbance reader (BioTek, ELX 808, Winooski, VT). Percent cytotoxicity levels were calculated using the equation (Treated LDH level–spontaneous LDH level) divided by (Maximum LDH level—spontaneous LDH level) multiplied by one hundred. Average percent cytotoxicity levels were calculated for each ethnic population, normalized to the L haplogroup, and then compared using a two-tailed t-test. P ≤ 0.05 was deemed statistically significant. Cytotoxicity levels were analyzed, each treatment condition was run in quadruplicate and the entire experiment was repeated three separate times.

RNA isolation, quantitative-real time PCR (qRT-PCR)

L, [A+B], and D cybrids were plated in six-well plates (500,000 cells/well) and incubated for 24 h. Cybrid cells were then treated with culture media containing either 0 or 40 µM of cisplatin for an additional 48 h. Trypsinized cells were pelleted, and RNA was isolated according to the manufacturer’s protocol (RNeasy Kit, Qiagen, Valencia, CA). Following RNA quantification (Nanodrop 1000, Thermoscientific, Wilmington, DE), the cDNA was synthesized from 2 µg of RNA (QuantiTect Reverse Transcription Kit, Qiagen), and used for qRT-PCR (StepOnePlus instrument; Applied Biosystems, Carlsbad, CA). Proprietary, predesigned and validated SYBR Green-based primer assays were used (QuantiTect Primer Assays, Qiagen). Qiagen, the source for the primer assays does not release the sequence of their primers and an internal reference gene number (rs#) is not provided, however, the identity of each primer assay can be identified via the gene’s GenBank accession number (NM) (https://geneglobe.qiagen.com/product-groups/quantitect-primer-assays). Cancer-related nuclear genes were selected as they are typically the target for many chemotherapeutic treatments and presently have medications that target them. They were examined to understand how variations in mtDNA haplogroups affect retrograde signaling (mitochondria to nucleus) in cells that only vary in their mtDNA. The genes analyzed included ALK (Anaplastic Lymphoma Receptor Tyrosine Kinase, NM_004304), BRCA1 (DNA Repair Associated, NM_007294), EGFR (Epidermal Growth Factor Receptor 1, NM_005228), and ERBB2/HER2 (Erb-b2 Receptor Tyrosine Kinase 2, NM_004448) genes. Similar to the study done by Patel et al. target cycle thresholds (Ct) values were first compared to the Ct values of reference genes, and then, comparisons between untreated and cisplatin-treated values (ΔΔCt) were evaluated for statistical significance. Fold differences were quantified using the equation 2^(ΔΔCt), and the values for each sample were normalized to that of the untreated. This was done separately for each haplogroup in order to identify a starting base expression level and observe the gene expression fold difference after cisplatin treatment (Tables 3–4) (Patel et al., 2019).

Table 3:

L, [A+B], and D mtDNA haplogroup cybrids each displayed unique overall results in cell and gene expression studies.

Results from the cell and gene expression studies were congregated into one table. As illustrated, each haplogroup had a unique result when observed in large as a summation of the assays and gene expression results. As demonstrated from the cell and gene expression studies, L cybrids had decreased cellular metabolic activity and ALK gene expression after cisplatin treatment (p = 0.0117, p < 0.0001 respectively). [A+B] cybrids had decreased cellular metabolic activity, ALK gene expression, and EGFR gene expression after cisplatin treatment (p = 0.0285, p = 0.0001, p = 0.0246, respectively). D cybrids had decreased cellular metabolic activity after cisplatin treatment (p = 0.0001). ^∗p < 0.05; ^∗∗p < 0.01; ^∗∗∗p < 0.001.

DOI: 10.7717/peerj.9908/table-3

Statistical analysis

Statistical analysis of the data were performed by ANOVA (GraphPad Prism, version 5.0). Newman-Keuls multiple-comparison or the two-tailed t-tests were used to compare the data within each experiment. Results were normalized to that of the average untreated-L cybrid for each cell study. P ≤ 0.05 was considered statistically significant. Error bars representing standard deviation (SD) as well as standard error of the mean (SEM) have been graphed for each assay (Patel et al., 2019).

Creation of L, A, B, and D cybrids

Figure 1 represents the cybrid cell creation process resulting in cells with identical nuclei but differing mtDNA haplogroups. The mtDNA haplogroups for each subject and cybrid cell line were verified using polymerase chain reaction (PCR) along with restriction enzyme digestion and mtDNA sequencing as described previously in the study conducted by Patel et al. on H and J mtDNA haplogroup cybrid cells (L cybrids, n = 7; [A+B] cybrids, n = 4; and D cybrids, n = 3) (Patel et al., 2019).

Sequencing of mtDNA from L, A, B, and D cybrids

The entire mtDNA from the L, A, B, and D cybrids were sequenced using Next Generation Sequencing (NGS) technology. Table 1 summarizes the mtDNA haplogroup, age and gender for the subjects in this study. Figures 2–4 show individual mtDNA SNPs that define the haplogroup classification of each cybrid. The private SNPs are those that do not define the L, A, B, or D haplogroups (non-haplogroup defining). The unique SNPs are those SNPs not listed in http://www.mitomap.org or other programs (Figs. 2–4, Table 1) (Patel et al., 2019). Values in the cell and gene expression studies are averages and do not represent any particular SNP variant within one haplogroup. They are instead representative of the haplogroup-defining SNPs common among all cybrids within a haplogroup. SNP variants associated with the L, [A+B], and D mtDNA haplogroups as well as the location of the mtDNA nucleotide change, the base pairs changed, their associated rs number, and any associated pathogenesis found in the literature were identified and assessed. Non-synonymous changes present in some but not all haplogroups tested were found at m.4824 A > G, rs ND; m.5178 C > A, rs28357984; m.5442 T > C, rs3020601; m.7146 G > A, rs372136420; m.8701 G > A, rs2000975; m.8794 C > T , rs2298007; m.10398 G > A, rs2853826; and m.13105 G > A, rs2853501 (http://www.hmtvar.uniba.it and http://www.phylotree.org) (Table 2).

Relative cellular metabolic activity of untreated L, A+B, and D cybrids

The cellular metabolic activity measured with the MTT assay is representative of the cell viability within the cybrid cultures. The untreated cybrids, representing different maternal ethnic populations (African (L), Hispanic [A+B], or Asian (D) mtDNA haplogroups), had statistically similar relative cellular metabolic activities. All values were normalized to the average of the untreated-L cybrids and displayed as the mean ±  SEM (Fig. 5A). The distribution of each value was also presented in a dot-plot graph representing mean ±  SD (Fig. 5B). The untreated-L cybrids (102.4% ±  7.0% SEM; SD = 21.0%, n = 7) had similar relative cellular metabolic activity when compared to the untreated-D cybrids (111.7% ±4.3% SEM; SD = 11.3%, n = 3, p = 0.311) and the untreated-[A+B] cybrids (95.0% ±  8.5% SEM; SD = 20.9%, n = 4, p = 0.512). Untreated-[A+B] cybrids showed similar cellular metabolic activity compared to the untreated-D cybrids (p = 0.0933). Experiments were analyzed in quadruplicate, and the entire experiment was repeated twice.

Relative cellular metabolic activity of L, A+B, and D cybrids after cisplatin treatment

We found significantly reduced cellular metabolic activity after cisplatin treatment in the L, [A+B], and D cybrids compared to their respective untreated controls (Figs. 6A 6B). All values were normalized to the average of the untreated-L cybrids and displayed as the mean ±  SEM (Fig. 6A). The distribution of each value was also presented in a dot-plot graph representing mean ±  SD (Fig. 6B). The cisplatin-treated-L cybrids showed the greatest reduction in relative cellular metabolic activity compared to the untreated-L cybrids (75.0% ±  6.7% SEM; SD = 20.0%, n = 7; 102.4% ±7.0% SEM; SD = 21.0%, n = 7, respectively) (p = 0.0117). The cisplatin-treated-D cybrids showed a significant decrease in cell viability compared to the untreated-D cybrids (86.9% ±  1.6% SEM; SD = 4.3%, n = 3; 111.7% ±  4.3% SEM; SD = 11.3%, n = 3, respectively) (p = 0.0001). The cisplatin-treated-[A+B] cybrids also had a statistically significant decrease in cellular metabolic activity compared to their respective untreated-[A+B] cybrids (70.3% ±  4.5% SEM; SD = 11.1%, n = 4; 95.0% ±  8.5% SEM; SD = 20.9%, n = 4, respectively) (p = 0.0285) (Figs. 6A and 6B). Each treatment was run in quadruplicate and the entire experiment was repeated twice.

Mitochondrial membrane potential (ΔΨm) after cisplatin treatment

Mitochondrial membrane potentials for untreated and treated cybrid cells were measured. All values were normalized to the average of the untreated-L cybrids and displayed as the mean ±  SEM (Fig. 7A). The distribution of each value was also presented in a dot-plot graph representing mean ±  SD (Fig. 7B). Results indicated that there was no significant difference in mitochondrial membrane potential in the cisplatin-treated cybrids compared to the untreated cybrids. Cisplatin-treated-L cybrids had a relative mitochondrial membrane potential of 108.4% ±  8.2% SEM; SD = 21.8% versus untreated-L cybrids (101.0% ±  11.2% SEM; SD = 29.5%, n = 7, p = 0.602). Cisplatin-treated-[A+B] displayed 136.5% ±  33.9% SEM; SD = 67.8% versus untreated-[A+B] cybrids (147.5% ±  42.0% SEM; SD = 84.0%, n = 4, p = 0.845). The cisplatin-treated-D cybrids were observed to have a relative mitochondrial membrane potential of 115.0% ±  3.5% SEM, SD = 7.0% versus the untreated-D cybrids (127.3% ±  20.1% SEM; SD = 40.3%, n = 3, p = 0.571) (Figs. 7A and 7B). Each treatment was run in quadruplicate and the entire experiment was repeated twice.

Mitochondrial membrane potential (ΔΨm) assay (IncuCyte^® live cell analyzer with ARPE-19 cells)

After cisplatin treatment, the cybrids showed decreased metabolic activity but the mitochondrial membrane potentials remained unchanged, which was unexpected. Therefore, we conducted additional experiments using the ARPE-19 cell line that allowed for comparison of growth/proliferation over 40 h in cultures treated with 20 µM or 40 µM cisplatin (n = 7 per treatment condition). All values were displayed as the mean ±  SD. Over the first 8 h of culture, the untreated and cisplatin-treated ARPE-19 cells showed similar rates of growth (Fig. 8B). By 10 h the untreated ARPE-19 cells showed an increasing slope of growth compared to the 20 µM and 40 µM cultures that were stagnant in their growth. By 40 h, the untreated had a 2-fold higher number of cells than cisplatin-treated cultures. (Untreated 1969 ±  115 standard deviation (SD)), versus 20 µM (1023 ±  81 SD, p = 0.0006) and 40 µM (945 ±  74 SD, p = 0.0001) (At 40 h, the ΔΨm levels were similar between the untreated AREP-19 (1.8993 ±  0.145 SD) versus 20 µM (2.0698 ±  0.161 SD, p = NS) and 40 µM (2.0520 ±  0.162 SD, p = NS) cisplatin-treated ARPE-19 cultures (Fig. 8A).

Lactate dehydrogenase cytotoxicity after cisplatin treatment

The cytotoxicity levels of each of the cybrids were measured with the lactate dehydrogenase (LDH) assay. All values were normalized to the average of the untreated-L cybrids and displayed as the mean ±  SEM (Fig. 9A). The distribution of each value was also presented in a dot-plot graph representing mean ±  SD (Fig. 9B). When normalized to the L cybrids, there were significantly higher LDH cytotoxicity levels in L cybrids compared to [A+B] cybrids (111.5% ±  16.5% SEM; SD = 40.5%, n = 7 versus 41.8% ±  9.8% SEM; SD = 16.9%, n = 4, respectively) (p = 0.0270). The D cybrids trended lower LDH levels but they were not significant (48.9% ±  20.2% SEM; SD = 35.0% , n = 3, p = 0.0576) compared to the L cybrids. LDH cytotoxicity levels between the [A+B] and D cybrids were not statistically different (p = 0.766) (Figs. 9A and 9B). The entire experiment was repeated three separate times.

Gene expression levels in L, [A+B], and D cybrids after cisplatin treatment

ALK (Anaplastic Lymphoma Kinase): Transcription levels for L cybrids and the combined [A+B] cybrids dropped significantly after cisplatin treatment (0.18 fold ±  0.097, n = 7, p < 0.0001; 0.17 fold ±  0.095, n = 4, p = 0.0001, respectively), while the D cybrids had no significant change in ALK gene expression (n = 3, p = 0.290) (Tables 3–4).

BRCA1 (Breast Cancer 1): Gene expression levels for all cybrids (L, [A+B], and D) did not significantly change after treatment with cisplatin compared to untreated-control cybrids (n = 7, n = 4, n = 3, p = 0.356, p = 0.294, p = 0.158, respectively) (Tables 3–4).

EGFR (Epidermal Growth Factor Receptor 1): After cisplatin treatment, cybrids containing L and D mtDNA haplogroups had no significant change in EGFR gene expression compared to their respective untreated-controls (n = 7, p = 0.301 and n = 3, p = 0.218, respectively). However, cisplatin-treated cybrids containing [A+B] mtDNA haplogroups significant decreased in EGFR gene expression levels when compared to untreated-[A+B] cybrids (0.50 fold ±0.17, n = 4, p = 0.0246) (Tables 3–4).

ERBB2/HER2 (Erb-b2 Receptor Tyrosine Kinase 2): Similar to the BRCA1 gene expression levels, ERBB2/HER2 gene expression levels were not significantly different after cisplatin treatment in any of the cybrid cell lines (L, [A+B], and D) tested (n = 7, n = 4, n = 3; p = 0.839, p = 0.379, p = 0.325, respectively) (Tables 3–4).

Sequencing of the entire mtDNA

The NGS technology used to analyze the mtDNA from each of the L, A, B, and D cybrids showed that the majority of SNPs identified were haplogroup defining. The private SNPs (non-haplogroup defining), unique SNPs (not listed in http://www.mitomap.org), and heteroplasmy SNPs were found in individual cybrids and not throughout all of the L, A, B, or D cybrids. Similar to the study comparing cybrids with H and J mtDNA haplogroups, this suggests that the differential retrograde signaling between the L, A, B, and D mtDNA haplogroups is due to the accumulation of the haplogroup defining SNPs rather than a single mutation or private SNP (Patel et al., 2019). These data are also consistent with another cybrid study which used allelic discrimination and Sanger sequencing to identify the mtDNA haplogroups (Atilano et al., 2015). NGS technology was used for deep sequencing of the mtDNA (ranging from 1,000 to 100,000, with an average depth of 30,000), which allowed for low-level heteroplasmy to be reliably identified (Patel et al., 2019). This method helps distinguish artifact from low-level heteroplasmy because both strands of mtDNA are independently sequenced in both directions. The role of mtDNA haplogroups in retrograde signaling is still under investigation, and many pathways may still be unidentified.

Haplogroup-defining SNPs were analyzed and the nucleotide changes between the L, [A+B], and D cybrids were identified (Table 2). Non-synonymous changes (causing a change in amino acid) present in some but not all haplogroups tested were found at eight locations. Related pathogenesis to these SNP variants have been associated with diseases such as asthenospermia, atherosclerosis, breast cancer, cardiomyopathy, esophageal cancer, fertilization failure, Fuch’s endothelial corneal dystrophy (FECD), glaucoma, diabetes, hypertension, infertility, irritable bowel syndrome, myocardial infarction, neuropathy, osteoarthritis, osteosarcoma, Parkinson’s, mobility impairment, neuromuscular disorders, schizophrenia, somatic lung cancer, and thyroid cancer (http://www.hmtvar.uniba.it and http://www.phylotree.org) (Table 2). One could speculate that these SNP variants may play a role in the differential susceptibility to cisplatin in L, [A+B], and D cybrids because the cybrids have identical nuclei but vary in their mtDNA. Additional studies are needed to better understand the relationship between mtDNA haplogroups and drug susceptibility-related cellular pathways.

Variations within mtDNA in L, A+B and D cybrids are associated with a differential decline in relative cellular metabolic activity after treatment with cisplatin

After 48 h, the untreated L (African maternal descent), [A+B] (Hispanic maternal descent), and D (Asian mitochondrial descent) cybrids had similar levels of cellular metabolic activities.

However, in response to cisplatin treatment, there was a differential decline in relative cellular metabolic activity (L >D >[A+B] cybrids) compared to their respective untreated cultures with the [A+B] cybrids showing the smallest decrease in metabolic activity. One can speculate that if a person with mtDNA containing a more resistant SNP variant, such as haplogroup [A+B], was treated with identical doses of cisplatin compared to an individual with haplogroup L mitochondria, the first individual might experience signs of resistance to the drug, while the latter would experience a significant drop in cellular metabolic activity in the targeted cells. These data suggest that an individual’s mtDNA haplogroup profile may play a role in their response to cisplatin treatment, thereby affecting its efficiency and severity when used in chemotherapy. Understanding how pharmacogenomics play a role in patient treatment, particularly, how the patient’s mtDNA SNP profile is associated with variations in drug susceptibility, would allow future treatment programs to focus more on an individual’s genetic background in order to achieve more effective results (Van Gisbergen et al., 2015). Further studies will be required to better understand the relationship between mtDNA SNP variants and differences in drug susceptibility.

Cisplatin treatment induced apoptosis without significantly disrupting mitochondrial membrane potential

Data from the JC-1 assay indicated that when treated with cisplatin, mitochondrial membrane potential in all cybrids was maintained. Surprisingly, L, [A+B], and D cybrids did not show any significant change in mitochondrial membrane potential after cisplatin treatment. OXPHOS related ATP production is dependent on mitochondrial membrane potential, and a drastic decrease or increase in membrane potential in the intermembrane space results in less efficient ATP production via ATP synthase. Cisplatin is known to disrupt mitochondrial membrane potential by inducing the synthesis of pro-apoptotic proteins such as Bcl-2 family proteins, Bax and Bak, which form porous defects within the outer membrane of the mitochondria and result in the release of apoptotic factors such as Cytochrome c into the cytosol, ultimately resulting in cell death (Pabla et al., 2008). However, a failure of the mitochondrial membrane potential to decrease in the case of cybrids containing L, [A+B], or D mtDNA haplogroups, suggests that a decrease in cell viability leading to apoptosis was induced by a mechanism that did not disrupt the proton gradient within the intermembrane space of the mitochondria. In a previous study, it was observed that somatic mtDNA mutations affect apoptosis without affecting ROS levels or OXPHOS (Trifunovic et al., 2005). Additionally, a study on renal cells demonstrated that neither mitochondrial membrane potential nor OXPHOS related ATP production significantly changed at any time during cisplatin exposure (Cummings & Schnellmann, 2002).

The ARPE-19 experiments were conducted to verify the cybrid results obtained from the JC-1 mitochondrial membrane potential assays. The ARPE-19 experiments showed the 20 µM and 40 µM cisplatin prevented cell proliferation, indicating growth inhibition and decreased cell health but did not cause elimination of the cells. Surprisingly the levels for mitochondrial membrane potential (ΔΨm) were similar in the untreated and cisplatin-treated cultures. This assay relies on the ratio of red (healthy) and green (dead cells) to indicate change in ΔΨm. The red fluorescence is proportional to the potential but diminished in apoptotic or necrotic cells. In ARPE-19 cultures, the cisplatin caused cells to stain more brightly with both red and green intensity, so the final ratio was unchanged (Fig. 8A). This suggests that the cisplatin is not lowering the mitochondrial membrane potential, which agrees with Kleih and colleagues who observed that ovarian cancer cells treated with cisplatin also maintained their mitochondrial membrane potential (Cocetta, Ragazzi & Montopoli, 2019; Kleih et al., 2019). Similarly, English et al. found that cisplatin-treated neuronal cells could maintain and restore their mitochondrial membrane potential through the uptake of cisplatin-resistant mitochondria from astrocytes (English et al., 2020).

An alternative pathway leading to apoptosis as a result of cisplatin treatment may be induced by endoplasmic reticulum (ER) stress. A previous study conducted by Gisbergen et al. demonstrated that ER stress-related apoptosis was induced as a result of cisplatin treatment (Van Gisbergen et al., 2015). In that study, cisplatin treatment caused ER stress, which activated Caspase-12 cleavage and consequently resulted in apoptosis. Additionally, in a study conducted by Cumming et al. another ER-associated protein leading to apoptosis was identified. Inhibition of Ca²⁺-independent phospholipase A2 (ER-iPLA2) resulted in an increase of cisplatin-induced apoptosis in primary rabbit proximal tubular cultures (Cummings, McHowat & Schnellmann, 2004). In a previous study conducted on ARPE cybrid cells containing either H or J mtDNA haplogroups, J cybrids showed a significant decline in mitochondrial membrane potential after cisplatin treatment, while H cybrids did not (Patel et al., 2019). A decrease in mitochondrial membrane potential would suggest a decrease in OXPHOS related ATP production and an increase in glycolytic ATP production. As indicated by Stewart et al. (2011) ρ⁰ cells depleted of mitochondria via Ethidium Bromide (EtBr), rely completely on glycolysis for energy production, indicating that OXPHOS inhibition is inversely related to glycolytic ATP production. From the study conducted on H and J cybrids, along with the current study, one can speculate that the unique composition of an individual’s mtDNA plays a role in their susceptibility to cisplatin treatment. Additional studies are needed to better understand the relationship between mtDNA variants and drug susceptibility.

Cybrids with dissimilar mtDNA haplogroups vary in LDH cytotoxicity after cisplatin treatment

Lactate Dehydrogenase cytotoxicity levels measured by the LDH assay illustrated varying results for each of the groups examined. The L cybrids exhibited the highest normalized LDH cytotoxicity levels, which were statistically greater than the [A+B] but not the D cybrids (p = 0.0270 and p = 0.0576, respectively). Additionally, statistical analysis between the [A+B] cybrids and the D demonstrated that the two groups did not have statistically different LDH cytotoxicity levels (p = 0.766) (Figs. 9A and 9B). Cisplatin-induced cytotoxicity is primarily by the mediation of covalent purine adducts within the nuclear (n) DNA that result in disruption of transcription and replication, consequently leading to cell death (Dasari & Tchounwou, 2014). One can speculate that nDNA adduct formation leading to extracellular cytotoxicity is less prevalent in cybrids containing [A+B] mtDNA haplogroups compared to cybrids containing L mtDNA haplogroups. In a previous study conducted by Kohno et al. on cancer cells, it was determined that the prevalence of guanine base sequences within the mtDNA was a determining factor for the ability of cisplatin to bind effectively and form adducts within the DNA, therefore supporting the idea that cisplatin-induced cytotoxicity depended on the unique profile of the individual’s mtDNA (Kohno et al., 2015). Variations in cisplatin-induced cytotoxicity between cybrids with different mtDNA may suggest that mtDNA SNP variants play a key role in the extent to which adducts form. Additional studies are needed to better understand the relationship between mtDNA haplogroups and DNA adduct formation.

Variations within mtDNA are associated with modulation of retrograde signaling, resulting in varying nuclear gene expression levels