Immunometabolism is emerging as a critical determinant of cancer pathophysiology. In this study, we explored the contributions of macrophage-expressed lactate dehydrogenase-A (LDH-A) to tumor formation in a K-Ras murine model of lung carcinoma. Myeloid-specific deletion of LDH-A promoted accumulation of macrophages with a CD86high and MCP-1high M1-like phenotype that suppressed tumor growth. This phenotypic effect was accompanied by reduced VEGF expression and angiogenesis, diminished numbers of PD-L1+ cancer cells, increased numbers of CD3+ T cells, and activation status of CD8+ T cells. Furthermore, it was associated with more pronounced antitumor T-cell immunity via induction of IL17 and IFNγ-producing CD8+ T (Tc17 and Tc1) cells, likely via suppression of lactate-driven PD-L1 expression. Our results suggest that expressions of LDH-A and lactate by macrophage in the tumor microenvironment are major drivers of T-cell immunosuppression, strongly supporting the concept of targeting stromal LDH-A as an effective strategy to blunt tumoral immune escape. Cancer Res; 77(13); 3632–43. ©2017 AACR.

The balance between Mø-dependent tumor promotion and elimination is likely dependent on the state of functional polarization of macrophage (Mø) populations within the tumor microenvironment (TME). Mø can exist in at least two states of activation: M1 skewed Mø function to promote cytotoxic inflammatory responses, while M2 polarized Mø support wound healing–like functions, suppress immune inflammatory responses, and stimulate expression of VEGF, which facilitates angiogenesis and tumor growth. The metabolic properties of the TME are considerably different from the surroundings of normal tissue, typically characterized by lower glucose and oxygen and increased acidic conditions (low pH; ref. 1). One of the characteristic changes in cellular physiology induced by transformation is described as the Warburg effect, where glucose metabolism within tumor cells amplifies anaerobic glycolysis and decreases oxidative phosphorylation, leading to a characteristic increase in lactate production (2). Hypoxia not only drives the expression of lactate dehydrogenase-A (LDH-A), but also other genes, such as VEGF, programmed death-ligand 1 (PD-L1; ref. 3), glucose transporter (GLUT1), and many glycolytic enzymes. In contrast, hypoxia blocks monocyte chemotactic protein-1 (MCP-1) expression (4). In addition, the adoption of this fermentative glycolytic state by cancer cells results in an increased lactate secretion into the TME (5), potentially influencing gene expression profiles in nontransformed cells populating the TME.

The objective of this study was to investigate the role of the metabolic enzyme, LDH-A, that catalyzes the conversion of pyruvate to lactate, in regulating the polarization status of Mø and effector responses of T cells in tumor evolution and homeostasis using the spontaneous K-Ras–driven carcinoma model. LDH-A is upregulated in non–small cell lung carcinoma as well as other cancers and is correlated with poor clinical outcome (6). Although generation of lactate as the end product of fermentative glycolysis has long been recognized as a hallmark of cancer cell metabolism, the physiologic consequences of increased expression of LDH-A and/or lactate accumulation in the TME have been relatively unexplored. The combination of increased lactate production and reduced tumor perfusion leads to acidification of the TME and tumor progression, resulting in an increasingly immunosuppressive milieu (7–9). Several studies have suggested that lactic acid plays an integral role in facilitating the growth and invasiveness of tumors via altering host immune system. Indeed, cancer cell–derived lactate or exogenously supplemented lactate leads to reprogramming of Mø toward M2 phenotype (10), contributing to the immunosuppressive TME. Emerging evidence indicates that changes in glucose metabolism and the pentose phosphate pathway influence Mø polarization (8, 10, 11). In particular, the M2-polarizing signal (IL4) promotes higher oxygen consumption and reduced flux of the glycolytic pathway as measured by oxygen consumption rate (OCR) and extracellular acidification rate, respectively. In contrast, M1-skewing stimuli, such as by lipopolysaccharide (LPS), result in decreased OCR and increased glycolysis (12). Strong activation of the immune system against tumor may accelerate M1 activation and metabolic switch to aerobic glycolysis (13). However, these studies did not address a role of endogenous LDH-A in Mø and its influence on the tumor immune stroma. Recent studies showed that lactate secreted by tumor cells activates IL23 in Mø, which increases the levels of IL17 secretion by Th17 T cells (14). IL17 has a dual effect on tumor growth by promoting cytotoxic T-cell responses but also facilitating angiogenesis (15). Th17 cells accumulate within the tumor stroma and activate VEGF production from fibroblasts to induce new blood vessel formation (16). However, Th17 cells in established tumors were found to be effective in eliminating cancer cells (17). Th17/Tc17 cells significantly suppressed the growth of established melanoma, whereas Tc1 cells induced long-term tumor regression through secretion of IFNγ (18).

In this study, we show for the first time that Mø-expressed LDH-A is critical for Mø skewing and Mø-dependent activation of CD8+ T cells in tumors. Mø-specific deletion of LDH-A in mice bearing K-Ras–driven tumors leads to smaller tumors and lower angiogenesis while reversing immunosuppression in the TME as reflected by lower levels of VEGF and PD-L1 expression, respectively.

Animals

All animals were held under specific pathogen-free conditions.

K-Ras model.

FVB/N-Tg(teto-K-Ras2)12Hev/J lung adenocarcinoma mice (K-Ras) and K-Ras:LDH-Aflfl-CreERTME mice were described previously (19). LysM-Cre mice were from The Jackson Laboratory. All experiments were approved by the IACUC at Beth Israel Deaconess Medical Center (Boston, MA). Tamoxifen treatment (10 mg/kg in corn oil, three consecutive injections, intraperitoneally) was used in K-Ras:LDH-Aflfl-CreERTME to delete LDH-A after the tumor was established.

LLC model.

C57/Bl6 mice were used for implantation of 0.5 × 106 Lewis lung carcinoma (LLC) cells into the left flank. On day 4, wild-type CD8+ T cells cocultured for 2 days with antigen-presenting cells (APC) isolated from LDH-A−/− or LDH-A+/+ mice in mixed leukocytes reaction (MLR) assay (see description below) were harvested and the MLR splenocyte cultures intravenously injected (2 × 106) into the tumor-bearing mice as described previously (20). Growth of tumors was monitored by caliper.

Cell culture

Primary bone marrow–derived macrophages (BMDM) were isolated, differentiated, and cultured as described previously (21). Briefly, bone marrow cells were isolated from the mouse femurs by flushing with RPMI medium (Thermo Fisher Scientific) supplemented with Antibiotic-Antimycotic (Life Technologies). Isolated cells were differentiated with mouse recombinant M-CSF (ProSpec) at a final concentration of 20 ng/mL in RPMI medium supplemented with 15% FCS (Atlanta Biologicals) plus antibiotics and antifungal agents for 5 days in M-CSF medium. Fresh M-CSF medium was added to cells at day 3 of culture. Following 5 days differentiation, macrophages were skewed toward M1-like or M2-like using: IFNγ (10 ng/mL) + LPS (100 ng/mL) or IL4 (10 ng/mL) for 3 days, respectively. Cells were washed once with PBS and used for subsequent experiments. LDH-A inhibitor (compound 1, GSK, 10 μmol/L; ref. 19) was applied at the time of M1/M2 skewing of BMDM. Sodium nitroprusside (Calbiochem) as a nitric oxide donor was used at 10 or 100 μmol/L for 24 hours. DMSO was used as a vehicle control. LLC cells and B16-F10 melanoma cells were purchased from ATCC (purchased in 2013) and were in stored in passage 3 (used in experiments in up to passage 10), tested for mycoplasma (MycoAlert Substrate Kit) in 2015, and cultured in DMEM medium (Life Technologies) supplemented with 10% FBS and antibiotics. RAW and RAW inos−/− macrophage cell lines were purchased from ATCC (>10 years ago) and were described previously (22). These cells were kept in liquid nitrogen at passage 3 to 4 and were not tested for mycoplasma. LLC and RAW cells were cocultured in the transwell system (0.4-μm pores) without direct contact. LLCs were seeded in the upper chamber, and their survival was measured at 48 hours after coculture with RAW cells by staining with crystal violet (23).

T-cell isolation and MLR

T cells were isolated from the spleens using MACS Cell Separation CD3e MicroBead Kit following the manufacturer's instructions (Miltenyi Biotec). CD3 cells were selected as a negative flow-through, and CD3+ cells were enriched on the column. CD3+ cells were stained with carboxyfluorescein diacetate succinimidyl ester (CFSE; Molecular Probes, Invitrogen) following the manufacturers' protocol. CD8+ T cells were isolated from the spleens using MACs CD8 MicroBead Kit (Miltenyi Biotec) following the manufacturer's instructions. The appropriate ratio of CD3+ or CD8+ cells and APC from LDH-Aflfl-CreERTME and LDH-Aflfl mice injected with tamoxifen 2 to 3 weeks before the experiment to induce deletion was 2:1. Cells were cultured in the presence of IL2 (50 μg/mL), anti-Dynabeads mouse T-Activator CD3/CD28 (Thermo Fisher Scientific; ratio: 1:5, beads:cells), and LPS (100 ng/mL) for 2 to 3 days.

Flow cytometry

After harvesting and washing with PBS, cells were stained with PE anti-mouse CD197 (clone 4B12, eBioscience), APC anti-mouse CD86 (clone GL-1, Biolegend), PE anti-MMR (anti-CD206; clone FAB2535P, R&D Systems), PE anti-PD-L1 (clone 10F.9G2, BioLegend), or IgG control antibody (BioLegend, R&D Systems) for 30 minutes at room temperature. After washing with 1× PBS, cells were immediately analyzed using a Caliber flow cytometer (Becton Dickinson). Percentage of gated cells was calculated and analyzed using CellQuest Pro software (Becton Dickinson).

MLR cultures were harvested and stained with APC anti-CD8 (BioLegend) and PE anti-CD4 (BioLegend). CFSE staining (Invitrogen, Molecular Probes) was performed in CD8+ or CD4+ T-cell subsets following the manufacturers' protocol at 2 to 3 days after coculture was started.

IHC and hematoxylin and eosin

Lung samples were formalin- or Zn fixed, followed by paraffin embedding and immunostaining of 5-μm sections as described previously (21). Hematoxylin and eosin (H&E) staining was performed as reported before (23). Sections were stained with the following antibodies: anti-MCP-1 (BD Biosciences), anti-CD86 (BD Biosciences), anti-CD3 (Epitomics), anti-PD-1 (clone 29F.1A12, BioLegend), anti-PD-L1 (clone 10F.9G2, BioLegend), anti-granzyme B (Abcam), and anti-VEGF (Thermo Fisher Scientific).

Immunofluorescence staining

For immunofluorescence staining, lungs were frozen in freezing medium and cut on the cryotome in 6-μm sections and dried. Tissue sections were then fixed with 2% PFA followed by permeabilization with 0.5% Triton X-100. Sections were then incubated for 30 minutes in a blocking buffer containing 7% horse serum (Vector Laboratories) in PBS. A primary antibody was then applied overnight at 4°C. Sections were then incubated with biotin-labeled secondary antibody (1.5 μg/mL in PBS; Vector Laboratories) or fluorescently labeled secondary antibodies for 1 hour at room temperature. The images were captured using a Fluorescence Microscope (Zeiss). The following antibodies were used: anti-F4.80 (BioLegend), anti-CD31 (BD Pharmingen, BD Biosciences), anti-iNOS (Santa Cruz Biotechnology), anti-CD69 (eBioscience), anti-IFNγ (BioLegend), anti-CD45 (eBioscience), anti-CK18 (Millipore), anti-epithelial antigen (Abcam), anti-CD11c (eBioscience), P-histone H3 (Cell Signaling Technology), anti-perforin (eBioscience), anti-CD8, and anti-IL17 (eBioscience).

Immunoblotting

Lysates were prepared in ice-cold lysis buffer [50 mmol/L Tris-HCl (pH 7.4), 50 mmol/L sodium fluoride, 150 mmol/L NaCl, 1% Nonident P40, 0.5 mol/L EDTA (pH 8.0)] and the protease inhibitor cocktail Complete Mini (Roche). Samples were centrifuged at 14,000 × g at 4°C for 20 minutes and supernatants were collected. Protein concentrations of supernatants were measured using a Bicinchoninic Acid Protein Assay Kit (Thermo Fisher Scientific). Twenty-five micrograms of each protein sample was then electrophoresed on NuPAGE 4% to 12% Bis-Tris Gel (Life Technologies) in NuPAGE MES SDS running buffer (Life Technologies) for 90 minutes at 100 V. The membranes were blocked with 5% nonfat dry milk in 1× TBS (Boston BioProducts) for 1 hour and then probed with primary antibodies (diluted at 1:1,000 in 1× TBS with 5 % nonfat milk) overnight at 4°C. Membranes were then washed in 1× TBS buffer and incubated with horseradish peroxidase–conjugated secondary antibodies at a dilution of 1:5,000 in 1× TBS with 5 % nonfat milk for 1 hour at room temperature. Bands were visualized using Super Signal West Pico chemiluminescent substrate (Thermo Fisher Scientific) or Femto Maximum Sensitivity Substrate (Thermo Fisher Scientific), followed by exposure to the autoradiography film (BioExpress). The following antibodies were used: anti-LDH-A and anti-HO-1 (Abcam), anti-iNOS (Santa Cruz Biotechnology), anti-HIF-1α (Santa Cruz Biotechnology), anti-VEGF (Thermo Fisher Scientific), and anti-β-actin (Sigma Aldrich).

Real-time PCR

RNA was isolated from MLR cocultures using RNeasy Mini Kit (Qiagen) following the manufacturer's protocol. Cell pellets were suspended in the RLT lysis buffer, and total RNA was purified on the columns and eluted in RNase-free water. One microgram of RNA was used for cDNA synthesis using iScript cDNA Synthesis Kit (Bio-Rad). Quantitative PCR was performed using iTaq SYBR Green Supermix with ROX (Bio-Rad) with the following primers: IL17: forward (F) 5′-CTCAGACTACCTCAACCGTTC-3′; reverse (R) 5′-TGAGCTTCCCAGATCACAAG-3′, IL10: F 5′-CCAAGCCTTATCGGAAATGA-3′; R 5′-TTTTCACAGGGGAGAAATCG-3′; TGFβ: F5′-AGAGGTCACCCGCGTGCTAA-3′; R 5′-TCCCGAATGTCTGACGTATTGA-3′; granzyme B: F 5′-CCTCCTGCTACTGCTGAC-3′; R 5′-GTCAGCACAAAGTCCTCTC-3′; perforin: F 5′-GAGAAGACCTATCAGGACCA-3′; R 5′-AGCCTGTGGTAAGCATG-3′; IFNγ: F 5′-TCAAGTGGCATAGATGTGGAAGAA-3′; R 5′-TGGCTCTGCAGGATTTTCATG-3′, Bax (24), p21 F 5′-AGCCTGAAGACTGTGATGGG-3′ R 5′-AAAGTTCCACCGTTCTCGG-3′, Bcl-XL F 5′-CAGAGCAACCGGGAGCTGGT-3′, R 5′-GGATCCAAGGCTCTAGGTGG-3′.

Luminex and ELISA

Luminex was performed at the Human Immune Monitoring Center at Stanford University (Stanford, CA). VEGF DuoSet ELISA Kits were purchased from R&D Systems and used as per the manufacturers' directions.

Statistical analysis

All data are presented as mean ± SD unless otherwise indicated. Statistical analysis was performed using the Student t test or one-way ANOVA, followed by the post hoc Tukey test, using GraphPad Prism and Excel software. Differences between groups were rated significant at values of P < 0.05.

Conditional deletion of LDH-A in Mø reduces the growth of K-Ras lung tumor growth

Conditional deletion of LDH-A in mice using a tamoxifen-inducible Cre in the lung cancer model led to lower tumor burden (19) and early regression of the tumor nodules. Importantly, reduction in tumor size in LDH-A–deleted K-Ras lung carcinomas was associated with higher levels of iNOS in the lung tissues, suggesting changes in the TME, such as macrophage skewing toward an M1 antitumor phenotype (Supplementary Fig. S1A and S1B). This prompted us to ask whether LDH-A played a role in promoting the M1 skewing of myeloid cell polarization in the TME. IHC demonstrated increased numbers of CD86+ myeloid cells (consistent with M1 Mø polarization) in the LDH-A–deleted K-Ras lung carcinomas (Fig. 1A and B). Furthermore, we observed higher levels of monocyte chemoattractant protein (MCP-1/CCL2) in these tumors (Fig. 1C and D). As both effects indicated the role of LDH-A in myeloid cells, we tested whether Mø-expressed, rather than tumor cell expressed, LDH-A is critical for this control of lung tumor growth. We therefore generated LysM-Cre:LDH-Aflfl:K-Ras mice to specifically direct deletion in Mø (Fig. 1E). When measuring tumor growth development, we found that K-Ras–mutant mice lacking Mø-expressed LDH-A exhibited a growth pattern similar (Fig. 1F and G) to that observed in K-Ras–mutant mice with global deletion of LDH-A (19). This strongly suggests a major contribution of LDH-A in myeloid cells to the regulation of tumor growth (Fig. 1F and G). As is the case with mice bearing global deletion of LDH-A, these mice also exhibited higher number of MCP-1–positive infiltrating myeloid cells in the lung tumors (Fig. 1H and I).

Deletion of LDH-A in Mø suppresses VEGF expression

A more in depth analysis performed on ex vivo bone marrow–derived Mø showed that Mø LDH-A deletion resulted in significant decreases in the expression of VEGF and IP-10 and in increased secretion of MCP-1 and G-CSF among cytokines tested by Luminex assay (Fig. 2A; Table 1). Importantly, VEGF was the only analyzed growth factor for which the lower expression found in the LDH-A−/− mice was rescued by lactate supplementation (Fig. 2B; Table 1). These results demonstrated that regulation of IL12p40, TNF, and MCP-3 was dependent on lactate, but not on LDH-A, while increase in levels of IL10, G-CSF, and MCP-1 as well as decreases in the levels of IL6, IP-10, and TGFβ upon deletion of LDH-A in M1-skewed macrophages were attributed to LDH-A presence but independent of lactate (Table 1). Furthermore, we confirmed lower levels of VEGF in the supernatants of M1 macrophages lacking LDH-A (Fig. 2B and C) and associated with high levels of iNOS in these cells (Supplementary Fig. S1C). We found that VEGF levels are induced in the supernatants after lactate supplementation of Mø from LDH-A+/+ or LDH-A−/− mice (Fig. 2B). Moreover, BMDMs isolated from mice lacking LDH-A in myeloid compartment (LDH-Aflfl:LysM-Cre) and M1-polarized generated lower levels of VEGF as compared with the controls (Fig. 2D). Similarly, low VEGF and HIF-1α levels and high expression of iNOS were detected in peritoneal macrophage and APCs upon deletion of LDH-A (Fig. 2E). iNOS expression in Mø however was not indispensable for direct inhibition of cancer cell growth (Supplementary Fig. S2A).

Consistent with our in vitro finding, K-Ras mice lacking Mø expression of LDH-A exhibited reduced VEGF (Fig. 2F and G) and CD31 staining (Fig. 2H and I) in their lungs. We confirmed a lower expression of VEGF and CD31+ staining indicative of low angiogenesis in K-Ras tumors with conditional deletion of LDH-A in Mø as compared with K-Ras tumors with wild-type LDH-A in Mø (Fig. 2J–L).

Deletion of LDH-A amplifies skewing of Mø toward M1 phenotype

LPS/IFNγ- (M1) or IL4 (M2)–skewed Mø express LDH-A with a slightly higher level in M1 Mø (Fig. 3A). To evaluate the role of LDH-A in myeloid cells skewing toward M1 or M2 phenotype, we assessed the levels of CD86/CD197 and the macrophage mannose receptor (MMR, aka CD206) in nonpolarized, M1 and M2 polarized cultures of BMDM from Cretm-LDH-Aflfl (LDH-A−/−), and control mice treated with tamoxifen. There were no differences detected in the levels of these markers in the nonpolarized in either LDH-A–deficient or replete Mø cultures (Fig. 3B and C; Supplementary Fig. S2B and S2C). Likewise, the LDH-A did not appear to affect the ability of IL4 to drive Mø differentiation into an M2 phenotype. However, deletion of LDH-A resulted in higher expression of the M1 markers CD86 and CD197 reflecting enhanced M1 skewing (Fig. 3B). Furthermore, pharmacologic inhibition of LDH-A in Mø by LDH-A inhibitor (compound 1, GSK; ref. 19) also increased the M1 skewing of treated Mø without affecting the development of the M2 phenotype, recapitulating the results obtained with LDH-A genetic depletion (Fig. 3D and E).

To further elaborate on the mechanism in the TME upon deletion of LDH-A, we evaluated iNOS, a marker of activated M1-skewed macrophages in K-Ras tumors as early as 3 weeks after deletion of LDH-A (Fig. 3F and G). This revealed a significant increase in infiltrating iNOS-positive macrophages in the tumor stroma in mice lacking LDH-A (Fig. 3F and G). Furthermore, heme oxygenase-1 (HO-1), a marker of alternatively activated M2 macrophages, was decreased in polarized BMDMs lacking LDH-A (Fig. 3H) and in K-Ras tumors with deletion of LDH-A (Supplementary Fig. S3A).

Deletion of LDH-A in Mø increases T-cell numbers in the immunocompetent TME by blocking PD-L1 expression

Macrophage activation can affect T-cell function, suggesting that the observed inhibition of tumor growth in LysM-Cre:LDH-Aflfl:K-Ras mice might be due to differences in intratumoral T-cell activity. Therefore, we evaluated the number of T cells and the subset of immunosuppressive T regulatory cells (Treg; Foxp3+) in the TME in K-Ras lung tumors with and without Mø-expressed LDH-A (Fig. 4A–C). Tumors from mice lacking LDH-A showed higher numbers of CD3+ T cells and a lower ratio of Foxp3+ to CD3+ T cells compared with the LDH-A–expressing K-Ras tumors (Fig. 4A). However, the numbers of Foxp3+ cells did not differ based on LDH-A expression, suggesting a specific increase in infiltrating effector T cells into the tumor bed (Supplementary Fig. S3B). These data along with the observation of lower TGFβ levels in Mø with deletion of LDH-A (Table 1) suggest reduced immunosuppression in the TME and subsequent infiltration of higher number of F4.80+ myeloid cells and CD8+ T cells producing IFNγ and granzyme B (Supplementary Fig. S4B–S4F). Notably, there was no difference in early activation marker CD69, perforin, or IL17 in the tissues (Supplementary Fig. S4A, S4D, and S4G), which might reflect the analysis of tumors at 13 weeks after deletion of LDH-A in myeloid cells.

As we have demonstrated that LDH-A may control Mø-dependent antitumoral effects of T cells in the TME in our models (Fig. 4), we further defined the levels of immunosuppressive markers PD-1 and PD-L1 in the TME. Interestingly, lack of LDH-A in Mø (LysM-Cre:LDH-Aflfl:K-Ras mice) led to increased number of PD-1+ cells (Fig. 4D), which indicate either activated or exhausted T cells (25, 26). This is in contrast to the virtual absence of the PD-L1 staining in the TME in K-Ras tumors lacking LDH-A as compared with PD-L1 positivity in K-Ras tumors from wild-type mice (Fig. 4D). This suggests lack of PD-L1/PD-1 engagement on T cells, and it might therefore constitute a signal for T-cell activation rather than exhaustion. We showed that PD-L1 is expressed predominantly on tumor cells as it overlapped with epithelial antigen as well as cytokeratin 18 (Fig. 4E). There was no PD-L1 expression in CD45 or CD11c+ cells (Fig. 4E). To elaborate on the mechanisms of LDH-A–mediated PD-L1 regulation in cancer cells, we used B16-F10 melanoma and LLC cells supplemented with lactate or with stable knockdown of LDH-A. We showed that lactate promoted PD-L1 expression in lung cancer cell line (Fig. 4F) and B16-F10 melanoma (data not shown). Furthermore, decreased PD-L1 expression upon knockdown of LDH-A in B16-F10 melanoma cells was observed (Fig. 4G and H). The effect of lactate and LDH-A was not recapitulated by NO donor, suggesting an NO-independent mechanism of PD-L1 regulation (Supplementary Fig. S5), likely mediated by lactate-induced HIF-1α (Fig. 2E; ref. 3). These data postulate a general mechanism of PD-L1 induction by LDH-A/lactate in various tumor types and its critical role for control of tumor growth.

Deletion of LDH-A in APC amplifies antitumor CD8+ T cells responses

To test whether Mø LDH-A expression was playing a role in directing T-cell activation, we performed autologous MLR, in which LDH-A wild-type or deficient splenic APCs, which include Mø, were used to activate CD3+ cultures that included both CD8+ and CD4+ lymphocytes (Fig. 5A and B). Deletion of LDH-A in APC did not significantly affect CD8+ T-cell proliferation (Fig. 5A and B) and had no effect on CD4+ T-cell proliferation (data not shown). There was no difference in T-cell proliferation upon deletion of LDH-A in T cells (Fig. 5A and B).

To further evaluate the role of LDH-A, a key regulator of Mø phenotype that dictates their capacity to mediate antitumor responses of CD8+ T cells in the TME, we have performed MLR of CD8+-enriched T-cell population in the presence of APC from LDH-A wild-type or knockout (Fig. 5C and D). We showed no significant difference in the proliferation of wild-type CD8+ T cells in the presence of APCs isolated from knockout mice (Fig. 5C). Further, deletion of LDH-A in APC did not influence the levels of granzyme B, perforin, TGFβ, and IFNγ (Supplementary Fig. S6A–S6D). Interestingly, we detected higher levels of IL17 and IL10 mRNA expression in the CD8+ T-cell MLR cocultures in the presence of APC from LDH-A−/− mice (Fig. 5D and E). Importantly, there were no differences in proapoptotic gene (Bax) and antiapoptotic (p21, Bcl-XL; Supplementary Fig. S6E–S6G) but slightly higher number of live cells were present within the CD8+ T-cell population stimulated with APC from LDH-A−/− mice (Supplementary Fig. S6H and S6I).

The higher IL17 and IL10 cytokine levels were associated with higher proportions of CD68+ macrophages (Supplementary Fig. S7A) and lower PD-L1 expression (Supplementary Fig. S7B) in the splenic APC from LDH-A−/− mice compared with wild-type littermates.

To test the function of CD8+ T cells activated by APC isolated from LDH-A knockout or wild-type mice, we infused the CD8+ T cells and APCs from the MLR cocultures (as in Fig. 5C) to the mice bearing LLC tumors. We showed a significant decrease in tumor progression in the model of LLC upon infusion of CD8+ T cells activated by APC from LDH-A–deficient mice (Fig. 5G), suggesting more effective antitumor response of these cells. These slower growing tumors in mice infused with CD8+ T cells activated by APC from LDH-A–deficient mice showed lower rate of proliferation as demonstrated by decreased P-histone H3 staining (Fig. 5H and I).

We conclude that CD8+ T cells derived from MLR with APC from LDH-A–deficient mice are more efficient in suppressing tumor growth than CD8+ T cells activated by wild-type APC (Fig. 6).

This study allows us to conclude that blockade of LDH-A in myeloid cells restores the immunocompetent TME by reversing Mø phenotype and boosting Tc17 cell–dependent immunity. Therefore higher expression of LDH-A in the stroma of patients with lung cancer may indicate a metabolic adaptation to evade immune responses and allow for angiogenesis and tumor growth. Our previous work showed that global deletion of LDH-A inhibits tumor growth (13). In the current study, we investigated the role of LDH-A in specific immune compartments and its influence on cancer growth. Given the published data that lactate drives M2 phenotype (10), we asked whether LDH-A plays a role in the TME by influencing Mø polarity and their interaction with lymphocytes.

Our work is the first report to show regulation of the TME and tumor growth by depletion of LDH-A in myeloid cells and Mø, which in part is dependent on its enzymatic activity to generate lactate. LDH-A in Mø modulates the balance between pro- and antitumor immune responses. Furthermore, Mø metabolism within the tumor may play critical roles in tumor evolution. Part of the effects seen upon deletion of LDH-A might be lactate dependent, but some may be driven by metabolic or gene expression changes in the cells independently of lactate generation. Furthermore, these differential effects of lactate and LDH-A indicate that suppression of LDH-A in the TME may provide additional benefits of not only suppressing cancer cell growth fueled by lactate, but also in regulating immune cell functions in the TME. Our studies indicate that even in the absence of lactate production, LDH-A modulates Mø function to generate immunomodulatory cytokines. We found that lack of LDH-A amplifies Mø skewing toward M1-like phenotype and conditional deletion of LDH-A under the LysM-Cre promoter leads to suppression of tumor. Our studies are consistent with the work published by Dr. Medzhitov's laboratory that showed that lactate drives M2 polarization of Mø (3). Unlike in the presence of lactate, we did not see any effect on M2 skewing upon deletion of LDH-A, rather a direct effect on M1 skewing. The differential effect of LDH-A and lactate on M2 polarization as well as cytokine release indicates additional mechanism of LDH-A in addition to its enzymatic activity.

The deletion of myeloid LDH-A in the TME may result in reduced in immunosuppressive environment (i.e., lower PD-L1) and promotion of T-cell antitumor activity through induction of Tc17 and Tc1 effector cells. Our studies indicate that deletion of LDH-A in Mø promotes M1 skewing to boost Tc17 cells and the expression of IFNγ in the TME, which might be one of the mechanisms explaining regression of established tumors observed in prior published studies using LDH-A−/−:K-Ras compared with the K-Ras wild-type mice. We speculate that Tc1 (IFNγ) effector function is the result of the early orchestration driven by Tc17 (IL17) in the absence of LDH-A in the TME. Therefore, both Tc1 and Tc17 are likely important in mediating antitumor effects in our model.

IL6, IP-10, and TGFβ are decreased in LDH-A−/− and are regulated independently of lactate in the TME. In contrast, anti-inflammatory cytokine IL10 is upregulated in the absence of LDH-A in M1-polarized macrophages. These cytokines are strongly linked to T-cell function. IL6 deficiency leads to reduced number of T cells in the lung (27), while TGFβ has been known for a long time as an immunosuppressive cytokine and was described to differentiate naïve CD4 T cells into inducible Treg (iTreg) cells. IL6 together with TGFβ upregulates expression of the transcription factor retinoic acid receptor–related orphan nuclear receptor γt, which is required for IL17 expression and differentiation of Th17 T cells in mouse (28). Interestingly, IL6 and TGFβ can drive IL10 expression in Th17 cells, which limits bystander inflammatory damage of these cells (29). Macrophage-expressed LDH-A may influence the local production of TGFβ in the TME, thus potentially increasing the number of Tregs in the tumor. We have not seen a significant difference in the expression of Foxp3 (marker of Tregs); however, the Foxp3+/CD3+ T-cell ratio was significantly lower in the K-ras:LysM-Cre:LDH-Aflfl mice, with deletion of LDH-A in myeloid cells suggesting that lower TGFβ levels may dampen immunosuppression in the TME while fostering infiltration of activated CD8+ T cells producing IFNγ. Therefore, suppression of IL6, IP-10, or TGFβ in the absence of LDH-A might be a significant contributor to the reconstitution of host immune responses against cancer cells. IL10 is upregulated in association with other cytokines, including those that are proinflammatory. We speculate that LDH-A deletion may favor production of IL17 by “regulatory” IL10-producing cells. Furthermore, IP-10/CXCL10 is highly expressed by PSCs in the presence of pancreatic cancer cells, and its expression correlates with infiltration by Tregs and poor survival (30, 31).

Several groups have shown that MCP-1/CCL2 is required for efficient recruitment of immunosuppressive myeloid and T cells into the TME (32, 33). However, as in our model, all myeloid cells [including myeloid-derived suppressor cells (MDSC)] are LDH-A deficient, they will likely acquire anticancer M1 phenotype in the TME. Therefore, infiltration of a higher number of myeloid cells may trigger more efficient anticancer immunity. We believe that presence of LDH-A–deficient macrophages and infiltrating myeloid cells and their polarization toward M1 cells in the TME might be the key for boosting the immune responses. Therefore, increase in CCL2 in the absence of LDH-A likely serves as a positive feedback loop to promote infiltration of higher number of immune cells in the TME. Indeed, high lactate levels are linked to higher numbers of myeloid suppressor cells (13).

A recent report demonstrated that deletion of LDH-A in Th1 leads to lower expression of IFNγ, suggesting that aerobic glycolysis is required for IFNγ transcription (34). We did not find any difference in IFNγ expression by T cells in the MLR cocultures upon deletion of LDH-A in APC; however, we demonstrated higher IFNγ levels in the TME upon deletion of LDH-A in myeloid cells. Similarly, we found higher IL17 levels in in vitro cultures, whereas there was no difference in the tissues. These data suggest that a complexity of cell–cell interactions in the TME and that not T cell–APC interaction alone may be required for anticancer response observed in tumors lacking LDH-A. Furthermore, this may reflect the differential kinetics in cytokine production between the in vitro (early late) and in vivo (late responses) experiments.

IL17 and IL10 have dual effects on tumor growth. Part of the antitumoral effects of IL17 is due to activation of cytotoxic CD8+ T cells. Importantly, we found that CD8+ T cells in the presence of APC from LDH-A–deficient animals express higher levels of IL17 and IL10. These IL17-producing CD8+ T cells (Tc17) were found to be highly efficient at ablating tumors (35). It is likely that Tc17 cells are mediating the effects of APC/myeloid-expressed LDH-A on delayed tumor growth. Furthermore, infusion of these T cells into the mice with LLC tumors resulted in significantly lower growth of tumors and suppression of proliferation. This indicates that T cells activated by APC lacking LDH-A have high functional potential. Presence of mixed population of splenocytes, including CD8+ T cells and APC, allows for more effective antitumor responses due to potential activation of donor and recipient T cells. Activated splenocytes mix was previously reported to suppress tumor growth when applied in combination with IL17 (20). Accordingly, our data showed that CD8+ T cells stimulated with LDH-A−/− APC produce higher level of IL17, and these combinations significantly suppressed tumor growth.

IL17 not only activates cytotoxic T cells but also promotes the maturation of APCs (36). IL10 is known for its effects on expansion of Tregs but can also directly activate tumor-resident CD8+ T cells without de novo infiltration from secondary lymphoid organs (37). These data indicate that LDH-A deletion may lead to better immune responses due to higher expression of IL17 in CD8+ T cells. Infiltrating T cells showed higher expression of PD-1, a marker of activation or exhaustion as well as IFNγ, a potent antitumoral factor. On the basis of the fact that myeloid LDH-A–deficient mice have low PD-L1 expression in the TME, we expect that these T cells have high cytotoxic activity. Whether LDH-A modulates immunotherapy responses is a question of current investigations.

Lactate promotes M2 polarization (10), whereas LDH-A influences M1 skewing. Therefore, there is need for the signal from LPS or damage tissues (DAMP) for Mø skewing to observe LDH-A's effects on M1 phenotype. As the majority of the tumor-associated Mø are M2-like phenotype, deletion of LDH-A protein is important to both: (i) block lactate generation; and (ii) trigger immune-promoting cytokine release upon deletion of LDH-A protein.

Our data indicate that infiltrating macrophages after deletion of LDH-A in the TME show higher levels of iNOS locally. It is possible that iNOS expressing Mø can directly kill cancer and other cells via generation of NO (38). However, we did not see NO to mediate antitumoral effects of LDH-A inhibition in myeloid cells in our transwell system. Others (39) showed that killing of cancer cells involves both TNF and NO and needs contact-dependent delivery of these mediators from the macrophage to its target. It is possible that LDH-A–deficient Mø require direct contact with cancer cells to induce cell death, and this will be addressed in further studies.

High LDH-A levels are one of the strongest biomarkers of resistance in several cancer types. We recognize that our findings may have a potential clinical application considering development of new drugs that target LDH-A. LDH-A inhibitors may work through immunomodulation of TME and suppression of angiogenesis. Angiogenesis is a key component of tumor progression and is driven by the Warburg effect. Lactate in the TME might drive VEGF and EC mobilization/differentiation to build new vessels. We speculate that antiangiogenic drugs may potentially affect LDH-A–mediated macrophage protumoral activity by suppressing VEGF expression. However, as LDH-A seems to be important for many other cytokines as well as activities in other cells, including tumor cells, the potential cross-talk of antiangiogenic drugs and LDH-A expression needs to be further investigated. Previous data showed that the frequency of MDSCs in the spleens of mice carrying LDH-A–depleted tumors was significantly reduced (13). We demonstrated that LDH-A–expressing myeloid cells akin to cancer cells might affect the clinical efficacy of immune checkpoint inhibitors. We found that lactate and LDH-A expression promotes PD-L1 expression, indicating a newly discovered cross-talk between the target of immunotherapy and tumor cell metabolism.

HIF1α transcription factor was shown to regulate PD-L1 and both are targets of hypoxia (3). Indeed, lack of LDH-A in myeloid cells resulted in lower HIF1α expression, suggesting a potential mechanism of PD-L1 regulation in responses to LDH-A/lactate axis.

In summary, we present a novel observation that LDH-A deletion in myeloid cells amplifies antitumor immunity through induction of Tc17 cells, regulates angiogenesis, and PDL1 expression. We speculate that the use of LDH-A inhibitors may be a novel approach to improve the efficacy of checkpoint inhibitors.

No potential conflicts of interest were disclosed.

Conception and design: P. Seth, B. Wegiel

Development of methodology: B. Wegiel

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): P. Seth, A. Hedblom, M. Vuerich, H. Xie, M. Li, B. Wegiel

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): A. Hedblom, B. Wegiel

Writing, review, and/or revision of the manuscript: P. Seth, A. Hedblom, M.S. Longhi, B. Wegiel

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): E. Csizmadia, M. Li

Study supervision: P. Seth, B. Wegiel

Other (contributed to histology and IHC work on different conditions of tissues with different methods): E. Csizmadia

We acknowledge the fruitful discussion and comments on the manuscript from Dr. Kenneth Swanson.

Our studies were supported in part by funding from R01 DK104714, R21 CA169904, American Heart Association10SDG2640091, start-up funds from Department of Surgery at BIDMC (to B. Wegiel), and by R01 CA152330, R01 GM098453, CDMRP W81XWH-15-1-0686, and seed funds from BIDMC (to P. Seth).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Plathow
C
,
Weber
WA
. 
Tumor cell metabolism imaging
.
J Nucl Med
2008
;
49
:
43S
63S
.
2.
Warburg
O
.
The metabolism of tumors
.
London, United Kingdom
:
Arnold Constable
; 
1930
.
3.
Noman
MZ
,
Desantis
G
,
Janji
B
,
Hasmim
M
,
Karray
S
,
Dessen
P
, et al
PD-L1 is a novel direct target of HIF-1alpha, and its blockade under hypoxia enhanced MDSC-mediated T cell activation
.
J Exp Med
2014
;
211
:
781
90
.
4.
Negus
RP
,
Turner
L
,
Burke
F
,
Balkwill
FR
. 
Hypoxia down-regulates MCP-1 expression: implications for macrophage distribution in tumors
.
J Leukoc Biol
1998
;
63
:
758
65
.
5.
Mauro
C
,
Leow
SC
,
Anso
E
,
Rocha
S
,
Thotakura
AK
,
Tornatore
L
, et al
NF-kappaB controls energy homeostasis and metabolic adaptation by upregulating mitochondrial respiration
.
Nat Cell Biol
2011
;
13
:
1272
9
.
6.
Gatenby
RA
,
Gillies
RJ
. 
Why do cancers have high aerobic glycolysis?
Nat Rev Cancer
2004
;
4
:
891
9
.
7.
Fang
JS
,
Gillies
RD
,
Gatenby
RA
. 
Adaptation to hypoxia and acidosis in carcinogenesis and tumor progression
.
Semin Cancer Biol
2008
;
18
:
330
7
.
8.
Fischer
K
,
Hoffmann
P
,
Voelkl
S
,
Meidenbauer
N
,
Ammer
J
,
Edinger
M
, et al
Inhibitory effect of tumor cell-derived lactic acid on human T cells
.
Blood
2007
;
109
:
3812
9
.
9.
Gottfried
E
,
Kreutz
M
,
Mackensen
A
. 
Tumor-induced modulation of dendritic cell function
.
Cytokine Growth Factor Rev
2008
;
19
:
65
77
.
10.
Colegio
OR
,
Chu
NQ
,
Szabo
AL
,
Chu
T
,
Rhebergen
AM
,
Jairam
V
, et al
Functional polarization of tumour-associated macrophages by tumour-derived lactic acid
.
Nature
2014
;
513
:
559
63
.
11.
Gottfried
E
,
Kunz-Schughart
LA
,
Ebner
S
,
Mueller-Klieser
W
,
Hoves
S
,
Andreesen
R
, et al
Tumor-derived lactic acid modulates dendritic cell activation and antigen expression
.
Blood
2006
;
107
:
2013
21
.
12.
Haschemi
A
,
Kosma
P
,
Gille
L
,
Evans
CR
,
Burant
CF
,
Starkl
P
, et al
The sedoheptulose kinase CARKL directs macrophage polarization through control of glucose metabolism
.
Cell Metab
2012
;
15
:
813
26
.
13.
Husain
Z
,
Huang
Y
,
Seth
P
,
Sukhatme
VP
. 
Tumor-derived lactate modifies antitumor immune response: effect on myeloid-derived suppressor cells and NK cells
.
J Immunol
2013
;
191
:
1486
95
.
14.
Shime
H
,
Yabu
M
,
Akazawa
T
,
Kodama
K
,
Matsumoto
M
,
Seya
T
, et al
Tumor-secreted lactic acid promotes IL-23/IL-17 proinflammatory pathway
.
J Immunol
2008
;
180
:
7175
83
.
15.
Murugaiyan
G
,
Saha
B
. 
Protumor vs antitumor functions of IL-17
.
J Immunol
2009
;
183
:
4169
75
.
16.
Numasaki
M
,
Fukushi
J
,
Ono
M
,
Narula
SK
,
Zavodny
PJ
,
Kudo
T
, et al
Interleukin-17 promotes angiogenesis and tumor growth
.
Blood
2003
;
101
:
2620
7
.
17.
Muranski
P
,
Boni
A
,
Antony
PA
,
Cassard
L
,
Irvine
KR
,
Kaiser
A
, et al
Tumor-specific Th17-polarized cells eradicate large established melanoma
.
Blood
2008
;
112
:
362
73
.
18.
Yu
Y
,
Cho
HI
,
Wang
D
,
Kaosaard
K
,
Anasetti
C
,
Celis
E
, et al
Adoptive transfer of Tc1 or Tc17 cells elicits antitumor immunity against established melanoma through distinct mechanisms
.
J Immunol
2013
;
190
:
1873
81
.
19.
Xie
H
,
Hanai
J
,
Ren
JG
,
Kats
L
,
Burgess
K
,
Bhargava
P
, et al
Targeting lactate dehydrogenase–a inhibits tumorigenesis and tumor progression in mouse models of lung cancer and impacts tumor-initiating cells
.
Cell Metab
2014
;
19
:
795
809
.
20.
Zhou
P
,
Zhang
Q
,
Zhao
Y
,
Sha
H
,
Cao
X
,
Wang
Y
. 
IL-17 promoted the inhibition of medulloblastoma in mice by splenocyte injection
.
Eur J Med Res
2015
;
20
:
98
.
21.
Wegiel
B
,
Baty
CJ
,
Gallo
D
,
Csizmadia
E
,
Scott
JR
,
Akhavan
A
, et al
Cell surface biliverdin reductase mediates biliverdin-induced anti-inflammatory effects via phosphatidylinositol 3-kinase and Akt
.
J Biol Chem
2009
;
284
:
21369
78
.
22.
Wegiel
B
,
Gallo
D
,
Csizmadia
E
,
Roger
T
,
Kaczmarek
E
,
Harris
C
, et al
Biliverdin inhibits Toll-like receptor-4 (TLR4) expression through nitric oxide-dependent nuclear translocation of biliverdin reductase
.
Proc Natl Acad Sci U S A
2011
;
108
:
18849
54
.
23.
Wegiel
B
,
Gallo
D
,
Csizmadia
E
,
Harris
C
,
Belcher
J
,
Vercellotti
GM
, et al
Carbon monoxide expedites metabolic exhaustion to inhibit tumor growth
.
Cancer Res
2013
;
73
:
7009
21
.
24.
Wang
H
,
Lee
SS
,
Dell'Agnello
C
,
Tchipashvili
V
,
d'Avila
JC
,
Czismadia
E
, et al
Bilirubin can induce tolerance to islet allografts
.
Endocrinology
2006
;
147
:
762
8
.
25.
Agata
Y
,
Kawasaki
A
,
Nishimura
H
,
Ishida
Y
,
Tsubata
T
,
Yagita
H
, et al
Expression of the PD-1 antigen on the surface of stimulated mouse T and B lymphocytes
.
Int Immunol
1996
;
8
:
765
72
.
26.
Sharpe
AH
,
Wherry
EJ
,
Ahmed
R
,
Freeman
GJ
. 
The function of programmed cell death 1 and its ligands in regulating autoimmunity and infection
.
Nat Immunol
2007
;
8
:
239
45
.
27.
Dienz
O
,
Rincon
M
. 
The effects of IL-6 on CD4 T cell responses
.
Clin Immunol
2009
;
130
:
27
33
.
28.
Ivanov
II
,
McKenzie
BS
,
Zhou
L
,
Tadokoro
CE
,
Lepelley
A
,
Lafaille
JJ
, et al
The orphan nuclear receptor RORgammat directs the differentiation program of proinflammatory IL-17+ T helper cells
.
Cell
2006
;
126
:
1121
33
.
29.
McGeachy
MJ
,
Bak-Jensen
KS
,
Chen
Y
,
Tato
CM
,
Blumenschein
W
,
McClanahan
T
, et al
TGF-beta and IL-6 drive the production of IL-17 and IL-10 by T cells and restrain T(H)-17 cell-mediated pathology
.
Nat Immunol
2007
;
8
:
1390
7
.
30.
Lunardi
S
,
Jamieson
NB
,
Lim
SY
,
Griffiths
KL
,
Carvalho-Gaspar
M
,
Al-Assar
O
, et al
IP-10/CXCL10 induction in human pancreatic cancer stroma influences lymphocytes recruitment and correlates with poor survival
.
Oncotarget
2014
;
5
:
11064
80
.
31.
Lunardi
S
,
Lim
SY
,
Muschel
RJ
,
Brunner
TB
. 
IP-10/CXCL10 attracts regulatory T cells: implication for pancreatic cancer
.
Oncoimmunology
2015
;
4
:
e1027473
.
32.
Chang
AL
,
Miska
J
,
Wainwright
DA
,
Dey
M
,
Rivetta
CV
,
Yu
D
, et al
CCL2 produced by the glioma microenvironment is essential for the recruitment of regulatory T cells and myeloid-derived suppressor cells
.
Cancer Res
2016
;
76
:
5671
82
.
33.
Pena
CG
,
Nakada
Y
,
Saatcioglu
HD
,
Aloisio
GM
,
Cuevas
I
,
Zhang
S
, et al
LKB1 loss promotes endometrial cancer progression via CCL2-dependent macrophage recruitment
.
J Clin Invest
2015
;
125
:
4063
76
.
34.
Peng
M
,
Yin
N
,
Chhangawala
S
,
Xu
K
,
Leslie
CS
,
Li
MO
. 
Aerobic glycolysis promotes T helper 1 cell differentiation through an epigenetic mechanism
.
Science
2016
;
354
:
481
4
.
35.
Majchrzak
K
,
Nelson
MH
,
Bailey
SR
,
Bowers
JS
,
Yu
XZ
,
Rubinstein
MP
, et al
Exploiting IL-17-producing CD4+ and CD8+ T cells to improve cancer immunotherapy in the clinic
.
Cancer Immunol Immunother
2016
;
65
:
247
59
.
36.
Hirahara
N
,
Nio
Y
,
Sasaki
S
,
Minari
Y
,
Takamura
M
,
Iguchi
C
, et al
Inoculation of human interleukin-17 gene-transfected Meth-A fibrosarcoma cells induces T cell-dependent tumor-specific immunity in mice
.
Oncology
2001
;
61
:
79
89
.
37.
Emmerich
J
,
Mumm
JB
,
Chan
IH
,
LaFace
D
,
Truong
H
,
McClanahan
T
, et al
IL-10 directly activates and expands tumor-resident CD8(+) T cells without de novo infiltration from secondary lymphoid organs
.
Cancer Res
2012
;
72
:
3570
81
.
38.
Santoni
M
,
Cascinu
S
,
Mills
CD
. 
Altering macrophage polarization in the tumor environment: the role of response gene to complement 32
.
Cell Mol Immunol
2015
;
12
:
783
4
.
39.
Duerksen-Hughes
PJ
,
Day
DB
,
Laster
SM
,
Zachariades
NA
,
Aquino
L
,
Gooding
LR
. 
Both tumor necrosis factor and nitric oxide participate in lysis of simian virus 40-transformed cells by activated macrophages
.
J Immunol
1992
;
149
:
2114
22
.