Partial inhibition of mitochondrial-linked pyrimidine synthesis increases tumorigenic potential and lysosome accumulation

The correlation between mitochondrial function and oncogenesis is complex and is not fully understood. Here we determine the importance of mitochondrial-linked pyrimidine synthesis for the aggressiveness of cancer cells. The enzyme dihydroorotate dehydrogenase (DHODH) links oxidative phosphorylation to de novo synthesis of pyrimidines. We demonstrate that an inhibition of DHODH results in a respiration-independent significant increase of anchorage-independent growth but does not affect DNA repair ability. Instead, we show an autophagy-independent increase of lysosomes. The results of this study suggest that inhibition of mitochondrial-linked pyrimidine synthesis in cancer cells results in a more aggressive tumor phenotype.


Introduction
With the multiple essential roles of mitochondria within proliferation, apoptosis, metabolic adaptation and Ca2 + homeostasis (Reichert and Neupert, 2004), alterations of mitochondrial function have unsurprisingly been demonstrated to play critical roles in malignant transformation, cancer progression (Porporato et al., 2018), metastasis (Denisenko et al., 2019), immune evasion (Klein et al., 2020;Møller et al., 2020) and chemoresistance (Guerra et al., 2017).One of the better studied mitochondrial functions in this context is mitochondrial oxidative phosphorylation (OxPhos).However not all aspects of this pathway have received equal amounts of attention.We have previously correlated OxPhos depletion with decreased and imbalanced levels of cytosolic deoxyribonucleoside triphosphates (dNTP) and increased chromosomal instability (Desler et al., 2007).In this study, we describe the specific effect of inhibition of mitochondrial-linked synthesis of deoxypyrimidines on cancer cells.
The enzyme dihydroorotate dehydrogenase (DHODH) is an integral protein of the inner mitochondrial membrane and is functionally dependent on a functional OxPhos (Bader et al., 1998).DHODH is a ratelimiting enzyme for the de novo synthesis of pyrimidines and is responsible for the conversion of dihydroorotate to orotate.Orotate is further converted to uridine monophosphates, UTP and CTP, and ultimately dTTP and dCTP.Dihydroorotate oxidation by DHODH is coupled to respiratory ubiquinone reduction, and a lack of oxygen, presence of electron transport chain (ETC) inhibitors or mutations in ETC complex III or IV, impairs activity of DHODH and consequently results in a decrease in the de novo synthesis of pyrimidine ribonucleotides (Beuneu et al., 2000;Löffler et al., 1997).
Strict regulation of dNTP levels is important to maintain genomic stability, and a disturbance of this balance can induce a variety of genetic changes such as base substitutions, frameshift mutations, chromosomal alterations, delay of replication fork progression, and DNA replication, as well as increase the frequency of fragile sites (Bebenek and Kunkel, 1990;Jacky et al., 1983;Ke et al., 2005;Kunz and Kohalmi, 1991;Pai and Kearsey, 2017).
Here, we demonstrate that a specific inhibition of DHODH with brequinar reverts to a decrease in especially deoxypyrimidines and to an increased anchorage-independent cell growth.We demonstrate that supplementation with uridine restores deoxypyrimidine levels and reverts anchorage-independent cell growth to that of untreated controls.These results show that inhibition of mitochondrial-mediated pyrimidine synthesis can contribute to a more aggressive phenotype in cancer cells.We do not detect decrease in DNA repair efficiency but do see a significant accumulation of lysosomes, a phenotype that has recently been correlated to genomic instability and chromosome alterations (Almacellas et al., 2020).

Cell culture
The human MDA-MB-435 melanoma cell line and the human 143B osteosarcoma thymidine kinase 1 negative cell line were kind gifts from Dr. Keshav Singh (University of Alabama at Birmingham.United States).Thymidine kinase 1 deficient cell line were used as a model where the effect on de novo pyrimidine nucleotide synthesis could not be compensated by the salvage pathway.Cells were maintained in DMEM with Glutamax (Invitrogen, GIBCO) supplemented with 10% (v/v) fetal bovine serum (BioChrom AG), and 1% (v/v) Penicillin/Streptomycin (Invitrogen, GIBCO).Cell lines were cultured at 37 • C in a humidified atmosphere containing 5% CO 2 .Cell concentration was determined in a Casy Cell Counter (Model TT, Innovatis).For longer periods of incubation, ATP synthase was inhibited with 0.5 µM oligomycin (Sigma-Aldrich), DHODH was inhibited by 0.5 µM brequinar (Sigma-Aldrich).DMSO was used as a mock control.When mentioned, cells were supplemented with 16 µM uridine (Sigma-Aldrich) and 2 mM pyruvate (Sigma-Aldrich).Unless otherwise stated, incubation time with inhibitors was 6 h.

XF24 Microplate-based determination of mitochondrial respiration
Respiration of MDA-MB-435 and 143B cells was measured using an XF24 Extracellular Flux Analyzer (Agilent Technologies).MDA-MB-435 and 143B cells were seeded at a density of 6.5 × 10 4 and 5 × 10 4 cells per well in an Agilent Seahorse cell culture microplate and grown overnight to ensure adhesion.Cells were incubated with oligomycin, brequinar, or mock control as previously described, washed twice and finally resuspended in Seahorse assay medium (Agilent Technologies) supplemented with 10 mM glucose, 1 mM pyruvate, 2 mM glutamine and adjusted to pH 7.4.The oxygen consumption rate (OCR) and the extracellular acidification rate (ECAR) were measured while successively adding the ATP synthase inhibitor oligomycin (1.0 µM), the uncoupler carbonyl cyanide-4-(triflouromethoxy)phenylhydrazone (FCCP) (1 µM), and the ETC complex III inhibitor antimycin A (2 µM).This allowed the determination of basal ATP turnover (difference in OCR before and after addition of oligomycin), reserve respiratory capacity (difference in OCR before addition of oligomycin and after addition of FCCP), and glycolytic reserve capacity (difference in ECAR before and after addition of oligomycin).

Quantification of Whole-Cell ATP levels
The ATP content was determined using the Vialight MDA Plus kit (Lonza) luciferase-based assay according to the manufacturer's instructions.The luminescence levels were quantified on a Microbeta2 scintillation counter (Perkin Elmer).

Quantification of whole cell dNTP pools
The whole cell dNTP content of MDA-MB-435 and 143B cells was determined after exposure with oligomycin, brequinar, or mock control, with and without pyruvate and uridine supplement, using the DNA polymerase extension assay as previously described (Desler et al., 2007).

Cell cycle analysis by flow cytometry
For cell cycle analysis of oligomycin, brequinar, and mock-treated cells, 5 × 10 5 cells were washed and fixed in 2 ml ice-cold 70% ethanol.Cells were washed with PBS, 0.1 mg / ml of RNase A (Sigma-Aldrich) was added and cells were stained with 0.4 mg/ml of propidium iodide (Sigma-Aldrich).DNA content was measured using a Becton Dickinson FACSCalibur fluorescence-activated cell sorter (FACS), Cell QuestPro software and Verity ModFit software.
2.6.Soft agar assay for anchorage-independent cell growth 5 × 10 4 143B cells were seeded in each well of a 6-well plate, in a layer of DMEM containing 0.35% low-melting point agarose (GibcoBRL, Life Technologies) and layered on top of 0.5% agar and overlayered with 3 ml of growth medium.Agarose solutions as well as growth media was complemented with oligomycin, brequinar or mock control, with and without pyruvate and uridine supplement.Cells were grown for three weeks during which growth media with inhibitor or mock DMSO and supplements was exchanged every three days.Colonies were stained with crystal violet and colonies were counted under an inverted bright field microscope.

Clonogenic survival by colony formation assay
Exponentially growing 143B cells were seeded in 6-well dishes.After cell adhesion they were incubated with oligomycin, brequinar or mock control.After initial incubation, cells were UVB irradiated (0, 25, 50 or 75 J/m 2 ) by a 254 nm germicidal UVB lamp or incubated with methyl methanesulfonate (MMS) (0, 250, 375 or 500 µM for 2 h).UVB exposures were measured with a UVX-meter (UVP, Jena Company).Cells were fixed with 75% methanol and 25% acetic acid and stained with 1 % crystal violet.Surviving colonies (≥50 cells) were counted and normalized to the plating efficiency of unexposed control cells.

Alkaline comet assay
The alkaline comet assay was used to measure DNA damage in 143B cells incubated with oligomycin, brequinar, or mock control and subsequently exposed to 0 or 250 µM MMS for 2 h.The presence of DNA strand breaks was quantified in cells after 0, 1 and 4 h of repair.2500 cells were embedded in 0.75 % low-melting point agarose and attached on GelBond® films (Medinova Scientific, Hellerup, Denmark), using Lab-Tek II chambers (Nunc, Life Technologies) as molds.Cells were lysed (2.5 M NaCl, 10 mM Tris-Base, 100 mM EDTA-Na 2 , 1% Nasarcosinat, 10% DMSO, 1% Triton X-100, pH = 10) for 1 h at 4 • C in the dark.The Gelbond films were washed three times with water and submerged in cold electrophoresis buffer (0.3 M NaOH, 1 mM EDTA-Na 2 , pH ≥ 13.2) for 40 min and subsequently electrophoresed for 20 min (300 mA and 1 V/cm).After electrophoresis, samples were submerged twice in neutralization buffer (0.4 M Tris, pH = 7.5) for 5 min and washed once with ultrapure water before they were fixed in 96% ethanol.The gels were air dried overnight, dyed with SYBR Gold (Invitrogen), and analyzed under a Nikon fluorescence microscope at x40 magnification.From each gel, randomly selected comets (nucleoids) were analyzed using Komet analysis software (Andor Technologies) to calculate comet Olive tail moment.Mean Olive tail moment was calculated from the comet Olive tail moment of 50 randomly selected cells and each data point consists of a triplicate of mean Olive tail moments.This value was used as a quantitative index of single-and doublestranded DNA breaks.

Transmission electron microscopy
Lysosomal content was quantified by Transmission Electron Microscopy (TEM).143B cells were seeded and grown on Thermanox™ coverslips to near confluency, brequinar or mock-treated for 6 h, fixated with 2.5% glutaraldehyde solution (Sigma) and sliced in 100 nm sections on microtome.Images were acquired from same slice depth, on a 40KV − 100 kV CM 100 TEM.Cells were selected at random in low magnification from each sample and compared using automatically acquired stitched images from high magnification.Images were stitched, lysosome numbers counted, total cell area and total lysosome area measured in Photoshop CS6.

Statistics
Data are plotted as mean ± standard deviation.Comparisons between groups were analyzed with two-tailed unpaired Student's t tests (Fig. 1CED performed with GraphPad Prism (v9.00).

Inhibition of ATP synthase or DHODH decreases levels of dNTP
To differentiate the effects of an inhibition of DHODH and a complete inhibition of ETC, we incubated MDA-MB-435 and 143B cells with the DHODH inhibitor brequinar (0.5 µM) or the ATP synthase inhibitor oligomycin (0.5 µM) for 6 h.Subsequently, we quantified the resulting whole cell levels of dNTP and showed that both inhibitors had an effect.Inhibition of DHODH was shown to predominantly decrease pyrimidine deoxyribonucleotide levels, while inhibition of ATP synthase decreased both pyrimidine and purine levels of deoxyribonucleotides in the two cell lines (Fig. 1).
For MDA-MB-435 cells, the whole cell contents of dNTP of untreated cells were 53, 56, 42 and 31 pmol/10Λ6 cells for dTTP, dATP, dCTP and dGTP respectively.In MDA-MB-435 cells, inhibition of DHODH led to a significant decrease in dTTP and dCTP levels to 69% and 59% of the mock control (p = 0.0009 and p = 0.0009, respectively; n = 3), a small but significant increase of 9% of dATP levels compared to the mock control (p = 0.0318; n = 3), and no effect on dGTP levels (Fig. 1A).Inhibition of ATP synthase mediated a significant decrease in dTTP, dATP, dCTP and dGTP levels to 9%, 18%, 15% and 23% of the mock control (p < 0.001, p < 0.001, p < 0.001 and p < 0.0001 respectively; n = 3).For 143B cells, the whole cell contents of dNTP of untreated cells were 34, 48, 30 and 22 pmol/10Λ6 cells for dTTP, dATP, dCTP and dGTP respectively.In 143B cells, inhibition of DHODH mediated a significant decrease in predominantly pyrimidine levels of dTTP and dCTP  to 38% and 43% of the mock control (p < 0.001 and p = 0.003 respectively; n = 3), but also, to a smaller extent, purine levels dATP and dGTP to 83% and 67% of the mock control (p = 0.038 and p = 0.034 respectively; n = 3) (Fig. 1B), while inhibition of ATP synthase results in a significant decrease of dTTP, dATP dCTP and dGTP levels to 64%, 84%, 58% and 76% of the mock control (p = 0.01, p = 0.02, p = 0.006 and p < 0.001 respectively; n = 3).When supplementing brequinarexposed 143B cells with uridine and pyruvate, dNTP levels of DHODH inhibited cells were rescued, while the effect of ATP synthase inhibition in oligomycin-exposed cells further decreases dATP, dCTP and dGTP levels (Fig. 1B).
Cytosolic dNTP levels are reduced in nonproliferating cells compared to proliferating cells due to reduced activity of key enzymes of both the de novo and salvage pathways of nucleotide synthesis (Chabes et al., 2003;Munch-Petersen et al., 1995).Inhibition of the ATP synthase with 0.5 µM oligomycin has negligible effect on the cell cycle distribution of MDA-MB-435 or 143B cells compared to mock-exposed controls.Inhibition of DHODH with 0.5 µM brequinar, has no effect on distribution of cells in S-phase demonstrating that the altered dNTP levels of cells incubated with oligomycin and brequinar are not caused by fewer cells in S-phase (Table 1).Incubating cells with brequinar, resulted in an accumulation of cells in G1 and a decrease of cells in G2/M phase, suggesting that brequinar affects the DNA damage checkpoint dynamics of the cells (Chao et al., 2017).

Inhibition of DHODH by brequinar does not affect OxPhos
The activity of DHODH is fully dependent on complex II, III and IV, and a deficiency of OxPhos, therefore disables DHODH function (Beuneu et al., 2000;Fang et al., 2013;Löffler et al., 1997).In contrast, an inhibition of DHODH has been shown to impair overall OxPhos activity (Klotz et al., 2019;Xuan et al., 2018;Yu et al., 2019).To evaluate any residual effect of brequinar exposure on overall OxPhos activity, we measured OCR in MDA-MB-435 and 143B cells using the Seahorse XF Extracellular Flux Analyzer after exposure to oligomycin, brequinar, or mock control.Cellular ATP turnover and reserve respiratory capacity were determined as attributes of OxPhos (Fig. 2A-D).When measuring cells incubated with 0.5 µM oligomycin for 6 h, no additional decrease in ATP turnover was detectable, indicating that the ATP synthase was completely inhibited (data not shown).On the contrary, 6 h incubation with 0.5 µM brequinar did not influence ATP turnover or reserve respiratory capacity (Fig. 2A-D) when compared to mock-treated cells.This demonstrates that inhibition of DHODH by 0.5 µM brequinar does not affect OxPhos of MDA-MB-435 or 143B cells.
ECAR was also measured during addition of mitochondrial inhibitors (Fig. 2E).The difference in ECAR before and after the addition of oligomycin is a measure of the compensatory regulation of glycolysis in response to an inhibition of the OxPhos.The measure is defined as the glycolytic reserve capacity, and although it is not a complete measure of glycolysis, the lack of difference between cells incubated with 0.5 µM brequinar for 6 h, and the mock control indicates that no additional compensation of the glycolysis rate is measured after treatment with brequinar.The absence of effect of brequinar on overall bioenergetics is evident by unaltered whole-cell levels of ATP (Fig. 2F).

Inhibition of DHODH but not ATP synthase promotes Anchorageindependent growth
We have previously correlated mitochondrial-induced downregulation of cytosolic levels of dNTP with chromosomal instability (Desler et al., 2007).Therefore, we wanted to investigate whether inhibition of ATP synthase or DHODH would have a comparable effect on aggressiveness.We analyzed anchorage-independent cell growth in 143B cells by soft agar clonogenic assay.Loss of anchorage-dependent cell-cycle checkpoint is an important characteristic of many tumor cells (Cremona and Lloyd, 2009;Freedman and Shin, 1974).The soft agar clonogenic assay showed a significant 29% increase in colony formation in 143B cells incubated with brequinar (p < 0.001; n = 15) demonstrating an anchorage-independent induced phenotype.This phenotype was normalized when cells were supplemented with uridine and pyruvate, indicating that anchorage-independent growth was correlated with decreased levels of dNTPs (Fig. 3).In contrast, when OxPhos was inhibited by oligomycin an 65% decrease of colony formation (p < 0.0001; n = 6) was observed, demonstrating a reduced anchorage-independent phenotype of these cells.When supplemented with uridine and pyruvate, colony formation was still significantly suppressed to 33% lower than mock treated cells (p < 0.0001; n = 6).

Inhibition of ATP synthase but not DHODH reduces cell survival in response to MMS damage
To determine whether induced imbalances of the dNTP pool sensitize cells to different types of DNA damage or not, survival of 143B cells was assessed by clonogenic formation after increasing UVB irradiation and increasing MMS concentrations.Exposure to 0.5 µM oligomycin or brequinar resulted in a 25-30% reduction in viability compared to mock (Data not shown).After UVB irradiation, only significant survival differences were present between cells treated with brequinar and mock at the highest dose of nonlethal UVB irradiation of 75 J/m 2 , where the survival of the mock control is 32.7% and the survival of cells exposed to brequinar is 17.3% compared to unirradiated cells (p = 0.0026; n = 4) (Fig. 4A).After MMS exposure, there was no difference in survival of brequinar-exposed 143B cells compared to mock-control (Fig. 4B).Consistently, when oligomycin-exposed cells were exposed to increasing levels of UVB irradiation, the same trend was evident, where the only difference between oligomycin and mock-treated cells is apparent after exposure to 75 J/m 2 UVB irradiation, where the survival of the mock control is 32% and the survival of oligomycin-exposed cells is 2.6% compared to unirradiated cells (p = <0.0001;n = 4) (Fig. 4C).In contrast, 143B cells incubated with oligomycin are significantly more sensitive to MMS exposure than mock-treated cells since survival at all concentrations of MMS exposure is 27%-32% lower (p = 0.0054 -p < 0.0001; n = 6) (Fig. 4D).This demonstrates that an inhibition of DHODH by brequinar have very small effect on sensitivity towards UVB and MMS exposure, while inhibition of OxPhos by oligomycin results in a sensitivity of the cells towards alkylating DNA damage of MMS.

Mitochondrial-induced dNTP imbalance does not impair repair efficiency of MMS-induced DNA strand breaks.
To investigate whether exposure to brequinar or oligomycin affects the ability of cells to repair DNA damage, 143B cells were preincubated with inhibitors and assayed directly or exposed to 250 µM MMS for a period of 2 h and allowed to repair for a period of 4 h.After 0, 1, and 4 h of repair, the presence of DNA strand breaks was quantified using the  alkaline comet assay.There was no significant difference in the presence of DNA strand breaks between mock control, oligomycin, and brequinar exposed cells.The abundance of DNA breaks in mock-treated cells decreased to 77% of initial levels after 1 h of repair and 49% of initial levels after 4 h of repair, indicating that the 143B cell line is repairproficient (Fig. 5).When the repair efficiency of cells exposed to oligomycin and brequinar with mock-treated cells was compared, no significant differences were apparent, demonstrating that inhibition of ATP synthase nor DHODH and the ensuing decrease of dNTP levels do not affect the repair rate of MMS-induced DNA breaks.

Inhibition of DHODH increases the number of lysosomes
To evaluate the overall effect of DHODH inhibition, we investigated the general morphology of DHODH-inhibited cells by TEM.We noticed a significant increase of lysosomes.In images acquired by electron microscopy, the electron dense acidic lysosomes appear as dark spheres, with a diameter of anything between 0.1 and 1.2 µm.We totaled a significant 3-fold increase in the number of lysosomes in cells incubated with brequinar when compared to mock-treated controls (p < 0.001; n = 20) (Fig. 6A).The level of lysosomal activity reflects cellular activities, such as waste disposal, secretion, plasma membrane repair, mitosis, endocytosis and energy metabolism.There were no differences in the size of the lysosomes of cells incubated with brequinar and mock treated cells.Similarly, cell size was not affected by brequinar exposure (Fig. 6BC).This demonstrates an increased need to process and degrade molecules that can have intracellular or extracellular origin.

Lysosome formation is independent from LC3B-mediated autophagy
To better understand the role of the increased number of lysosomes, we wanted to examine whether an increase in autophagic activity was responsible for the increase in lysosomal formation in DHODH-inhibited cells.We measured autophagic activity by quantifying the protein levels of LC3B-I and LC3B-II.During the course autophagy the LC3B-II/LC3B-I ratio correlates with the number of autophagosomes.As the autophagosome fuses with a lysosome LC3B-I, levels decrease.We found no significant change between LC3B-II/LC3B-I ratio of DHODH nor ATP synthase-inhibited cells (Fig. 7 and supplementary Fig. 1).This demonstrate that the observed lysosome formation is not a result of a need to process intracellular molecules via autophagy and instead suggest an extracellular origin of molecules being processed by the increased number of lysosomes.

Discussion
The link between mitochondrial function and oncogenesis is extremely complex and multifaceted.In this study we successfully inhibit the DHODH with no residual effect on the overall OxPhos function.This allows us to focus on the DHODH-specific consequences of mitochondrial dysfunction.We demonstrate that the cellular levels of dNTP are declining by disabling OxPhos by inhibiting ATP synthase.However, we also demonstrate that a specific inhibition of DHODH reduces especially levels of deoxypyrimidines.This result supports our previous finding correlating OxPhos dysfunction with decreased and imbalanced cellular levels of dNTP (Desler et al., 2007), and substantiates our hypothesis that the mechanism responsible for this relationship is linked to the pyrimidine de novo pathway and its dependence on DHODH (Desler et al., 2010).
When supplemented with uridine and pyruvate, the effect on the dNTP levels of DHODH inhibited cells, but not ATP synthase inhibited cells, is normalized.Uridine replenishes pyrimidines via the salvage pathway whereas pyruvate serves as substrate for both OxPhos and glycolysis.We conclude that the pathway linking DHODH inhibition to dNTP regulation is different from the pathway linking inhibition of ATP synthase to dNTP regulation.Whereas oligomycin inhibits ATP synthase and, in turn, the ETC and, thereby, also DHODH, we show that brequinar inhibits DHODH but no other aspects of OxPhos.

Decrease of cellular dNTP levels mediated by inhibitions of DHODH and ATP synthase does not impair repair efficiency of UV and MMSinduced DNA damages
We wanted to investigate the direct effects on genomic stability of DHODH and ATP synthase-induced decreases in cellular dNTP levels.We therefore investigated the response of DHODH and ATP synthase inhibited cells to various DNA damaging agents.MMS is an alkylating agent that modifies DNA by methylation of both adenine and guanine to 3-methyladenine and 7-methylguanine, respectively, causing base mispairing and replication block (Beranek, 1990).MMS-induced lesions are repaired primarily by short patch base excision repair, but also by homologous recombination and DNA lesion bypass (Drabløs et al., 2004).During short patch base excision repair, only a single damaged nucleotide is excised and replaced (Dianov et al., 1992).In contrast, UVB primarily modifies DNA by inducing pyrimidine dimers (Kielbassa et al., 1997), which are primarily repaired by nucleotide excision repair in a process where approximately 30 bases are removed and replaced (Friedberg, 2001).In this study, the survival of DHODH and ATP synthase-inhibited cells after exposure to MMS and UVB was used to assess the effect of reduced cellular levels of dNTP on the fidelity of repair of DNA lesions.For cells exposed to UVB, only the highest levels of irradiation resulted in a significant change in survival between mocktreated cells and cells inhibited by DHODH and ATP synthase, indicating that DNA lesions induced by UVB irradiation up to 50 J/m 2 do not sensitize cells to decreased levels of dNTP.In this work, UVB irradiation exceeding 50 J/m 2 resulted in a survival rate of mock treated cells that was 77% lower than that of nonirradiated cells.At this dose, UVB irradiation also results in the formation of nitric oxide and increased mitochondrial superoxide formation and can induce cytotoxicity and apoptosis (Gonzalez Maglio et al., 2005).We can therefore not conclude the effect seen at these levels of UVB irradiation is solely related to impaired DNA repair.When assessing survival after exposure to MMS, no differences were observed between DHODH-inhibited cells and mock-treated cells.In contrast a significant difference in survival is present between ATP synthase inhibited and mock-treated cells at all levels of MMS exposure.If alkylated bases are left unrepaired, the risk of single-and double-stranded DNA breaks increases (Drabløs et al., 2004).However, we did not find any significant differences in the abundance of single-and double-stranded DNA breaks after MMS treatment between ATP synthase-inhibited and mock-treated cells.This suggests that effects of inhibition of ATP synthase are not mediated through changes of repair efficiency and, therefore, not specifically related to ATP synthase.Our results do not exclude the erroneous incorporation of nucleotides into DNA due to an imbalance of whole-cell dNTP levels.Such an event would also potentially promote the aggressiveness of cancer cells.

DHODH-mediated decrease of cellular dNTP levels induces aggressiveness
Full inhibition of the DHODH by inhibitors and a resulting depletion of the nucleotide pool, has been demonstrated to have antiproliferative effects on cancer cells (reviewed in (Madak et al., 2019)) and several attempts have been made to develop inhibitors of DHODH for cancer therapy, however none to date have gained FDA approval.In this study, the inhibition of DHODH is partial rather than complete, as reflected in dNTP levels that are imbalanced rather than depleted.This is a more relevant model for DHODH activity following decreasing OxPhos activity, as a full inhibition of DHODH only will be apparent after complete loss of OxPhos activity, and even though many cancers do have significantly reduced their OxPhos activity, a complete loss is not often seen (reviewed in (Zheng, 2012)).
Imbalance of the cytosolic dNTP pool has been demonstrated to decrease the genetic stability of newly transformed cells and has been suggested as a very important initiating step of cancer development (Bester et al., 2011;Saldivar et al., 2012).In this study tumorigenicity, as measured by anchorage-independent growth in soft agar, was significantly increased in DHODH inhibited cells but decreased in ATP synthase inhibited cells.This is in accordance with our previous study, where genomic instability was demonstrated as an effect of mitochondrial dysfunction and decreased cellular levels of dNTP (Desler et al., 2007).When dNTP synthesis is supplemented with uridine and pyruvate, the effects of DHODH inhibition but not ATP synthase inhibition are normalized.This trend was also evident with regard to normalization of dNTP levels and led us to believe that the effects of DHODH inhibition directly impair the cellular levels of dNTP while inhibition of ATP synthase results in a general lack of energy in the cell.Such an ATP starvation results in a multifaceted cell stress response (Yalamanchili et al., 2016), where the decrease of dNTP synthesis is likely attributed to decreased activity of all ATP dependent components of the de novo and salvage DNA synthesis pathways as well as the overall response to a general energy starvation.

DHODH-mediated decrease in cellular dNTP levels induces LC3Bindependent lysosomal accumulation
We observed a significant increase of lysosome formation in DHODH inhibited cells.Lysosomes are involved in the disposal and recycling of cellular macromolecules and extracellular material (Yim and Mizushima, 2020).Furthermore, lysosome function has been related to different aspects of transformation and cancer aggressiveness.Due to increased energy demand, boosted lysosomal biogenesis is a hallmark of malignant transformation (Miranda-Gonçalves et al., 2018;Perera et al., 2019) and an increase of lysosomal hydrolases in human cancers correlate with higher risk of reoccurrence and poor prognosis (Kallunki et al., 2013;Kuester et al., 2008).Furthermore, enhanced invasiveness of cancer cells is related to increased lysosomal exocytosis (Davidson and Vander Heiden, 2017;Wu et al., 2020) and correct lysosomal function has been related to chromosomal segregation, where impairment of lysosomes increases mitotic errors and promotes chromosomal instability (Almacellas et al., 2020).
Cellular biomolecules to be digested and recycled are fed to lysosomes by autophagy, while extracellular material is delivered by pinocytosis or phagocytosis (Yim and Mizushima, 2020).In this study, we did not observe an increase in autophagic activity, measured by the presence of LC3B proteins.This suggests that the accumulation of lysosomes in DHODH-inhibited cells is not caused by an increase in the number of intracellular biomolecules to be disposed of but rather from an uptake of extracellular material.This phenotype is similar to that of transforming cells boosting lysosomal function to cover an energy demand (Miranda-Gonçalves et al., 2018;Perera et al., 2019).We, therefore, suspect that the accumulation of lysosomes is an indicator of increased tumorigenic potential of the cells and is likely to be an attempt from the cells to retrieve more extracellular material for enhanced cell growth.Whereas an elucidation of the specific mechanism responsible for the lysosomal response is outside the scope for this work, we believe it is an interesting topic worthy of future investigations.
Due to the multifaceted role of the mitochondria in the cell, their functions are often modulated by transforming cells and tumors to gain an advantage.Full or enhanced function of DHODH has been suggested as a prerequisite for cancer development and progression, since pyrimidine depletion blocks proliferation (Boukalova et al., 2020;Zhou et al., 2021).A complete inhibition of DHODH activity by leflunomide or ablation of TFAM, a mitochondrial transcription factor responsible for expression of mitochondrial DNA-encoded subunits of the OXPHOS complexes, suppressed UVB-induced skin tumor progression (Hosseini et al., 2018).Furthermore, pharmacological inhibition of DHODH interferes with melanoma tumor growth and in small cell lung cancer models (Li et al., 2019;White et al., 2011).In this study, inhibition of DHODH results in a decrease but not a depletion of deoxyribonucleotides.At this level of DHODH inhibition, we demonstrate an increase in anchorage-independent growth and an accumulation of lysosomes.We further demonstrate that this phenotype is not associated with a reduction in DNA repair activity and, therefore, likely not the cause of dNTP starvation.Whereas a complete inhibition of the DHODH would be an endpoint of OxPhos dysfunction, a partly inhibition of DHODH is the result of the decreasing OxPhos activity leading up to this event.

Conclusions
We conclude that the oncogenic effects elicited by the DHODH is complex, where a partial inhibition of OxPhos can promote aggressiveness, but a full inhibition will retard proliferation.This again underscores the complex modulatory relationship between mitochondrial function and tumor development.

Fig. 2 .
Fig. 2. Respiration is not affected by brequinar inhibition of DHODH.Respiratory properties of 143B and MDA-MB-435 cells incubated with 0 (mock control) and 0.5 µM brequinar.A. Oxygen consumption rates of cells exposed to brequinar or mock control.During the stress test, cells are supplemented with oligomycin, FCCP, and antimycin A in consecutive order.B. Basal respiration is determined as initial resting consumption of oxygen.C. ATP turnover is measured as a decrease in oxygen consumption after the addition of oligomycin.D. Reserve respiratory capacity is determined as the difference between basal respiration and respiration after FCCP addition.E. Glycolytic capacity is measured after the addition of oligomycin.F. Whole-cell ATP levels of cells exposed to brequinar and mock control.

Table 1
Cell cycle distribution of MDA-MB-435 and 143B cell incubated with oligomycin and brequinar.