LDH-A inhibition, a therapeutic strategy for treatment of hereditary leiomyomatosis and renal cell cancer

The genetic basis for the hereditary leiomyomatosis and renal cell cancer syndrome is germ-line inactivating mutation in the gene for the Krebs/tricarboxylic acid cycle enzyme, fumarate hydratase (FH), the enzyme that converts fumarate to malate. These individuals are predisposed to development of leiomyomas of the skin and uterus as well as highly aggressive kidney cancers. Inhibition of FH should result in significant decrease in oxidative phosphorylation necessitating that glycolysis followed by fermentation of pyruvate to lactate will be required to provide adequate ATP as well as to regenerate NAD+. Moreover, FH deficiency is known to up-regulate expression of hypoxia-inducible factor (HIF)-1α by enhancing the stability of HIF transcript. This leads to activation of various HIF-regulated genes including vascular endothelial growth factor and glucose transporter GLUT1 and increased expression of several glycolytic enzymes. Because lactate dehydrogenase-A (LDH-A), also a HIF-1α target, promotes fermentative glycolysis (conversion of pyruvate to lactate), a step essential for regenerating NAD+, we asked whether FH-deficient cells would be exquisitely sensitive to LDH-A blockade. Here, we report that hereditary leiomyomatosis and renal cell cancer tumors indeed overexpress LDH-A, that LDH-A inhibition results in increased apoptosis in a cell with FH deficiency and that this effect is reactive oxygen species mediated, and that LDH-A knockdown in the background of FH knockdown results in significant reduction in tumor growth in a xenograft mouse model. [Mol Cancer Ther 2009;8(3):626–35]


Introduction
Many cancer cells show enhanced glycolysis (1). In recent years, the molecular basis for this effect has started to be elucidated. In particular, up-regulation of hypoxia-inducible factor (HIF), which occurs as a consequence of hypoxia as well as from alterations in certain oncogenes or mutations in tumor suppressor genes, results in the increased transcription of genes involved in glucose transport, glucose metabolism, lactate formation, and lactate export from cells (2 -6). Moreover, HIF activation results in decreased activity of the pyruvate dehydrogenase complex (7,8). Collectively, these studies provide a partial explanation for the Warburg effect, which states that, even under aerobic conditions, cancer cells preferentially undergo glycolysis followed by fermentation (9). Fumarate hydratase (FH) plays an important role in the mitochondrial tricarboxylic acid cycle by catalyzing the conversion of fumarate to malate. The primary function of the Krebs/ tricarboxylic acid cycle is oxidation of pyruvate to generate ATP. The genetic basis for the hereditary leiomyomatosis and renal cell cancer (HLRCC) syndrome is germ-line inactivating mutation in the FH gene (10 -12). Inhibition of FH and presumably of oxidative phosphorylation should necessitate that fermentative glycolysis will be required to provide ATP as well as regenerate NAD + for glycolysis. The pseudohypoxic drive in HLRCC involving fumaratedependent HIF activation as a result of competitive HIF prolyl hydroxylase inhibition by accumulated fumarate is likely to lock FH-deficient cells into a glycolytic pathway for sustaining the metabolic energy requirements (13).
In the present study, we wanted to ask whether FHdeficient cells do rely on glycolysis for growth and/or proliferation and whether they would be sensitive to inhibition of fermentative glycolysis. The fermentative reaction is regulated by lactate dehydrogenase (LDH), which in most tissue occurs in two major isoforms LDH-A and LDH-B. LDH-B is a homotetramer of LDH-H and favors the conversion of lactate to pyruvate and is also known to be upregulated in many cancers (14). LDH-A is a heterotetramer of LDH-M and favors conversion of pyruvate to lactate and is up-regulated in many solid tumors that have high HIF expression (3,15). LDH-A is an attractive target for cancer therapy because its expression is largely relegated to skeletal muscle. It is not present in red cells, in which glycolysis followed by fermentation is an obligatory process for energy generation. Moreover, it is well known that humans with LDH-A deficiency only show myoglobinuria under intense anaerobic exercise (16,17) and individuals with complete lack of LDH-A or LDH-B subunit have been documented with no apparent increase in hemolysis (16,18,19). Therefore, it is likely that potential LDH-A inhibitors might show relatively modest systemic toxicity.

Materials and Methods
Patient Samples and Characteristics: HLRCC and von Hippel-Lindau Phenotypic/Genotypic Assessment and Tumor Sample Procurement Patients with suspected or confirmed HLRCC and von Hippel-Lindau (VHL) were evaluated in the Urologic Oncology Branch of the National Cancer Institute, where they underwent a comprehensive clinical evaluation for phenotype assessment. Each of the 10 patients with HLRCC-associated kidney cancer came from a family with confirmed FH germ-line mutation: 6 had surgery at the National Cancer Institute and the other 4 had surgery at outside institutions. The clear cell kidney cancers came from VHL patients with confirmed germ-line mutation of the VHL gene; all 6 had kidney cancer surgery done at the National Cancer Institute. The LDH-A staining in VHL and HLRCC tumor samples was compared with nonmalignant kidney tissue. The pathology of all of the renal tumors was reviewed by coauthor M.J.M.
Immunohistochemistry Slides (5 Am) form formalin-fixed, paraffin-embedded tissue samples were used for immunohistochemistry. Slides were deparaffinized in three changes of xylene for 5 min each followed by rehydration in graded alcohols. Antigen retrieval was achieved by heating the slides in Tris-EDTA (pH 9.0) in a microwave oven at 95jC for 20 min. Endogenous peroxidase activity was inhibited by incubation in 3% hydrogen peroxide in methanol for 10 min. Sections were then incubated for 1 h at room temperature with sheep anti-human LDH-A antibody (1:30,000; Abcam) or isotype-matched IgG as negative controls. After several washes in Tris, samples were further incubated with rabbit anti-sheep IgG (1:1,000; Millipore) as a secondary antibody for 30 min at room temperature. Slides were then incubated with goat anti-rabbit horseradish peroxidase polymer (Envision PO System; DAKO) for 30 min followed by 3,3 ¶-diaminobenzidine as chromogen for 5 min. The sections were counterstained with Mayer's hematoxylin and permanently mounted. Staining for LDH-A in HLRCC and VHL-null tumors was evaluated in four high-power fields and samples were scored as strongly positive (3+), moderately positive (2+), weakly positive (1+), or negative (0). Staining pattern (membranous, cytoplasmic, or nuclear) was also recorded.

Chemicals and Reagents
Cell culture medium was purchased from the American Type Culture Collection. Lentiviral short hairpin RNA (shRNA) constructs were purchased from Sigma-Aldrich, and the viral power kit and packaging cell line 293FT were purchased from Invitrogen. Sheep polyclonal LDH-A antibodies were purchased from Abcam, and FH mouse monoclonal was purchased from Novus. Rabbit polyclonal for activated caspase-3 and activated poly(ADP-ribose) polymerase (PARP) were purchased from Cell Signaling. The assay kit for measuring ATP was purchased from Roche. Lactate measurement kit was purchased from CMA microdialysis. Cell cycle reagents and Annexin V assay kits were purchased from Guava Technologies. Athymic mice for tumor implantation were purchased from Charles River Laboratories. A549 cells were maintained in Ham's F-12 with glutamine and sodium pyruvate with 10% FCS (Sigma). RCC4 and RCC4 + VHL cells were gift from Celeste Simon (University of Pennsylvania School of Medicine). Sequences of shRNA used in this study were LDH-A CCACCATGATTAAGGGTCTTT, LDH-A1 GATC-TGTGATTAAAGCAGTAA, FH63 CGCTGAAGTAAAC-CAGGATTA, and FH65 CCCAACGATCATGTTAATAAA.
Generation of FH-, LDH-A-, and FH/LDH-A-Deficient Cell Lines A549 cells were infected separately with empty shRNA vector control, three different FH shRNAs, and LDH-A lentiviral particles as described (20). Briefly, Recombinant lentiviral particles were produced by transient transfection of 293T cells according to standard protocol. Briefly, subconfluent 293FT cells were cotransfected with 3 Ag shRNA plasmid and 9 Ag viral power packaging mix (an optimized proprietary mix of three plasmids, pLP1, pLP2, and pLP/VSVG; Invitrogen) using Lipofectamine 2000 (Invitrogen). After 16 h, culture medium was switched to regular growth medium and cells were allowed to incubate for additional 48 h. Conditioned cell culture medium containing recombinant lentiviral particles was harvested and frozen. A549 cells were treated with above cell culture supernatant containing lentiviral particles for 24 h. These cells were then selected in puromycin (Sigma-Aldrich) to generate stable cell lines encoding empty vector, FH, and LDH-A shRNA. The selected cell lines were validated for diminished LDH-A and FH expression by Western blot analysis. Briefly, total cellular proteins were separated by SDS-PAGE and electrotransferred to polyvinylidene fluoride membranes and immunoblotted with anti-LDH-A (Abcam) or anti-FH (Novus) antibody for overnight at 4jC. After washing with TBS-Tween 20, the membrane was incubated with secondary antibody of choice for 30 min. The protein bands were detected using SuperSignal West Pico Chemiluminescent substrate (Pierce). For generation of FH/LDH-A clones, A549 cells were infected with FH and LDH-A shRNA and selected with puromycin and validated with Western blot analysis with FH and LDH-A antibodies for double knockdown. For tumor implantation, double stable FH/LDH-A cells were generated by larger-scale infection and tested for the lack of FH/LDH-A expression before implantation. LDH-A-deficient RCC4 cells were generated in a similar manner as described above and validated for LDH-A inhibition by Western blot analysis.

Proliferation Assay
Control, FH-deficient, and FH/LDH-A-deficient cell lines were plated in 60 mm dish at a density of 1 Â 10 5 per dish in Ham's F-12 supplemented with 10% fetal bovine serum for 24 h at 37jC in a 5% CO 2 incubator. After 24, 48, 72, 96, and 120 h initial plating, cells were scraped, washed with PBS, and resuspended in 1 mL HBSS and counted in the presence of trypan blue. All samples were assayed in duplicate to generate proliferation curves as described (21).

Lactate Accumulation
We measured lactate accumulation in control and LDH-A-deficient cell lines as described (22). Briefly, the cell culture conditioned supernatant was used to measure lactate levels by a simple calorimetric analysis based on the reduction of the tetrazolium salt INT in a NADHcoupled enzymatic reaction to formazan, which is watersoluble and exhibits an absorption maximum at 492 nm. The intensity of the red color formed is proportional to the lactate concentration in the conditioned medium. All samples were assayed in three independent experiments in triplicate. The average value of the absorption readings was used for graphical representation.
LDH-A Enzyme Activity LDH-A activity was measured in the direction of reduction of pyruvate to lactate by monitoring the changes in the absorbance of NADH at 340 nm. Briefly, the reaction mixture contained 50 mmol/L potassium phosphate (pH 7.4) and pyruvate (200 Amol/L). The reaction was initiated by addition of cell extract prepared in potassium phosphate and NADH (250 Amol/L), and decrease in absorbance at 340 nm was monitored every 4 min.
Intracellular ATP Measurements Intracellular ATP levels in control, FH-deficient, and FH/ LDH-A-deficient cells were measured according to the manufacturer's instructions and as described before (7). In brief, cell lysates were collected and luminescence was measured using a luminescence reader (Molecular Devices) and normalized for protein concentration.
Annexin V Apoptosis Assay Apoptosis was measured by Guava PCA-96 Nexin (Guava Technologies) as per the manufacturer's protocol. Briefly, cells were harvested and resuspended in 1Â Nexin buffer with Annexin V-PE and Nexin 7-amino-actinomycin D. The cells were allowed to incubate for 15 min and analyzed in the Guava flow cytometer.

Invasion Assay
The invasiveness of cells was analyzed using Falcon BioCoat Matrigel chambers (Becton Dickinson) with 6.4 mm diameter and 8 Am pore size. Invasiveness of cells expressing control, FH, and FH/LDH-A shRNA was assayed as described previously (23,24). Briefly, 4 Â 10 4 cells were seeded on the upper side of a Matrigel invasion chamber. After 24 h, the cells on the upper surface of the filter were removed using cotton swabs, and the invaded cells were fixed with 20% solution of 4% paraformaldehyde and stained with crystal violet. Incorporated dye was extracted with 0.05 mol/L sodium phosphate (pH 4.5) in 50% ethanol and the absorbance of released dye was read at 550 nm. The data represent average of two independent experiments done in triplicate.

Oxygen Consumption and Extracellular Acidification Rate Analysis
The XF24 extracellular flux analyzer (Seahorse Biosciences) is a fully integrated 24-well instrument that measures in real-time the uptake and release of metabolic endproducts. Each XF24 assay well contains a disposable sensor cartridge, embedded with 24 pairs of fluorescent biosensors (oxygen and pH), coupled to fiber-optic waveguides. The wave guides deliver light ray at various excitation wavelengths (oxygen = 532 nm, pH = 470 nm) and transmit a fluorescent signal (oxygen = 650 nm, pH = 530 nm) to a set of highly sensitive photodetectors. This technology was used to measure oxygen consumption expressed in pmol/min and extracellular acidification rate, expressed in mpH/min in control, FH-and FH/LDH-Adeficient cell lines as described (25).
Reactive Oxygen Species Measurements Intracellular reactive oxygen species (ROS) production was measured by staining with 2 ¶,7 ¶-dichlorodihydrofluorescein diacetate (H 2 DCFDA; Invitrogen). H 2 DCFDA is a cell-permeant indicator for ROS that is nonfluorescent until removal of the acetate groups by intracellular esterases and oxidation occurs within the cell. The procedure for measuring ROS was carried out as described earlier (7), with minor modification. Briefly, cells were loaded with 5 Amol/L H 2 DCFDA for 1 h, washed in PBS, and incubated in fresh medium for 30 min. The cells were then subjected to fluorescence-activated cell sorting analysis to visualize the fluorescence.
Tumor Implantation Control and FH65/LDH-A-deficient cells (3.0 Â 10 6 ) were s.c. implanted in eight athymic mice. The implantation was done such that each mouse had FH65 cells on the right flank and FH65/LDH-A-deficient cells on the left flank. The control shRNA cells were implanted in a separate set of athymic mice. Tumors were measured every 5 days and tumor volume was calculated as described before (26). Tumor lysates were prepared by homogenization of tumor tissues in lysis buffer, separated by SDS-PAGE, electrotransferred to polyvinylidene fluoride membranes, immunoblotted with anti-LDH-A and anti-FH antibody, and normalized by actin as a loading control.
Statistical Analysis Student's t test was used to evaluate the statistical significance of the results.

Enhanced Expression of LDH-A in HLRCC
Because FH deficiency up-regulates HIF-1a (13), and LDH-A is a known HIF-1a target, we asked whether expression of LDH-A is significantly enhanced in HLRCC tumors. Previous microarray profiling data in FH-deficient uterine fibroids suggest that LDH-A expression may be increased compared with normal myometrium (27). As shown in Fig. 1A, LDH-A expression is abundant in HLRCC tumors and almost negligible in surrounding normal tissue. This expression pattern largely overlaps with HIF expression in the corresponding section from the same tumor (Fig. 1B). This result is almost universal in HLRCC tumor samples that were sequence validated for FH mutations (Fig. 1C and D). Based on these results, we wanted to ask whether inhibition of LDH-A in a FH-deficient background would have significant effect on the survival or the rate of proliferation of FH-deficient cells in vitro.
Generation and Validation of FH-Deficient Cell Lines as a Surrogate Cell Culture Model for HLRCC Because there is no established HLRCC cell line that can be used to study metabolic alterations and signaling events associated with FH deficiency, we used the A549 lung cancer cell line as a surrogate model system to generate FH-deficient cell lines. Our choice of the cell line to create HLRCC surrogate model system in cell culture was based on the previous publication that suggested that these cells carry wild-type VHL gene, as is the case in HLRCC patients, and inhibition of endogenous FH expression in these cells results in enhanced glycolysis (13). Western blot analysis with anti-FH antibody shows that cell lines FH63 and FH65, which express shRNAs targeting two different regions within the FH mRNA, have diminished FH protein expression compared with the cell line expressing control shRNA (Fig. 2A). The FH-deficient cells compared with control cells have comparable rates of proliferation (Fig. 2B). However, in agreement with the study by Isaacs et al. (14), FH-deficient cells compared with control cells have increased lactate levels (Fig. 2C), enhanced HIF-1a expression in normoxic conditions (Fig. 2F), and showed increased expression of two HIF-1a target genes, GLUT1 and vascular endothelial growth factor (VEGF; Fig. 2D and E). Because FH deficiency mimics the pseudohypoxic state that should favor fermentative glycolysis regulated by LDH-A, a HIF-1a target, we also found increased HIF-1a expression in FH-deficient cells and a 2.0-fold increase in LDH-A enzyme activity in FH-deficient cells compared with empty vector control or wild-type cells as measured by loss of NADH in the presence of pyruvate (Fig. 2G).

Generation and Validation of LDH-A-Deficient Cell Lines in a FH-Deficient Background
To test whether inhibition of LDH-A in FH-deficient cells would result in reduced proliferation, we generated

LDH-A-deficient cells in a FH-deficient background and will refer to these cells as FH/LDH-A-deficient cells in this article. Western blot analysis with anti-FH and anti-LDH-A antibodies shows that cell lines FH63/LDH-A and FH65
/ LDH-A have diminished FH and LDH-A protein expression compared with the control cell line or FH onlydeficient cell lines (Fig. 3A and B). As expected, FH65 shRNA cells show increased lactate formation as measured by lactate accumulation and real-time measurements in a XF-24 metabolite analyzer compared with control cells, but FH/LDH-A-deficient cells show reduced lactate formation ( Fig. 3C and D).

In vitro Characterization of FH/LDH-A-Deficient Cell Lines
Next, we asked whether cell lines deficient in FH/LDH-A expression would show altered cell survival/growth. We first characterized these stable cell lines in various in vitro assays. As shown in Fig. 4A, all FH/LDH-A-deficient cell lines display a significant proliferation defect compared with cells with LDH-A knockdown in the background of wild-type FH gene. The FH/LDH-A-deficient cells also show significantly higher levels of apoptosis compared with control shRNA cells (Fig. 4B). Moreover, induction of apoptosis is also shown by Western blot analysis of caspase-3 cleavage and PARP cleavage in FH/LDH-Adeficient cell lines (Fig. 4C). These data could also be extended to growth in three dimensions: we have noted that, in a soft agar growth assay, LDH-A-deficient cells on a FH-deficient background generate smaller colonies compared with FH-deficient cells, suggesting lower tumorigenic potential (data not shown).
As metastatic potential is one of the hallmarks of HLRCC tumors (28,29), we asked whether a cell line deficient in FH would display a more invasive phenotype as assessed by a Matrigel invasion assay. We found that FH-deficient cells have significantly enhanced invasive potential in this assay and this invasive behavior can be dialed back to the same level as cells expressing control shRNA (Fig. 3E) by LDH-A inhibition. Moreover, this invasive ability correlated with lactic acid generation ( Fig. 3C and D).
To determine the effect on ATP generation of LDH-A inhibition in FH-deficient cells, ATP levels were measured. In comparison with the control cells or FH alone-deficient cells, FH/LDH-A-deficient cells show significant ATP depletion (f50-60% reduction; Fig. 5A). This result is in agreement with the substantially reduced proliferation rate observed in our various growth assays. We were unable to show increased NADH/NAD + ratio (data not shown) in FH/LDH-A-deficient cells. Furthermore, we also noted an increased rate of oxygen consumption in FH/LDH-A-deficient cells, suggesting that LDH-A inhibition in FH-deficient background may be forcing the cells to increase oxidative phosphorylation (Fig. 5B).

Enhanced LDH-A Expression Occurs in Renal Tumors with VHL Deficiency
VHL deficiency is a common feature of clear cell renal carcinoma (30,31). It also results in HIF-1a up-regulation (32). We would therefore expect enhanced LDH-A expression in VHL-deficient renal cancer tumor tissue samples. This was indeed found to be the case (Fig. 6A). Moreover, we have also shown that inhibition of LDH-A by two different shRNAs in the VHL-deficient cell line RCC4 results in increased apoptosis ( Fig. 6B and C). RCC4 cells reconstituted with wild-type VHL lacks LDH-A expression as a result of diminished HIF-1a expression (33); hence, these cells cannot be used to generate LDH-A-deficient cell lines for comparison studies between VHL-deficient and VHL-reconstituted RCC4 cells. Our analysis of VHLdeficient tumors and in vitro data suggest that LDH-A is overexpressed in VHL-deficient cells and its inhibition results in apoptosis.

Increased ROS Generation in FH/LDH-A-Deficient Cells Correlates with Apoptosis
Based on our previous results, we asked whether enhanced oxidative phosphorylation due to LDH-A knockdown in FH-deficient cells might result in increased formation of ROS. H 2 DCFDA is a fluorogenic probe commonly used to detect cellular ROS. We found enhanced ROS levels as measured by H 2 DCFDA in FH/LDH-Adeficient cells compared with control or FH-deficient cells (Fig. 5C). Furthermore, this ROS generation could be blocked by the antioxidant N-acetylcysteine and resulted in partial rescues from apoptosis as measured by cleaved PARP (Fig. 5D). These data suggest that ROS generation may be partly implicated in the increased apoptosis seen in FH/LDH-A-deficient cells.

In vivo Characterization of FH/LDH-A-Deficient Cell Lines
To ask whether FH/LDH-A-deficient cell lines have significantly diminished potential for tumor formation, we implanted these cells into the flanks of athymic mice. We found that FH65/LDH-A knockdown cells show marked growth inhibition (Fig. 7A) compared with cells expressing control or FH shRNA. We also show representative tumor tissue from control, FH-deficient, and FH/ LDH-A-deficient tumors assayed for FH and LDH-A

Discussion
The major findings in this report are the following: (a) HLRCC tumors overexpress LDH-A, (b) LDH-A inhibition results in increased apoptosis via ROS production in an A549 surrogate FH knockdown cell line, and (c) inhibition of fermentative glycolysis in this cell line results in significant reduction in growth in a xenograft mouse model.
The common feature of tumor cells is their reliance on fermentative glycolysis, a phenomenon coined as the Warburg effect (1). Although the mechanism that may control this metabolic shift is not clearly defined, many  lines of investigations suggest that HIF expression may be central to Warburg effect both by induction of fermentative glycolysis and by diminishing the activity of pyruvate dehydrogenase (7,8). FH inactivation should mimic this state, as it results in increased expression of HIF-1a and up-regulation of HIF-dependent glycolysis (13,35). This pseudohypoxic drive is similar to that described in VHL-deficient cells as a result of the HIF accumulation in the absence of the appropriate ubiquitina-tion process that targets HIF. The switch to the fermentative glycolysis should therefore be a common feature in tumors with VHL, FH, and succinate dehydrogenase deficiency, resulting from HIF-1a stabilization, and we would expect to see increased LDH-A expression in these tumors. Indeed, we have shown this to be the case in two of these tumor types (VHL-and FH-deficient tumors).
The functional consequences of inhibiting LDH-A have been explored both in vitro and in vivo. In the background of FH deficiency, LDH-A knockdown cells proliferate slower, undergo apoptosis, and are less invasive. The underlying basis for these effects remains to be fully defined. FH/LDH-A-deficient cells did show f50% to 60% reduction in intracellular ATP. We were surprised that the levels were not more profoundly depressed because both fermentative glycolysis and tricarboxylic acid cycle were expected to be compromised in FH/LDH-A cells. It is possible that intracellular ATP levels are being sustained by leakiness in our surrogate system, by LDH-B overactivity or expression compensating for LDH-A inhibition, or by diminished ATP consumption. We are currently exploring whether inhibition of both isoforms of LDH may result in significant ATP depletion. This might be technically challenging, as it will involve a triple knockdown to create such a cell line.
A correlation among apoptosis, ROS, and mitochondrial respiration has been reported in several cancers (36). Studies have suggested that HIF-1a-dependent mitochondrial repression may provide survival benefit by decreasing the risk associated with apoptotic cell death (37). Moreover, reduced ROS levels and decreased apoptosis result from forced expression of HIF-1a in oral squamous cell carcinoma (38). We therefore asked if increased ROS production was mediating apoptosis. This appears to be  the case, at least partially, because N-acetylcysteine was able to decrease PARP cleavage. We also found that oxidative phosphorylation was enhanced in FH/LDHdeficient cells as evidenced by increased oxygen consumption. One effect of this might be to increase electron flow in oxidative phosphorylation and thereby account for increased ROS production. However, in preliminary observations, we did not find increased superoxide formation as measured by the Mitosox reagent (data not shown), so the source of increased ROS production remains to be defined. It is also possible that various antioxidant systems in these cells have also been affected accounting for increased ROS.
Another effect of the increased rate of oxygen consumption observed in FH/LDH-A-deficient cells could be to maintain the NADH/NAD + ratio. Because LDH-A catalyzes regeneration of NAD + from NADH, we expected to observe an increased NADH/NAD + ratio in FH/LDH-deficient cells. The competitive phenomenon between mitochondrial NADH/NAD + transport system and LDH-A for NADH consumption is driven by LDH-A as a result of its fast enzymatic kinetic properties and is thought to be a potential factor for decreased mitochondrial respiration (39). Our failure to note an increase in the NADH/NAD + ratio may be partly due to a decrease in the LDH-A-driven competitive reaction that drives more NADH into mitochondria.
Because hypoxia regulates transcription factors HIF-1 and HIF-2, and in general it is difficult to design inhibitors targeting transcription factors, we have investigated whether targeting downstream targets of HIF-1 and HIF-2 may be therapeutically advantageous. Indeed, VEGF, a HIF-1a-induced gene product, has been successfully targeted for renal cell carcinoma therapy and our in vitro data showing enhanced VEGF expression in these cells supports the use of VEGF inhibitors in HLRCC. Our in vitro and in vivo data support the notion that targeting LDH-A, another HIF-1a target, may be a viable strategy for treating this disease as well. Moreover, high lactate levels are associated with poor prognosis in advanced renal cell carcinoma (40) and head and neck cancer (41). It is likely that inhibition of LDH-A by increasing the extracellular pH (42,43) may reduce the metastatic ability of these advanced cancer cells as extrapolated from our in vitro invasion assay results. Also, HLRCC patients may benefit from combined antiangiogenic and LDH-A inhibitor therapy as might patients with any tumors that have up-regulated HIF.
Our findings add to a growing literature that suggest that metabolism plays a key role in tumorigenesis and is linked either directly or indirectly to hallmarks of cancer that are involved in initiation and proliferation of tumors. In conclusion, our data support the hypothesis that inhibition of fermentative glycolysis might serve as a therapeutic strategy for HLRCC. Therefore, the development of inhibitors of LDH-A makes eminent sense for the treatment of HLRCC and a potentially large group of cancers in which the Warburg effect is operational, perhaps as many as 60% to 80% of tumor types.

Disclosure of Potential Conflicts of Interest
No potential conflicts of interest were disclosed.