An Inherited Heteroplasmic Mutation in Mitochondrial Gene COI in a Patient with Prostate Cancer Alters Reactive Oxygen, Reactive Nitrogen and Proliferation

Mitochondrial DNA (mtDNA) mutations have been found in many cancers but the physiological derangements caused by such mutations have remained elusive. Prostate cancer is associated with both inherited and somatic mutations in the cytochrome c oxidase (COI) gene. We present a prostate cancer patient-derived rare heteroplasmic mutation of this gene, part of mitochondrial respiratory complex IV. Functional studies indicate that this mutation leads to the simultaneous decrease in cytochrome oxidation, increase in reactive oxygen, and increased reactive nitrogen. These data suggest that mitochondrial DNA mutations resulting in increased reactive oxygen and reactive nitrogen generation may be involved in prostate cancer biology.


Introduction
e mitochondrion contains the only functional DNA outside the eukaryotic cell nucleus and mutations in this genome have been linked to a vast array of human disease including pediatric neurologic disease, degenerative muscular disease, and blindness in the pediatric population and more recently to chronic diseases of adults including diabetes, Alzheimer's dementia, Parkinson's disease, cardiovascular disease, and cancer [1]. Because of the critical role the mitochondria play in normal human cell physiology, severe mutations that cause dramatic derangements of mitochondrial function can be fatal [2]. Indeed, recent evidence indicates that severe mutations are culled from the germ line through depletion of mutant germ cells in the ovary [3]. Severe mutations can exist and even be passed between generations if they exist in a state of heteroplasmy where some copies of the genome contain the mutation and some are wild type. For this reason, one indication that a mitochondrial DNA (mtDNA) mutation is potentially pathogenic is that it exists in a heteroplasmic state, though this is not strictly necessary in all cases.
Prostate cancer is one example of an adult disease tightly linked to ageing for which there is evidence linking both inherited and somatic mtDNA mutations to disease [4]. Mutations in the mitochondrial cytochrome c oxidase subunit I (COI) gene are found at disproportionately high rates in prostate cancer patients compared to either the general population or rigorously de�ned controls without prostate cancer [5]. A population-based case-control study of African Americans found 2 single nucleotide polymorphisms in this BioMed Research International gene signi�cantly associated with prostate cancer ( ) and in strong linkage disequilibrium with each other ( 6) [6]. MtDNA mutations are easily detected in early stage prostate cancer tissues and can also be found in the serum and urine from the same individuals [7]. MtDNA content has been linked to androgen responsiveness of prostate cancer cells in vitro [8] and analysis of clinical samples suggests that mtDNA content is higher in prostate cancer compared to adjacent normal prostatic tissue in the same individual [9]. A speci�c mtDNA deletion mutation may have utility in identifying adult cancer not apparent in clinical prostate biopsies [10].
While multiple independent investigators have con-�rmed mtDNA mutations in prostate cancer, there is little understanding of the cell biologic and biochemical consequences of speci�c prostate cancer-associated mtDNA mutations. We investigated the mtDNA from a patient with prostate cancer and found a heteroplasmic missense mutation in the mitochondrial COI gene that impairs the oxidation of cytochrome c (respiratory complex IV inhibition) and increases the generation of both reactive oxygen species (ROS) and reactive nitrogen species (RNS).

Cytochrome c Oxidase Subunit I (COI) Gene Sequencing.
e mtDNA region encompassing COI was ampli�ed using a forward primer starting at nucleotide position (np) 5772 (5 � AGG TTT GAA GCT GCT TCT TC 3 � ) and a reverse primer ending at np 7600 (5 � CGC TGC ATG TGC CAT TAA GA 3 � ). e template was denatured at 95 ∘ C for 7 min and primers extended for 35 cycles of 94 ∘ C for 1 min, then 55 ∘ C for 1 min, and 72 ∘ C for 1 min. Both strands of the COI polymerase chain reaction (PCR) product were cyclesequenced using the slip primers in the forward direction starting at np 6080 (5 � TCT ACA ACG TTA TCG TCA CA 3 � ) and at np 6930 (5 � TGC AGT GCT CTG AGC CCT AG 3 � ) and in the reverse direction starting at np 6340 (5 � CTA GGT GTA AGG AGA AGA TG 3 � ) and at np 7150 (5 � GAT TTA CGC CGA TGA ATA TG 3 � ). e templates were denatured at 96 ∘ C and primers extended in the presence of "Big Dye Terminators" for 25 cycles of 96 ∘ C for 10 sec, then 55 ∘ C for 5 sec, and 60 ∘ C for 4 min. e reactions were chilled to 4 ∘ C, and the excess dye terminators removed by Centri-Sep Columns. e trace �les were determined using an Applied Biosystems (ABI) Prism 3100 genetic analyzer, analyzed using Sequencher gene analysis soware v 4.1 (Gene Codes, Ann Arbor, MI), and interpreted within the context of MITOMAP (http://www.mitomap.org/). Full mtDNA sequencing was performed as described above using primer sets that span the full length of the mitochondrial DNA [11].

Epstein-Barr Transformation.
Lymphocytes were isolated from whole blood by centrifugation and diluted with phosphate buffered saline (PBS). Red cells were lysed by the addition of H 2 O. Aer 20 s, osmolarity was restored with 10x concentrated PIPES (piperazine-N, � -bis-[2-ethanosulfonic-acid]) buffer, centrifuged, and the pellet was resuspended in RPMI with fetal bovine serum (FBS) and incubated for 45 min in 5% CO 2 at 37 ∘ C. e nonadherent cells were washed and collected and lymphocytes pelleted. Lymphocytes were infected with the B95-8 strain of Epstein-Barr Virus (EBV) and cultured in RPMI-1640 medium supplemented with 10% heat-inactivated fetal bovine serum, 2 mM l-glutamine, 100 U/mL penicillin, and 100 mg/mL streptomycin, at 37 ∘ C in a humid atmosphere saturated with 5% CO 2 until cryopreservation or use [12].

Mitochondrial Preparation and Isolation.
Mitochondria were isolated by cell fractionation and centrifugation. All procedures were carried out on ice. Cells were pelleted and washed in cold PBS and resuspended in 10 volumes of isolation buffer (250 mM sucrose, 10 mM HEPES, 1 mM EDTA, pH 7.35). Cell lysis was performed with 5 passes of a hand held Dounce homogenizer and centrifuged at 500 g (4 ∘ C) for 5 min. e supernatant was reserved and the pellet was resuspended in isolation buffer, rehomogenized, and centrifuged. is process was repeated a third time and supernatants combined and centrifuged 12,000 g at 4 ∘ C for 10 min to pellet mitochondria which were then washed in isolation buffer and repelleted three times. Following the �nal wash, mitochondria were resuspended in reaction buffer (1 mL/mL of original cell pellet), and frozen until assayed.

Cytochrome c Oxidase
Activity. Cytochrome c oxidase activity of isolated mitochondrial preparations was measured by dual wavelength spectrophotometry (550/540 nm) as previously described [13]. Brie�y, mitochondria were suspended in reaction buffer (250 mM Sucrose, 10 mM HEPES, 1 mM MgCl 2 , pH 7.35). An aliquot of mitochondria suspension (∼300 ug mitochondrial protein) was added to the cuvette containing reaction buffer plus ∼5 uM reduced cytochrome c in a �nal volume of 700 uL. e oxidation of cytochrome c was recorded over time and the cytochrome c oxidase speci�c activity was calculated. Mitochondrial protein was determined by the Bradford method [14].

Cybrid Formation.
Mitochondrial donors (patient lymphoblasts) were enucleated by short term incubation with Actinomycin D, which was subsequently removed from the culture medium by centrifugation and washing. ese cytoplasts were then rescued by polyethylene glycol (PEG)induced fusion with 143B rho zero cells. Fusions are monitored by phase contrast microscopy and isolated by ring cloning. Clones were expanded and genotyped to assure that the donor mtDNA had been incorporated (sequencing) and that a single nucleus was present (Karyotyping).

Flow Cytometric Assay for ROS and RNS. Peroxides
Assay: Lymphoblasts were grown in suspension in �asks. When cells were in growth phase, 1 × 10 6 cells per sample were removed, pelleted, and resuspended in 2 M CM-DCFDA (5-(and -6)-chloromethyl-2 � ,7 � -dichlorodihydro-�uorescein diacetate) (Life Technologies, Grand Island, N�). Cells were placed in a dark room at room temperature while gently rocking for 30 minutes (CM-DCFDA). Following this incubation, samples were kept on ice until counting on a Becton Dickinson FACScan �ow cytometer. 10,000 cells were counted and analyzed by FlowJo 7.6.4 to compare mean values of DCF �uorescence intensity. All samples were repeated in triplicate. Some sample tubes also contained 10 M FCCP (Mesoxalonitrile 4-tri�uoromethoxyphenylhydrazone) and were preincubated with the lymphoblasts prior to the addition of CM-DCFDA. Cybrid cell lines were also treated with CM-DCFDA. Brie�y, adherent cells were grown to 70% con�uency, washed with PBS, and cells were removed from the plate by the addition of trypsin-EDTA (Mediatech, Inc, Manassas, VA). Cells were pelleted at 400 g for 5 minutes at room temperature and the cells were resuspended in CM-DCFDA and assay continued as described above. Mitochondrial Superoxide Assay: Media (RPMI 1640 containing 10% fetal bovine serum) was replaced on cells in growth phase with media containing 5 M MitoSOX (Life Technologies). Cells were incubated in the dark at 37 ∘ C, 5% CO 2 for 10 minutes. Following this incubation, cells were harvested by trypsin digestion, pelleted, and resuspended in HANKS Buffer containing 10% FBS. Samples were kept on ice until counting on a Becton Dickinson LSRII �ow cytometer as described above. NO Assay: Media (RPMI 1640 containing 10% fetal bovine serum) was replaced on cells in growth phase with media containing 4 M 4-amino-5-methylamino-2 � ,7 � -di�uoro�uorescein diacetate (DAF-FM DA) (Life Technologies). Cells were incubated in the dark at 37 ∘ C, 5% CO 2 for 30 minutes, DAF-FM DA was removed and replaced with media and incubated in the dark at 37 ∘ C, 5% CO 2 for a additional 20 minutes. Following this incubation, cells were harvested by trypsin digestion, pelleted, and resuspended in HANKS Buffer containing 10% FBS. Samples were kept on ice until counting on a Becton Dickinson LSRII �ow cytometer as described above. Time between dye loading and reading was 30 minutes and all samples were treated identically. e DAF experiment was done in the presence of serum.

Cell Proliferation Assay.
Cell were plated in 6 well plates, 18,000 cells per well in triplicate for each time point and grown in RPMI 1640 (Mediatech) containing 15% Fetal Bovine Serum. Cells were harvested aer 1, 2, 3, or 4 days and cell number was determined using FluoReporter Blue Fluorometric dsDNA Quantitation Kit (Life Technologies) according to manufacturer's protocol.

Reverse Transcription and PCR.
In order to determine which isoform of Nitric Oxide Synthase (NOS) is present in our cybrid cell lines we performed RT-RCR on RNA from the cybrid cell lies and a positive control. cDNA was obtained from the RNA of 6124WT, 6124Mut, and BT474 (positive control) by reverse transcription using Advantage RT for PCR from Clontech (a Takara Bio Company, Mountain View, CA) according to protocol. e presence of NOS1, NOS2, and NOS3 were examined. Using the Amplitaq Gold Kit (Life Technologies), reactions were as follows: cDNA, 1XBuffer, 1.5 mM MgCl 2 , 0.2 mM each dNTPs, 150 nM each forward and reverse primers for NOS1 (Forward 5 � CGA CAC CAC TAG CAC TTA CCA G 3 � ; Reverse 5 � CAG ACT CGG AAG TCG TGC TTG 3 � ), NOS2 (Forward 5 � TCG GCT GCA GAA TCC TTC ATG A 3 � ; Reverse 5 � CAT TGT CTT GCG CAT CAG CAT AC 3 � ) or NOS3 (Forward 5 � GAA GCA CCT GGA GAA TGA GCA G 3 � ; Reverse 5 � CTT CAC TCG CTT CGC CAT CAC 3 � ) and 5 units of Taq for a �nal volume of 100 uL. Primers were designed to cross two introns and amplify all major isoforms of NOS1. e PCR reaction included an initial cycle of 95 ∘ C for 10 min followed by 40 cycles of 95 ∘ C for 30 sec and 61 ∘ C for 30 sec, 72 ∘ C for 1 min. A 3% agarose gel was run aer PCR. e presence of NOS1, NOS2, and NOS3 were indicated by the presence of a 261 nt band, 288 nt or a 274 nt band, respectively.

Mice.
Male nu/nu mice, 6-8-weeks old, were purchased from Charles River Laboratories (Wilmington, MA) and housed in ventilated cages under sterile conditions. For surgical manipulation, mice were anesthetized with an intramuscular injection of a mixture of ketamine hydrochloride, Xylazine, and Acepromazine.

In Vivo Tumor Study.
Mice were injected subcutaneously in the neck with 3 separate 6124WT clones for a total of 29 mice and 3 separate 6124Mut clones for a total of 30 mice. 2.5 × 10 5 cells, resuspended in PBS, were injected per mouse. Mice were checked daily for tumor growth. When tumors were visually observed they were measured twice a week and the tumor volumes calculated using the formula [length × (width) 2 ]/2 [15]. Experiment was repeated 2 additional times. Mice were sacri�ced when the tumors reached 10% of body weight. Tumor tissues were dissected for further study.

Mutation Analysis.
A 60-year old patient underwent radical prostatectomy aer presenting to his primary care physician with an elevated-serum-prostate-speci�c antigen (PSA = 5.5 ng/mL) that prompted a prostate biopsy revealing a Gleason 6 prostatic adenocarcinoma. Histopathologic examination of the radical prostatectomy specimen revealed a Gleason's score 6 (moderately differentiated-grade G2) conventional (acinar) prostatic adenocarcinoma involving ∼5% of the le lobe of the prostate with negative surgical margins (pathologic stage pT2a). Follow-up of over 7 years with history, physical examinations and serum PSA determinations found the patient to be without evidence of prostate cancer recurrence. ere was no family history of prostate cancer and the patient was otherwise healthy. DNA from the prostate tissue was the �rst specimen to be sequenced from this patient and this demonstrated a heteroplasmic point mutation at nucleotide position (np) 6124 of the mitochondrial genome. e sequencing chromatogram demonstrated approximately equal proportions of the wild type base (T) and mutant base (C) (Figure 1(a)). e mutation changes amino acid 74 from the nonpolar, methionine (sulfur side chain) to the polar, threonine (hydroxyl side chain). is is a highly conserved amino acid, being methionine in 95% of all species ( ) for which the sequence has been determined. is conservation index (CI) of 95 is at the very upper end of what is observed for "adaptive mutations" (CI = 85 ± 9%) and signi�cantly above the CI of "neutral polymorphisms" (CI = 23 ± 5%), making it highly likely to be functional in terms of mitochondrial physiology [1]. Another indicator of the uniqueness of this base and amino acid change is that this patient remains the only example of this base change in MITOMAP and mutation at this nucleotide position has not been found in the 2704 complete mtDNA sequences reported in the online database mtDB [16], where 2704 out of 2704 individuals have the wild type base (T). Subsequent sequencing of patient peripheral blood DNA, lymphoblast cell lines, and laser capture microdissection of various cellular compartments of the prostate all revealed similar levels of BioMed Research International 5 heteroplasmic mutation at this base, con�rming inheritance in the germ line and maintenance of the mutation in the prostate. e entire mitochondrial DNA sequence for this patient was determined and all changes from the revised Cambridge reference sequence are shown in Table 1. e mutation at n.p. 6124 was the only heteroplasmic mutation and the only mutation that had not been previously found in the 2407 complete mtDNA sequences published in the online database mtDB.

Cytochrome Oxidase Enzyme Activity.
In order to study the biochemical consequences of this mutation we studied reactive oxygen production and enzymatic activity of respiratory complex (RC) IV (cytochrome oxidoreductase). e COI polypeptide forms the catalytic core of this enzyme. Cytochrome oxidation was measured in mitochondria isolated from the patient's lymphoblast cell line and compared to other lymphoblast cell lines from two unrelated, individuals' lymphoblast lines, sequence proven to be wild type at the COI locus (Figure 1(b)). e mutation is associated with a 29% decrease in COI activity.

Reactive Oxygen.
Mutations in the COI polypeptide that lead to decreased RCIV activity could potentially lead to increased reactive oxygen [17]. In order to determine if this was occurring, we assayed the lymphoblast cell line of the patient for increased peroxide levels using dichloro�uorescin (DCF) and compared the results to the two-wild type patients described above. ere was a marked increase (1.75 fold) in DCF �uorescence in the lymphoblast cell line containing the mutation compared to wildtype (Figure 1(c), gray bars). e ma�ority of the increase in DCF �uorescence can be attributed to the mitochondria as demonstrated when oxidative phosphorylation is inhibited with FCCP (white bars), DCF �uorescence levels are decreased to similar levels. More detailed analysis of the effect of the 6124 mutation on ROS generation was obtained in the cybrid experiments, results highlighted below ("Cybrid ROS and RNS").

Cybrid Generation.
In order to eliminate the potential confounding effects of the nuclear genome we made cytoplasmic hybrids (cybrids) that combined either pure mutant or pure wild type genomes from this patient with a stable nuclear background (143 B cells). e resultant pair of cybrids thus have the exact same nucleus (different from the patient), and the exact same mitochondrial DNA sequence except for the single base mutation at n.p. 6124. MtDNA genotypes were sequence veri�ed in cybrids.

Cybrid
In Vitro Proliferation. ree cybrids 6124 wild type clones and three 6124 mutant clones were analyzed for proliferation in culture. All three mutant 6124 cybrid clones grew faster than the three wild type clones with an average doubling time of 1.57 ± 0.12 days compared to 2.81 ± 0.56 days, mutant to wild type, respectively, ( 0.0 8) ( Figure  2). T 1: Patient mitochondrial DNA mutations. Patient peripheral blood, lymphoblast, and cybrid mtDNA was sequenced in its entirety. All changes from rCRS are shown as is the region in which the change occurs and the amino acid alteration when applicable. AI (allelic index) is the measure of the frequency of the mutation when compared to mtDB-Human Mitochondrial Genome Database [16]. Cybrid mtDNA was identical to peripheral blood with the exception of n.p. 6124 at which point the cybrids were determined to be either homoplasmic mutant or wild type.

Cybrid ROS and RNS.
We then compared multiple wild type and mutant clones for reactive oxygen species (ROS) and reactive nitrogen species generation using �ow cytometry and ROS and RNS sensitive �uorescent dyes. In order to study the cellular peroxide levels in cells that were either entirely mutant or entirely wild type at this base, cybrid cell lines were made and multiple clones assayed by DCF �uorescence. Overall, 6 individual mutant clones showed a consistent increase in peroxide levels when compared to 5 individual wildtype clones (data not shown). When averaging the relative level of �uorescence in all wildtype clones compared to mutant clones, the mutation is associated with a statistically signi�cant ( ) increase in DCF �uorescence (Figure 3(a)). Similarly, in order to determine mitochondrial superoxide levels, there was a slight overall decrease in mitochondrial superoxide in the 6124Mut cell lines when compared to 6124WT ( ) (Figure 3(b)). A substantial and signi�cant increase was observed in nitric oxide (NO) levels in the same 6124Mut and 6124WT cell lines (Figure 3(c)). When treated with DAF-FM diacetate, 6124Mut cell lines showed an average of 5.2-fold increase in NO levels compared to 6124WT ( ). For comparison, NO was also measured cybrids containing 8993WT or Mut. 8993Mut cells had 1.8-fold higher NO when compared to 8993WT ( ). Finally, there were no differences in the overall levels of hydroxyl radicals and peroxynitrite anions as measured by hydroxyphenyl�uorescein (HPF) (Figure 3(d)).

Nitric Oxide Synthases (NOS) in Cybrids Cell
Lines. In order to determine the possible source of the increased NO, we harvested RNA from the 6124WT and 6124Mut cybrids cell lines, followed by reverse transcription and PCR. ere were no detectable levels of NOS1 or NOS3 observed (data not shown). However, NOS2 was demonstrated to be in both WT and Mut cell lines (Figure 4). e breast cancer cell line BT474 is used as a robustly positive control for NOS2 RNA.
3.8. Apoptosis. PAPR (poly ADP ribose polymerase) cleavage was analyzed to determine if changes in cellular proliferation were in part due to changes in apoptosis. Western blot analysis of PARP cleavage demonstrates that PARP cleavage was reduced in 6124Mut cybrids compared to 6124WT ( Figure 5).
3.9. In Vivo Proliferation. 6124Mut cybrid cell lines grew faster in vivo when compared to 6124WT cybrids cells ( Figure  6). ere was some variability within each group as to the number and size of tumor, but at all time points there was a statistically signi�cant increase in the size of tumors from 6124Mut cybrids cell lines ( at day 32). e animals injected with cybrids showed no systemic symptoms in response to tumor injection or growth.

Discussion
e mitochondrially encoded COI gene was �rst implicated in cancer biology in 1998 when a somatically acquired chain termination mutation was reported in colon cancer by the Vogelstein group at Johns Hopkins [18]. In 2005 an analysis of 260 patients with prostate cancer revealed that 31 (12%) had an inherited mutation in COI compared to <2% of no-cancer controls. Similarly, a case-control analysis of African-American men revealed that two COI gene single nucleotide polymorphisms (SNPs) (T6221C and T7389C) were signi�cantly associated with prostate cancer ( < ) and in strong linkage disequilibrium with each other ( 2 > 6). It is therefore likely that COI gene mutations predispose individuals to the development of prostate cancer. For this reason we have sequenced the COI gene from 482 prostate cancer patients and identi�ed missense mutations in 116 (24.1%) [19]. e patient reported in this paper is part of that cohort.
e biochemical analysis of mtDNA mutations requires that viable cells are obtained from patients with such mutations and that potentially confounding nuclear events are controlled. is is made possible by the combination of capturing clinically relevant mutations in patient lymphoblast cell lines and the subsequent formation of cybrids with a common nuclear background. e cybrid formation process has the further advantage of allowing the mutant base to be studied separately from the wild type base. is paper documents that a prostate cancer-associated COI mutation affects the normal functioning of respiratory complex (RC) IV (cytochrome oxidase) in at least two distinct ways: decreasing the rate of mitochondrial cytochrome c oxidation and increasing the rate of ROS and NO generation in intact cells. It therefore seems likely that both of these effects are a direct consequence of the mutation. As Figure 3 depicts, not all ROS are affected equally by the 6124 mutation. In particular, peroxides are elevated, superoxide is depressed and there is no difference in hydroxyl radicals. One possible explanation for the decrease in superoxide levels in the presence of abundant superoxide dismutase that results in the rapid conversion of superoxide to hydrogen peroxide. Hydrogen peroxide is stable and able to accumulate while superoxide anion is transient and highly reactive. ). For comparison, 143B cybrids cell lines containing a separate patient mtDNA with either a mutation at position 8993 (8993Mut) or wild type 8993 (8993WT) is shown. Error bars represent the standard error of the mean of 9 data points each 6124WT, 6124Mut, and 3 data points each 8993WT and 8993Mut. (d) Hydroxyl radicals and peroxynitrite anions as measured by hydroxyphenyl�uorescein (HPF) remain unchanged. Error bars represent the standard error of the mean of 9 data points each 6124WT and 6124Mut.
e possible tumorigenic effects of increases in ROS are well known and include (at least) an increased rate of DNA mutations and ROS-induced promitogenic signaling [20][21][22]. ese are common �ndings in prostate cancer and other solid tumors [23,24]. It is likely that the cumulative effect over a lifetime of a tonic increase in cellular ROS increases the chances of malignant transformation. It is uncertain whether the prostate is more susceptible to this in�uence than other   organs, but the association of COI mutations with disease suggests some speci�city. e possible mechanism by which a decreased efficiency of oxidative phosphorylation is related to malignant transformation is less obvious. It is probably relevant to the so called "Warburg Effect" wherein tumor cells exhibit defective oxidative phosphorylation and increased glycolysis as a primary means of ATP generation [25]. It is possible that inherited or somatically acquired mtDNA mutations are partially or wholly responsible for this effect. It is possible that the decrease in oxidative phosphorylation is not in and of itself causally related to the increased risk of cancer, but that the relation to cancer risk is conferred predominantly by the increased ROS generation and that compromises to oxidative phosphorylation are merely a bystander effect of no direct consequence to tumorigenesis. is remains to be determined.
e possible tumorigenic effects of increases in RNS are also well known. Nitric oxide (NO) is generated enzymatically by synthases (NOS), which oxidize L-arginine to Lcitrulline. e inducible form, iNOS, is present in a variety of cell types. Over the past 20 years iNOS expression has been associated with various human tumors including: breast, brain, lung, prostate, colorectal, and melanoma [26][27][28][29][30][31][32]. NO action is concentration-dependent. Increased cGMPmediated ERK phosphorylation is associated with low levels of NO in cancer cells whereas HIF-1a stabilization is associated with intermediate levels [33]. ough NO has been shown to induce apoptosis it can transcriptionally enhance MMP-1 via the ERK and p38 MAPK pathways resulting in tumor progression and also can result in the overproduction of VEGF [33]. Although NO levels are transient, iNOS generated NO �uctuations varying from seconds to days can in�uence the antiapoptotic�proapoptotic effects of NO [33].
We have demonstrated that the T6124C (M74T) mutation was inherited in a heteroplasmic state by a patient that developed prostate cancer and that the mutation not only causes increased ROS and nitric oxide but also induces increased cellular proliferation, decreased apoptosis, and increased in vivo tumor growth. ese are quintessential characteristics of the malignant phenotype. ese results are concordant with other studies that demonstrate a profound effect of mtDNA mutations on tumor cell biology. In 2008 Ishikawa demonstrated that mtDNA mutations determined the metastatic potential of lung cancer cell lines independent of the characteristics of the cancer cell nucleus. Speci�cally, high-metastatic cell lines lost their metastatic potential when the mtDNA was replaced with mtDNA from low-metastatic cell lines and low-metastatic cell lines acquired high-metastatic capabilities when mtDNA from high metastatic lines were inserted [34]. e metastatic capacity could be eliminated by treatment with antioxidants indicating the central importance of ROS-induced signaling in metastasis. Similarly, an ATP6 (ATP synthase subunit 6) missense mutation in the mitochondrial genome (T8993G) causes increased ROS in prostate cancer cell lines and increased tumor growth [5].

Conclusions
Mitochondrial DNA from a prostate cancer patient with an inherited-heteroplasmic-mtDNA mutation in COI, the catalytic core of mitochondrial respiratory complex IV was studied. In the laboratory, this mutation was found to simultaneously decrease the activity of the respiratory complex as measured by the rate of cytochrome c oxidation and to increase the rate of mitochondrial reactive oxygen generation. Other mutation-induced biochemical changes included increased generation of nitric oxide. Cells harboring the mutation proliferated faster in vitro and caused increased tumor growth in vivo. ese �ndings suggest a possible molecular substrate (mtDNA mutation) for the "Warburg effect" of anaerobic metabolism exhibited by some tumors and increased cellular reactive oxygen, a common �nding in solid tumors.