A key challenge in the field of T-cell immunotherapy for cancer is creating a suitable platform for promoting differentiation of effector cells while at the same time enabling self-renewal needed for long-term memory. Although transfer of less differentiated memory T cells increases efficacy through greater expansion and persistence in vivo, the capacity of such cells to sustain effector functions within immunosuppressive tumor microenvironments may still be limiting. We have therefore directly compared the impact of effector versus memory differentiation of therapeutic T cells in tumor-bearing mice by introducing molecular switches that regulate cell fate decisions via mTOR. Ectopic expression of RAS homolog enriched in brain (RHEB) increased mTORC1 signaling, promoted a switch to aerobic glycolysis, and increased expansion of effector T cells. By rapidly infiltrating tumors, RHEB-transduced T cells significantly reduced the emergence of immunoedited escape variants. In contrast, expression of proline-rich Akt substrate of 40 kDa (PRAS40) inhibited mTORC1, promoted quiescence, and blocked tumor infiltration. Fate mapping studies following transient expression of PRAS40 demonstrated that mTORC1low T cells made no contribution to initial tumor control but instead survived to become memory cells proficient in generating recall immunity. Our data support the design of translational strategies for generating heterogeneous T-cell immunity against cancer, with the appropriate balance between promoting effector differentiation and self-renewal. Unlike pharmacologic inhibitors, the genetic approach described here allows for upregulation as well as inhibition of the mTORC1 pathway and is highly selective for the therapeutic T cells without affecting systemic mTORC1 functions. Cancer Res; 75(13); 2641–52. ©2015 AACR.

A key challenge within the T-cell immunotherapy field is identifying a suitable strategy for inducing potent effector responses to eliminate cancerous cells without compromising the development of T-cell memory required for long-term immune surveillance. Recent fate mapping studies of immune responses have demonstrated a remarkable heterogeneity in the responses of individual naïve CD8+ T cells and their progeny following a first and subsequent exposure to antigen (1, 2). The initial clonal burst is composed primarily of short-lived effector cells that undergo multiple divisions, whereas cells that have low levels of expansion during the primary response acquire memory potential and contribute to the recall responses after subsequent secondary or tertiary antigen challenge; in this linear model of differentiation, memory cells have been stimulated the least during initial antigen encounters but upon rechallenge are highly efficient in generating new waves of effector cells as well daughter memory cells through self-renewal (1, 2).

In the adoptive immunotherapy setting, an inverse co-relation between T-cell differentiation and tumor immunity has been described; adoptive transfer of fully differentiated end stage effector cells was less effective than transfer of less differentiated T cells (3). The inefficiency of fully differentiated effector cells was linked to poor in vivo expansion and survival (3), while less differentiated and memory-stem cells expanded and showed long-term persistence (4–7). Although these observations highlight the importance of transferring cells with the potential to expand and persist in vivo, the capacity of such cells to sustain effector functions in tumor-bearing hosts may still be limiting. Indeed, there is substantial evidence that the tumor microenvironment actively limits the efficacy of T-cell therapies by preventing effector differentiation (8, 9). Thus, therapeutic strategies designed to enhance memory potential must also be evaluated in relation to their impact upon effector functions.

In this study, we used overexpression of RAS homolog enriched in brain (RHEB) and proline-rich Akt substrate of 40 kDa (PRAS40), two key molecules involved in positive and negative regulation of the mTORC1 pathway, respectively (10, 11), to explore the relative roles of effector function and memory potential in a murine model of adoptive immunotherapy of cancer. The results revealed that overexpression of RHEB in tumor-specific T cells increased mTORC1 activity, promoted earlier tumor infiltration, and prevented disease progression resulting from immune escape. In contrast, transient overexpression of PRAS40 decreased mTORC1 activity and increased subsequent recall immunity. Together, the data indicate that RHEB- and PRAS40-mediated tuning of mTORC1 can be used to enhance protective effector function and memory differentiation of tumor-specific T cells in vivo.

Mice

Experiments were performed in accordance with Home Office Guidance (Animals Scientific Procedures Act 1986) and approved by the local ethics committee. B6, B6.PL-Thy1a/CyJ (B6 Thy1.1), B6.SJL-Ptprca Pepcb/BoyJ (B6 CD45.1), and Thy1.1+ C57BL/6 Firefly Luciferase transgenic mice were bred in-house.

Cell lines

EL4-NP stably expressing the influenza A virus nucleoprotein (NP) was a kind gift from Dr. B. Stockinger (National Institute for Medical Research, London).

CD8+ T-cell transduction

Purification of CD8+ T cells and retroviral transduction was performed as described previously (12).

Oxygen consumption rate and extracellular acidification rate measurements

CD8+ T cells were transduced with VC, RHEB, or PRAS40 vectors and selected using the mouse CD19 reporter. For oxygen consumption rate (OCR) and extracellular acidification rate (ECAR) measurements, the Seahorse XF-24 metabolic extracellular flux analyzer (Seahorse Bioscience) was used. T cells were resuspended in serum-free unbuffered RPMI-1640 medium (Sigma-Aldrich) and plated onto Seahorse cell plates (6 × 105 cells per well in 5 replicate wells) coated with Cell-Tak (BD Biosciences). An “In-Seahorse” activation protocol (13) was used to monitor changes in metabolism before and after T-cell activation with anti-mouse CD3 antibody (145-2C11; 1 μg/mL; BD Biosciences), which was directly applied onto plated cells via the instrument's multi-injection port 80 minutes after the experiment was initiated.

EL4-NP tumor challenge

B6 mice were sublethally irradiated (5.5 Gy) and inoculated with 1 × 106 EL4-NP (s.c.). Tumors were allowed to grow for 5 days before i.v. injection of 0.1 to 1 × 105 TRHEB, TPRAS, or TVC (Thy-1.1+). Tumor-free mice were rechallenged with 1 × 106 (50 Gy) irradiated EL4-NP (i.p.). Bioluminescent imaging of T-cell infiltration of tumors was performed as described previously (12).

Doxycycline induction in vitro and in vivo

In vivo induction of transgene expression in cells transduced with pSERS vectors was achieved by treatment with 2 mg/mL doxycycline in drinking water (Sigma-Aldrich).

Methods for flow cytometry and retroviral vector construction are outlined in the Supplementary Materials.

Overexpression of RHEB and PRAS40 modulates T-cell metabolism and functions

To address the role of increased or reduced mTORC1 signaling in gene-engineered T cells, we transduced C57BL/6 (B6) purified CD8+ splenocytes with retrovirus encoding codon-optimized Tcrα and Tcrβ genes for the F5 TCR (that recognizes the influenza-derived pNP366 peptide presented by H2-Db), together with retrovirus encoding Rheb or Pras40 or a vector control construct to produce T cells referred to as TRHEB, TPRAS, and TVC, respectively (Fig. 1A). To test the effect of overexpressing RHEB and PRAS40 on mTORC1 signaling following TCR activation, we stimulated transduced CD8+ T cells with cognate antigen (EL4 thymoma cells expressing influenza NP protein) or with CD3/CD28 beads. T-cell stimulation triggers mTORC1 kinase activity, resulting in phosphorylation of direct targets including the eukaryotic translation initiation factor 4E-binding protein 1 (p-4E-BP1) and the ribosomal S6 kinase, which in turn phosphorylates ribosomal S6 proteins (p-S6; refs. 14, 15). We observed increased levels of p-S6 in TVC within 1 hour of stimulation with EL4-NP thymoma cells (Fig. 1B). TRHEB showed augmented levels of p-S6, both at baseline and after antigen stimulation, whereas TPRAS showed diminished phosphorylation at baseline and following antigen stimulation (Fig. 1B). At 24 hours after stimulation, TRHEB displayed increased, whereas TPRAS cells showed reduced p-S6 levels compared with vector control T cells (Fig. 1B and C). We did not find increased levels of p-4E-BP1 in TRHEB, whereas diminished phosphorylation of the TORC1 target protein was seen in TPRAS (Fig. 1C). Although overexpression of RHEB increased p-S6 levels at baseline and after antigen stimulation, this did not alter the sensitivity to inhibition by rapamycin, with similar IC50 values observed for TRHEB, TPRAS, and TVC (Fig. 1D). These data show that overexpression of RHEB and PRAS40 respectively enhances or suppresses mTORC1 signaling in CD8+ T cells.

Figure 1.

Overexpression of RHEB increases whereas PRAS40 decreases the phosphorylation of mTORC1 targets. A, retroviral vector maps. LTR, long terminal repeat; IRES, internal ribosome entry site; W-PRE, Woodchuck hepatitis post-transcriptional regulatory element. B, time course of ribosomal protein S6 phosphorylation (p-S6) of TVC, TRHEB, or TPRAS stimulated in vitro with anti-CD3/CD28 antibodies. Flow cytometry histograms show CD8+ GFP+ gated events (representative of two independent experiments). C, p-S6 and p-4E-BP1 staining in TVC, TRHEB, and TPRAS 24 hours after coculture with target EL4-NP cells (top, gated on CD8+ GFP+ and CD19+, marking the F5-TCR) and summary plots (bottom). Summary plots show the median fluorescence intensity (MFI) of p-S6 or 4E-BP1 in TRHEB and TPRAS divided by the median fluorescence intensity of TVC from the respective experiment. Horizontal bars represent median. *, P = 0.04; **, P = 0.005; ***, P < 0.001; n.s., not significant (one sample t test against a theoretical mean of 1). D, inhibitory effect of rapamycin upon S6 phosphorylation 24 hours after coculture with target EL4-NP cells. Dashed gray line, median fluorescence intensity of unstimulated cells. Data shown are representative of two independent experiments. Data points were fitted to a sigmoidal dose response curve with R2 values of 0.98, 0.99, and 0.94 for TVC, TRHEB, and TPRAS, respectively. IC50 (±95% confidence intervals) were 345 pmol/L (130–923), 288 pmol/L (139–558), and 347 pmol/L (68–1,759) for TVC, TRHEB, and TPRAS, respectively.

Figure 1.

Overexpression of RHEB increases whereas PRAS40 decreases the phosphorylation of mTORC1 targets. A, retroviral vector maps. LTR, long terminal repeat; IRES, internal ribosome entry site; W-PRE, Woodchuck hepatitis post-transcriptional regulatory element. B, time course of ribosomal protein S6 phosphorylation (p-S6) of TVC, TRHEB, or TPRAS stimulated in vitro with anti-CD3/CD28 antibodies. Flow cytometry histograms show CD8+ GFP+ gated events (representative of two independent experiments). C, p-S6 and p-4E-BP1 staining in TVC, TRHEB, and TPRAS 24 hours after coculture with target EL4-NP cells (top, gated on CD8+ GFP+ and CD19+, marking the F5-TCR) and summary plots (bottom). Summary plots show the median fluorescence intensity (MFI) of p-S6 or 4E-BP1 in TRHEB and TPRAS divided by the median fluorescence intensity of TVC from the respective experiment. Horizontal bars represent median. *, P = 0.04; **, P = 0.005; ***, P < 0.001; n.s., not significant (one sample t test against a theoretical mean of 1). D, inhibitory effect of rapamycin upon S6 phosphorylation 24 hours after coculture with target EL4-NP cells. Dashed gray line, median fluorescence intensity of unstimulated cells. Data shown are representative of two independent experiments. Data points were fitted to a sigmoidal dose response curve with R2 values of 0.98, 0.99, and 0.94 for TVC, TRHEB, and TPRAS, respectively. IC50 (±95% confidence intervals) were 345 pmol/L (130–923), 288 pmol/L (139–558), and 347 pmol/L (68–1,759) for TVC, TRHEB, and TPRAS, respectively.

Close modal

Consistent with the role for mTORC1 in promoting cell growth (16), nonstimulated TRHEB cells were larger in size than TVC cells, whereas TPRAS were smaller (Fig. 2A). Furthermore, even in the absence of stimulation, TRHEB had reduced surface expression of the lymph node homing molecule CD62L compared with control cells consistent with the inhibitory effect of mTORC1 upon CD62L transcription (17), whereas TPRAS had increased CD62L expression (Fig. 2B). Although we could detect no differences in cell viability between the groups (data not shown), we consistently observed reduced expression of activated caspase-3 in TPRAS compared with controls whereas expression was unchanged in TRHEB (Supplementary Fig. S1). Upon stimulation in vitro, both TRHEB and TVC showed similar upregulation of the activation markers, CD69 and CD25 (Supplementary Fig. S2). When we examined cell functions, TRHEB expressed similar levels of intracellular IFNγ as control cells following stimulation (Fig. 2C), but demonstrated increased specific cytotoxicity against EL4-NP cells (Fig. 2D). In contrast, TPRAS showed both reduced intracellular IFNγ and decreased cytotoxic functions relative to controls (Fig. 2C and D).

Figure 2.

The effect of overexpression of RHEB and PRAS40 upon phenotype, function, and metabolism of CD8+ T cells. A, cell size of TVC, TRHEB, and TPRAS 3 days after transduction. Flow cytometry histograms show representative forward scatter (FSC) and pooled summary data. Horizontal bars represent median. **, P = 0.004; ***, P < 0.001 (one sample t test against a theoretical mean of 1). B, representative flow cytometry plots of surface expression of CD62L 3 days after transduction and pooled summary data showing the difference in CD62L+% (Δ%) between TRHEB and TPRAS and the TVC from the respective experiment. Horizontal bars represent median. ***, P < 0.001; **, P < 0.008 (one sample t test against a theoretical mean of 0). C, representative flow cytometry histograms of IFNγ generation 4 hours after coculture with EL4-NP cells and pooled summary data showing the difference in IFNγ+% (Δ%) between TRHEB and TPRAS and TVC from the respective experiment. Horizontal bars represent median. ***, P < 0.002; n.s., not significant (one sample t test against a theoretical mean of 0). D, cytotoxicity of flow sorted transduced T cells against target EL4-NP-GFP cells (parental cells transduced with pMP71-NP-IRES-GFP). T cells were cocultured for 48 hours with a 1:1 mix of target EL4-NP (GFP+) and nontarget EL4 (GFP) at varying effector-to-target (E:T) ratios and cytotoxicity measured. Graph shows one representative assay of three independent experiments. E, mean ± SEM ECAR and OCR before and after stimulation with anti-CD3 antibody (representative of two independent experiments).

Figure 2.

The effect of overexpression of RHEB and PRAS40 upon phenotype, function, and metabolism of CD8+ T cells. A, cell size of TVC, TRHEB, and TPRAS 3 days after transduction. Flow cytometry histograms show representative forward scatter (FSC) and pooled summary data. Horizontal bars represent median. **, P = 0.004; ***, P < 0.001 (one sample t test against a theoretical mean of 1). B, representative flow cytometry plots of surface expression of CD62L 3 days after transduction and pooled summary data showing the difference in CD62L+% (Δ%) between TRHEB and TPRAS and the TVC from the respective experiment. Horizontal bars represent median. ***, P < 0.001; **, P < 0.008 (one sample t test against a theoretical mean of 0). C, representative flow cytometry histograms of IFNγ generation 4 hours after coculture with EL4-NP cells and pooled summary data showing the difference in IFNγ+% (Δ%) between TRHEB and TPRAS and TVC from the respective experiment. Horizontal bars represent median. ***, P < 0.002; n.s., not significant (one sample t test against a theoretical mean of 0). D, cytotoxicity of flow sorted transduced T cells against target EL4-NP-GFP cells (parental cells transduced with pMP71-NP-IRES-GFP). T cells were cocultured for 48 hours with a 1:1 mix of target EL4-NP (GFP+) and nontarget EL4 (GFP) at varying effector-to-target (E:T) ratios and cytotoxicity measured. Graph shows one representative assay of three independent experiments. E, mean ± SEM ECAR and OCR before and after stimulation with anti-CD3 antibody (representative of two independent experiments).

Close modal

T-cell activation induces a shift from oxidative respiration to aerobic glycolysis, also known as the Warburg effect, a process regulated by mTORC1 (14). We took real-time measurements of ECAR and OCR, which are molecular surrogates for glycolysis and respiration, respectively, before and after TCR stimulation with monoclonal anti-CD3 antibody. Compared with control cells, baseline ECAR was increased in TRHEB but decreased in TPRAS (Fig. 2E). TCR stimulation resulted in rapid increases in ECAR for the three cell types, but it was augmented in TRHEB and diminished in TPRAS compared with TVC. Baseline oxygen consumption was also higher for TRHEB and lower for TPRAS; following stimulation, TRHEB demonstrated a rapid decrease in OCR accentuating the shift from oxidative to glycolytic metabolism, whereas this rapid shift was not seen for TVC and TPRAS (Fig. 2E).

Taken together, these data demonstrate that overexpression of RHEB enhances T-cell effector differentiation as evidenced by increased cell growth, a switch to aerobic glycolysis and greater cytotoxicity, whereas overexpression of PRAS40 induces largely reciprocal effects.

The effect of RHEB and PRAS40 upon the kinetics of T-cell expansion and function in vivo

To assess how overexpression of RHEB and PRAS40 affects T-cell performance in vivo, we performed a competition assay where TVC were mixed in a 1:1 ratio with either TRHEB or TPRAS and injected into sublethally irradiated B6 mice (Fig. 3A). Two days later, mice were vaccinated with irradiated EL4-NP cells to provide antigenic stimulation, and the frequency of transferred cells in peripheral blood was monitored weekly. The two competing T-cell populations were distinguished by the expression of congenic markers Thy1.1 and Thy1.2 (Fig. 3A). Vaccination with EL4-NP resulted in an increased frequency of F5 TCR+ T cells in peripheral blood followed by a sharp decline. During the initial expansion, TRHEB cells accumulated rapidly with a significant advantage over control cells as reflected by a high TRHEB:TVC ratio on day 7 (Fig. 3B). However, this was followed by an accentuated and rapid contraction phase with a persistent disadvantage for TRHEB at all later time points (TRHEB:TVC ratio <1.0). In the other experimental arm, TPRAS demonstrated a relative failure to accumulate in peripheral blood after vaccination, resulting in a constant low TPRAS:TVC ratio (Fig. 3B).

Figure 3.

The effect of overexpression of RHEB and PRAS40 upon in vivo expansion, cytokine production, and phenotype of CD8+ T cells. A, experiment layout. B, competitive expansion between TRHEB (top) or TPRAS (bottom) and TVC in peripheral blood after vaccination with irradiated EL4-NP. Graphs show mean ± SEM accumulation in peripheral blood (left) and ratio against TVC (right). Data shown are representative of two independent experiments (n = 3–4 mice/group each experiment). Blood frequency, TRHEB, or TPRAS versus TVC; *, P < 0.05 (two-tailed t test). TRHEB: TVC and TPRAS: TVC ratios; *, P < 0.05 (one sample t test against a theoretical mean of 1). C–F, experimental layout as in A with responses analyzed at day 8 after vaccination. Data are from two independent experiments. C, surface expression of CD62L. **, P = 0.002; ***, P = 0.001 (two-tailed Wilcoxon matched pairs signed rank test). D, intracellular expression of T-bet and Eomes. Representative flow cytometry plots and summary data. *, P = 0.04; n.s., not significant (two-tailed Wilcoxon matched pairs signed rank test). E, proliferation of transduced CD8+ T cells in vivo. Mice were injected with BrdUrd on day 7 and BrdUrd incorporation measured by flow cytometry. ***, P = 0.001 (two-tailed Wilcoxon matched pairs signed rank test). F, cytokine production of transduced T cells following 4-hour stimulation ex vivo with 1 μmol/L pNP366. Representative flow cytometry plots showing IFNγ and TNFα intracellular staining (left) and summary graph for IFNγ production (right) with data from one of two independent experiments with similar results (n = 6 mice/condition). *, P = 0.03; n.s., not significant (two-tailed Wilcoxon matched pairs signed rank test).

Figure 3.

The effect of overexpression of RHEB and PRAS40 upon in vivo expansion, cytokine production, and phenotype of CD8+ T cells. A, experiment layout. B, competitive expansion between TRHEB (top) or TPRAS (bottom) and TVC in peripheral blood after vaccination with irradiated EL4-NP. Graphs show mean ± SEM accumulation in peripheral blood (left) and ratio against TVC (right). Data shown are representative of two independent experiments (n = 3–4 mice/group each experiment). Blood frequency, TRHEB, or TPRAS versus TVC; *, P < 0.05 (two-tailed t test). TRHEB: TVC and TPRAS: TVC ratios; *, P < 0.05 (one sample t test against a theoretical mean of 1). C–F, experimental layout as in A with responses analyzed at day 8 after vaccination. Data are from two independent experiments. C, surface expression of CD62L. **, P = 0.002; ***, P = 0.001 (two-tailed Wilcoxon matched pairs signed rank test). D, intracellular expression of T-bet and Eomes. Representative flow cytometry plots and summary data. *, P = 0.04; n.s., not significant (two-tailed Wilcoxon matched pairs signed rank test). E, proliferation of transduced CD8+ T cells in vivo. Mice were injected with BrdUrd on day 7 and BrdUrd incorporation measured by flow cytometry. ***, P = 0.001 (two-tailed Wilcoxon matched pairs signed rank test). F, cytokine production of transduced T cells following 4-hour stimulation ex vivo with 1 μmol/L pNP366. Representative flow cytometry plots showing IFNγ and TNFα intracellular staining (left) and summary graph for IFNγ production (right) with data from one of two independent experiments with similar results (n = 6 mice/condition). *, P = 0.03; n.s., not significant (two-tailed Wilcoxon matched pairs signed rank test).

Close modal

In a similar experiment, we determined the phenotypic and functional profiles of TRHEB and TPRAS at 8 days following vaccination. As demonstrated in vitro, splenic TRHEB had reduced CD62L expression whereas TPRAS showed heightened expression compared with controls (Fig. 3C). Expression of the terminal differentiation marker, killer cell lectin-like receptor subfamily G member 1, was similar in TRHEB and TVC, whereas TPRAS demonstrated reduced expression (Supplementary Fig. S3). Consistent with published data for the expression of the T-box transcription factors in effector cells (18), TRHEB were characterized by the increased number of cells that possessed high intracellular expression of T-bet but low expression of Eomesodermin (Eomes), while the frequency of T-bethigh Eomeslow cells was lower in TPRAS and TVC (Fig. 3D). We evaluated in vivo proliferation by measuring BrdUrd incorporation. As shown in Fig. 3E, TRHEB demonstrated greater proliferation than control cells, whereas TPRAS were hypoproliferative. Upon ex vivo restimulation with antigen, TRHEB produced similar levels of IFNγ and TNFα to TVC, whereas TPRAS consistently produced less (Fig. 3F).

When T cells were tracked over a longer period following vaccination (to day 135), we found that TPRAS in the spleen and lymph node retained high expression levels of CD62L and CD127, markers associated central memory differentiation (Fig. 4A). In contrast, both TVC and TRHEB populations had a greater proportion of CD62LlowCD127low cells. Thus, although the ratio of TPRAS to TVC was low when considered at a whole population level, the frequencies of cells with a central memory phenotype were similar (Fig. 4B). In contrast, at both whole population and central memory level, the ratios of TRHEB to TVC were <1.0 (Fig. 4B).

Figure 4.

Increased central memory differentiation of CD8+ T cells expressing PRAS40. A, surface expression of CD62L and CD127 in transduced CD8+ T cells in spleen (Sp) and lymph nodes (LN) 135 days after vaccination (experiment layout as in Fig. 3A). Summary graph showing data pooled from two independent experiments. *, P = 0.03; n.s., not significant (two-tailed Wilcoxon matched pairs signed rank test). B, ratio of TRHEB (top) and TPRAS (bottom) against TVC in whole population or within the CD62L+CD127+ population.

Figure 4.

Increased central memory differentiation of CD8+ T cells expressing PRAS40. A, surface expression of CD62L and CD127 in transduced CD8+ T cells in spleen (Sp) and lymph nodes (LN) 135 days after vaccination (experiment layout as in Fig. 3A). Summary graph showing data pooled from two independent experiments. *, P = 0.03; n.s., not significant (two-tailed Wilcoxon matched pairs signed rank test). B, ratio of TRHEB (top) and TPRAS (bottom) against TVC in whole population or within the CD62L+CD127+ population.

Close modal

Taken together, these data show that overexpression of RHEB drives the differentiation and rapid proliferation of short-lived T-bethigh effectors in secondary lymphoid organs. On the other hand, constitutive overexpression of PRAS40 induces quiescence and the acquisition of a central memory phenotype.

TRHEB demonstrate enhanced early infiltration and antitumor effects

To address how mTORC1 activity affects control of tumor by CD8+ T cells, we injected EL4-NP cells into the flanks of sublethally irradiated B6 mice and 5 days later adoptively transferred TVC, TRHEB, and TPRAS bearing the F5 TCR, or mock-transduced T cells. Tumors grew rapidly in mice receiving mock-transduced T cells with 100% lethality observed before day 20, while TVC controlled tumor growth in 60% of mice by day 60 (Fig. 5A). Adoptive therapy with TRHEB improved tumor protection and resulted in 85% tumor-free survival by day 60. Consistent with their overall quiescence, TPRAS were completely unable to control tumor growth and behaved like mock-transduced cells with 100% lethality before day 20 (Fig. 5A). Measurement of tumor size indicated that TRHEB reduced tumor burden more rapidly than TVC (Fig. 5B). The reisolation of tumors that grew progressively in mice treated with TRHEB and TVC revealed that F5 TCR-transduced CD8+ T cells were unable to recognize these re-isolates, although the same T cells efficiently recognized the parental EL4-NP tumor cells resulting in antigen-specific IFNγ production (Fig. 5C). The reisolated tumors continued to express H2-Db, indicating that the observed immune escape was not due to loss of MHC class I surface expression (Fig. 5C).

Figure 5.

Overexpression of RHEB increases whereas PRAS40 impairs antitumor effects of CD8+ T cells. A, sublethally irradiated B6 mice were subcutaneously inoculated with 1 × 106 EL4-NP cells and 5 days later, received 0.1 to 1 × 105 TVC, TRHEB, TPRAS, or mock-transduced cells intravenously. Kaplan–Meier plots of survival are pooled from 5 independent experiments (n = 37 TVC, n = 28 TRHEB, n = 10 TPRAS, and n = 18 mock). Statistical significance tested by log-rank (Mantel–Cox). B, tumor size over time. Graphs of a representative tumor challenge experiment showing mean ± SEM tumor size (n = 5 mice/group). C, tumors were isolated from mice unable to control tumor growth and cultured for 7 days before analysis. Histograms (left) show surface expression of H-2Db on parental EL4-NP line (par.) and EL4-NP isolated from mice treated with TVC or TRHEB cell. Gray filled histograms show isotype control staining. F5-TCR–transduced CD8+ T cells were cocultured with EL4 (antigen negative), parental EL4-NP line or reisolated EL4-NP, and stained for IFNγ 4 hours later. Flow cytometric plots show intracellular IFNγ staining (right). D, mice bearing EL4-NP tumors for 5 days received firefly luciferase+ TRHEB, TPRAS, TVC, or mock-transduced CD8+ T cells. At designated time points, 150 μg/g luciferin was administered and bioluminescence measured as an estimate of T-cell accumulation. Representative image at day 6 after T-cell injection (left), with pooled data for day 6 (top right) and time course over 14 days (bottom right) from two independent experiments (n = 15 mice for TRHEB, n = 10 mice for TPRAS, n = 12 mice for TVC, and n = 7 mice for mock). *, P = 0.01; **, P = 0.03 (two-tailed Mann–Whitney test).

Figure 5.

Overexpression of RHEB increases whereas PRAS40 impairs antitumor effects of CD8+ T cells. A, sublethally irradiated B6 mice were subcutaneously inoculated with 1 × 106 EL4-NP cells and 5 days later, received 0.1 to 1 × 105 TVC, TRHEB, TPRAS, or mock-transduced cells intravenously. Kaplan–Meier plots of survival are pooled from 5 independent experiments (n = 37 TVC, n = 28 TRHEB, n = 10 TPRAS, and n = 18 mock). Statistical significance tested by log-rank (Mantel–Cox). B, tumor size over time. Graphs of a representative tumor challenge experiment showing mean ± SEM tumor size (n = 5 mice/group). C, tumors were isolated from mice unable to control tumor growth and cultured for 7 days before analysis. Histograms (left) show surface expression of H-2Db on parental EL4-NP line (par.) and EL4-NP isolated from mice treated with TVC or TRHEB cell. Gray filled histograms show isotype control staining. F5-TCR–transduced CD8+ T cells were cocultured with EL4 (antigen negative), parental EL4-NP line or reisolated EL4-NP, and stained for IFNγ 4 hours later. Flow cytometric plots show intracellular IFNγ staining (right). D, mice bearing EL4-NP tumors for 5 days received firefly luciferase+ TRHEB, TPRAS, TVC, or mock-transduced CD8+ T cells. At designated time points, 150 μg/g luciferin was administered and bioluminescence measured as an estimate of T-cell accumulation. Representative image at day 6 after T-cell injection (left), with pooled data for day 6 (top right) and time course over 14 days (bottom right) from two independent experiments (n = 15 mice for TRHEB, n = 10 mice for TPRAS, n = 12 mice for TVC, and n = 7 mice for mock). *, P = 0.01; **, P = 0.03 (two-tailed Mann–Whitney test).

Close modal

To determine the extent to which manipulation of mTORC1 activity influenced the capacity of adoptively transferred CD8+ T cells to infiltrate tumors, we transduced T cells from Firefly Luciferase+ donors and measured their accumulation at the tumor site by bioluminescence at timed intervals following injection (Fig. 5D). TRHEB infiltrated tumors significantly earlier than TVC with the median bioluminescence value being 2-fold greater at day 6, whereas TPRAS infiltration was significantly delayed with bioluminescence values ∼3-fold lower than controls. Furthermore, in contrast with their behavior in blood and secondary lymphoid organs, TRHEB numbers expanded at the tumor site over the course of the experiment. Together, these experiments showed that overexpression of RHEB increased the capacity of therapeutic T cells to rapidly infiltrate tumors and reduced the development of progressively growing immune escape variants, whereas constitutive overexpression of PRAS40 abolished protective tumor immunity.

Long-term immune memory is impaired following constitutive overexpression of RHEB or PRAS40

To test how constitutive overexpression of RHEB or PRAS40 would affect the capacity of F5 TCR-transduced T cells to mount a recall response, we performed the competition experiment as outlined in Fig. 3 by cotransferring 1:1 mixes of either TRHEB with TVC or TPRAS with TVC into sublethally irradiated mice and then vaccinating them with EL4-NP cells on day 2. At day 40, we repeated the vaccination and tracked the response of transduced T cells in peripheral blood. As shown in Supplementary Fig. S4, TRHEB mounted a reduced but measurable recall response compared with TVC. Despite increased frequencies of TPRAS with a memory phenotype at day 40, no recall response was detectable in peripheral blood following revaccination, indicating that quiescence was permanently induced in this population (Supplementary Fig. S4). In mice that had previously eradicated subcutaneous EL4-NP tumors and were later revaccinated, we observed a similar relative reduction in recall responsiveness of TRHEB (Supplementary Fig. S4); no long-term survivors were observed in tumor-bearing recipients of TPRAS making analysis of memory responses in this context not possible. Taken together, these data show that fixing mTORC1 activity at high or low levels reduces recall responsiveness of CD8+ T cells.

Inducible PRAS40 expression during initial antigen encounter increases recall immunity

We reasoned that transient interruption of mTORC1 activity during the initial period of antigen exposure, followed by re-enablement of function at the time of recall, would increase recall immunity and the potential for durable immune surveillance. We therefore developed an inducible system for PRAS40 expression, by adapting a pSERS retroviral vector incorporating a Tet-ON “all-in-one” inducible system for regulating gene expression (Fig. 6A; ref. 19). The vector was modified to encode PRAS40 marked with GFP (iPRAS) or GFP alone (iGFP), under the control of a minimal promoter regulated by a Tet operon (TetO), which renders the promoter transcriptionally silent. The vector also encodes the reverse tetracycline–controlled transactivator (rtTA2-M2), which is marked by the Q8 surface tag (a 16-residue sequence of CD34 linked to a CD8 stalk; ref. 20) under the control of the constitutively active human PGK promoter. In the presence of doxycycline, rtTA2-M2 binds the TetO and promotes transcription of the gene of interest. Comparable dose-dependent induction of GFP expression was observed for both iPRAS and iGFP vectors with a maximum induction achieved with 1 μg/mL doxycycline (Supplementary Fig. S5). To assess the effect of inducible PRAS40 expression on mTORC1 signaling, CD8+ T cells transduced with the F5 TCR and either iPRAS (TiPRAS) or iGFP (TiGFP) were treated with 1 μg/mL doxycycline for 24 hours, or left untreated, before being stimulated with EL4-NP cells (Fig. 6B). In the absence of doxycycline, no GFP expression was detected in TiGFP and TiPRAS, and both cell types responded to stimulation by increasing the levels of p-S6 equally. Treatment with doxycycline induced GFP expression in both groups, but GFP+ cells from the TiPRAS group showed reduced levels of p-S6 compared with the GFP+ cells from the TiGFP group and the GFP (un-induced) fraction of both groups (Fig. 6B).

Figure 6.

Inducible PRAS40 expression during initial antigen encounter increases recall immunity. A, pSERS doxycycline-inducible vector maps. LTR, long terminal repeat; TetO7, tetracycline operator; FMD-2A, foot-and-mouth disease 2A sequence; hPGK, human phosphoglycerate kinase; rtA2-M2, optimized reverse tetracycline-controlled transactivator; W-PRE, Woodchuck hepatitis post-transcriptional regulatory element; P2A, picornavirus 2A sequence. B, TiGFP or TiPRAS were split in two lots, one receiving 1 μg/mL doxycycline (Dox) and the other left untreated. Twenty-four hours later, CD8+ T cells were stimulated with anti-CD3/CD28 microbeads for further 24 hours and cells stained for p-S6. Flow cytometry plots shown are representative of three independent experiments and show GFP expression and p-S6 levels of both GFP+ and GFP populations. C, sublethally irradiated B6 mice were inoculated with 1 × 106 EL4-NP subcutaneously and divided in two cohorts, one receiving 2 mg/mL doxycycline in the drinking water and the other left untreated. Five days after tumor injection, mice received a 1:1 mix of TiGFP and TiPRAS that could be tracked with the congenic markers Thy1.1 and CD45.1, respectively. Doxycycline treatment was stopped at day 30 after tumor injection. At day 45, mice were rechallenged with irradiated EL4-NP (R). Tumor size during experiment (middle); data are representative of two independent experiments (n = 3 mock and n = 5 for +Dox and no Dox groups). Mean ± SEM TiPRAS:TiGFP ratio in peripheral blood after rechallenge at day 45 (right). Summary data are pooled from two independent experiments (n = 8 mice for no Dox and n = 6 mice for +Dox). *, P = 0.013 (two-tailed t test).

Figure 6.

Inducible PRAS40 expression during initial antigen encounter increases recall immunity. A, pSERS doxycycline-inducible vector maps. LTR, long terminal repeat; TetO7, tetracycline operator; FMD-2A, foot-and-mouth disease 2A sequence; hPGK, human phosphoglycerate kinase; rtA2-M2, optimized reverse tetracycline-controlled transactivator; W-PRE, Woodchuck hepatitis post-transcriptional regulatory element; P2A, picornavirus 2A sequence. B, TiGFP or TiPRAS were split in two lots, one receiving 1 μg/mL doxycycline (Dox) and the other left untreated. Twenty-four hours later, CD8+ T cells were stimulated with anti-CD3/CD28 microbeads for further 24 hours and cells stained for p-S6. Flow cytometry plots shown are representative of three independent experiments and show GFP expression and p-S6 levels of both GFP+ and GFP populations. C, sublethally irradiated B6 mice were inoculated with 1 × 106 EL4-NP subcutaneously and divided in two cohorts, one receiving 2 mg/mL doxycycline in the drinking water and the other left untreated. Five days after tumor injection, mice received a 1:1 mix of TiGFP and TiPRAS that could be tracked with the congenic markers Thy1.1 and CD45.1, respectively. Doxycycline treatment was stopped at day 30 after tumor injection. At day 45, mice were rechallenged with irradiated EL4-NP (R). Tumor size during experiment (middle); data are representative of two independent experiments (n = 3 mock and n = 5 for +Dox and no Dox groups). Mean ± SEM TiPRAS:TiGFP ratio in peripheral blood after rechallenge at day 45 (right). Summary data are pooled from two independent experiments (n = 8 mice for no Dox and n = 6 mice for +Dox). *, P = 0.013 (two-tailed t test).

Close modal

To test the effect of inducible PRAS40 overexpression in vivo, we inoculated sublethally irradiated B6 mice with subcutaneous EL4-NP tumors and divided the recipients into two cohorts, one receiving 2 mg/mL doxycycline in the drinking water and one left untreated; 5 days later, we adoptively transferred a 1:1 mix of TiPRAS with TiGFP (Fig. 6C). The tumors were controlled to a similar extent in both the doxycycline (day 0–day 30) and no doxycycline groups (Fig. 6C). To test recall immunity, we rechallenged mice in both groups with irradiated EL4-NP cells on day 45. Relative to controls, we observed significantly greater expansion of TiPRAS relative to TiGFP under conditions where they had been exposed to doxycycline during the initial response, whereas expansion of TiPRAS and TiGFP was similar in recipient mice that had not received doxycycline (Fig. 6C).

In contrast with the constitutive pMP71-PRAS40 vector, where all transduced T cells were uniformly GFPhigh and mTORC1 was inhibited to a similar extent (Supplementary Fig. S6), we found that doxycycline induced the GFP reporter in vitro at varying levels that were proportional to the Q8 tag (Fig. 7A). Consistent with the inhibitory effects of PRAS40, S6 phosphorylation was inversely related to GFP expression in the TiPRAS group, with GFPhigh cells showing the greatest levels of inhibition and GFPlow cells showing intermediate inhibition, whereas GFPneg cells behaved like controls. When we examined the fate of F5 TCR+ T-cell populations from each group in vivo in the presence of doxycycline, we consistently observed that expression of both the Q8 tag (marking cells transduced with the Tet-On vector) and the GFP reporter was much lower in TiPRAS than TiGFP (Fig. 7B). Taken together, these data show that the initial response mediated by TiPRAS was characterized by in vivo selection for cells with low- rather than high-level inhibition of mTORC1.

Figure 7.

Relationship between level of mTORC1 inhibition and memory differentiation. A, plots show expression of Q8 and GFP used to delineate 6 gates (1, not-transduced; 2, transduced notinduced; 3–6, increasing levels of GFP expression). Histograms show level of p-S6 within each gate from 1 to 6. Gray histogram, unstimulated cells. Flow cytometry plots from same experiment shown in Fig. 6B. B, flow cytometry plots showing expression of Q8 tag (marks cells transduced with inducible vector) and GFP (marks induction of gene of interest) in TiGFP and TiPRAS 17 days after tumor injection (left). Top (no Dox) and bottom (+Dox) are pairs from the same individual. Pooled summary data from two independent experiments showing the proportion of TiGFP and TiPRAS cells that are Q8+ (middle) and the proportion of Q8+ that are GFP+ (right). ***, P < 0.0001 (two-tailed Mann–Whitney test). C, proportion of Q8+ for TiPRAS cells over time in mice receiving doxycycline or untreated in same experiment as shown in Fig. 6C. Left, representative plots showing Q8 expression at days 10 and 70 following T-cell injection. Right, summary data are shown as mean ± SEM from one representative experiment of three independent experiments performed (n = 4–7 mice). *, P < 0.0001 for TiPRAS + Dox and not significant for TiPRAS no Dox (two-tailed Mann–Whitney test).

Figure 7.

Relationship between level of mTORC1 inhibition and memory differentiation. A, plots show expression of Q8 and GFP used to delineate 6 gates (1, not-transduced; 2, transduced notinduced; 3–6, increasing levels of GFP expression). Histograms show level of p-S6 within each gate from 1 to 6. Gray histogram, unstimulated cells. Flow cytometry plots from same experiment shown in Fig. 6B. B, flow cytometry plots showing expression of Q8 tag (marks cells transduced with inducible vector) and GFP (marks induction of gene of interest) in TiGFP and TiPRAS 17 days after tumor injection (left). Top (no Dox) and bottom (+Dox) are pairs from the same individual. Pooled summary data from two independent experiments showing the proportion of TiGFP and TiPRAS cells that are Q8+ (middle) and the proportion of Q8+ that are GFP+ (right). ***, P < 0.0001 (two-tailed Mann–Whitney test). C, proportion of Q8+ for TiPRAS cells over time in mice receiving doxycycline or untreated in same experiment as shown in Fig. 6C. Left, representative plots showing Q8 expression at days 10 and 70 following T-cell injection. Right, summary data are shown as mean ± SEM from one representative experiment of three independent experiments performed (n = 4–7 mice). *, P < 0.0001 for TiPRAS + Dox and not significant for TiPRAS no Dox (two-tailed Mann–Whitney test).

Close modal

Although outcompeted during the initial response, we reasoned that TiPRAS with higher levels of mTORC1 inhibition would be more efficient at transitioning to memory and would therefore contribute more to the recall response once the suppression of mTORC1 was relieved. This hypothesis could be tested in vivo because the Q8 tag identified TiPRAS populations that during initial antigen exposure in the presence of doxycycline had either low or high levels of mTORC1 inhibition and could therefore be used to map the fate of such cells at a later phase of the response. As shown in Fig. 7C, the proportion of Q8-expressing cells significantly increased from their initial low level among the F5 TCR+ population upon withdrawal of doxycycline and following antigenic rechallenge. In contrast, no changes in Q8 expression were noted in the group that received no doxycycline. Thus, TiPRAS memory recall responses were derived mainly from low frequency cells with higher levels of mTORC1 inhibition during initial antigen encounter.

By direct comparison in the same model systems, our data show for the first time a role for enhancement as well inhibition of intrinsic mTORC1 function in generating effective antitumor responses by T cells. Enhanced mTORC1 activity accelerated T-cell infiltration at the tumor site and exerted more aggressive immune pressure on the growing tumors, which was associated with a reduction in the selection of tumor escape variants driving progressive disease.

Our data also show that levels of RHEB are limiting for full mTORC1 activation and maximal effector differentiation following TCR activation. In sharp contrast with the peripheral blood and lymphoid organs where TRHEB expanded rapidly but were only short-lived, TRHEB infiltration of EL4-NP was sustained over the course of at least 2 weeks. These data highlight the need to evaluate T-cell responses both within tumors and secondary lymphoid organs where manipulation of mTORC1 activity may have distinct effects. Our findings are also in accordance with recent studies demonstrating that enhanced glycolytic capacity of effector CD8+ T cells is linked to increased antitumor immunity (21, 22). As antitumor T cells begin to infiltrate tumors, they encounter multiple factors likely to inactivate mTORC1 (8, 9). For example, tumor-associated myeloid cells upregulate the arginase pathway to deplete the microenvironment of arginine, leading to impaired T-cell functions (23, 24). In this context, we have observed that overexpression of RHEB increases resistance of proliferating T cells to arginine depletion in vitro (data not shown) indicative that TRHEB have greater fitness in contexts where mTORC1 activity would ordinarily be suppressed.

Failure of transferred T cells to persist or to be replenished through self-renewal may lead to relapse (25–27), particularly under conditions where transforming stimuli remain (e.g., from viruses) or where the initial response cannot eliminate all cancerous cells and long-term immune surveillance is required to keep them in check. Previous studies have demonstrated that using rapamycin to reduce mTORC1 signaling in a responding CD8+ T-cell population can improve memory potential and antiviral or antitumor immunity (18, 28), although the extent to which this strategy works while also maintaining the capacity for an effector response is highly variable and dependent upon the model systems studied (29–31). Furthermore, this strategy may be complicated by systemic “on target” effects of the drug, for example enhancement of Treg expansion (29–32). Our data show that constitutive mTORC1 inhibition through overexpression of PRAS40 increases formation of memory phenotype CD127+CD62L+ CD8+ T cells; however, the transduced cells enter a state of permanent quiescence incapable of generating effector cells either during the primary response or recall. Indeed, although knockdown of RAPTOR and MTOR (both components of the mTORC1 complex) has been shown to enhance memory differentiation of CD8+ T cells (28), we have found that such cells were incapable of expansion following transfer to tumor-bearing hosts (Supplementary Fig. S7). Thus, inhibition of mTORC1 with rapamycin may be most effective at enhancing antitumor CD8+ T-cell immunity when employed at low doses that do not ablate the initial expansion of effector cells (28) or when it is curtailed early following priming to maximize secondary effector responses (18, 31).

In contrast with the constitutive expression system where PRAS40 expression was uniformly high, PRAS40 was induced across a range of values using the Tet-ON system. Upon transfer in vivo, T cells with higher levels of PRAS40 were rapidly outcompeted by effector cells where the protein had been induced at lower levels. However, low-frequency memory precursor cells, characterized by higher levels of initial mTORC1 inhibition, survived and reconstituted the secondary response when mTORC1 signaling was no longer suppressed. Our findings are consistent with recent fate mapping studies for CD8+ T cells responding to vaccines (1, 2) and show clearly that cells destined for memory differentiation contribute less to the initial control of tumor growth. Furthermore, we have demonstrated for the first time an intrinsic requirement for dynamic rather than fixed tuning of the mTOR pathway to generate functional memory cells. It is possible that the recently described aptamer-based targeting of antigen-activated CD8+ T cells with siRNAs that knockdown RAPTOR (29) enables a similar temporary interruption of mTORC1 signaling; in this case, non- or minimally targeted cells would constitute the initial effector response, whereas T cells with greater initial levels of inhibition would generate memory precursors capable of mounting recall responses as the effect of the introduced siRNA wanes. The importance of dynamic tuning of mTORC1 has parallels with other strategies to induce memory where effector CD8+ T-cell differentiation is blocked only during initial priming ex vivo, for example by interrupting glycolytic flux (22) or Akt activation (33).

In summary, our data support the design of translational strategies that generate heterogeneous T-cell immunity with the appropriate balance between self-renewal and effector differentiation. Precise clinical application of this approach will require further dissection of how mTOR functions of adoptively transferred T cells are modulated in the tumor environment of individual cancer types and at different stages of disease.

M. Pule is CSO at Autolus Ltd, reports receiving a commercial research grant from Cellectis, and has ownership interest in patents in the field of T-cell engineering. H. Stauss is a consultant and share holder at Cell Medica, and consultant at Sofinnova. No potential conflicts of interest were disclosed by the other authors.

Conception and design: P. Veliça, M. Zech, H. Stauss, R. Chakraverty

Development of methodology: P. Veliça, M. Zech, M. Pule, H. Stauss, R. Chakraverty

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): P. Veliça, M. Zech, S. Henson, A. Holler, T. Manzo, R. Pike, P. Santos e Sousa

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): P. Veliça, M. Zech, T. Manzo, M. Pule, H. Stauss, R. Chakraverty

Writing, review, and/or revision of the manuscript: P. Veliça, M. Zech, M. Pule, H. Stauss, R. Chakraverty

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): M. Zech, L. Zhang, B. Schiedlmeier

Study supervision: H. Stauss, R. Chakraverty

This study was supported by Cancer Research UK (C34276/A11995 to H. Stauss and R. Chakraverty); Leukaemia and Lymphoma Research UK (12006 to R. Chakraverty; 13004 to H. Stauss); European Union Framework 6 Integrated Project (LSHC-CT-2005-018914 to H. Stauss).

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.
Buchholz
VR
,
Flossdorf
M
,
Hensel
I
,
Kretschmer
L
,
Weissbrich
B
,
Graf
P
, et al
Disparate individual fates compose robust CD8+ T cell immunity
.
Science
2013
;
340
:
630
5
.
2.
Gerlach
C
,
Rohr
JC
,
Perie
L
,
van Rooij
N
,
van Heijst
JW
,
Velds
A
, et al
Heterogeneous differentiation patterns of individual CD8+ T cells
.
Science
2013
;
340
:
635
9
.
3.
Gattinoni
L
,
Klebanoff
CA
,
Palmer
DC
,
Wrzesinski
C
,
Kerstann
K
,
Yu
Z
, et al
Acquisition of full effector function in vitro paradoxically impairs the in vivo antitumor efficacy of adoptively transferred CD8+ T cells
.
J Clin Invest
2005
;
115
:
1616
26
.
4.
Gattinoni
L
,
Zhong
XS
,
Palmer
DC
,
Ji
Y
,
Hinrichs
CS
,
Yu
Z
, et al
Wnt signaling arrests effector T cell differentiation and generates CD8+ memory stem cells
.
Nat Med
2009
;
15
:
808
13
.
5.
Hinrichs
CS
,
Borman
ZA
,
Cassard
L
,
Gattinoni
L
,
Spolski
R
,
Yu
Z
, et al
Adoptively transferred effector cells derived from naive rather than central memory CD8+ T cells mediate superior antitumor immunity
.
Proc Natl Acad Sci U S A
2009
;
106
:
17469
74
.
6.
Hinrichs
CS
,
Borman
ZA
,
Gattinoni
L
,
Yu
Z
,
Burns
WR
,
Huang
J
, et al
Human effector CD8+ T cells derived from naive rather than memory subsets possess superior traits for adoptive immunotherapy
.
Blood
2011
;
117
:
808
14
.
7.
Klebanoff
CA
,
Gattinoni
L
,
Torabi-Parizi
P
,
Kerstann
K
,
Cardones
AR
,
Finkelstein
SE
, et al
Central memory self/tumor-reactive CD8+ T cells confer superior antitumor immunity compared with effector memory T cells
.
Proc Natl Acad Sci U S A
2005
;
102
:
9571
6
.
8.
Baitsch
L
,
Fuertes-Marraco
SA
,
Legat
A
,
Meyer
C
,
Speiser
DE
. 
The three main stumbling blocks for anticancer T cells
.
Trends Immunol
2012
;
33
:
364
72
.
9.
Palazon
A
,
Aragones
J
,
Morales-Kastresana
A
,
de Landazuri
MO
,
Melero
I
. 
Molecular pathways: hypoxia response in immune cells fighting or promoting cancer
.
Clin Cancer Res
2012
;
18
:
1207
13
.
10.
Duran
RV
,
Hall
MN
. 
Regulation of TOR by small GTPases
.
EMBO Rep
2012
;
13
:
121
8
.
11.
Wang
H
,
Zhang
Q
,
Wen
Q
,
Zheng
Y
,
Lazarovici
P
,
Jiang
H
, et al
Proline-rich Akt substrate of 40 kDa (PRAS40): a novel downstream target of PI3k/Akt signaling pathway
.
Cell Signal
2012
;
24
:
17
24
.
12.
Ahmadi
M
,
King
JW
,
Xue
SA
,
Voisine
C
,
Holler
A
,
Wright
GP
, et al
CD3 limits the efficacy of TCR gene therapy in vivo
.
Blood
2011
;
118
:
3528
37
.
13.
Gubser
PM
,
Bantug
GR
,
Razik
L
,
Fischer
M
,
Dimeloe
S
,
Hoenger
G
, et al
Rapid effector function of memory CD8+ T cells requires an immediate-early glycolytic switch
.
Nat Immunol
2013
;
14
:
1064
72
.
14.
Powell
JD
,
Pollizzi
KN
,
Heikamp
EB
,
Horton
MR
. 
Regulation of immune responses by mTOR
.
Annu Rev Immunol
2012
;
30
:
39
68
.
15.
Xu
X
,
Ye
L
,
Araki
K
,
Ahmed
R
. 
mTOR, linking metabolism and immunity
.
Semin Immunol
2012
;
24
:
429
35
.
16.
Fingar
DC
,
Salama
S
,
Tsou
C
,
Harlow
E
,
Blenis
J
. 
Mammalian cell size is controlled by mTOR and its downstream targets S6K1 and 4EBP1/eIF4E
.
Genes Dev
2002
;
16
:
1472
87
.
17.
Sinclair
LV
,
Finlay
D
,
Feijoo
C
,
Cornish
GH
,
Gray
A
,
Ager
A
, et al
Phosphatidylinositol-3-OH kinase and nutrient-sensing mTOR pathways control T lymphocyte trafficking
.
Nat Immunol
2008
;
9
:
513
21
.
18.
Rao
RR
,
Li
Q
,
Odunsi
K
,
Shrikant
PA
. 
The mTOR kinase determines effector versus memory CD8+ T cell fate by regulating the expression of transcription factors T-bet and Eomesodermin
.
Immunity
2010
;
32
:
67
78
.
19.
Heinz
N
,
Schambach
A
,
Galla
M
,
Maetzig
T
,
Baum
C
,
Loew
R
, et al
Retroviral and transposon-based tet-regulated all-in-one vectors with reduced background expression and improved dynamic range
.
Hum Gene Ther
2011
;
22
:
166
76
.
20.
Philip
B
,
Kokalaki
E
,
Mekkaoui
L
,
Thomas
S
,
Straathof
K
,
Flutter
B
, et al
A highly compact epitope-based marker/suicide gene for easier and safer T-cell therapy
.
Blood
2014
;
124
:
1277
87
.
21.
Doedens
AL
,
Phan
AT
,
Stradner
MH
,
Fujimoto
JK
,
Nguyen
JV
,
Yang
E
, et al
Hypoxia-inducible factors enhance the effector responses of CD8(+) T cells to persistent antigen
.
Nat Immunol
2013
;
14
:
1173
82
.
22.
Sukumar
M
,
Liu
J
,
Ji
Y
,
Subramanian
M
,
Crompton
JG
,
Yu
Z
, et al
Inhibiting glycolytic metabolism enhances CD8+ T cell memory and antitumor function
.
J Clin Invest
2013
;
123
:
4479
88
.
23.
Norian
LA
,
Rodriguez
PC
,
O'Mara
LA
,
Zabaleta
J
,
Ochoa
AC
,
Cella
M
, et al
Tumor-infiltrating regulatory dendritic cells inhibit CD8+ T cell function via L-arginine metabolism
.
Cancer Res
2009
;
69
:
3086
94
.
24.
Rodriguez
PC
,
Quiceno
DG
,
Zabaleta
J
,
Ortiz
B
,
Zea
AH
,
Piazuelo
MB
, et al
Arginase I production in the tumor microenvironment by mature myeloid cells inhibits T-cell receptor expression and antigen-specific T-cell responses
.
Cancer Res
2004
;
64
:
5839
49
.
25.
Brentjens
RJ
,
Riviere
I
,
Park
JH
,
Davila
ML
,
Wang
X
,
Stefanski
J
, et al
Safety and persistence of adoptively transferred autologous CD19-targeted T cells in patients with relapsed or chemotherapy refractory B-cell leukemias
.
Blood
2011
;
118
:
4817
28
.
26.
Dudley
ME
,
Yang
JC
,
Sherry
R
,
Hughes
MS
,
Royal
R
,
Kammula
U
, et al
Adoptive cell therapy for patients with metastatic melanoma: evaluation of intensive myeloablative chemoradiation preparative regimens
.
J Clin Oncol
2008
;
26
:
5233
9
.
27.
Louis
CU
,
Savoldo
B
,
Dotti
G
,
Pule
M
,
Yvon
E
,
Myers
GD
, et al
Antitumor activity and long-term fate of chimeric antigen receptor-positive T cells in patients with neuroblastoma
.
Blood
2011
;
118
:
6050
6
.
28.
Araki
K
,
Turner
AP
,
Shaffer
VO
,
Gangappa
S
,
Keller
SA
,
Bachmann
MF
, et al
mTOR regulates memory CD8 T-cell differentiation
.
Nature
2009
;
460
:
108
12
.
29.
Berezhnoy
A
,
Castro
I
,
Levay
A
,
Malek
TR
,
Gilboa
E
. 
Aptamer-targeted inhibition of mTOR in T cells enhances antitumor immunity
.
J Clin Invest
2014
;
124
:
188
97
.
30.
Goldberg
EL
,
Smithey
MJ
,
Lutes
LK
,
Uhrlaub
JL
,
Nikolich-Zugich
J
. 
Immune memory-boosting dose of rapamycin impairs macrophage vesicle acidification and curtails glycolysis in effector CD8 cells, impairing defense against acute infections
.
J Immunol
2014
;
193
:
757
63
.
31.
Li
Q
,
Rao
R
,
Vazzana
J
,
Goedegebuure
P
,
Odunsi
K
,
Gillanders
W
, et al
Regulating mammalian target of rapamycin to tune vaccination-induced CD8(+) T cell responses for tumor immunity
.
J Immunol
2012
;
188
:
3080
7
.
32.
Wang
Y
,
Sparwasser
T
,
Figlin
R
,
Kim
HL
. 
Foxp3+ T cells inhibit antitumor immune memory modulated by mTOR inhibition
.
Cancer Res
2014
;
74
:
2217
28
.
33.
van der Waart
AB
,
van de Weem
NM
,
Maas
F
,
Kramer
CS
,
Kester
MG
,
Falkenburg
JH
, et al
Inhibition of Akt signaling promotes the generation of superior tumor-reactive T cells for adoptive immunotherapy
.
Blood
2014
;
124
:
3490
500
.