Fer and FerT Govern Mitochondrial Susceptibility to Metformin and Hypoxic Stress in Colon and Lung Carcinoma Cells

Aerobic glycolysis is an important metabolic adaptation of cancer cells. However, there is growing evidence that reprogrammed mitochondria also play an important metabolic role in metastatic dissemination. Two constituents of the reprogrammed mitochondria of cancer cells are the intracellular tyrosine kinase Fer and its cancer- and sperm-specific variant, FerT. Here, we show that Fer and FerT control mitochondrial susceptibility to therapeutic and hypoxic stress in metastatic colon (SW620) and non-small cell lung cancer (NSCLC-H1299) cells. Fer- and FerT-deficient SW620 and H1299 cells (SW∆Fer/FerT and H∆Fer/FerT cells, respectively) become highly sensitive to metformin treatment and to hypoxia under glucose-restrictive conditions. Metformin impaired mitochondrial functioning that was accompanied by ATP deficiency and robust death in SW∆Fer/FerT and H∆Fer/FerT cells compared to the parental SW620 and H1299 cells. Notably, selective knockout of the fer gene without affecting FerT expression reduced sensitivity to metformin and hypoxia seen in SW∆Fer/FerT cells. Thus, Fer and FerT modulate the mitochondrial susceptibility of metastatic cancer cells to hypoxia and metformin. Targeting Fer/FerT may therefore provide a novel anticancer treatment by efficient, selective, and more versatile disruption of mitochondrial function in malignant cells.


Introduction
Metastasis is the main cause of death among cancer patients, and targeting the metastatic process is therefore one of the major challenges in current cancer therapy. Increasing evidence suggests that migrating cancer cells and primary tumor cells utilize distinct metabolic pathways [1][2][3][4]. For instance, analysis of gene expression signatures in an orthotopic breast cancer model indicated that circulating tumor cells are enriched in factors that regulate mitochondrial respiration and biogenesis compared with primary and metastatic lesions [2]. Furthermore, the mitochondria of cancer cells are reprogrammed and modified compared to normal cells. For example, the mitochondrial glutamine uptake machinery which propels the tricarboxylic acid (TCA) cycle is upregulated in cancer cells [5,6]. Notably, other functional pathways including the electron transport chain (ETC) are also altered in the mitochondria of malignant cells. Specifically, complex I (Comp. I) mutations which could serve as metabolic determinants of malignant-cell sensitivities to glucose limitation are frequently observed in many cancers [7]. Two constituents that associate with the reprogrammed Comp. I in malignant but not normal somatic cells are the intracellular tyrosine kinase Fer and its sperm-and cancer-specific truncated variant, FerT [8,9]. Fer bears a kinase domain residing in its C-terminal part that is preceded by an SH2 domain and a 414 aa long N terminal tail bearing three coiled-coil domains and an FCH (Fps/Fes/Fer/CIP4 Homology) motif [10][11][12]. The genomic Fer locus is located on chromosome 5q21 [13] and contains two genes: fer and ferT, which encode two distinct kinases: Fer and FerT. An intronic promotor in the fer gene directs the expression of the FerT protein, which consequently gains a unique 43 aa long N terminal tail [14]. While Fer is expressed in all somatic cells except for pre-T, pre-B, and naïve T cells [15], the expression of FerT is normally restricted to spermatocytes, spermatids, and sperm cells [8,16]. However, FerT is also found in various subcellular compartments of malignant cells, and together with Fer, it associates with Comp. I of the mitochondrial ETC only in spermatogenic and cancer cells [8]. Fer was shown to regulate breast cancer cell adhesion, migration, and anoikis resistance and to be necessary for tumor growth and metastasis in mice [11,[16][17][18][19]. Unlike Fer, not much is known about the regulatory roles of FerT in cancer cells. A better understanding of the roles of Fer and FerT in metabolic reprogramming of malignant cells may open new avenues for efficiently and selectively targeting the reprogrammed mitochondria of metastatic cells.
Metformin, a guanidine derivative initially extracted from the plant Galega afficinalis (French lilac), has been used as a glucose-lowering medication in humans for more than 60 years [20]. Metformin exerts its primary effect at the molecular level as an inhibitor of oxidative phosphorylation (Oxphos) by reversibly inhibiting NADH dehydrogenase (mitochondrial ETC-Comp. I) activity, resulting in reduced ATP production [21][22][23]. The AMP-activated protein kinase (AMPK) is also a key molecular mediator through which metformin exerts its anticancer effects [24]. A meta-analysis on diabetic cancer patients treated with metformin reported a significant reduction in mortalities for various cancers [25][26][27]. These findings motivated the inclusion of metformin in numerous anticancer therapeutic combinations [28,29]. However, it turned out that the efficacy and therapeutic impact of metformin depends on the site and type of cancer [30]. Furthermore, it was shown that cancer cells, which are insensitive to low glucose supplementation, are also moderately sensitive to metformin treatment [7]. Thus, there is a profound importance in further unraveling regulatory factors that control the moderate susceptibility of cancer cells to metformin therapy. Since metformin targets the reprogrammed mitochondrial ETC of malignant cells, we sought to decipher the roles of Fer and FerT in modulating the susceptibility of cancer cell's mitochondria to metformin-evoked stress. In this study, we show that Fer and its cancer-specific variant, FerT, are novel regulators of mitochondria vulnerability to mitochondrial stresses like metformin treatment and onset of hypoxic conditions.

Generating SW/H∆Fer/FerT and SW/H∆Fer Cells
Knockout of the fer and ferT genes by the CRISPR-Cas9 paired nickases system was carried out according to the manufacturer's instructions (Sigma) using the Cas9 D10A mutant fused to GFP [31] and a pair of gRNAs targeting a common region of Fer and FerT in Exon 12-gRNA1: TATTCTGGGAATTGCACCATGG and gRNA2: GAGAGAGTCATGGGAAA CCTGG-or a pair of gRNAs targeting a specific region of Fer in Exon 6-gRNA1: GCTTTG TCGTATCGTTCCTTGG and gRNA2: TTGCACAATCAGTATGTATTGG. SW620 and H1299 cells were transfected with three plasmids using Lipofectamine 2000 (Invitrogen-Cat. 11668019) according to the manufacturer's guides. Cells expressing the GFP-Cas9 were sorted by fluorescence-activated cell sorting (FACS)-FACSAriaIII (Becton Dickinson Biosciences, San Jose, CA 95131, USA). Authentication of the parental SW620 and H1299 cells, and all their derived clones was performed at the Genomic Center of Biomedical Core Facility, the Technion, Israel. All cell lines were authenticated using short tandem repeat (STR) profiling (please see the Supplementary Materials) within the last three years, and all experiments were performed with mycoplasma-free cells.

Subjecting Cells to Hypoxia
SW620, SW∆Fer/FerT, SW∆Fer, H1299, H∆Fer/FerT, and H∆Fer cells were grown for 48 h in glucose reach medium or in glucose-deprived medium (DMEM without glucose, Biological Industries) supplemented with 2 mM L-glutamine at 37 • C with 5% CO 2 . The cells were subjected to anaerobic culture jar containing CO 2 -generating envelope (GasPak EZ, BD Biosciences, BD260001) for an additional 24 h. These conditions reduced the oxygen level in the jar to 1% within 30 min.

Immunoblot Analysis
Whole-cell lysates were prepared as described before [8]. In brief, 30 µg protein lysates from each sample were resolved by 10% SDS-PAGE and analyzed by western blotting using polyclonal anti-Fer/FerT antibodies selectively directed toward the common SH2 domain of the two proteins [8]. Reacting protein bands were visualized using a horseradish peroxidase (HRP)-conjugated secondary antibody to rabbit or mouse IgG (Jackson,111-035-144, 115-035-062, respectively) in conjunction with a western blot (WB) chemiluminescence reagent (Pierce Cat. 34080).

Cell Death Analysis
Cells (5 × 10 5 ) were seeded in 6-cm cell culture dishes in glucose-deprived medium (DMEM without glucose, BI, 01-057-1A) supplemented with 2 mM L-glutamine for 48 h. Cells were then treated with metformin at the indicated concentration for 24 h. Cells were stained with annexin V-FITC and propidium iodide (PI) using Annexin V-FITC Apoptosis Detection Kit (Biovision Cat. K101-100) following the manufacturer's instructions. Staining was quantified by FACS ARIAIII. All data were analyzed using FlowJo software (FlowJo LLC, Ashland, Oregon, USA).

Quantification of Cellular ATP and NAD + Levels
Cells were suspended in 0.5-mL cold perchloric acid (PCA) solution in 1.5-mL tubes. The mixture was incubated on ice for 15 min and was then centrifuged at 13,000× g for 2 min to remove the precipitate. The supernatant was neutralized followed by incubation with NaOH for 15 min on ice. After another centrifugation at 10,000× g for 2 min, the supernatants were taken for chromatographic analysis as described before [8].

Determination of Mitochondrial Activity
Oxygen consumption rate (OCR) was monitored as an indicator for mitochondrial respiration activity and was measured with an XF24 Extracellular Flux Analyzer using XF Cell Mito Stress Test kit according to the manufacturer's instructions (Seahorse, XF Cell Mito Stress Test Kit, 103015-100, Agilent, Santa Clara, CA 95051, USA). Cell seeding number was optimized to 100,000 cells/well for SW620 cells. For determining mitochondrial susceptibility to stress cues, cells were plated into XF24 plates in glucose-deprived media for 48 h followed by metformin treatment (5 mM) for 16 h. For determining mitochondrial basal rate activities, mitochondrial respiratory chain drugs were added, following the Mito Stress kit specifications. One micromole of oligomycin was used to block ATPlinked oxygen consumption, 1 µM of carbonilcyanide p-triflouromethoxyphenylhydrazone (FCCP) was used as an uncoupling agent to obtain maximal respiration, and 0.5 µM of rotenone/antimycin A was applied to inhibit complexes I and III, thereby arresting all mitochondrial respiration. OCR was measured 3 consecutive times following the injection of each drug and was normalized to protein content.

Statistical Analysis
Statistical analysis was performed using the paired and unpaired Student's t-tests, with p < 0.05 being considered significant. The results are depicted as mean ± standard error (±SE) of the mean for n given samples.
For statistical analysis, percent change values were divided by 100 and log-transformed. Mean percent change of treatments and/or cell type were compared to the control group using one-sample t-tests against a mean of 0, and correction for multiple testing was applied using the false discovery rate (FDR) procedure. Treatments and/or cell types were compared using one-way or two-way ANOVA tests, followed by Tukey's post hoc analysis. The normality of residuals assumption was assessed with residuals plots.

Fer/FerT Deficiency Exacerbates Susceptibility of SW620 and H1299 Cells to Metformin
To decipher the roles of Fer and FerT in modulating mitochondria susceptibility to stress cues in cancer cells, we initially focused on metastatic SW620 colon cancer (CC) cells, which express both Fer and FerT [16] ( Figure 1A). We generated Fer-and FerTdeficient SW620 cells (SW∆Fer/FerT) using the modified CRISPR-Cas9 mutated knockout system [31,32]. Expression analysis of Fer and FerT revealed efficient knockout of the fer and ferT genes in 4 SW∆Fer/FerT clones ( Figure 1A); SW∆Fer/FerT clones #1 and #3 were selected for further studies.
To test the impact of Fer and FerT absence on the sensitivity of SW620 cells to the mitochondrial targeting effect of metformin, SW∆Fer/FerT cells were cultured in a medium devoid of glucose and supplemented with glutamine. Under these conditions, glycolysis was halted and mitochondrial Oxphos became the main energy-generation process. While parental SW620 cells were only mildly affected by either 5-or 10-mM metformin treatment, SW∆Fer/FerT cells were profoundly affected by metformin and their viability was significantly impaired after 24 h of treatment with 10-mM metformin ( Figure 1B). Annexin V/PI staining analysis revealed that the main form of death induced by metformin was primarily late apoptosis ( Figure 1C).
To examine whether metformin affects mitochondrial function in the treated metastatic cells, the level of ATP was determined under conditions in which Oxphos prevails (low glucose/high glutamine). This revealed a significant decrease in cellular ATP levels in metformin-treated SW∆Fer/FerT cells compared to metformin-treated parental SW620 cells ( Figure 1D). Since metformin was shown to inhibit the activity of the ETC Comp. I type I NADH dehydrogenase (ubiquinone) [21][22][23] with which Fer and FerT associate in SW620 cells [8], we compared NAD + levels in SW620 and SW620∆Fer/FerT cells. This parameter inversely reflects the mitochondrial Comp. I activity level. While NAD + levels in SW620 cells were barely affected even after 24 h of metformin treatment, NAD + levels were drastically decreased 16 and 24 h posttreatment in SW∆Fer/FerT cells ( Figure 1E).
To check the generality of our findings, we turned to knocking out the fer and ferT genes in other colon cancer cell lines. However, these attempts failed in several colon cancer cell lines derived from different stages of the disease (Supplementary Materials Figure  S1). Of the other cancer cells lines tested, we managed to knockout the fer and ferT genes in metastatic NSCLC H1299 cells (Figure 2A). As seen with SW∆Fer/FerT cells, Fer-and FerT-deficient H1299 (H∆Fer/FerT) cells were significantly more sensitive to metformin treatment than the parental H1299 cells (Figure 2B-D).

FerT Governs the Vulnerability of SW620 Cells to Metformin
To discern the roles of Fer and FerT in modulating the sensitivity of SW620 cells to metformin, we selectively knocked-out the fer gene while preserving ferT and the encoded FerT protein. Multiple SW∆Fer clones that failed to express Fer but maintained FerT levels similar to parental SW620 cells were generated, two of which (#10, and #12, henceforth referred to as SW∆Fer 1 and SW∆Fer 2, respectively) were selected for further studies ( Figure 3A).  We then compared the sensitivity of parental SW620, SW∆Fer/FerT, and SW∆Fer cells to metformin. Unlike SW∆Fer/FerT cells (SW∆Fer /FerT 1 and 3), which exhibit increased sensitivity to 10 mM metformin, SW∆Fer cells (SW∆Fer 1 and 2) expressing FerT but not Fer maintained moderate sensitivity to metformin like parental SW620 cells ( Figure 3B).
The decreased sensitivity to metformin of SW∆Fer cells was reflected by a decreased level of late apoptosis and necrosis induced by metformin in these cells in comparison to SW∆Fer/FerT cells ( Figure 3C).
To examine whether the reduced sensitivity of SW∆Fer cells to metformin is also reflected by moderate downregulation of mitochondrial functions, we compared cellular ATP levels in SW620, SW∆Fer/FerT, and SW∆Fer cells. While this revealed a significant decrease in cellular ATP in metformin-treated SW∆Fer/FerT cells, the ATP and NAD+ levels in treated SW∆Fer cells, which express FerT, were similar to that measured in parental SW620 cells (Figure 3D,E). Coinciding with the above findings, mitochondrial respiratory activity was higher in metformin-treated SW∆Fer cells compared with SW∆Fer/FerT cells and matched the mitochondrial respiration rate in SW620 cells ( Figure 3F). Thus, FerT modulates the sensitivity of SW620 cells to metformin. Notably, unlike SW620 cells, in H1299 cells, the expression of FerT alone ( Figure 4A, clones H∆Fer, *2, and *4) showed only marginal resumption of the moderate sensitivity of these cells to metformin ( Figure 4B-E), suggesting that Fer is required for resuming moderate sensitivity of these cells to metformin.

FerT Governs The Sensitivity of SW620 Cells to Hypoxic Stress
The modulatory role of FerT on the sensitivity of SW620 cells to metformin prompted us to examine the role of this cancer-specific Fer variant in governing malignant-cell sensitivity to hypoxia, another stress affecting mitochondria in solid tumors. SW620 and SW∆Fer/FerT cells were grown in either glucose-reach or glucose-deprived medium supplemented with glutamine to promote glycolysis or mitochondrial Oxphos, respectively. The cells were subjected to hypoxic stress for 24 h, after which their viability was determined. While no obvious difference between the survival of SW620 and SW∆Fer/FerT cells could be seen under high glucose (DMEM) (Figure 5A), glucose deprivation (glutamine only) of SW∆Fer/FerT cells rendered them significantly more susceptible to hypoxia ( Figure 5A). Notably, the expression of FerT alone in the absence of Fer, endowed the SW∆Fer cells, with minor sensitivity to hypoxia exhibited by the parental SW620 cells ( Figure 5B). Accordingly, FerT maintained moderate mitochondrial sensitivity to hypoxic stress as reflected by the similar ATP and NAD + levels measured in hypoxic SW∆Fer and the parental SW620 cells ( Figure 5C,D). As was seen for SW∆Fer/FerT cells, Fer-and FerT-deficient H1299 cells (H∆Fer/FerT) exhibited increased sensitivity to hypoxic stress. However, unlike the SW620 cells, the expression of FerT alone did not resume moderate sensitivity to hypoxia exhibited by the parental H1299 cells (Figure 6A,B).
Hence, while FerT governs the sensitivity of SW620 cells to hypoxic stress, the presence of Fer is required for maintaining the moderate mitochondrial susceptibility of H1299 cells to hypoxic cue.

Discussion
Several clinical studies have shown that metformin can attenuate the proliferation and growth of breast, prostate, and endometrial cancer or malignant tumors [33][34][35][36][37]. However, even patients that initially respond to metformin acquire resistance to this drug [38]. Thus, there is an eminent need to extend our understanding of the molecular pathways that promote resistance to metformin. A pharmacodynamics window study of metformin in breast cancer patients revealed that tumors with upregulated Oxphos genes showed impaired response to metformin [38]. These findings imply and substantiate the notion that upregulation of the mitochondrial Oxphos could reduce the sensitivity of cancer cells to metformin [7]. This could coincide with the observation that one of the primary targets of metformin anticancer activity is inhibition of mitochondrial Comp. I function [21]. Furthermore, a recent work by Liu et al. comparing the metabolite profile of ten ovarian tumor samples from untreated and ten metformin-treated patients demonstrated decreases in the levels of some TCA cycle intermediates [39]. Hence, metformin interferes also with mitochondria functioning in vivo, thereby restricting the survival of cancer cells. Upregulation of Oxphos through Oxphos gene induction or the reprogramming of mitochondrial ETC in malignant cells could therefore endow cancer cells with reduced sensitivity and acquired resistance to the metformin growth suppressive effects [7,40]. In the current work, we showed that the intracellular tyrosine kinase Fer and its cancer-specific variant, FerT, which associate with Comp. I and potentiate its activity in malignant cells under stress-imposed conditions [8], play this alleviating role. The absence of the two enzymes is shown to increase mitochondrial vulnerability and susceptibility of metastatic cancer cells to metformin. It should be noted that the metastatic cell lines studied in the current work are insensitive to low glucose growth conditions (Supplementary Materials Figure S2), and their moderate susceptibility to metformin coincides with a previous study indicating that insensitivity to low glucose is linked to a low sensitivity of cancer cells to metformin [7]. Thus, our findings define Fer and FerT as novel supporters of the impaired susceptibility of low-glucose insensitive metastatic cells, to metformin treatment.
The recruitment of Fer and the sperm-and cancer-specific FerT to the reprogrammed ETC Comp. I of malignant cells seems to also enable mitochondria functioning under metabolic stress conditions encountered by abnormally growing cancer cells. One such metabolic challenge is hypoxic stress evoked in solid tumors, which outgrow their vasculature, and in metastatic cells that detach from the primary tumor and begin a dissemination process in the patient's body [41,42]. The fact that Fer and FerT decrease the susceptibility of metastatic malignant cells to hypoxic stress may have translational ramifications on the development of new anti-metastatic therapeutic approaches directed toward Fer and FerT. Of note is the fact that, while in SW620 cells FerT alone could direct reduced susceptibility of the cells to hypoxia and metformin, in the H1299 NSCLC cells, FerT alone did not have this significant impact. Thus, in H1299 cells, either Fer may have a dominant modulatory role on cell sensitivity to hypoxia and metformin or a collaborative functioning of Fer and FerT is required for exerting their susceptibility-alleviating effects on these cells. Interestingly, H1299 clones lacking both Fer and FerT, or Fer alone exhibited similar mitochondrial activity (Supplementary Materials Figure S3). While our study does not exclude the possibility that Comp. I-associated Fer can also decrease the sensitivity of SW620-cell mitochondria to metformin and hypoxia, our findings do indicate that the cross-regulatory roles of Fer and FerT may differ among distinct metastatic cell types. However, since Fer and FerT share a common kinase domain [8], targeting the kinase domain of the two enzymes with a synthetic molecule should increase the sensitivity of cancer cells to hypoxia and biguanides. Late-phase clinical trials investigating metformin as an anticancer drug are underway. Combining metformin with a specific Fer/FerT inhibitor may potentiate the therapeutic efficacy of this commonly used antidiabetic and anticancer compound.
Supplementary Materials: The following are available online at https://www.mdpi.com/2073-440 9/10/1/97/s1, Figure S1: Colon cell lines derived from different stages of the disease failed to lose the expression of fer and ferT genes. Figure S2: SW620 and H1299 cells are insensitive to low-glucose growth conditions. Figure S3: Similarly impaired mitochondrial activity in H∆Fer/FerT and H∆Fer clones.

Conflicts of Interest:
The authors declare that they have no conflicts of interest with the contents of this article. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript; or in the decision to publish the results.