Correspondence to: Norimitsu Inoue, Department of Molecular Genetics, Osaka Medical Center for Cancer and Cardiovascular Diseases, 1-3-3 Nakamichi, Higashinari-ku, Osaka 537-8511, Japan, Tel.: 81-6-6972-1181, ext. 4101, Fax: +81-6-6973-5691, E-mail: firstname.lastname@example.org
The activation of oncogenic signaling pathways induces the reprogramming of glucose metabolism in tumor cells and increases lactic acid secretion into the tumor microenvironment. This is a well-known characteristic of tumor cells, termed the Warburg effect, and is a candidate target for antitumor therapy. Previous reports show that lactic acid secreted by tumor cells is a proinflammatory mediator that activates the IL-23/IL-17 pathway, thereby inducing inflammation, angiogenesis and tissue remodeling. Here, we show that lactic acid, or more specifically the acidification it causes, increases arginase I (ARG1) expression in macrophages to inhibit T-cell proliferation and activation. Accordingly, we hypothesized that counteraction of the immune effects by lactic acid might suppress tumor development. We show that dichloroacetate (DCA), an inhibitor of pyruvate dehydrogenase kinases, targets macrophages to suppress activation of the IL-23/IL-17 pathway and the expression of ARG1 by lactic acid. Furthermore, lactic acid-pretreated macrophages inhibited CD8+ T-cell proliferation, but CD8+ T-cell proliferation was restored when macrophages were pretreated with lactic acid and DCA. DCA treatment decreased ARG1 expression in tumor-infiltrating immune cells and increased the number of IFN-γ-producing CD8+ T cells and NK cells in tumor-bearing mouse spleen. Although DCA treatment alone did not suppress tumor growth, it increased antitumor immunotherapeutic activity of Poly(IC) in both CD8+ T cell- and NK cell-sensitive tumor models. Therefore, DCA acts not only on tumor cells to suppress glycolysis but also on immune cells to improve the immune status modulated by lactic acid and to increase the effectiveness of antitumor immunotherapy.
Many types of immune cells infiltrate tumors. Although these immune cells were classically thought to attack and eliminate tumors, recent studies indicate that they actually induce inflammation within tumors, thereby promoting tumor progression by inducing angiogenesis and tissue remodeling within the tumor microenvironment and tumor invasion and metastasis.[1, 2] Furthermore, immune cells such as tumor-associated macrophages (TAMs), myeloid-derived suppressor cells (MDSCs) and regulatory T cells (Tregs), all of which have potent suppressive effects on anticancer immune responses, are also recruited to tumors.[3, 4]
We previously showed that lactic acid secreted by tumor cells enhances the production of IL-23 by monocytes/macrophages stimulated with Toll-like receptor (TLR) ligands. Furthermore, lactic acid acts on monocytes/macrophages to increase antigen-dependent production of IL-17A by effector/memory CD4+ T cells, but suppresses differentiation from naïve to Th17 T cells and inhibits CD4+ T-cell proliferation. IL-23 is an inflammatory cytokine overproduced by tumors, which induces angiogenesis and reduces the infiltration of cytotoxic T lymphocytes into the tumor microenvironment to facilitate tumor growth.[7-9] Therefore, we propose that lactic acid functions as a proinflammatory mediator rather than just a terminal metabolite of anaerobic glycolysis. Recent reports show that lactic acid is an immunosuppressive factor that inhibits the function of dendritic cells (DCs) and cytotoxic T cells.[10, 11] Diclofenac is reported to downregulate lactic acid production and to counteract local immunosuppression by affecting intratumor DCs and Tregs. Lactic acid also acts on endothelial cells and fibroblasts to provide an environment that promotes tumor growth and motility.[13-17] Furthermore, many types of tumor cell produce large amounts of lactic acid; indeed, high concentrations of lactic acid are correlated with distant metastasis and poor prognosis of head and neck, cervical and colorectal cancers. Thus, increasing evidence suggests that lactic acid is a tumor-derived mediator that modulates the tumor microenvironment to promote tumor progression.[13-15]
Cells usually import glucose into the cytoplasm, where it is metabolized to pyruvate. Under aerobic conditions, pyruvate is converted to acetyl-CoA by the pyruvate dehydrogenase (PDH) complex in the mitochondria and metabolized to CO2, H2O and energy metabolites through the tricarboxylic acid cycle. Tumor cells show increased glucose uptake and lactate production even under normoxic conditions, resulting in a state known as the “Warburg effect.”[19, 20] This reprogramming of glucose metabolism is mediated by activation of canonical oncogenic signaling pathways, including the phosphatidylinositol 3-kinase-AKT pathway, c-Myc, hypoxia-inducible transcriptional factor-1 α, AMP-activated protein kinase and mammalian target of rapamycin.[21, 22] Therefore, glycolytic enzymes that are specifically and commonly activated in cancer cells are expected to be good targets for antitumor chemotherapy. We hypothesized that agents targeting glycolysis in tumor cells may also inhibit the activation of IL-23/IL-17 inflammatory pathway by lactic acid in immune cells to improve the immune status in tumor-bearing patients.
Therefore, this study examined the effects of antiglycolysis agents on immune responses in tumor-bearing mice. We show that dichloroacetate (DCA), which is an inhibitor of PDH kinases (PDKs), is one candidate drug that improves the immune status of tumor-bearing individuals, and enhances the effects of antitumor immunotherapy.
Material and Methods
C57BL/6J mice were purchased from CLEA Japan (Tokyo, Japan). OT-II and OT-I mice were kindly provided by Dr. W. R. Heath (The Walter and Eliza Hall Institute of Medical Research) and Dr. Hiroshi Kosaka (Osaka University), respectively. Ly5.1-positive congenic C57BL/6J mice were also kindly provided by Dr. Hiromitsu Nakauchi (University of Tokyo). All mice were maintained under specific pathogen-free conditions at the Osaka Medical Center animal facility. All animal experiments were performed in accordance with institutional guidelines and approved by the Animal Care and Use Committee of the Osaka Medical Center for Cancer and Cardiovascular Diseases.
Splenocytes and B16 melanoma cells were cultured in RPMI1640 supplemented with 10% heat-inactivated fetal bovine serum (FBS), 100 U/ml penicillin and 100 μg/ml streptomycin. E.G7-OVA (EG7) cells, which are ovalbumin-expressing EL4 thymoma cells, were cultured in RPMI1640 supplemented with 10% FBS, 10 mM Hepes (pH 7.3), 1 mM sodium pyruvate, 50 μM 2-mercaptoethanol, 100 U/ml penicillin and 100 μg/ml streptomycin. Cell cultures were incubated at 37°C under a 5% CO2 atmosphere.
l-Lactic acid, sodium DCA, sodium oxamate, α-cyano-4-hydroxycinnamic acid (αCHCA) and mitomycin C were purchased from Sigma-Aldrich (St Louis, MO), hydrochloric acid (HCl) and sodium lactate were from WAKO Pure Chemical (Osaka, Japan), OVA323–339 peptide (OT-II peptide) was from Bio Synthesis (Lewisville, TX), OVA257–264 peptide (OT-I peptide) was from MBL (Nagoya, Japan) and phorbol-12-myristate 13-acetate (PMA) and ionomycin were from Merck Millipore (Darmstadt, Germany). Bacillus Calmette–Guérin cell wall skeleton (BCG-CWS) was prepared as described previously.
Cytokine production analysis
OT-II splenocytes (5 × 105 cells) were stimulated with 400 ng/ml OT-II peptide in the presence or absence of 15 mM l-lactic acid for 4 days. Production of mouse IFN-γ or mouse IL-17A was analyzed with ELISA kits according to the manufacturer's instructions (R&D Systems, Minneapolis, MN).
Generation of bone marrow-derived macrophages
Bone marrow-derived macrophages (BMDMs) were induced as previously described. The harvested BMDMs (2 × 105 cells) were treated with 15 mM lactic acid in the presence of DCA for 12 hr (real-time reverse transcription [RT]-PCR) or 24 hr (arginase activity).
Real-time RT-PCR was performed as previously described. The following TaqMan probes and primer sets were used: Il10, Mm00439614_m1; Il12b, Mm99999067_m1; Il17a, Mm00439619_m1; Il23a, Mm00518984_m1; Arg1, Mm00475988_m1; Nos2, Mm00440502_m1; Vegfa, Mm01281449_m1 and 18S ribosomal RNA (rRNA), 4352930E. The relative expression of each cytokine gene was normalized to that of the 18S rRNA and calculated using the ΔΔCt method.
Arginase activity was measured as previously described. One unit of enzyme activity is defined as the amount of enzyme that catalyzes the formation of 1 μmol urea per minute (min).
BMDMs generated by treatment with M-CSF for 4 days were further treated with or without DCA in the presence of 15 mM lactic acid and M-CSF for 24 hr. The harvested BMDMs (2.5 × 105) were cocultured with OT-I splenocytes (5 × 105). For the carboxyfluorescein diacetate succinimidyl ester (CFSE) proliferation assay, splenocytes were labeled with 5 mM CFSE (Dojindo, Kumamoto, Japan) and stimulated with 400 ng/ml OT-I peptide for 3 days. The stimulated cells were stained with phycoerythrin (PE)-conjugated anti-mouse CD8α antibody (53-6.7; eBioscience, San Diego, CA) and the proliferation of CD8+ T cells was evaluated in terms of CFSE dilution by fluorescence-activated cell sorting (FACS) analysis.
Invivo tumor immunotherapy
The backs of C57BL/6 female mice were shaved and subcutaneously injected (on Day 0) with 200 μl of 3 × 106 EG7 or 6 × 105 B16 cells in phosphate-buffered saline (PBS). DCA was added to the drinking water (0.3 g/l) from Day 7. This dose of DCA was adjusted to yield a daily dose similar to that used clinically (clinical dose, 25–100 mg/kg/day).[30-32] Tumor volume was measured as previously described. Poly(I:C) (GE Healthcare, Little Chalfont, England) was intraperitoneally injected at a dose of 25 μg per mouse on Days 8, 11 and 14.
Measurement of lactate concentration in tumors
On Day 10, isolated tumors were homogenized in PBS (1 ml per 100 mg tumor) and centrifuged for 5 min at 1,500 rpm. The concentrations of lactate in the obtained supernatant were measured using a Determiner LA Kit (Kyowamedics, Tokyo, Japan).
Purification of tumor-infiltrating immune cells
Tumor-infiltrating immune cells were obtained from tumors formed by EG7 cells in Ly5.1-positive C57BL/6 mice or B16 cells in Ly5.2-positive C57BL/6 mice (Day 13 or 14). The tumors were chopped into small pieces using a razor and treated with 0.05 mg/l collagenase I, 0.05 mg/l collagenase IV (Life Technologies, Carlsbad, CA), 0.025 mg/l hyaluronidase and 0.01 mg/ml DNase I (Roche, Penzberg, Germany) in Hanks' balanced salt solution at 37°C for 10 min as described previously. The cell debris was removed by filtration with a 40-μm cell strainer (352340, BD biosciences, San Jose, CA). Immune cells were purified using a biotinylated anti-mouse CD45.1 antibody (for EG7) or an anti-mouse CD45.2 antibody (for B16) and anti-Biotin antibody MACS MicroBeads according to the manufacturer's instruction (Miltenyi Biotec, Gladbach, Germany).
Intracellular cytokine staining
On Days 15–18, splenocytes were isolated from EG7 tumor-bearing mice and the cells (1 × 107 cells) were cocultured with mitomycin C-treated EG7 cells (1 × 106 cells) in six-well plates. After coculture for 3 days, cells were stimulated with 50 ng/ml PMA and 750 ng/ml ionomycin in the presence of 1 μg/ml brefeldin A for 5 hr. Cells were stained with a PE-conjugated anti-mouse CD8α antibody, or an anti-mouse NK1.1 antibody (PK136; eBioscience), followed by fixation and permeabilization with the Cytofix/Cytoperm plus Kit (BD Biosciences) and intracellular staining with APC-labeled anti-IFN-γ antibody (XMG1.2; eBioscience). The fluorescence intensity of the cells was analyzed using a FACSCalibur system (BD Biosciences).
For measurements of cytokine levels, gene expression levels and tumor volumes, the data represent the mean values ± standard deviation (SD; for cytokines and gene expression) or the mean values ± standard error (SE; for tumor volumes). Statistical significant differences between groups were analyzed by the Student's t-test in all experiments. p < 0.05 was considered significant.
Lactic acid induces the expression of arginase I
We previously showed that, in a coculture consisting of OT-II CD4+ T cells and C57BL/6 CD11b+ cells or BMDMs, lactic acid enhances the OT-II peptide-dependent expression of IL-23p19 and IL-17A. Therefore, microarray gene expression analysis was performed to identify other genes induced by lactic acid in the coculture consisting of OT-II CD4+ T cells and CD11b+ cells. We found that lactic acid not only increased the expression of IL-23p19 (8.2-fold) and IL-17A (3.9-fold) but also increased the expression of arginase I (ARG1) (4.9-fold) in the presence of OT-II peptide. ARG1 is produced by myeloid cells, including TAMs and MDSCs, and meditates the inhibition of T-cell proliferation and activation.[3, 34, 35] To confirm the upregulation of ARG1, BMDMs were cocultured with OT-II CD4+ T cells in the presence of OT-II peptide and lactic acid, and the expression of ARG1 was measured. Lactic acid was used at 15 mM, which is the concentration that effectively enhances IL-23p19 expression. In addition to the increased expression of IL-23p19 and IL-17A, the expression of ARG1 was enhanced by lactic acid stimulation (Fig. 1a). Interestingly, the expression of ARG1 was enhanced by lactic acid even in the absence of the OT-II peptide. Therefore, we hypothesized that downregulation of lactate metabolism or lactate production might suppress the activation of the IL-23/IL-17 pathway, reduce the expression of ARG1 within the tumor microenvironment and improve immune status to induce more effective antitumor immune responses.
Suppression of glycolysis regulates lactic acid-dependent IL-17A and IFN-γ production
First, to examine the effect of several agents that regulate glucose metabolism or lactate secretion on cytokine production by immune cells, we investigated IL-17A and IFN-γ production by OT-II splenocytes stimulated with sodium oxamate (a lactate dehydrogenase inhibitor), αCHCA (a inhibitor of monocarboxylate transporters) or DCA, in the presence of OT-II peptide and lactic acid. All these agents reduced lactic acid-enhanced IL-17A production in a dose-dependent manner, but did not reduce IL-17A production in the absence of lactic acid (Fig. 1b). On the other hand, sodium oxamate and αCHCA also reduced IFN-γ production involved in antitumor immunity, whereas DCA did not influence IFN-γ production (Fig. 1c). These results suggest that DCA would be a suitable agent for the improvement of antitumor immunoreactivity.
DCA inhibits the activity of PDKs, which phosphorylate a subunit of PDH, PDH-E1α, at three serine residues (Ser232, Ser293 and Ser300), leading to the promotion of PDH activity. We confirmed that DCA inhibited the phosphorylation of the three serine residues in PDH-E1α in OT-II splenocytes. Lactic acid also remarkably suppressed the phosphorylation of PDH-E1α at Ser232, which was further decreased by treatment with DCA plus lactic acid (Supporting Information Fig. S1). In contrast, lactic acid treatment had only a minor effect on the phosphorylation of PDH-E1α at Ser293 and Ser300.
DCA suppresses the lactic acid-mediated expression of IL-23p19, IL-17A, ARG1 and inducible NO synthase in macrophages
TAMs and MDSCs are known to affect immune status in the tumor microenvironment. To analyze the effect of DCA on macrophages, we examined whether DCA suppresses the expression of IL-17A, IL-23p19 and ARG1 in a coculture consisting of OT-II CD4+ T cells and BMDMs (Fig. 2a). The lactic acid-mediated increase in IL-17A, IL-23p19 and ARG1 expression was suppressed by DCA in a dose-dependent manner (Figs. 2a and 2b). Next, in the absence of CD4+ T cells, we examined the effects of lactic acid and DCA on the expression of IL-23p19, ARG1 and other inflammatory genes (inducible NO synthase [iNOS], IL-12p40, VEGF-A and IL-10) in BMDMs. Lactic acid stimulation resulted in an increase in the expression of ARG1 (22.0-fold) and iNOS (29.7-fold) in BMDMs, which was significantly suppressed by DCA (Fig. 2c). The expression of both IL-12p40 (3.1-fold) and VEGF-A (3.2-fold) was increased significantly by lactic acid but the increases in the levels of these factors were lower than the increases in the levels of ARG1 and iNOS; however, the increase of IL-12p40 and VEGF-A expression by lactic acid was not suppressed by DCA (Fig. 2c). IL-10 expression was not affected by either lactic acid or DCA (Fig. 2c). DCA also suppressed the induction of IL-23p19 expression upon costimulation with BCG-CWS (a ligand for TLR2 and 4) and lactic acid (Fig. 2d). Furthermore, we evaluated arginase enzymatic activity in macrophages. The stimulation of macrophages with lactic acid resulted in an increase in arginase activity, which was suppressed by DCA (Fig. 2e). Because DCA promotes PDH activity, DCA may affect the concentration of lactic acid in the culture supernatant, leading to a decrease in ARG1 expression. Therefore, we measured the concentration of lactic acid in the culture supernatants after stimulation with lactic acid in the presence of DCA. The concentration of lactic acid in the culture supernatant was unaffected, regardless of the DCA concentration (Fig. 2f). These results suggest that DCA directly suppresses lactic acid-induced ARG1 expression without decreasing lactic acid levels in the surrounding environment.
DCA suppresses the increased expression of ARG1 induced by BCG-CWS but not by IL-4
TLR ligands and IL-4 induce the expression of ARG1 in macrophages. To analyze the effect of DCA on lactic acid-independent expression of ARG1, BMDMs were stimulated with BGG-CWS or IL-4 in the presence or absence of DCA. DCA suppressed the increased expression of ARG1 induced by BCG-CWS but not by IL-4 (Fig. 3a).
Acidification induces the increased expression of ARG1 but not IL-23p19
Our previous reports showed that lactate anions are cotransported with protons into cells via monocarboxylate transporters to induce higher expression of IL-23p19 within the cells, because only lactic acid but not sodium lactate or HCl increased IL-23p19 expression.[5, 6] We examined whether the transportable lactate anion was also responsible for the induced expression of ARG1 in BMDMs. The expression of ARG1 in BMDMs was increased by 10 mM HCl as well as by 15 mM lactic acid. The increased expression of ARG1 was suppressed by DCA (Fig. 3b). By contrast, HCl did not increase the expression of IL-23p19 in BMDMs stimulated with BCG-CWS (Fig. 3c). Sodium lactate did not increase the expression of either ARG1 or IL-23p19 in BMDMs (Figs. 3b and 3c). These results suggest that the ARG1 expression is induced by acidification, whereas IL-23p19 expression induced by the TLR ligand is increased by the transportable lactate anion.
CD8+ T-cell proliferation is suppressed by macrophages pretreated with lactic acid and increased by the addition of DCA
TAMs and MDSCs upregulate expression of ARG1, which leads to the suppression of T-cell growth and activation. We tested whether CD8+ T-cell proliferation is suppressed by macrophages pretreated with lactic acid or lactic acid plus DCA. After lactic acid or lactic acid plus DCA-pretreated BMDMs were washed with medium to remove lactic acid and DCA, the BMDMs were cocultured with CFSE-labeled OT-I splenocytes. Lactic acid-pretreated BMDMs were more effective in suppressing the proliferation of CD8+ T cells than untreated BMDMs (Fig. 4; mean fluorescence intensity [MFI]: 960 vs. 713, respectively). However, upon addition of DCA, lactic acid-pretreated BMDMs showed increased proliferation of CD8+ T cells. Furthermore, treatment of BMDMs with higher concentrations of DCA further increased the proliferation of CD8+ T cells compared to the proliferation of CD8+ T cells cocultured with untreated BMDMs (Fig. 4; MFI: 637 vs. 713, respectively). Similar results with lactic acid-pretreated BMDMs were observed in HCl-pretreated BMDMs (Supporting Information Fig. S2). Moreover, BMDMs pretreated with DCA alone increased the proliferation of CD8+ T cells compared to the proliferation of CD8+ T cells cocultured with untreated BMDMs (Supporting Information Fig. S2). These results suggest that it is the acidification caused by lactic acid that results in macrophages acquiring the ability to inhibit CD8+ T-cell proliferation, and that DCA treatment converts macrophages into cells that promote CD8+ T-cell proliferation.
DCA decreases ARG1 expression and arginase activity in tumor-infiltrating immune cells
To examine the effects of DCA on immune responses in tumor-bearing mice, EG7 or B16 cells were implanted into mice. DCA was then given orally via the drinking water (0.3 g/l) ad libitum from Day 7. In vitro, DCA effectively suppressed lactic acid production in both EG7 and B16 cells (Supporting Information Fig. S3). By contrast, in mice treated with DCA for 3 days, the concentration of lactic acid decreased significantly within EG7 tumors, but not within B16 tumors (Fig. 5a).
Tumor-infiltrating immune cells and splenocytes obtained from EG7 or B16 tumor-bearing mice were analyzed by FACS. The results showed that the populations of F4/80+CD11b+, Gr-1+CD11b+, CD11c+, CD4+, CD8+, NK1.1+, and CD4+Foxp3+ cells in DCA-treated and -untreated mice were the same (data not shown). We next analyzed gene expression in tumor-infiltrating immune cells. The immune cells infiltrating the EG7 (Fig. 5b) and B16 tumors (Fig. 5c) showed significantly reduced expression of ARG1 upon DCA treatment. Moreover, the enzymatic activity of arginase in B16 tumor-infiltrating immune cells was also significantly reduced by DCA treatment (Fig. 5d). However, the expression of iNOS and IL-17A, which was induced by lactic acid stimulation and significantly inhibited by DCA in vitro, was unaffected by DCA treatment in immune cells infiltrating either tumor (Figs. 5b and 5c). No expression of IL-23p19 was detected in these immune cells (data not shown).
Furthermore, we analyzed gene expression in particular subtypes of tumor-infiltrating immune cells obtained from EG7 tumors (F4/80+ cells and Gr-1+ cells). Both F4/80+ cells and Gr-1+ cells showed significantly reduced expression of ARG1 upon DCA treatment (Supporting Information Fig. S4).
DCA treatment enhances IFN-γ production
To evaluate the effects of DCA treatment on immune responses in vivo, we examined IFN-γ-producing immune cells in spleen isolated from EG7 tumor-bearing mice. The splenocytes were stimulated with mitomycin C-treated EG7 cells for 3 days in vitro. The number of IFN-γ producing CD8+ T cells and NK cells significantly increased in DCA-treated mice (Figs. 5e and 5f). When splenocytes from EG7 tumor-bearing mice were stimulated with Poly(I:C) (a ligand for TLR3) in vitro, the IFN-γ production of splenocytes derived from DCA-treated mice was significantly higher than that of splenocytes from DCA-untreated mice (Fig. 5g). These results show that DCA increases the capacity of effector cells to produce IFN-γ in tumor-bearing mice and enhances Poly(I:C)-mediated IFN-γ production.
DCA enhances the effect of antitumor immunotherapy with Poly(I:C)
We expected that DCA would enhance the antitumor effects of Poly(I:C) in vivo. Because Poly(I:C) treatment suppresses the growth of EG7 and B16 tumors in a CD8+ T cell- and NK cell-dependent manner, respectively, we treated EG7 or B16 cells implanted in mice with a low dose of Poly(I:C) (25 μg per mouse) in the presence and absence of DCA (0.3 g/l in drinking water). Treatment with DCA alone did not suppress tumor growth in either tumor model. Poly(I:C) treatment alone decreased tumor growth in the EG7 tumor model, although this was not statistically significant; however, Poly(I:C) plus DCA significantly suppressed tumor growth (Fig. 6a). In the B16 tumor model, Poly(I:C) plus DCA significantly suppressed tumor growth compared with Poly(I:C) alone (Fig. 6b). Therefore, DCA enhanced the antitumor activity of Poly(I:C) in both the EG7 and B16 tumor models. These results suggest that DCA enhances both CD8+ T cell- and NK cell-mediated antitumor immunity.
This study showed that tumor-secreted lactic acid promotes not only the IL-23/IL-17 inflammatory pathway but also ARG1 expression in macrophages. These effects were suppressed by DCA. Furthermore, DCA treatment suppressed arginase activity in tumor-infiltrating immune cells and enhanced the effects of antitumor immunotherapy with Poly(I:C) in vivo.
ARG1 is an enzyme that metabolizes l-arginine into l-ornithine and urea, and is highly expressed by tumor-associated myeloid cells, including TAMs and MDSCs. Depletion of l-arginine from the tumor microenvironment via ARG1-dependent consumption causes suppression of T-cell activation and proliferation.[3, 35] Accordingly, ARG1 plays an important role in tumor immune escape mechanisms. In this study, we show that lactic acid-pretreated macrophages have increased ARG1 expression and can suppress antigen-specific CD8+ T-cell proliferation. Previous reports have shown that lactic acid inhibits the differentiation of monocytes into DCs and decreases cytokine release from DCs, monocytes and cytotoxic T cells. Here, we identify a new immunosuppressive property of lactic acid: the induction of ARG1 expression in macrophages. We propose that lactic acid acts on myeloid cells to promote tumor development in two independent ways: suppression of T-cell activation and proliferation via an increase in the expression of ARG1, and induction of inflammation through activation of the IL-23/IL-17 pathway.
Interestingly, when macrophages were transiently stimulated by lactic acid, they suppressed CD8+ T-cell proliferation in the absence of lactic acid. This indicates that macrophages are converted into immunosuppressive cells by lactic acid. Therefore, macrophages exposed to high levels of lactic acid within the tumor microenvironment acquire immunosuppressive activity and suppress T-cell proliferation in both the tumor microenvironment and secondary lymphoid tissues.
DCA facilitates PDH activity, leading to a metabolic shift from glycolysis to glucose oxidation. DCA is used clinically to treat patients with lactic acidemia, including PDH complex deficiency and mitochondrial encephalomyopathy.[31, 32, 41, 42] The potential of DCA as an anticancer therapy has been the subject of many clinical and laboratory studies.[43-49] DCA decreases mitochondrial membrane potential, increases the levels of mitochondrial H2O2, and activates potassium channels, leading to inhibition of proliferation and induction of apoptosis in cancer cell lines. However, most reports to date have evaluated the direct effects of DCA on cancer cells. To our knowledge, this is the first report to evaluate the effect of DCA on immune cells in tumor-bearing mice.
In vitro, DCA suppressed the IL-23/IL-17 pathway and lactic acid-induced ARG1 expression in splenocytes and macrophages, but had no effect in the absence of lactic acid. Macrophages pretreated with lactic acid plus DCA increased antigen-specific CD8+ T-cell proliferation compared with macrophages pretreated with lactic acid alone. DCA rescued the immunosuppressive phenotype of macrophages treated with lactic acid. In vivo, DCA treatment decreased ARG1 expression in tumor-infiltrating immune cells and increased the number of IFN-γ-producing CD8+ T cells and NK cells in tumor-bearing mouse splenocytes. Splenocytes derived from DCA-treated mice also increased IFN-γ production induced by Poly(I:C) stimulation. Therefore, DCA treatment may cancel the immunosuppressive effects of lactic acid on tumor-associated myeloid cells and increase antitumor immunoreactivity. In fact, DCA treatment enhanced both the CD8+ T cell- and NK cell-dependent effects of antitumor immunotherapy with Poly(I:C). Interestingly, macrophages pretreated with lactic acid and a high concentration of DCA enhanced CD8+ T-cell proliferation to a much greater extent than untreated macrophages. Thus, DCA may enhance the activity of macrophages that induce CD8+ T-cell proliferation, in addition to counteracting the immunosuppressive effects of lactic acid. This study did not provide any evidence that DCA suppresses the activation of the IL-23/IL-17 pathway in vivo. Therefore, it will be necessary to assess the effect of DCA on the IL-23/IL-17 pathway using other tumor models.
Because tumor cells increase glycolysis and produce a large amount of lactic acid via the Warburg effect, DCA effectively suppressed lactic acid production in tumor cells in vitro, as reported previously.[43, 45, 48] In EG7-bearing mice, DCA also decreased the concentration of lactic acid within tumors in vivo, resulting in loss of the effects of lactic acid on myeloid cells. In contrast, DCA suppressed the effects of lactic acid on macrophages without affecting the level of lactic acid in the surrounding environment in vitro. In B16-bearing mice, DCA did not decrease the concentration of lactic acid within tumors in vivo, but decreased ARG1 expression in tumor-infiltrating immune cells. These results suggest that DCA synergistically suppresses the effects of lactic acid in vivo by two independent mechanisms. First, DCA decreases lactic acid levels in tumor cells, which thereby indirectly suppresses ARG1 expression in tumor-infiltrating myeloid cells. Second, DCA directly acts on myeloid cells to suppress ARG1 expression in these cells.
HCl and lactic acid promoted ARG1 expression and the inhibition of CD8+ cell proliferation by macrophages, suggesting that acidification and not the lactate anion itself is the causative factor. Moreover, DCA suppressed these effects of HCl and lactic acid. As tumor-secreted lactic acid predominantly contributes to acidification within tumors, it is suggested that ARG1 expression in tumor-infiltrating myeloid cells is mainly induced by lactic acid. A previous report indicated that both lactic acid and acidification are involved in the suppression of tumor necrosis factor (TNF) secretion by human monocytes. Acidification also reduces cytokine production such as that of TNF by mouse macrophage via a member of the ovarian cancer G-protein coupled receptor 1 family, T-cell death-associated gene 8. Thereby, not only the lactate anion but also acidification plays an important role in the determination of immune status in tumor-bearing hosts.
We have not yet clarified the molecular mechanisms by which lactic acid induces IL-23 and ARG1 expression in myeloid cells or how DCA suppresses these effects. Different mechanisms probably mediate lactic acid-induced IL-23 and ARG1 expression in myeloid cells as acidification increased the expression of ARG1 but not that of IL-23. DCA also suppressed ARG1 expression induced by lactic acid and the TLR ligand but not that induced by IL-4. A previous report showed that the TLR ligand and IL-4 activate different signaling pathways to induce ARG1 expression. Therefore, ARG1 expression induced by lactic acid and IL-4 could be induced via the activation of different signaling pathways. DCA suppressed the phosphorylation of PDH-E1α at all three serine residues, whereas the degree of phosphorylation induced by lactic acid was different at each residue. Although further studies will be necessary to elucidate how lactic acid regulates phosphorylation, this result suggests that lactic acid increases IL-23p19 and ARG1 expression without the need for direct activation of PDK, which is inhibited by DCA. It is possible that DCA might directly affect other targets involved in the lactic acid signaling pathway or inhibit the IL-23p19 and ARG1 expression independently of the lactic acid signaling pathway. Because anticancer therapies targeting lactic acid production and the lactic acid signaling pathway are now receiving increasing attention, it will be important to elucidate the molecular mechanisms underlying lactic acid signaling pathways.
The authors thank Dr. W. R. Heath (The Walter and Eliza Hall Institute of Medical Research), Dr. Kosaka (Osaka University) and Dr. Nakauchi (University of Tokyo) for providing OT-II, OT-I and Ly5.1-positive C57BL/6 mice, respectively. The authors also thank Toshiko Yasuda and Hiroko Murakami for technical assistance. This work was supported by Grant-in-Aid for Scientific Research (C) from Japan Society for the Promotion of Science (grant number 21590326, N.I.) and Grant-in-Aid for Cancer Research from the Ministry of Health, Labour and Welfare of Japan (N.I.). No potential conflicts of interest were disclosed.