Restricting Tumor Lactate Metabolism using Dichloroacetate Improves T Cell Functions

Hosein Rostamian Tehran University of Medical Sciences Mohammad Khakpoor-Kooshe Tehran University of Medical Sciences Leila Jafarzadeh Tehran University of Medical Sciences Elham Masoumi Ilam University of Medical Sciences Keyvan Fallah-Mehrjardi Tehran University of Medical Sciences Mohammad Javad Tavassolifar Tehran University of Medical Sciences John M. Pawelek Yale School of Medicine Hamid Reza Mirzaei Tehran University of Medical Sciences Jamshid Hadjati (  hajatij@tums.ac.ir ) Tehran University of Medical Sciences


Introduction
Cancer immunotherapy through adoptive cellular therapy (ACT) and its derivative, chimeric antigen receptor (CAR) T cells, has shown clinical effectiveness for hematological malignancies and immunogenic tumors such as melanoma but the e cacy of ACT for solid tumors to date has been limited and challenges still remain. This may in part be due to the microenvironment immunosuppressive nature of tumor cells [1][2][3]. Metabolites produced by the cancer cells can inhibit T cells in the tumor microenvironment, e.g. high concentrations of lactate and extracellular acidosis are typical features of tumors [4,5].
Lactic acid accumulation in tumors is a by-product of hypoxia which occurs when tumors switch to an anaerobic metabolism. Also in the presence of oxygen some tumors undergo glycolysis, a phenomenon called the Warburg Effect [6]. Acidi cation by lactic acid promotes angiogenesis, immunosuppression and metastasis all of which are associated with poorer clinical outcome [7]. Lactic acid produced by highly glycolytic tumors has been shown to overcome immune surveillance by suppressing activation of NK and in ltrating T cells and inhibiting the proliferation and cytokine production of T lymphocytes [8,9].
Some studies have shown that lactate has a relation with the reactive oxygen species (ROS) system.
Lactate causes an increase in ROS production [10]. High levels of ROS can be cytotoxic agents because of their capability of destroying DNA and other subcellular structures. On the other hand, ROS were shown to be crucial second messengers for signaling of T cell receptor and T cell activation [11].
Strategies to impairment of Tumor cell glycolysis could improve immunosurveillance and tumor growth regulation [12]. and manipulation of the enzymes involved in tumor cell glycolysis might be a way to overcome immunosuppression [13]. Pyruvate Dehydrogenase Kinase (PDK) is a gatekeeper enzyme regulating metabolism of glucose in tumors. PDK inactivates the pyruvate dehydrogenase complex (PDC) through its phosphorylation. PDC converts pyruvate to acetyl-CoA, which is further metabolized in the mitochondria. Overexpression of PDK has been reported in several tumors and is associated with invasion, metastasis and chemotherapy drug resistance. High PDK expression contributes to a change in glucose metabolism towards glycolysis rather than oxidative phosphorylation [14]. Thus, PDK inhibition with the drug dichloroacetate (DCA) changes the cancer cell metabolism from glycolysis towards mitochondrial glucose oxidation and as a result reduces lactate levels [15].
In this study, we show that lactate can dampen T cell function through suppression of T cell proliferation, cytokine production, and TCR signaling. Lactate also suppressed the redox system of T cells and reduced production of both oxidant and antioxidant molecules. Our studies open new avenues to manipulate the metabolism of tumor cells by limiting tumor-derived lactate. Here, we tested the hypothesis of whether T cell function could be enhanced by pharmacological targeting of tumor glycolysis. DCA decreased the concentration of tumor lactate by suppression glucose metabolism of tumor cells leading to improvement of T cell function. T cell proliferation and cytokine production were increased in an in vitro co-culture test by pre-treating lymphoma cells with DCA. DCA also rescued the T cells from apoptosis. Therefore, DCA can overcome immunosuppression of lactate in the tumor microenvironment and could be useful for adoptive T cell immunotherapy.

Cell culture and media
The Raji cell line was acquired from the Iranian Biological Resource Center (IBRC). Cells were cultured In RPMI 1640 (Gibco, USA) supplemented with 10% Fetal bovine serum (FBS)(Gibco) and 1% penicillin/streptomycin (Sigma-Aldrich, USA) and incubated at 37°C in 5% CO 2 . FBS was heat-inactivated for 30 minutes at 56°C before use. CD19 expression on Raji cells was analyzed by ow cytometry utilizing APC-conjugated anti-human CD19 antibodies (Miltenyi Biotec, Germany) prior to the experiments.

Lactate production in tumor cell media
Raji cells were seeded into 48-well plates at 2×105 cells per well and They were treated with DCA (Alfa Aesar, USA) at 0.5 mM, 1 mM, 2 mM, 5 mM and 10 mM for 24 and 48h. To measure lactate production the cell supernatants were harvested and lactate was measured using a colorimetric and lactate assay kit (Greiner Diagnostic GmbH, Germany).

T cell proliferation and cytokine assay
Raji cells were treated with mitomycin C (25 µg/ml) (Sigma-Aldrich) for 30 minutes to prevent proliferation. Then mitomycin treated Raji cells (2×105 cells/well) were cultured in 24-well plates in the presence or absence of DCA (1 mM) for 24 h. The cell culture media containing DCA was removed and replenished with fresh RPMI 1640 complete media and they were co-cultured with CFSE-labeled T cells for 72h. In order to label T cells with Carboxy uorescein succinimidyl ester (CFSE) (Life Technologies, USA), the cells (1x10 6 cells/well) were treated with 2.5 µM CFSE dye, and then 4 ml of FBS was applied to quench the reaction. CFSE-labeled T cells were co-cultured with Raji cells at 1:1 ratios (2×10 5 cells/well) in RPMI1640 complete media without the presence of hIL-2 and anti-CD3/CD28 antibodies to induce the unspeci c proliferation of T cells. T cells (2×10 5 cells/well) also cultured without Raji cells in RPMI1640 complete media with and without lactic acid (20 mM) (Sigma-Aldrich) as a negative and positive control of lactic acid. T cells (2×10 5 cells/well) were also cultured without anti-CD3/CD28 antibodies as unstimulated T cells. Cells were stained with anti-CD3-APC (BioLegend) and after 72 hr, and CFSE dilution of CD3-gated lymphocytes was calculated by ow cytometry to evaluate their proliferation. T cell proliferation was determined based on the difference between the mean uorescence intensity of stimulated and unstimulated cells. To measure the amount secreted IL-2 and IFN-γ by T cells, the supernatant was harvested 24 hours after co-culture and assessed using ELISA.

Apoptosis assay, Annexin V staining
Raji cells (2×10 5 cells/well) were cultured in 24-well plates in the presence and absence of DCA (1 mM) for 24 h. Then the cell culture media containing DCA was removed and DCA-treated cells were co-cultured with T cells at 1:1 ratios (2×10 5 cells/well) for 48h. T cells (2×10 5 cells/well) were also cultured in the absence of Raji cells in RPMI1640 complete media with and without lactic acid (30 mM) as negative and positive controls of apoptosis. The apoptotic cells were measured using the Annexin V-FITC ( uorescein isothiocyanate)/PI (propidium iodide) apoptosis detection kit (MBR, Iran). Brie y, after 48h of incubation, the cells were harvested and centrifuged at 1500 rpm for 10 ' . Cells were collected and counted and Annexin-V-FITC/PI labeling was carried out according to the instructions of the manufacturer (MBR, Iran).
2.6. Ros assay ROS and superoxide detection assay kits (ab139476, USA) were used to evaluate the production level of intracellular ROS. T cells (2×10 5 cells/well) cultured in RPMI1640 complete media with and without lactic acid (20 mM). After harvesting and washing, the cells were incubated with permeable green probe (reacts with hydroxyl radicals (HO), hydrogen peroxide, peroxynitrite (ONOO ), peroxyradical (ROO) and nitric oxide (NO)) and orange probe (in particular reacts with superoxide (O2 )) at 37°C for 30 ' . The level of the antioxidant molecule GSH was measured with a GSH assay kit (ab112132, USA). After harvesting and washing, the cells were incubated at 24°C with thiol green dye for 20 minutes. Finally, the cells were analyzed by ow cytometry. Based on the difference between the mean uorescence intensity of lactic acid treated and untreated cells, the production of ROS/superoxide and GSH was determined.

Real-Time PCR
T cells (2×10 5 cells/well) were cultured in RPMI1640 complete media with and without 20 mM lactic acid. According to the manufacturer's instruction, total RNA was extracted from T cells using RNX-plus solution (RN7713C, Sinaclon, Iran) to determine levels of gene expression of NADPH oxidase subunit, gp91phox, and antioxidant enzymes including SOD1, SOD2, Nrf2, and CAT. Purity and concentration of RNA concentration were assessed using NanoDrop (Thermo Fisher). To eliminate genomic DNA the isolated RNA was treated with DNase I (Fermentas, USA). cDNA was then synthesized by a cDNA synthesis kit (Thermo Fisher Scienti c, USA). Real-time PCR was performed using 2x SYBR Green qPCR Mix plus (ROX) (Ampliqone, Denmark) on an ABI step one plus real-time PCR system (Applied Biosystem). Relative expression levels of these genes were normalized by 18s rRNA as a housekeeping gene and calculated by the 2 − ΔΔCt method. The primers sequences are listed in the supplementary data.

Flow cytometric analysis
All samples were acquired and analyzed on a BD FACS Calibur (BD Biosciences, USA) with FlowJo software (v7.6.1). All experiments were conducted in triplicate and repeated three times.

Statistical analysis
Statistical analysis was performed using Prism 7 software. Comparisons between treatment groups were conducted by independent t-test and ANOVA with Tukey's post hoc test. Differences were accepted statistically signi cant when P < 0.05.

Proliferation and cytokine secretion was suppressed by lactic acid
We investigated the effect of lactic acid on various properties of T cells that have a role in T cell activation. A crucial characteristic of a successful immune response is T cell proliferation and cytokine secretion. We hypothesized that lactate has an immunosuppressive impact on the proliferation of T cells.
To determine the effects of lactic acid we compared the suppressive impact of media containing lactic acid and media without lactic acid on the proliferation of stimulated human T cells [ gure 1]. The lactic acid concentration was set to 20 mM, which matches the lactate concentration experienced by T cells in previously published studies in human and murine tumors [16][17][18]. Representative ow cytometry data of the CFSE staining are presented in Fig. 1B. As expected, lactic acid treated T cells exhibited signi cantly lower rates of proliferation compared to the control group [ gure 1C].
Next, we explored the effects of lactic acid on the production of T cell cytokines. T cells were stimulated with anti-CD3 and anti-CD28 antibodies and the level of IL-2 and IFN-γ was assessed by ELISA in the supernatants of cell culture after 24h [ gures 1D and 1E]. Reductions in IL-2 and IFN-γ secretion were observed in T cells cultured with lactate. Finally, our data indicated that lactic acid inhibits the activation of T cells by limiting their proliferation and cytokine production.

Lactic acid decreased ROS, Superoxide and Glutathione in T Cells
We next asked how lactic acid may affect the cellular redox state of T cells as a factor responsible for

Gene Expression of Antioxidant Molecules decreased in T cells
We next studied the effects of lactate on T cell oxidative stress molecules and the balance between oxidants (NOX-gp91phox) and antioxidants (SOD1, SOD2, Nrf2, and CAT). Gene expression patterns in lactate treated T cells were also examined. There was a signi cant decrease in the expression of gp91phox in T cells cultured in lactic acid compared to that the control group [ gure 2F]. A similar trend was observed in gene expression of the antioxidant molecules SOD1, SOD2, Nrf2, and CAT and they were signi cantly reduced in lactate-treated cells compared to control T cells [ gure 2F].

DCA reduced lactate production of tumor cells
High-lactate and low-glucose environments, such as seen in the tumor microenvironment, are immunosuppressive-particularly in the case of effector T cells. To manipulate tumor metabolic conditions and to overcome tumor immunosuppressive effects we used DCA to target the production of lactate in tumor cells. To assess the effects of DCA on lactate secretion of tumor cells we treated Raji cells with various concentrations of DCA and measured lactate concentrations in the supernatants of tumor cells after 24h and 48h. Two time-points were chosen because we speculated it takes time to observe DCA effects on the tumor cells. Our data shows DCA-treated tumor cells signi cantly decreased lactate production of lymphoma cells in comparison to untreated cells [ gure 3]. We also found that DCA had suppressive effects within 24 hours and treating cells for 48h is not required, in addition, higher dosages of DCA increased the suppressive effects. Doses higher than 2mM showed no further reduction in lactate levels as they were toxic to the cells (unpublished). We thus chose a dose of 1mM DCA to continue our study. At this dose the tumor cells appeared healthy and their ability to produce lactate was reduced, supporting our conclusion that DCA inhibits lactate production by tumor cells.

Treating tumor cells with DCA improves the proliferation capacity of T cells
To examine the function of the T cells co-cultured with tumor cells with regard to reduce lactate production we explored the proliferative response of T cells after they were exposed to lactate-inhibited tumor cells. T cells were stimulated with anti-CD3 and CD28 antibodies, labeled with CFSE dye and then co-cultured with equivalent numbers of DCA treated and untreated Raji cells for 72h. T cells were also cultured in the presence and absence of lactic acid (20 mM) as a positive and negative control for lactate [ gure 4]. Representative ow cytometry data of the CFSE staining are presented in Figs. 4B and 4C. As expected DCA signi cantly decreased the suppressive effect of lactate secreted by lymphoma cells on the T cell proliferative responses [ gure 4D]. However the DCA-treated group did not proliferate as well as those cultured in plain media but the rate of proliferation by the DCA-treated group was increased by 25% compared to the DCA-negative group. Raji cells were also treated with mitomycin C for 30 ' to prevent proliferation. We were concerned whether mitomycin C could change the glucose metabolism of tumor cells and dampen their lactate secretion. To address this we compared the concentrations of lactate in the supernatants of mitomycin C-treated and untreated-Raji cells. The data displayed no difference in the lactate production between the two groups [supplementary data 2].

Increased T cell cytokine secretion by inhibition of lymphoma cells lactate production
Different groups of T cells were evaluated for cytokine production. ELISAs were used to examine IL-2 and IFN-γ concentrations, main cytokines are employed for cell therapy. We examined cytokine concentrations in the supernatants of T cells co-cultured with DCA-treated Raji cells [ gure 5]. DCA treatment led to enhanced production of IFN-γ, and IL-2 by T cells compared to the DCA untreated group [ gures 5A and 5B]. Consistent with our previous ndings DCA could not rescue cytokine production as well as the T cell media-only group. Our results showed that IFN-γ secretion is more vulnerable to lactate compared to IL-2 [ gure 5C].
Lactate secreted by tumor cells in the DCA-negative group suppressed IFN-γ secretion nearly 3 times more than IL-2 and DCA can nearly doubles the IFN-g secretion and decrease the IL-2/IFN-g ratio to 1.5.

DCA rescued T cells from lactate
For evaluation of apoptosis induced by tumor-derived lactate we explored the exposure of cell surface phosphatidylserine using Annexin V/PI double staining [ gure 6]. Representative results of ow cytometry data of the Annexin V/PI staining are presented in Fig. 6B. T cells cultured with lactic acid (30mM) served as a positive control for apoptosis because this concentration of lactate was cytotoxic for T cells [unpublished]. Flow cytometry analyses revealed a signi cant positive effect for DCA on T cell viability. The percentage of viable T cells in the DCA-treated group was nearly twice that in the DCA-untreated group [ gure 6C]. The Annexin V/PI assay indicated that lactate induced late-stage apoptosis in T cells and the difference of late apoptotic and necrotic cells was not signi cant between the groups.

Discussion
A positive correlation has been reported between high levels of lactate and tumor progression in a variety of tumors [18,19]. The relationship between lactate levels in the tumor microenvironment and T cell activation is a new concept in this context [20]. A signi cant decline in CTL cytolytic activity was observed in the low PH of the Tumor microenvironment in the tumor-bearing mice [21,22]. In our study, we found that lactic acid decreased the function of T cells in vitro and it has an immunosuppressive impact on the proliferation of T cells. Prior studies have shown the inhibiting effect of lactate on effector T cell proliferation [23]. One of the known mechanisms that tumor cells utilize to limit T cell proliferation is lactate elevation which results in the blocking of glyceraldehyde 3-phosphate dehydrogenase (GAPDH) and 3-phosphoglycerate dehydrogenase, leading to the depletion of subsequent glycolytic intermediates including the 3-phosphoglycerate derivative serine which is known to be an essential factor for proliferation of T cells [16]. We have observed that lactic acid could also impair the IL-2 and IFN-γ secretion of T cells. Consistent with previous ndings it has been shown that lactic acid reduces IL-2, IFNγ, and granzyme B expression in human T cells [24]. Brand et al. proposed that that intracellular acidi cation restricts NFAT regulation, a signi cant transcription factor involved in IFN-γ transcriptional control. Besides, acidi cation can also disrupt the translocation of NFAT to the nucleus [20,25].
We assessed oxidative stress in T cells because it was previously shown that ROS perform as messengers for T cell receptor signaling in the steady-state and upon antigen recognition. Therefore, ROS play a critical role in T cell activation [26][27][28]. Here we investigated the production of ROS, superoxide and intracellular levels of GSH in T cells treated with lactate. We also examined gene expression of NOX-gp91phox as an oxidant molecule and SOD1, SOD2, Nrf2 and CAT as antioxidants. Since under normal conditions the levels of endogenous ROS are tightly regulated by different antioxidant systems inside the cell [29]. Signi cantly lower production of both oxidants and antioxidants was seen in the lactate-treated T cells. The levels of gene expression paralleled this. Our observations were not supported in a recent report showing a rapid and striking elevation of intracellular ROS which was caused by the exposure of activated CD4 + T cells to lactate [30]. However in that report the levels of ROS in T cells were measured in the presence of 10 mM of sodium lactate at three-time points. They assessed ROS at 5, 10 and 30 ' ' after the exposure of T cells to lactate. The levels of the ROS showed a downward trend from the rst-time point to the third-time point. We investigated the amount of oxidant and antioxidant molecules after culturing T cells for 24h in the presence of 20mM of lactate. The duration of our test and lactate concentrations were different from those in the recent study [30]. The mechanism(s) through which lactate interrupts the redox system of T cells remains for future research.
A promising therapeutic strategy is to target the glycolysis pathway of tumor cells as the impairment of glucose metabolism could cause defects in tumor cells growth and survival [31,32] It further decreases their lactate secretion and acidi cation of the tumor microenvironment that impairs the T and NK cells' anti-tumor immune responses [20,22,33]. Consequently, reducing the amounts of intratumoral lactate and acidi cation improves immunosurveillance potentially the effectiveness of cancer immunotherapies [12,[34][35][36].
In recent years DCA which already is used for the treatment of lactic acidosis has been considered as an anticancer agent [37,38]. DCA targets cancer cells and inhibits pyruvate dehydrogenase kinase, the inhibitor of pyruvate dehydrogenase. Therefore, DCA alters the metabolism of tumors from glycolysis towards oxidative phosphorylation [39]. Activation of PDH induces pyruvate mitochondrial oxidation and limits the metabolic advantage of tumor cells. Besides, DCA could prevent acidosis in the tumor microenvironment by decreasing lactate secretion and thus leading to inhibition of tumor growth [40,41]. The direct effects of DCA on cancer cells have been tested in most studies to date but here we have focused on evaluating the effects of DCA on tumor-derived lactic acid and its impact on T cells. Our results indicate that DCA can restore the T cell proliferative response and cytokine production from the suppressive effect of tumor-derived lactic acid. DCA also reduced apoptosis in T cells and preserved their viability. These data are in the line with the previous study in which diclofenac promoted anti-tumor response of T cell by reprogramming tumor glycolysis and inhibiting their lactic acid production [36]. Activation, viability, and effector functions of T cells were maintained in vitro following diclofenac treatment. They also showed that treatment of tumor cells with diclofenac caused an increase the in vitro anti-PD-1-mediated T cell killing of tumor cells. Diclofenac also enhanced the response to the anti-PD-1 blockade in tumor-bearing mice [36], as there was a negative correlation between response to anti-PD-1 therapy and metabolic genes overexpression [42]. To better understand the impact of DCA on checkpoint therapy we suggest further studies on using the combination of DCA and immune checkpoint inhibitors to treat tumors are warranted.

Conclusion
Taken together our results suggest that the accumulation of lactic acid in the tumor microenvironment restricts T cell responses and could interfere with the success of T cell therapy. For this reason, blocking microenvironment acidi cation prior to immunotherapy could strengthen the anti-tumor responses. It has been shown that DCA supports anti-tumor responses of T cells by metabolic reprogramming of tumor cells. DCA reduced the lactate production of tumor cells and preserved T cell activation. Tumor metabolic alteration illustrates a promising strategy to develop novel immunotherapies or improve the existing ones.
Our future research will be to evaluate DCA therapy in combination with adoptive cellular therapy beginning with murine models in vivo.

Declarations
Ethics approval and consent to participate All experimental protocols were approved by Tehran University of Medical Sciences Ethics committee. All methods were also carried out in accordance with relevant guidelines and regulations and Informed consent was obtained from all participants.

Not applicable
Availability of data and materials The datasets generated and analyzed during the current study are available from the corresponding author on reasonable request. Lactic acid inhibited T cell Proliferation and cytokine secretion. Human T cells were labeled with CFSE dye, stimulated for 72h with anti-CD3ε/CD28 antibody, and then cultured in media with or without lactic acid (20mM). After 24h the supernatants were collected and assessed for cytokine production by ELISA. Proliferation was analyzed by CFSE dilution through ow cytometry.