Neutralization of Tumor Acidity Improves Antitumor Responses to Immunotherapy

Abstract

Cancer immunotherapies, such as immune checkpoint blockade or adoptive T-cell transfer, can lead to durable responses in the clinic, but response rates remain low due to undefined suppression mechanisms. Solid tumors are characterized by a highly acidic microenvironment that might blunt the effectiveness of antitumor immunity. In this study, we directly investigated the effects of tumor acidity on the efficacy of immunotherapy. An acidic pH environment blocked T-cell activation and limited glycolysis in vitro. IFNγ release blocked by acidic pH did not occur at the level of steady-state mRNA, implying that the effect of acidity was posttranslational. Acidification did not affect cytoplasmic pH, suggesting that signals transduced by external acidity were likely mediated by specific acid-sensing receptors, four of which are expressed by T cells. Notably, neutralizing tumor acidity with bicarbonate monotherapy impaired the growth of some cancer types in mice where it was associated with increased T-cell infiltration. Furthermore, combining bicarbonate therapy with anti-CTLA-4, anti-PD1, or adoptive T-cell transfer improved antitumor responses in multiple models, including cures in some subjects. Overall, our findings show how raising intratumoral pH through oral buffers therapy can improve responses to immunotherapy, with the potential for immediate clinical translation.

Conflict of interest statement

Conflicts of Interest: R.J. Gillies is shareholder in and member of scientific advisory board of Health-Myne, Inc. S. Pilon-Thomas is grant recipient from Lion Biotechnologies, Inc.

©2015 American Association for Cancer Research.

Figures

Figure 1. Effect of pH on activation of T cells

T cells were activated at pH 7.4 for 48 hr with either gp10025–33 peptide (pmel T cells, panels A–D) or OVASIINFEKL peptide (OT-1 or OT-2 T-cells, panel E–F), or left unactivated in complete media (CM) followed by an incubation at either pH 6.6 or 7.4 for an additional 24 hr. (A) percent cell viability of CD8+ pmel T cells following the entire treatment as measured by flow cytometry; (B) percent of CD8+ pmel T cells that contained IFN-γ as measured by flow cytometry. (C) amount of IFN-γ secreted into the media by pmel T cells as measured by ELISA following the 2nd 24 hr incubation at pH 6.6 or 7.4; (D) IFN-γ production by pmel T cells after re-plating from pH 6.6 to 7.4 as measured by ELISA. Activated pmel T cells were cultured in CM for 24 hours at pH 6.6 or 7.4 (original culture). After 24 hours, cells were collected and replated at pH 7.4 with gp100 peptide (restimulation at pH 7.4). Culture supernatants were collected after 24h for IFN-γ measurement by ELISA. (E) IFN-γ secreted into the media by CD8+ OT-1 T cells measured by ELISA following a 2nd 24 hr incubation at pH 6.7 or 7.4. (F) shows the amount of IFN-γ secreted into the media by CD4+ OT-II T cells measured by ELISA following the 2nd 24 hr incubation at pH 6.7 or 7.4. Data shown as mean ± SD from a minimum of three independent experiments. P-values were determined by two-tailed Student’s t-test.

Figure 2. T cell characteristics

(A) IFN-γ mRNA levels are unaffected by pH. Levels of IFN-γ mRNA in pmel T cells following 48 hr stimulation at pH 7.4 and subsequent incubation at pH 7.4 or 6.6 for a subsequent 24 hr. Data were normalized to ribosomal protein 36B4 and expressed as mean ratio ± SD from three independent experiments; (B) Intracellular pH (pHin) of activated T cells is unaffected by incubation pH. Pmel T cells were activated as above, and then incubated at pH 6.6 or 7.4 for 24 hr, after which they were loaded with SNARF-1 AME, washed and resuspended in PBS (2,4) or high K/nigericin (1,3) at pH 6.6 (1,2) or 7.4 (3,4) and the 580/640 fluorescence ratio subsequently determined by flow cytometry. The ratio shows that the pHin values of cells in PBS at low and high pH were both ~7.2, and the ratios in nigericin show pH calibration points. (C). Forskolin inhibits T cell function. Activated pmel T cells were cultured with DMSO or forskolin in the presence of gp100 peptide. After 24h, intracellular IFN-γ was measured by flow cytometry. Data represent results from 3 separate experiments and p-values were determined using two-tailed Student’s t-test.

Figure 3. pmel T cell metabolism

Pmel T cells were activated with gp10025–33 peptide for 48 hr, followed by a subsequent 48 h incubation at either pH 6.7 or 7.4. Cells were subsequently metabolically profiled using a Seahorse XF-96 analyzer. (A) Representative results of a glucose stress tests (GST) which measures extracellular acidification rate (ECAR) following addition of glucose, oligomycin and 2-deoxy glucose; (B) rRepresentative results of a mitochondrial stress test (MST) which measures the oxygen consumption rate (OCR) in glucose-containing media following sequential additions of oligomycin, FCCP and Rotenone/Antimycin A. See Materials and Methods for description of GST and MST. (C) Acidosis inhibits T cell glycolysis. Data show the glucose-stimulated increase in ECAR; (D) Acidic cells are more oxidative. Data show the basal OCR for activated cells incubated at pH 6.7 or 7.4. Data represent results form 3 separate experiments, with 6 replicates per experiment. A two-tailed Student t-test was used to calculate statistical significance.

Figure 4. Buffer monotherapy in vivo

Animals bearing syngeneic melanoma tumors were treated with or without 200 mM ad lib NaHCO3 (bicarb) in their drinking water. (A) Bicarb raises extracellular pH. The extracellular pH levels were measured with microelectrode in B16 tumors treated with or without 200 mM sodium bicarbonate therapy. Four separate animals were investigated under each condition and a minimum of four measurements were made in each tumor along a single needle track (p<0.001); (B) Bicarb does not affect B16 tumor growth. Mice bearing B16 tumors were treated with or without buffer therapy, showing no effect on tumor growth (n = 8 mice/group); (C) Bicarb inhibits Yumm 1.1 tumor growth. Mice bearing Yumm1.1 tumors were treated with or without buffer therapy, showing effect on tumor growth with time (n = 8 mice/group); Panel D shows the resected tumor weights at the endpoint from this experiments (P<0.01) and Panel E shows the flow quantified relative numbers of CD8+ and CD4+ T cells infiltration into Yumm1.1 tumors (n = 8 mice/group; p<0.05 tap vs. bicarb CD8 cells; CD4 not significant).

Figure 5. Buffer therapy enhances efficacy of anti-immune therapy in B16 melanoma

C57BL/6 mice received tap water or tap water containing 200 mM NaHCO3 (bicarb) ad lib 3 days prior to tumor inoculation with 1×105 tumor cells injected s.c. in the left flank (n= 10 mice/group). Three days later, animals subsequently received i.p. injections of 20 mg/kg of anti–PD1 and/or anti-CTLA4 antibodies and continued to receive antibodies every 3 to 4 days until the end of the experiment. Tumors sizes were measured every 2–3 days and at the endpoint, tumors were excised, weighed and extracted for T cell repertoire or prepared for histology. (A, B) Bicarb improved CTLA4 therapy in B16 melanoma. Panel A shows growth of B16 tumors in mice treated with anti-CTLA4 antibody, with or without bicarb therapy. Tumor weights are shown in Panel B. (C, D) Bicarb improved PD1 therapy in B16. Panel A shows growth of B16 tumors in mice treated with anti-PD1 antibody, with or without bicarb therapy; tumor weights are shown in Panel D. (E) Bicarb does not add to combination checkpoint blockade. Panel shows growth of B16 tumors in mice treated with the combination of anti-CTLA4 and anti-PD1 antibodies with or without bicarb therapy. Differences between checkpoint combination with or without bicarbonate were not significant. (F) Bicarb improved PD1 therapy in Panc02. Data show weights of Panc02 tumors in mice (n=10 per group) treated with combination of anti-PD1 antibodies with or without bicarb therapy.

Figure 6. Effect of bicarb on adoptive T cell transfer

C57BL/6 mice (n = 10 per group) received tap water or tap water containing 200 mM NaHCO3 (bicarb) ad lib 3 days prior to tumor inoculation with 1×105 B16 cells injected s.c. in the left flank. Three days after inoculation, mice received a sub lethal dose (600 cGy) of total body irradiation (TBI) administered by an X-ray irradiator. For adoptive transfer experiments, T cells were isolated from the spleens of pmel mice and cultured in media containing 10 IU/ml of IL-2 and 5ug/ml of gp10025–33in vitro for 5 days. On day 4 following tumor injection, 5×106 T cells were injected intravenously. IL-2 (2.5e5 IU) was given i.p. following T cell injection, and continued every 12 hours for three days, for a total of six injections. Following this treatment, tumor size was measured and recorded every 2 days. (A) shows tumor growth after adoptive transfer of T cells or controls in combination with or without buffer therapy. Group mean differences between T cells vs. T cells + bicarb were not significant. However there was a survival advantage, as shown in the survival curve, (B) which had a log rank p = 0.002; (C) shows the percent of T cell persistence after adoptive transfer and buffer therapy (p<0.05).