Akt inhibition enhances expansion of potent tumor-specific lymphocytes with memory cell characteristics

Abstract

Adoptive cell therapy (ACT) using autologous tumor-infiltrating lymphocytes (TIL) results in complete regression of advanced cancer in some patients, but the efficacy of this potentially curative therapy may be limited by poor persistence of TIL after adoptive transfer. Pharmacologic inhibition of the serine/threonine kinase Akt has recently been shown to promote immunologic memory in virus-specific murine models, but whether this approach enhances features of memory (e.g., long-term persistence) in TIL that are characteristically exhausted and senescent is not established. Here, we show that pharmacologic inhibition of Akt enables expansion of TIL with the transcriptional, metabolic, and functional properties characteristic of memory T cells. Consequently, Akt inhibition results in enhanced persistence of TIL after adoptive transfer into an immunodeficient animal model and augments antitumor immunity of CD8 T cells in a mouse model of cell-based immunotherapy. Pharmacologic inhibition of Akt represents a novel immunometabolomic approach to enhance the persistence of antitumor T cells and improve the efficacy of cell-based immunotherapy for metastatic cancer.

Trial registration: ClinicalTrials.gov NCT00670748 NCT01319565.

Conflict of interest statement

Conflict-of-interest disclosure: The authors declare no competing financial interests.

©2014 American Association for Cancer Research.

Figures

Figure 1. Inhibition of Akt promotes expansion of human tumor-infiltrating lymphocytes (TIL) with enhanced expression of the memory-marker CD62

(A) FACS histogram and quantification of phosphorylation events at indicated residues during acute time points after CD3 stimulation either in presence or absence of pharmacologic inhibition of Akt. Grey shading represents unstimulated TIL. (B) FACS histogram and quantification of CD62L expression on CD4+ and CD8+ TIL isolated from 3 patients and expanded ex vivo at clinical-scaled with agonistic anti-CD3 (OKT3) antibody and irradiated allogeneic feeders with high dose IL-2 in the presence or absence of Akt inhibitor. (C) Scatter plot showing fold expansion at clinical-scale of human TIL from 3 patients cultured independently in triplicate. (D) Bar graph showing interferon-gamma (IFNγ) release by ELISA when either Akti-treated TIL or vehicle are co-cultured for 12 hours under following conditions: no tumor cells (TC), allogeneic TC, autologous TC, autologous TC with MHC-I blocking antibody, and OKT3 alone. Asterisks are used to indicate following P values: * = P<0.05, ** P<0.01, *** P<0.001, **** P<0.0001. Center bar indicates mean and error bars indicate SEM.

Figure 2. Inhibition of Akt promotes expansion of human tumor-infiltrating lymphocytes (TIL) with transcriptional signature of memory T cells

Human TIL isolated from 3 patients with advanced melanoma were stimulated ex vivo with agonistic anti-CD3 (OKT3) antibody and irradiated allogeneic feeders and expanded to therapeutic scale with high dose IL-2 in the presence or absence of Akt inhibitor. (A) Principle Component Analysis of microarray data from CD8+ TIL of 3 patients (in quadruplicate) either cultured with or without Akti. (B) Hierarchical cluster analysis of 2,602 identified differentially-expressed genes (pFDR<0.05) in TIL isolated from 3 patients and cultured in indicated treatment groups. (C) Bar graph showing fold expression of canonical “memory” and “effector”-associated genes from microarray analysis (E) Enrichment plots designated as “naïve” and “effector” from GSEA showing enhanced expression of genes upregulated in human naïve (CD62L+ CD45RA+) vs. effector memory (CD62L- CD45RO+) CD8 T cells. Normalized enrichment score (NES). False discovery rate (FDR).

Figure 3. Global metabolomic analysis shows Akt inhibition is associated with enhanced fatty-acid oxidation in human tumor-infiltrating lymphocytes (TIL)

(A) Principle component analysis of metabolome (362 biochemicals) of Akti-treated TIL that were stimulated ex vivo using agonistic anti-CD3 (OKT3) antibody and irradiated allogeneic feeders and expanded to clinical-scale for 30 days with high dose IL-2 in the presence or absence of Akt inhibitor. TIL were then isolated to analyze basal metabolic profile (in absence of restimulation) under basal cell culture conditions (B) Relative abundance of key metabolites in glycolytic pathway. Glucose; G-1-P, glucose-1-phosphate; G-6-P, glucose-6-phosphate; 3-PGC, 3-phosphoglycerate. (C) Relative abundance of key metabolites involved in lipid metabolism are shown. EPA, eicosapentaenoate 20:5n3. (D) Relative abundance of lysolipids in Akti-treated TIL versus vehicle. GPC glycerophosphorylcholine; GPE glycerophosphoethanolamine. Asterisks are used to indicate following P values: * = P<0.05, ** P<0.01, *** P<0.001, **** P<0.0001. Center bar indicates mean and error bars indicate SEM.

Figure 4. Therapeutic tumor-infiltrating lymphocytes (TIL) isolated from patients with melanoma have poor mitochondrial spare respiratory capacity that is augmented with pharmacologic inhibition of Akt

(A) Human TIL isolated from indicated patients were stimulated ex vivo using agonistic anti-CD3 (OKT3) antibody and irradiated allogeneic feeders and expanded to clinical-scale for 30 days with high dose IL-2 in the presence or absence of Akt inhibitor. An XF24 extracellular analyzer (Seahorse Bioscience) was used to measure oxygen consumption rates (OCR) of Akti-treated and vehicle TIL in real time under basal cell culture conditions (after 30 days in culture) and in response to indicated inhibitors at acute time points: FCCP, Carbonyl cyanide 4-(trifluoromethoxy) phenylhydrazone; R&A, rotenone and antimycin A. Spare respiratory capacity is maximal respiration (after FCCP) minus basal respiration. Data representative of 3 independent experiments for each patient.

Figure 5. Akt inhibition enhances persistence of human tumor-infiltrating lymphocytes (TIL) and murine cytotoxic T lymphocytes after adoptive transfer

(A) Representative FACS analysis and quantification of human Akti-treated and vehicle TIL that had been expanded to clinical-scale ex vivo and subsequently adoptively-transferred into NSG mice. Human TIL were isolated from indicated lymphoid and non-lymphoid organs of NSG mice 30 days after adoptive-transfer. Data representative of 5 biological replicates per treatment group. (B) FACS analysis of CD8+ T cells from age and gender-matched Thy1.1 (Akti) and Ly5.1 (vehicle) Pmel mice after contransfer into B6 mice and spleens harvested at the indicated time points. Data representative of 5 biologic replicates for each time point. (C) Representative FACS analysis of Akti-treated and vehicle T cells in lung and mesenteric lymph nodes (LN) at day 600. Asterisks are used to indicate following P values: * = P<0.05, ** P<0.01, *** P<0.001, **** P<0.0001. Center bar indicates mean and error bars indicate SEM.

Figure 6. Akt inhibition of cytotoxic T improves anti-tumor immunity of adoptively-transferred T cells in mouse model of cell-based immunotherapy for melanoma

(A) Representative FACS analysis and enumeration of Ly5.1 Pmel CD8+ T cells isolated from spleen and tumor microenvironment 5 days after adoptive-transfer into B16 melanoma-bearing mice. Donor T cells were stimulated in vitro with cognate peptide and cultured in IL-2. After 5 days, cells were restimulated with cognate peptide and irradiated B6 feeders for an additional 5 days prior to transfer. Akti was supplemented in media for entirety of in vitro culture. Data representative of 4 biological replicates per treatment group. (B) Representative FACS analysis of interferon gamma (IFNγ) and tumor necrosis factor alpha (TNFα) intracellular cytokine staining of adoptively-transferred Ly5.1 Pmel CD8+ T cells isolated from spleen of tumor-bearing mice (C) Treatment response of 2×106 Pmel CD8+ T cells adoptively-transferred (after same in vitro culture conditions described above) into mice bearing established B16 melanomas. Recipient mice were pretreated with 6Gy total body irradiation, adjuvant vaccine, and IL-2 in conjunction with cell therapy. Serial tumor measurements were obtained and tumor area calculated. Data representative of 10 biological replicates per treatment group. (D) Kaplan-Meier analysis of survival in tumor-bearing mice receiving adoptively-transferred cells treated with Akti (32 days) versus control (22 days; p <0.0001). Asterisks are used to indicate following P values: * = P<0.05, ** P<0.01, *** P<0.001, **** P<0.0001. Center bar indicates mean and error bars indicate SEM.