Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2026 Aug 22.
Published in final edited form as: Immunity. 2016 Feb 16;44(2):380–390. doi: 10.1016/j.immuni.2016.01.021

Distinct Signaling of Coreceptors Regulates Specific Metabolism Pathways and Impacts Memory Development in CAR T Cells

Omkar U Kawalekar 1, Roddy S O’Connor 2, Joseph A Fraietta 1, Lili Guo 3, Shannon E McGettigan 1, Avery D Posey Jr 1, Prachi R Patel 1, Sonia Guedan 1, John Scholler 1, Brian Keith 1, Nathaniel W Snyder 3,4, Ian A Blair 3, Michael C Milone 1,2,5, Carl H June 1,2,5,*
PMCID: PMC13498396  NIHMSID: NIHMS2129298  PMID: 26885860

SUMMARY

Chimeric antigen receptors (CARs) redirect T cell cytotoxicity against cancer cells, providing a promising approach to cancer immunotherapy. Despite extensive clinical use, the attributes of CAR co-stimulatory domains that impact persistence and resistance to exhaustion of CAR-T cells remain largely undefined. Here, we report the influence of signaling domains of coreceptors CD28 and 4–1BB on the metabolic characteristics of human CAR T cells. Inclusion of 4–1BB in the CAR architecture promoted the outgrowth of CD8+ central memory T cells that had significantly enhanced respiratory capacity, increased fatty acid oxidation and enhanced mitochondrial biogenesis. In contrast, CAR T cells with CD28 domains yielded effector memory cells with a genetic signature consistent with enhanced glycolysis. These results provide, at least in part, a mechanistic insight into the differential persistence of CAR-T cells expressing 4–1BB or CD28 signaling domains in clinical trials and inform the design of future CAR T cell therapies.

In Brief

Despite the successes of chimeric antigen receptor (CAR) T cells in treating hematologic malignancies, the impact of various CAR signaling domains remains unclear. Kawalekar et al. show that the choice of signaling domain can metabolically reprogram T cells to alter its mitochondrial biogenesis and persistence.

Graphical Abstract

graphic file with name nihms-2129298-f0001.webp

INTRODUCTION

Adoptive immunotherapy based on the infusion of genetically redirected autologous T cells has demonstrated promise for the treatment of both hematologic malignancies and solid tumors. Accordingly, multiple gain-of-function strategies to endow T cells with desired antigen receptors, based on either T cell receptors (TCRs) or chimeric antigen receptors (CARs), have been described (June et al., 2015). Among several proposed strategies, the use of CARs has shown potent effects in augmenting the immune response to cancers, particularly B cell malignancies (Brentjens et al., 2013; Grupp et al., 2013; Kalos et al., 2011). Although CAR T cell therapy can have a significant impact on disease clearance, the essential components of a clinically successful CAR and how they influence therapeutic efficacy remain largely undefined (Kalos and June, 2013).

CARs are synthetic molecules that combine the effector functions of T cells with the exquisite specificity of antibody binding domains. In their simplest form, these receptors consist of the TCR grafted to extracellular variable regions of an antibody (Eshhar et al., 1993; Kuwana et al., 1987). One advantage of antibody-based receptors is that they can recognize pre-defined tumor targets independent of antigen processing and major-histocompatibility-complex-restricted presentation, rendering a single design applicable to a wide range of patients. First-generation CARs consisting of the cytoplasmic domain of the Fc receptor γ chain or the CD3ζ signaling modules alone, often become anergic, and do not elicit potent T cell antitumor effects (Brocker, 2000; Kershaw et al., 2006; Lamers et al., 2006). This led to the development of second- and third-generation CARs that incorporate additional costimulatory cytoplasmic domains such as CD28, 4–1BB (CD137), ICOS, and OX40 either individually or in combination (Dotti et al., 2014; Sadelain et al., 2013). This modular design successfully recapitulates many aspects of natural costimulation and enhances proliferation and function of CAR T cells (Maus et al., 2014).

CD19-specific CAR T cells have shown encouraging clinical responses against various hematological malignancies, including chronic lymphocytic leukemia (CLL), acute lymphoblastic leukemia (ALL), and diffuse large B cell lymphoma. However, the success rates have been difficult to compare because of several variations in study design as well as differences in the single-chain variable antibody fragment (scFv), costimulatory domains, gene-transfer protocols, and interventions following CAR T cell infusion, among others. Trials conducted with CARs incorporating CD28 or 4–1BB costimulatory domains have shown similar initial response rates in patients with ALL (Brentjens et al., 2013; Lee et al., 2015; Maude et al., 2014). However, in CLL, the clinical efficacy of CAR T cells with 4–1BB costimulatory domains (Porter et al., 2015) appears superior to that of CD28 domains (Brentjens et al., 2011). The reported persistence of CD28-based CAR T cells in vivo is about 30 days (Brentjens et al., 2013; Lee et al., 2015), in contrast to the sustained expression and effector function of 4–1BB CAR T cells, which may exceed 4 years in some patients (Porter et al., 2015). In addition, the incorporation of 4–1BB signaling domains in certain CARs ameliorates exhaustion (Long et al., 2015). Another important consideration is that endogenous CD28 and members of the tumor necrosis factor receptor family (TNFR), such as 4–1BB, invoke distinct signaling cascades in T cells. CD28 leads to activation of the P13K-Akt pathway with downstream effects on glucose metabolism and increased glycolysis (Frauwirth et al., 2002). In contrast, endogenous 4–1BB signaling has been implicated in imparting long-term survival benefits to T cells (Sabbagh et al., 2008) and signaling pathways used by 4–1BB are distinct from CD28 (Martinez-Forero et al., 2013). Thus, a thorough understanding of the molecular signaling effects of CARs may in part explain the observed differences in clinical efficacy for CLL.

A challenge for the identification of optimal CAR designs has been the lack of a physiological in vitro model investigating the impact of CAR-based stimulation. Moreover, current gene-transfer protocols with retroviruses require concomitant activation of T cells via its endogenous TCR, potentially obscuring effects due to signaling through the CAR per se. In this report, we describe an approach enabling CAR expression in over 90% of the T cells without the need to activate the endogenous TCR. Stimulating the CAR T cells with cognate antigen-permitted identification of distinct effects on the differentiation and metabolism of primary human T cells. Interestingly, we find that CAR signaling domains can mediate metabolic reprograming while modifying bioenergetics and mitochondrial biogenesis. We found that 4–1BBζ CAR T cells demonstrate enhanced survival associated with an increased frequency of central memory T (Tcm) cells, mitochondrial biogenesis, and greater oxidative metabolism. In contrast, antigen stimulation of CD28ζ CAR T cells promoted effector memory differentiation and led to enhanced aerobic glycolysis.

RESULTS

BBζ CAR T Cells Show Increased Expansion and Survival Ex Vivo

We initially compared two CAR designs (Figure 1A) specific for either CD19 or mesothelin. The CARs were equipped with signaling domains comprised of either CD28 (Kochenderfer et al., 2009) or 4–1BB (Milone et al., 2009). These CARs were chosen because they have been tested extensively in clinical trials (Beatty et al., 2014; Kochenderfer et al., 2012; Lee et al., 2015; Maude et al., 2014; Maus et al., 2013; Porter et al., 2015). Both CAR constructs were expressed on >90% of CD4+ and CD8+ T cells at comparable mean fluorescence intensities (MFIs) (Figure 1B). We compared the effects of the CD28 and 4–1BB (referred to as 28ζ and BBζ) signaling domains on the differentiation and metabolic fate of T cells. CD4+ T cells were cultured in medium supplemented with 30 U/ml of human IL2. CD8+ T cells were cultured in medium supplemented with either 100 U/ml of human IL2 or 10 ng/ml IL7 and 10 ng/ml IL15, as indicated in the Supplemental Experimental Procedures. Approximately 24 hr after electroporation, CAR-T cells were stimulated with a bead-bound anti-idiotype-Fc to the FMC-63 scFv, which serves as a surrogate for cognate CD19 antigen. To ensure that the CAR T cells received uniform stimulation, we analyzed the surface expression of the activation molecule CD69 on day 1 after activation. CD69 is an inducible cell-surface glycoprotein that is a sensitive indication of lymphoid activation (Hara et al., 1986). Cells that received CAR-specific stimulation showed elevated expression of CD69 on day 1 that was similar on 28ζ and BBζ CAR T cells (Figure 1C). However, the proliferative potential of both CD4+ and CD8+ T cells bearing the BBζ CAR was extended through to at least day 20. In contrast, the proliferative phase of 28ζ CAR T cells was limited to 14 days (Figures 1D and S1, p < 0.01). CAR surface expression rapidly decreased following stimulation with cognate antigen (Figure S2). Importantly, cytokine receptor expression was comparable in both CAR groups (Figure S2), indicating that the proliferative differences between the different CAR T cells are not due to differences in cytokine receptor expression. In one donor, we observed over ten population doublings in the BBζ CAR T cell culture, expanding the starting culture of 4 × 106 cells to a calculated yield of over 5 × 109 in less than 4 weeks (Table 1). Interestingly, the BBζ CAR T cells persisted in culture for over 4 weeks in cytokine-supplemented medium following a single stimulation. In contrast, the proliferation and survival of the 28ζ CAR T cells was lower. Although proliferative capacity varied among donors, the trend remained consistent, in that BBζ CAR T cells displayed a higher proliferative capacity and persistence in comparison to the 28ζ CAR T cells (Figure S1, p < 0.01). Similar results were obtained with CARs directed against mesothelin (Figures 1E and S1; Table 1). For the remainder of this study, we focused on the effect of CAR design in CD8+ T cells.

Figure 1. 4–1BBζ Signaling Domain Provides a Survival and Proliferative Advantage to CD8+ T Cells In Vitro.

Figure 1.

Open in a new tab

(A) Schematics of the CAR constructs compared in this study. CARs contain a single-chain variable fragment of the FMC63 antibody that binds human CD19 or the SS1 scFv that binds human mesothelin. The transmembrane and intracellular domains are indicated.

(B) Flow cytometric analysis of cell-surface expression of the CARs on day 1 after electroporation in comparison to an electroporation only (mock) control. Right, mean fluorescence intensities (MFIs) of the CARs detected with an anti-idiotype reagent. Data are representative of independent experiments verified with cells from over 25 individual healthy human donors.

(C) CD69 expression measured on cell surface 24 hr after co-culture with cognate antigen.

(D) CD19 CAR T cell growth. CD4+ and CD8+ T cells were stimulated, as described in the Experimental Procedures. Data are representative of at least ten different healthy donors.

(E) Mesothelin CAR T cell growth of bulk CD8+ T cells (left) or naive (CD45RO-CD62L+CD8+) T cells (right). CAR T cells were stimulated with beads coated with mesothelin-Fc.

Table 1.

Population Doublings and Cell Yield of T Cells Expressing a CD19- or Mesothelin-Specific CAR Coupled to CD28 or 4–1BB Signaling Domain

Donor # CAR Number of Days in Culture Total Population Doublings Cell Yield ( × 106 cells)
1 CD19 28ζ 20 4.3 78.8
CD19 BBζ 22 5.0 128.0
2 CD19 28ζ 22 6.0 256.0
CD19 BBζ 28 7.2 588.1
3 CD19 28ζ 24 6.9 477.7
CD19 BBζ 30 10.3 5,042.8
4 SS1 28ζ 12 5.8 222.9
SS1 BBζ 24 8.8 1,782.9
5 SS1 28ζ 16 6.9 477.7
SS1 BBζ 24 8.4 1,351.2
6 SS1 28ζ 14 6.0 256.0
SS1 BBζ 22 8.4 1,351.2
Open in a new tab

Healthy donor T cells were electroporated with the indicated CAR and stimulated with anti-idiotype (CD19) or recombinant mesothelin Fc. The cells were counted and maintained in culture until at least two consecutive declines in cell numbers were observed. The last column shows the total number of cells obtained at the end of expansion, starting with 4 × 106 cells in each group. SS1 is a mesothelin-specific scFv.

BBζ CAR Signaling Promotes Enhanced Central Memory T Cell Subset

We hypothesized that the enhanced persistence of BBζ T cells was due to a relative preservation of cells with a more extensive proliferative capacity. To test the differentiation status of BBζ and 28ζ CAR-T cells, we used a standard panel of cell-surface markers associated with T cell differentiation. We assessed expression of CD45RO and CCR7, which are associated with Tcm cells. All cultures contained the same heterogeneous population of T cell subsets at day 0. After stimulation through the CAR, the proportion of CD45RO+CCR7+ cells was progressively enriched (Figure 2A). Notably, the enrichment of this Tcm cell population was higher in the BBζ CAR group in comparison to the 28ζ group (p < 0.01) and persisted through the end of culture (Figure 2B). In contrast, the 28ζ CAR cultures consistently yielded a higher proportion of effector-memory phenotype (Tem), identified as CD45RO+CCR7− cells.

Figure 2. 4–1BBζ Signaling Domain Leads to Enrichment of Tcm Subset, whereas 28ζ Promotes Tem Population.

Figure 2.

Open in a new tab

(A) Representative plots (from at least six donors) of cell-surface expression of CCR7 and CD45RO on CAR T cells at specified time points during culture. Cells shown have been pre-gated for live CD3+CD8+ T cells. Numbers shown are percentages of cells detected in each gate.

(B) Relative change of Tcm and Tem cell subsets in 28ζ and BBζ CD19 CAR T cell cultures. Absolute numbers of live cells were calculated for each population at the specified time points. The graphs show relative fold change of Tcm or Tem in BBζ CAR T cells normalized to 28ζ CAR T cells. Data are plotted as mean ± SEM (****, p < 0.0001, and **, p = < 0.01).

CAR Signaling Domains Reprogram T Cell Metabolism

Upon stimulation, CD8+ T cells undergo an ordered process involving proliferation and differentiation into effector and memory cells. Activation is associated with a biosynthetic and bioenergetics flux required to support T cell proliferation and function (Pearce and Pearce, 2013; Wang and Green, 2012). For example, naive and memory T cells rely primarily on the mitochondrial oxidation of free fatty acids for development and persistence (Pearce et al., 2009; van der Windt et al., 2012). In contrast, activated effector T cells shift to glycolysis or concurrently upregulate oxidative phosphorylation and aerobic glycolysis to fulfill the metabolic demands of proliferation (van der Windt et al., 2012).

Based on the distinct growth rates and differentiation of 28ζ and BBζ CAR T cells, we sought to explore the interconnection of cellular metabolism and CAR signaling. First, we examined the metabolic profiles of T cells expressing the two CARs at different time points after stimulation. Cell volume, a surrogate for cell mass, was found to be comparable after cognate antigen stimulation (Figure 3A). We measured the oxygen consumption rate (OCR) of 28ζ and BBζ CAR T cells before and 7 and 21 days after antigenic stimulation during log-phase proliferation. Basal OCR was measured, followed by serial additions of oligomycin (an inhibitor of ATP synthesis), carbonyl cyanide-p-trifluoromethoxyphenylhydrazone (FCCP; an uncoupling ionophore), and rotenone with antimycin A (blocking agents for complexes I and III of the electron transport chain, respectively) to discern the relative contributions of mitochondrial and non-mitochondrial mechanism of oxygen consumption (van der Windt et al., 2012). The OCR profiles were similar before antigen stimulation on day 0 (Figure 3B). After antigen stimulation, there was a ~10-fold increase in basal OCR in both groups of T cells on days 7 and 21 (Figure 3C). However, there was a robust increase in maximal respiratory capacity that was specific to the BBζ CAR T cells, following decoupling of the mitochondrial membrane using FCCP on both days 7 and 21 (Figure 3D). In contrast the maximal respiratory capacity of the 28ζ CAR T cells on days 7 and 21 was similar to what it was on day 0. To confirm that these differences in OCR were due to the signaling domains of the receptor, similar experiments were performed with mesothelin-specific CAR T cells. The mesothelin-BBζ CAR T cells exhibited an elevated basal and maximal respiratory capacity compared to the 28ζ CAR T cells on days 7 and 21 after stimulation with mesothelin (Figure S3). We also measured the extracellular acidification rate (ECAR) as a measurable surrogate for lactic acid production during glycolysis. Glycolysis involves a series of enzyme-catalyzed reactions culminating in the production of lactic acid. At physiologic pH, lactic acid dissociates into lactate and H+, which are exported extracellularly. ECAR levels were elevated in 28ζ cells in comparison to BBζ CAR T cells on days 7 and 21 (Figure 3E).

Figure 3. Effects of CAR Signaling Domain on Cellular Metabolism.

Figure 3.

Open in a new tab

(A) Effects of antigen stimulation on mean cell volume after stimulation of CD19 CAR CD8+ T cells expressing 28ζ and BBζ signaling domains with anti-idiotype.

(B) The oxygen consumption rates (OCRs) of 28ζ and BBζ CAR T cells at baseline (after electroporation of CAR mRNA and before stimulation) on day 0 and after stimulation on days 7 and 21 in culture under basal metabolic conditions and in response to mitochondrial inhibitors, as specified in the Experimental Procedures.

(C–E) Basal OCR levels (C), maximum respiratory levels (D), and basal ECAR levels (E) measured at days 7 and 21. Data are representative of at least five independent experiments performed with cells from at least five healthy human donors plotted as mean ± SEM (*, p < 0.05).

(F–H) Basal OCR levels measured for CAR T cells sorted for different memory phenotypes: central memory (CM; F), naive (N), and effector memory (EM; H). Data representative of at least three independent experiments performed with cells from at least three healthy human donors and plotted as mean ± SEM.

(I) Basal ECAR levels measured for the three different sorted memory subsets. Data are representative of at least three independent experiments performed with cells from at least three healthy human donors plotted as mean ± SEM (*, p < 0.05).

Several reports have shown that natural central memory differentiated T cells display elevated basal OCR and SRC in comparison to effector memory and terminally differentiated effector cells. These oxidative features suggest that an increased reliance on fatty acid oxidation (FAO) may be necessary for central memory differentiation and survival (Pearce et al., 2009; van der Windt et al., 2012). Because we saw a differential enrichment of memory phenotypes in the two CAR T cell groups in culture, we extended our analysis to uncover how individual memory subsets contribute to the metabolic properties of CAR T cells. Again, using CCR7 and CD45RO as phenotypic markers, we sorted the populations into CCR7+CD45RO−, CCR7+CD45RO+, and CCR7−CD45RO+ to define naive-like, Tcm cell, and Tem cell subpopulations, respectively. Metabolic flux revealed higher basal OCR and maximum respiratory capacity of the BBζ in the Tcm and Tn memory subtypes as compared to 28ζ CAR T cells (Figures 3F and 3G). As observed in past reports concerning effector cells, the basal OCR as well as the maximum respiratory levels remained low for the Tem cell subpopulations for both CAR groups (Figure 3H). On the other hand, the ECAR levels remained higher for Tcm and Tem cell subpopulations of cells obtained from the 28ζ CAR T cell culture (Figure 3I). In aggregate, these studies show that BBζ CAR T cells are metabolically distinct from 28ζ CAR T cells with the former displaying greater capacity for oxidative metabolism that might contribute to the enhanced central memory differentiation and persistence of BBζ CAR T cells.

28ζ and BBζ CAR T Cells Have Distinct Glycolytic and Fatty Acid Metabolism

To investigate whether the differences in the basal OCR in CAR T cells altered the fuel sources by which these cells satisfy their bioenergetic appetite, we proceeded to measure glucose uptake and fatty acid utilization rates in CAR T cells. At day 7 after stimulation, the cells were replated in fresh media. At different points (as indicated in Figure 4A), we measured the amount of residual glucose in the media and the lactate produced. 28ζ CAR T cells consumed glucose at a relatively quicker rate along with production of lactic acid. This is consistent with the greater ECAR we observed in 28ζ CAR T cells (Figures 3E and 3I).

Figure 4. Preferential Reliance on Glycolysis or Fatty Acid Oxidation by CAR T Cells.

Figure 4.

Open in a new tab

(A) Measurement of glucose uptake from extracellular media and lactate release into the media over a course of 48 hr.

(B) Percentage of labeled acetyl-CoA measured in T cells cultured with [13C16] palmitic acid to assess fatty acid uptake and breakdown.

(C) Relative mRNA expression levels of genes involved in glycolytic metabolism and lipid oxidation assessed in 28ζ and BBζ CAR T cells. Plot represents data from at least three independent experiments with cells obtained from four independent donors (**, p < 0.01; *, p < 0.05). Data are represented as mean ± SEM.

The increased OCR in BBζ CAR T cells prompted us to examine the fatty acid consumption rate in these cells. Using a heavy-carbon-labeled long-chain fatty acid (palmitic acid), we analyzed its uptake rate by measuring the levels of heavy-carbon-labeled acetyl-CoA. The catabolic process of β oxidation breaks down fatty acid molecules into acetyl-CoA in the mitochondria to feed the citric-acid cycle. We found that BBζ showed a higher percentage of labeled acetyl-CoA pool as compared to 28ζ CAR T cells (Figure 4B). This data suggest that BBζ CAR T cells, similar to CD8+ Tcm cells, extensively rely on catabolic pathways such as FAO to fuel their bioenergetic demands.

To gain insight into the mechanism leading to the metabolic differences conferred by distinct CAR signaling domains, we measured the expression of candidate genes that are implicated in glycolytic and lipid metabolism. We initially focused on two main enzymes implicated in glucose metabolism, Glut1 and PDK1. The cell-surface expression of Glut1, the transporter involved in glucose uptake, is induced following CD28 activation (Frauwirth et al., 2002). In certain contexts, including hypoxia, PDK1 inhibits the decarboxylation of pyruvate and entry of glucose derivatives into the tricarboxylic acid cycle (Düvel et al., 2010). Both Glut1 and PDK1 are induced to significantly higher levels in 28ζ cells relative to BBζ cells at day 7 (Figure 4C). Increased expression levels of Glut1 and PDK1, coupled with our earlier finding of increased ECAR, is consistent with enhanced glycolysis in 28ζ CAR T cells in comparison to their BBζ counterparts.

Two critical enzymes involved in the breakdown of glucose during the ATP-generating step of the glycolytic pathway are phosphoglycerate kinase (PGK) and glucose-6-phosphate dehydrogenase (G6PD). PGK transfers a phosphate group to ADP in order to facilitate ATP generation, whereas G6PD, an NADP+-dependent enzyme, catalyzes the oxidative phase of the pentose phosphate pathway. Given that these enzymes have an important role in glycolysis, we investigated their expression levels in CAR T cells on day 7. Interestingly, their levels were elevated in 28ζ CAR T cells relative to BBζ CAR T cells. Finally, we also examined the levels of solute carrier family 16 (SLC16A3), an exporter of the glycolysis byproducts lactic acid, and pyruvate. 28ζ CAR T cells showed higher levels of SLC16A3 mRNA in comparison to BBζ T cells, consistent with our hypothesis that 28ζ CAR T cells use increased glycolysis as a means to meet their metabolic demands. We also detected increased expression of VEGFA in 28ζ CAR T cells, which is an established target of the hypoxia-inducible factors (HIFs). Several genes involved in glycolysis are targets of HIF1α (Finlay et al., 2012), including Glut1 and PFK. Others have shown that HIF1A−/− T cells display impaired autoreactivity (Dang et al., 2011). Our findings add to the growing body of evidence implicating costimulation through CD28 and glycolytic reprogramming in effector differentiation.

Next, we sought to investigate genes associated with mitochondrial FAO after triggering CAR T cells. Increasing evidence has demonstrated a role for carnitine palmitoyl transferase (CPT1A) in regulating oxidative metabolism in CD8+ memory T cells (van der Windt et al., 2012). CPT1A is a metabolic enzyme that controls a rate-limiting step in mitochondrial FAO and promotes mitochondrial biogenesis. We observed significantly higher levels of CPT1A mRNA in BBζ CAR T cells in comparison to 28ζ CAR T cells. Additionally, mRNA levels of fatty acid binding protein (FABP5), which plays a critical role in long-chain fatty acid uptake, transport and metabolism were significantly upregulated in BBζ CAR T cells in comparison to 28ζ (Figure 4C). These findings suggest that 28ζ CAR T cells rely more on a glycolytic-based metabolism, whereas BBζ programs T cells to use fatty acids as the predominant energy source, which are characteristics of natural effector and memory T cells, respectively.

BBζ CAR T Cells Have Increased SRC

Mitochondrial SRC is a measure of how effectively protons can be shuttled into the mitochondrial intermembrane space upon cellular or mitochondrial stress (Mookerjee et al., 2010; Nicholls, 2009). SRC enhances survival and function of memory T cells by providing a contingency source of energy for cells exposed to metabolic stress including nutrient depletion, oxygen deprivation or under conditions of increased cellular activity. Increased SRC likely supports T cell function in a hostile tumor environment (Ferrick et al., 2008; Nicholls, 2009; Yadava and Nicholls, 2007). Memory CD8+ T cells, unlike effectors, maintain a substantial SRC (van der Windt et al., 2012). When comparing the SRC of the two CAR groups, we observed that BBζ CAR T cells maintained higher levels of SRC in comparison to 28ζ CAR T cells (Figure 5A). This is consistent with the metabolic characteristics of long-lived CD8+ memory cells, lending additional support to the hypothesis that BBζ signals support a metabolic reprogramming that contributes to long-lived memory-like T cells.

Figure 5. Enhanced SRC in 4–1BBζ CAR T Cells.

Figure 5.

Open in a new tab

(A) SRC measured as the ratio between the maximum OCR levels after treating cells with FCCP to the basal OCR levels while in culture. Data represent three independent donors tested (*, p < 0.05).

(B) Transmission electron microscopy of 28ζ and BBζ CAR CD8+ T cells imaged at two different time points. Scale bars represent 2 μm.

(C) Enumeration of the individual mitochondrion per cell. Data shown represent 20 individual randomly chosen cells (out of at least 75 cells analyzed per condition) represented as mean ± SEM (***, p < 0.001).

(D) Confocal images stained with Mitotracker (green), DAPI (blue), and a cell-membrane dye DiI (red). Scale bars represent 2 μm.

(E) Quantification of the percentage of cytoplasm occupied by mitochondria and measured as percentage of Mitotracker (green) within area enclosed by the cell membrane (red). Data represented as mean ± SEM from at least three images each at specified time points with at least 15 independent cells scored per image (****, p < 0.0001).

Given the role of mitochondrial density in oxidative metabolism (van der Windt et al., 2012), we next explored the possibility whether the increased SRC in BBζ CAR T cells was associated with an increase in mitochondrial mass. Using electron microscopy, we measured similar mitochondrial density between 28ζ and BBζ CAR-T cells at day 7 (Figures 5B and 5C). However, there was a substantial increase in mitochondrial mass in BBζ CAR T cells at days 14 (Figure 5B) and 21 (Figure S4) after antigen stimulation. Despite similar cell volumes (Figure 3A), we observed a significantly (p < 0.001) increased density of mitochondria in BBζ CAR-T cells. To confirm that BBζ CAR T cells have enhanced mitochondrial content, we also measured mitochondrial density using confocal microscopy (Figure 5D). BBζ CAR T cells showed an increased ratio of mitochondrial mass to total cell mass on days 14 and 21 (Figure 5E).

BBζ CAR T Cells Show Enhanced Mitochondrial Biogenesis

We speculated that specific signals from the 4–1BB signaling domain in the CAR structure supported mitochondrial biogenesis, thus endowing these cells with greater mitochondrial mass. However, in addition to quantitative differences in mitochondrial content, we examined whether qualitative differences in mitochondria might contribute to the differences in metabolic profiles between these CAR cells. We examined levels of certain mitochondrial genes encoded by the nuclear and the mitochondrial genome, namely mitochondrial transcription factor A (TFAM) and MTCO-1, respectively. Notably, BBζ cells had significantly enhanced mRNA expression of mitochondrial TFAM and mitochondrially encoded cytochrome c oxidase 1, the main subunit of the cytochrome c oxidase complex (Figure 6A).

Figure 6. Enhanced Mitochondrial Biogenesis in T Cells Expressing 4–1BBζ CARs Is Regulated at a Genetic Level.

Figure 6.

Open in a new tab

(A) Relative mRNA expression of mitochondrial cytochrome c oxidase 1 (MT-CO1) and mitochondrial transcription factor A (TFAM) in BBζ CAR T cells normalized to expression levels of 28ζ CAR T cells at specified time points. Data generated from at least three independent experiments with four independent donors (*, p < 0.05) represented as mean ± SEM.

(B) Normalized mRNA expression levels of nuclear respiratory factor 1 (NRF1) and GA binding protein (NRF2) in BBζ CAR T cells in comparison to 28ζ CAR T cells at specified time points. Data are generated from at least three independent experiments with four independent donors (*, p < 0.05) and represented as mean ± SEM.

To explore the role of 28ζ and BBζ costimulatory domains on the mitochondrial function in the context of CAR T cells, we measured gene expression of two transcription factors of mitochondrial genes, namely nuclear respiratory factor 1 (NRF1) and GA-binding protein (also known as NRF2). Whereas NRF1 regulates the expression of TFAM and coordinates mtDNA replication and expression, NRF2 has a role in the transcription of the OXPHOS components, mitochondrial import, and TFAM. Consistent with its enhanced oxidative features as seen by metabolic flux analyses and mitochondrial density, we found that BBζ CAR T cells had significantly higher expression of NRF1 and NRF2 in comparison to the 28ζ CAR T cell group (Figure 6B).

Taken together, these findings suggest increased mitochondrial content in BBζ CAR T cells in comparison to 28ζ CAR T cells, which strongly correlates with the increased SRC observed in these cells. Our findings are consistent with a model in which BBζ CAR signaling reprograms transcriptional networks supporting mitochondrial biogenesis and oxidative metabolism. Given the role of metabolic adaptation in enabling T cell memory and effector functions, the aforementioned oxidative features in BBζ CAR T cells most likely support central memory differentiation and T cell persistence.

DISCUSSION

We have uncovered significant differences in the differentiation and metabolic profiles of CAR T cells using CD28 or 4–1BB signaling domains. The predominant metabolic program in 28ζ CAR T cells is aerobic glycolysis, and, in BBζ CAR T cells, it is oxidative breakdown of fatty acids. Our studies provide evidence for plasticity in T cell metabolic reprogramming and, further, that the choice of CAR signaling domain can impact the subsequent fate of the T cells. The enhanced proliferation and persistence of BBζ over 28ζ CAR T cells observed in our studies mirrors the outcomes of CAR persistence observed in clinical studies (Brentjens et al., 2013; Brentjens et al., 2011; Lee et al., 2015; Porter et al., 2015). Our studies suggest that one mechanism for the differential persistence may be the metabolic reprograming of the CAR T cells to enhance either oxidative phosphorylation that is characteristic of memory cells or aerobic glycolysis that is characteristic of effector cells (MacIver et al., 2013; van der Windt et al., 2012).

Previous studies have shown that CD28 signaling initiates a cascade leading to enhanced surface expression of Glut1 and increased reliance on aerobic glycolysis (Frauwirth et al., 2002). In contrast, a TNFR pathway is required for the initiation of mitochondrial FAO and T cell memory development (Pearce et al., 2009). Although IL2 promotes effector differentiation and glycolysis in CD8+ T cells (Finlay et al., 2012; Liao et al., 2013; Pipkin et al., 2010), IL7 and IL15 have been implicated in the maintenance of memory T cells and increased mitochondrial biogenesis (Ku et al., 2000; Schluns and Lefrançois, 2003; van der Windt et al., 2012). Given that human CD8+ T survival is impaired in the absence of exogenous cytokines, IL7 and IL15 are necessarily present in our culture system. Although these extrinsic factors may play a significant role in stabilizing the metabolic profiles of T cells, we hypothesize that our system is largely governed by cell-intrinsic factors influenced by the two unique intracellular CAR signaling domains. This is further corroborated by the lack of differences in the cell-surface expression of these cytokine receptors, suggesting that the relative distinction in metabolic reprogramming between the two CARs cannot be solely mediated by the supplemented cytokines. Thus, our studies suggest that the ectopic expression of CD28 or 4–1BB signaling domains in CARs leads to a phenocopy of the natural T cell activation process. By extension, our studies suggest that the incorporation of various signaling modules may biosynthetically reprogram T cells to desired effector or regulatory functions. For example, we have recently found that the incorporation of the ICOS signaling domain in CARs promotes a Th17 cell differentiation program (Guedan et al., 2014).

One clinical application of our findings is that short-lived or long-lived CAR T cells can be created “at will.” This could extend the range of targets, depending on certain surface molecules where long-term CAR effects may not be tolerable due to potential off-tumor toxicity. In this case, a CD28 signaling domain would be expected to be superior. Another implication from our studies is that a mixture of CAR T cells expressing 4–1BB and CD28 domains may be superior to either CAR as a single population. We speculate this because the combination of CAR T cells would be expected to more completely mimic a natural immune response comprised of an early dominance of T effector cells, achieved with CD28 CARs having enhanced aerobic glycolysis in the cytoplasm, and T memory cells, achieved with 4–1BB CARs having enhanced mitochondrial oxidative phosphorylation.

Apart from cell intrinsic factors, there has been substantial interest in understanding the effects of nutrient consumption on T cell survival in the tumor microenvironment. T cells have substantial bioenergetic and biosynthetic challenges to survive and conduct effector functions. Our results indicate that BBζ CAR T cells have an increased capacity to generate mitochondrial mass. This increase in mitochondrial mass provides a survival advantage (van der Windt et al., 2013). We consistently saw a higher SRC in BBζ CAR T cells, and this mitochondrial respiratory capacity has been shown to be an important characteristic of natural CD8+ T cell memory development (van der Windt et al., 2012). The increased basal oxygen consumption of BBζ cells also suggests a preferential reliance on oxidative phosphorylation as the predominant energy generating mechanism to account for the metabolic demands required for enhanced CAR T cell proliferation. Furthermore, our data suggests that metabolism is a key mediator of CAR T cell survival and is influenced by the signaling induced by the costimulatory domain included in the CAR.

In summary, our results reveal a new role for CAR T cell engineering to control T cell metabolism as a key determinant of T cell effector and memory responses. Using synthetic biology, it is possible to shape the immune response to a desired balance of long-lived memory cells and short-lived effector cells. By extension, our studies should influence the design of engineered T effector or engineered T regulatory cells that resist exhaustion or have enhanced survival in hostile tumor and inflammatory microenvironments.

EXPERIMENTAL PROCEDURES

CAR Constructs and Generation of CAR-Encoding, In-Vitro-Transcribed RNA

For the purpose of these studies, CARs specific to the human CD19 or mesothelin antigen were used. Figure 1A shows the schematic of the CARs used in this study. All CARs contained the scFv against human CD19 (clone FMC-63) or the SS1 scFv against human mesothelin protein, wherever indicated (Hassan et al., 2002; Nicholson et al., 1997). The mesothelin CAR was previously described (Carpenito et al., 2009). The CD28ζ CAR consisted of the scFv linked in cis to the intracellular domains of CD28 and CD3ζ through the CD8α hinge and a CD28 transmembrane domain, as described previously (Milone et al., 2009). Similarly, the BBζ CAR contained the scFv linked to the 4–1BB intracellular portion and the CD3z domain through a CD8α hinge and transmembrane domain (Milone et al., 2009). For preparation of in-vitro-transcribed (IVT) RNA, the CAR-encoding gene constructs were subcloned into the pGEM.64A-based vector, as described previously (Zhao et al., 2010).

Isolation, Electroporation, and Expansion of Primary Human T Lymphocytes

Primary human T lymphocytes were obtained from anonymous healthy donors at the University of Pennsylvania Apheresis Unit. Using the BTX CM380 (Harvard Apparatus BTX) electroporation machine, we introduced the IVT RNA into the T cells at a ratio of 1 ug RNA per 106 cells. This technique was optimized to promote uniform CAR expression on the cell surface (Figure 1B). T cells were stimulated with magnetic beads coated with a recombinant anti-CD19 idiotype or mesothelin-Fc.

Flow Cytometry Analysis

Live cells were gated on live/dead aqua-negative and then gated for CD3−, CD4−, and CD8-positive events. Using markers for memory, CCR7, and CD45RO, we analyzed cells in culture and sorted them for the three different memory phenotypes using the BD FACSCalibur analyzer. Absolute T cell counts were determined with the aid of CountBright Absolute Counting Beads (Life Technologies) using the following formula: (numberofTcellsevents/numberofbeadevents)×numberofbeadsused.

Analysis of Metabolic Parameters

Mitochondrial function was assessed with an extracellular flux analyzer (Seahorse Bioscience). Individual wells of an XF24 (Figures 3B–3E) or XF96 (Figures 3F–3I) cell-culture microplates were coated with CellTak in accordance with the manufacturer’s instructions. The matrix was adsorbed overnight at 37°C, aspirated, air-dried, and stored at 4°C until use. Mitochondrial function was assessed on days 0, 7, and 21. To assay mitochondrial function, we centrifuged T cells at 1,200 × g for 5 min. Cell pellets were resuspended in XF assay medium (non-buffered RPMI 1640) containing 5.5 mM glucose, 2 mM L-glutamine, and1 mM sodium pyruvate and seeded at 1 × 106 cells per well. The microplate was centrifuged at 1,000 × g for 5 min and incubated in standard culture conditions for 60 min. During instrument calibration (30 min), the cells were switched to a CO2-free (37°C) incubator. XF24 and XF96 assay cartridges were calibrated in accordance with the manufacturer’s instructions. Cellular OCRs were measured under basal conditions and, following treatment with 1.5 μM oligomycin, 1.5 μM FCCP, and 40 nM rotenone, with 1μM antimycin A (XF Cell Mito Stress kit, Seahorse Bioscience).

Gene-Expression Analysis by RT-PCR

qRT-PCR was used to quantify expression levels of certain candidate genes. Total RNA from cells was used as a template to synthesize cDNA with a High Capacity RNA-to-cDNA Kit (Applied Biosystems). qRT-PCR was performed in triplicates with TaqMan Universal Master Mix on a ViiA 7 Real-Time PCR System as per the manufacturer’s instructions. mRNA levels of each candidate gene as quantified by the PCR system were normalized to a housekeeping gene, GAPDH. All probes used are commercially available (Applied Biosystems).

Glucose Uptake Assay

Cells at day 7 after stimulation were starved in PBS at room temperature for 30 min followed by incubation at 37°C in regular RPMI culture media supplemented with 11 mM glucose, 10% FCS, 100 U/ml penicillin, 100 μg/ml streptomycin sulfate, and 2 mM glutamax. 500 uL aliquots of cell culture was collected at indicated time points and spun down, and the supernatants were analyzed for glucose and lactate concentrations with the Nova BioProfile Analyzer (Nova Biomedical).

Palmitic Acid Uptake Assay

[13C16] palmitic acid was purchased from Sigma-Aldrich. All solvents for liquid chromatography mass spectrometry were Optima grade and purchased from Fisher Scientific. For palmitic acid-labeled isotope experiments, cells were cultured overnight in RPMI 1,640 without D-glucose or L-glutamine (Biological Industries) and supplemented with 10% charcoal-stripped FBS (GIBCO), 2 mM L glutamine (Life Technologies), 5.0 mM glucose, and 100 μM [13C16] palmitic acid.

Short-chain Acyl-CoA Extraction

Extractions were performed as described previously (Basu and Blair, 2012; Worth et al., 2014). In brief, lymphocytes were centrifuged at 1,200 rcf for 5 min. Cell pellets were resuspended in 750 μl of ice-cold 10% trichloroacetic acid and pulse-sonicated with a sonic dismembrator (Fisher Scientific). The samples were centrifuged at 15,000 rcf for 15 min, and the supernatants were purified by solid-phase extraction. In brief, Oasis HLB 1 ml (30 mg) solid-phase extraction columns were conditioned with 1 ml methanol followed by 1 ml of H2O. The supernatants were applied to the column and washed with 1 ml of H2O. The analytes were eluted in methanol containing 25 mM ammonium acetate, dried overnight in N2 gas, and resuspended in 50 μl of 5% 5-sulfosalicylic acid. 10 μl injections were applied in LC/ESI/MS/MS analysis.

LC/MS Analysis of Acyl-CoA Thioesters

Acyl-CoAs were separated with a Phenomenex Luna C18 reverse-phase high-performance liquid chromatography column (2.0 × 150 mm, 5 μm pore size) with 5 mM ammonium acetate in water as solvent A, 5 mM ammonium acetate in acetonitrile (ACN)/water (95:5, v/v) as solvent B, and ACN/water/formic acid (80:20:0.1, v/v) as solvent C, as described previously (Basu et al., 2011; Worth et al., 2014). A linear gradient was run as follows: 2% solvent B for 1.5 min, increased to 25% over 3.5 min, increased to 100% over 0.5 min, held for 8.5 min, and washed with 100% solvent C for 5 min before equilibration for 5 min. The flow rate was 200 μl/min. Samples were analyzed with an API 4000 triple-quadrupole mass spectrometer (Applied Biosystems) in the positive electrospray ionization (ESI) mode. Samples (10 μl) were injected with a LEAP autosampler (CTC Analytics AG) and maintained at 4°C. Data were analyzed with Analyst Version 1.4.1 software (AB SCIEX). The column effluent was diverted to the mass spectrometer from 8–23 min and to waste for the remainder of the run. The mass spectrometer operating conditions were as follows: ion spray voltage (5.0 kV), nitrogen as curtain gas (15 U), ion source gas 1 (8 U), ion source gas 2 (15 U), and collision-induced dissociation gas (5 U). The ESI probe temperature was 450°C, the declustering potential was 105V, the entrance potential was 10 V, the collision energy was 45 V, and the collision exit potential was 15 V. A loss of 507 Da was monitored for each acyl-CoA.

Microscopy

Cells at different time points were stained with DiI, Mitotracker green, and DAPI (Life Technologies) and fixed with 4% PFA before imaging on the Leica TSC SP8 confocal microscope. Captured images were analyzed with Fiji (ImageJ), and fluorescence emission was quantified as MFI. For transmission electron microscopy, the cells were prepared with Penn’s Electron Microscopy Resource Laboratory and imaged with the Jeol-1010 microscope.

Supplementary Material

Document S1. Supplemental Experimental Procedures and Figures S1–S4.

SUPPLEMENTAL INFORMATION

Supplemental Information contains Supplemental Experimental Procedures and four figures and can be found with this article online at http://dx.doi.org/10.1016/j.immuni.2016.01.021.

Highlights.

  • CAR T cells containing 4–1BB signaling domains have enhanced in vitro persistence

  • Increased central memory differentiation of CD8+ CAR T cells expressing 4–1BB

  • CD28-enforced metabolic reprograming enhances glycolytic metabolism

  • Mitochondrial biogenesis is selectively induced in 4–1BB CAR T cells

ACKNOWLEDGMENTS

This study was supported by grants from the NIH (5R01CA120409), National Cancer Institute (NCI T32CA009140 to J.A.F.), and Novartis. The authors would like to thank Li Liu, Felipe Bedoya, and Jos Melenhorst for the generous gift of the CD19 anti-idiotype; Yangbing Zhao, Daniel Powell, Jose Conejo-Garcia, and Andrew Wells for helpful discussions; Biao Zuo and Raymond Meade at Penn’s Electron Microscope Resource Laboratory; Xinyu Zhao (Jasmine) and Andrea Stout from CBD Microscopy Core Facility; Joyce Lee and Kathryn E. Wellen for expert technical assistance; and the Human Immunology Core for reagents. The authors also express their gratitude to Brian Dranka and David Ferrick from Seahorse Bioscience for technical assistance in the metabolic flux assays. O.U.K., J.A.F., S.E.M., A.D.P., S.G., J.S., M.C.M., and C.H.J. are inventors of intellectual property licensed by the University of Pennsylvania to Novartis. This work was supported in part by a grant from Novartis.

REFERENCES

  1. Basu SS, and Blair IA (2012). SILEC: a protocol for generating and using isotopically labeled coenzyme A mass spectrometry standards. Nat. Protoc. 7, 1–12. [Google Scholar]
  2. Basu SS, Mesaros C, Gelhaus SL, and Blair IA (2011). Stable isotope labeling by essential nutrients in cell culture for preparation of labeled coenzyme A and its thioesters. Anal. Chem. 83, 1363–1369. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Beatty GL, Haas AR, Maus MV, Torigian DA, Soulen MC, Plesa G, Chew A, Zhao Y, Levine BL, Albelda SM, et al. (2014). Mesothelin-specific chimeric antigen receptor mRNA-engineered T cells induce anti-tumor activity in solid malignancies. Cancer Immunol. Res. 2, 112–120. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Brentjens RJ, Rivière I, Park JH, Davila ML, Wang X, Stefanski J, Taylor C, Yeh R, Bartido S, Borquez-Ojeda O, et al. (2011). Safety and persistence of adoptively transferred autologous CD19-targeted T cells in patients with relapsed or chemotherapy refractory B-cell leukemias. Blood 118, 4817–4828. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Brentjens RJ, Davila ML, Riviere I, Park J, Wang X, Cowell LG, Bartido S, Stefanski J, Taylor C, Olszewska M, et al. (2013). CD19-targeted T cells rapidly induce molecular remissions in adults with chemotherapy-refractory acute lymphoblastic leukemia. Sci. Transl. Med. 5, 177ra38. [Google Scholar]
  6. Brocker T (2000). Chimeric Fv-zeta or Fv-epsilon receptors are not sufficient to induce activation or cytokine production in peripheral T cells. Blood 96, 1999–2001. [PubMed] [Google Scholar]
  7. Carpenito C, Milone MC, Hassan R, Simonet JC, Lakhal M, Suhoski MM, Varela-Rohena A, Haines KM, Heitjan DF, Albelda SM, et al. (2009). Control of large, established tumor xenografts with genetically retargeted human T cells containing CD28 and CD137 domains. Proc. Natl. Acad. Sci. USA 106, 3360–3365. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Dang EV, Barbi J, Yang HY, Jinasena D, Yu H, Zheng Y, Bordman Z, Fu J, Kim Y, Yen HR, et al. (2011). Control of T(H)17/T(reg) balance by hypoxia-inducible factor 1. Cell 146, 772–784. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Dotti G, Gottschalk S, Savoldo B, and Brenner MK (2014). Design and development of therapies using chimeric antigen receptor-expressing T cells. Immunol. Rev. 257, 107–126. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Düvel K, Yecies JL, Menon S, Raman P, Lipovsky AI, Souza AL, Triantafellow E, Ma Q, Gorski R, Cleaver S, et al. (2010). Activation of a metabolic gene regulatory network downstream of mTOR complex 1. Mol. Cell 39, 171–183. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Eshhar Z, Waks T, Gross G, and Schindler DG (1993). Specific activation and targeting of cytotoxic lymphocytes through chimeric single chains consisting of antibody-binding domains and the gamma or zeta subunits of the immunoglobulin and T-cell receptors. Proc. Natl. Acad. Sci. USA 90, 720–724. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Ferrick DA, Neilson A, and Beeson C (2008). Advances in measuring cellular bioenergetics using extracellular flux. Drug Discov. Today 13, 268–274. [DOI] [PubMed] [Google Scholar]
  13. Finlay DK, Rosenzweig E, Sinclair LV, Feijoo-Carnero C, Hukelmann JL, Rolf J, Panteleyev AA, Okkenhaug K, and Cantrell DA (2012). PDK1 regulation of mTOR and hypoxia-inducible factor 1 integrate metabolism and migration of CD8+ T cells. J. Exp. Med. 209, 2441–2453. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Frauwirth KA, Riley JL, Harris MH, Parry RV, Rathmell JC, Plas DR, Elstrom RL, June CH, and Thompson CB (2002). The CD28 signaling pathway regulates glucose metabolism. Immunity 16, 769–777. [DOI] [PubMed] [Google Scholar]
  15. Grupp SA, Kalos M, Barrett D, Aplenc R, Porter DL, Rheingold SR, Teachey DT, Chew A, Hauck B, Wright JF, et al. (2013). Chimeric antigen receptor-modified T cells for acute lymphoid leukemia. N. Engl. J. Med. 368, 1509–1518. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Guedan S, Chen X, Madar A, Carpenito C, McGettigan SE, Frigault MJ, Lee J, Posey AD Jr., Scholler J, Scholler N, et al. (2014). ICOS-based chimeric antigen receptors program bipolar TH17/TH1 cells. Blood 124, 1070–1080. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Hara T, Jung LK, Bjorndahl JM, and Fu SM (1986). Human T cell activation. III. Rapid induction of a phosphorylated 28 kD/32 kD disulfide-linked early activation antigen (EA 1) by 12-o-tetradecanoyl phorbol-13-acetate, mitogens, and antigens. J. Exp. Med. 164, 1988–2005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Hassan R, Lerner MR, Benbrook D, Lightfoot SA, Brackett DJ, Wang QC, and Pastan I (2002). Antitumor activity of SS(dsFv)PE38 and SS1(dsFv) PE38, recombinant antimesothelin immunotoxins against human gynecologic cancers grown in organotypic culture in vitro. Clin. Cancer Res. 8, 3520–3526. [PubMed] [Google Scholar]
  19. June CH, Riddell SR, and Schumacher TN (2015). Adoptive cellular therapy: a race to the finish line. Sci. Transl. Med. 7, 280ps7. [Google Scholar]
  20. Kalos M, and June CH (2013). Adoptive T cell transfer for cancer immunotherapy in the era of synthetic biology. Immunity 39, 49–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Kalos M, Levine BL, Porter DL, Katz S, Grupp SA, Bagg A, and June CH (2011). T cells with chimeric antigen receptors have potent antitumor effects and can establish memory in patients with advanced leukemia. Sci. Transl. Med. 3, 95ra73. [Google Scholar]
  22. Kershaw MH, Westwood JA, Parker LL, Wang G, Eshhar Z, Mavroukakis SA, White DE, Wunderlich JR, Canevari S, Rogers-Freezer L, et al. (2006). A phase I study on adoptive immunotherapy using gene-modified T cells for ovarian cancer. Clin. Cancer Res. 12, 6106–6115. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Kochenderfer JN, Feldman SA, Zhao Y, Xu H, Black MA, Morgan RA, Wilson WH, and Rosenberg SA (2009). Construction and preclinical evaluation of an anti-CD19 chimeric antigen receptor. J. Immunother. 32, 689–702. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Kochenderfer JN, Dudley ME, Feldman SA, Wilson WH, Spaner DE, Maric I, Stetler-Stevenson M, Phan GQ, Hughes MS, Sherry RM, et al. (2012). B-cell depletion and remissions of malignancy along with cytokine-associated toxicity in a clinical trial of anti-CD19 chimeric-antigen-receptor-transduced T cells. Blood 119, 2709–2720. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Ku CC, Murakami M, Sakamoto A, Kappler J, and Marrack P (2000). Control of homeostasis of CD8+ memory T cells by opposing cytokines. Science 288, 675–678. [DOI] [PubMed] [Google Scholar]
  26. Kuwana Y, Asakura Y, Utsunomiya N, Nakanishi M, Arata Y, Itoh S, Nagase F, and Kurosawa Y (1987). Expression of chimeric receptor composed of immunoglobulin-derived V regions and T-cell receptor-derived C regions. Biochem. Biophys. Res. Commun. 149, 960–968. [DOI] [PubMed] [Google Scholar]
  27. Lamers CH, Sleijfer S, Vulto AG, Kruit WH, Kliffen M, Debets R, Gratama JW, Stoter G, and Oosterwijk E (2006). Treatment of metastatic renal cell carcinoma with autologous T-lymphocytes genetically retargeted against carbonic anhydrase IX: first clinical experience. J. Clin. Oncol. 24, e20–e22. [DOI] [PubMed] [Google Scholar]
  28. Lee DW, Kochenderfer JN, Stetler-Stevenson M, Cui YK, Delbrook C, Feldman SA, Fry TJ, Orentas R, Sabatino M, Shah NN, et al. (2015). T cells expressing CD19 chimeric antigen receptors for acute lymphoblastic leukaemia in children and young adults: a phase 1 dose-escalation trial. Lancet 385, 517–528. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Liao W, Lin JX, and Leonard WJ (2013). Interleukin-2 at the crossroads of effector responses, tolerance, and immunotherapy. Immunity 38, 13–25. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Long AH, Haso WM, Shern JF, Wanhainen KM, Murgai M, Ingaramo M, Smith JP, Walker AJ, Kohler ME, Venkateshwara VR, et al. (2015). 4–1BB costimulation ameliorates T cell exhaustion induced by tonic signaling of chimeric antigen receptors. Nat. Med. 21, 581–590, 10.1038/nm.3838. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. MacIver NJ, Michalek RD, and Rathmell JC (2013). Metabolic regulation of T lymphocytes. Annu. Rev. Immunol. 31, 259–283. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Martinez-Forero I, Azpilikueta A, Bolaños-Mateo E, Nistal-Villan E, Palazon A, Teijeira A, Perez-Chacon G, Morales-Kastresana A, Murillo O, Jure-Kunkel M, et al. (2013). T cell costimulation with anti-CD137 monoclonal antibodies is mediated by K63-polyubiquitin-dependent signals from endosomes. J. Immunol. 190, 6694–6706. [DOI] [PubMed] [Google Scholar]
  33. Maude SL, Frey N, Shaw PA, Aplenc R, Barrett DM, Bunin NJ, Chew A, Gonzalez VE, Zheng Z, Lacey SF, et al. (2014). Chimeric antigen receptor T cells for sustained remissions in leukemia. N. Engl. J. Med. 371, 1507–1517. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Maus MV, Haas AR, Beatty GL, Albelda SM, Levine BL, Liu X, Zhao Y, Kalos M, and June CH (2013). T Cells Expressing Chimeric Antigen Receptors Can Cause Anaphylaxis in Humans. Cancer Immunol. Res. 1, 26–31. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Maus MV, Fraietta JA, Levine BL, Kalos M, Zhao Y, and June CH (2014). Adoptive immunotherapy for cancer or viruses. Annu. Rev. Immunol. 32, 189–225. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Milone MC, Fish JD, Carpenito C, Carroll RG, Binder GK, Teachey D, Samanta M, Lakhal M, Gloss B, Danet-Desnoyers G, et al. (2009). Chimeric receptors containing CD137 signal transduction domains mediate enhanced survival of T cells and increased antileukemic efficacy in vivo. Mol. Ther. 17, 1453–1464. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Mookerjee SA, Divakaruni AS, Jastroch M, and Brand MD (2010). Mitochondrial uncoupling and lifespan. Mech. Ageing Dev. 131, 463–472. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Nicholls DG (2009). Spare respiratory capacity, oxidative stress and excitotoxicity. Biochem. Soc. Trans. 37, 1385–1388. [DOI] [PubMed] [Google Scholar]
  39. Nicholson IC, Lenton KA, Little DJ, Decorso T, Lee FT, Scott AM, Zola H, and Hohmann AW (1997). Construction and characterisation of a functional CD19 specific single chain Fv fragment for immunotherapy of B lineage leukaemia and lymphoma. Mol. Immunol. 34, 1157–1165. [DOI] [PubMed] [Google Scholar]
  40. Pearce EL, and Pearce EJ (2013). Metabolic pathways in immune cell activation and quiescence. Immunity 38, 633–643. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Pearce EL, Walsh MC, Cejas PJ, Harms GM, Shen H, Wang LS, Jones RG, and Choi Y (2009). Enhancing CD8 T-cell memory by modulating fatty acid metabolism. Nature 460, 103–107. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Pipkin ME, Sacks JA, Cruz-Guilloty F, Lichtenheld MG, Bevan MJ, and Rao A (2010). Interleukin-2 and inflammation induce distinct transcriptional programs that promote the differentiation of effector cytolytic T cells. Immunity 32, 79–90. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Porter DL, Hwang WT, Frey NV, Lacey SF, Shaw PA, Loren AW, Bagg A, Marcucci KT, Shen A, Gonzalez V, et al. (2015). Chimeric antigen receptor T cells persist and induce sustained remissions in relapsed refractory chronic lymphocytic leukemia. Sci. Transl. Med. 7, 303ra139. [Google Scholar]
  44. Sabbagh L, Pulle G, Liu Y, Tsitsikov EN, and Watts TH (2008). ERK-dependent Bim modulation downstream of the 4–1BB-TRAF1 signaling axis is a critical mediator of CD8 T cell survival in vivo. J. Immunol. 180, 8093–8101. [DOI] [PubMed] [Google Scholar]
  45. Sadelain M, Brentjens R, and Rivière I (2013). The basic principles of chimeric antigen receptor design. Cancer Discov. 3, 388–398. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Schluns KS, and Lefrançois L (2003). Cytokine control of memory T-cell development and survival. Nat. Rev. Immunol. 3, 269–279. [DOI] [PubMed] [Google Scholar]
  47. van der Windt GJ, Everts B, Chang CH, Curtis JD, Freitas TC, Amiel E, Pearce EJ, and Pearce EL (2012). Mitochondrial respiratory capacity is a critical regulator of CD8+ T cell memory development. Immunity 36, 68–78. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. van der Windt GJ, O’Sullivan D, Everts B, Huang SC, Buck MD, Curtis JD, Chang CH, Smith AM, Ai T, Faubert B, et al. (2013). CD8 memory T cells have a bioenergetic advantage that underlies their rapid recall ability. Proc. Natl. Acad. Sci. USA 110, 14336–14341. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Wang R, and Green DR (2012). Metabolic checkpoints in activated T cells. Nat. Immunol. 13, 907–915. [DOI] [PubMed] [Google Scholar]
  50. Worth AJ, Basu SS, Snyder NW, Mesaros C, and Blair IA (2014). Inhibition of neuronal cell mitochondrial complex I with rotenone increases lipid β-oxidation, supporting acetyl-coenzyme A levels. J. Biol. Chem. 289, 26895–26903. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Yadava N, and Nicholls DG (2007). Spare respiratory capacity rather than oxidative stress regulates glutamate excitotoxicity after partial respiratory inhibition of mitochondrial complex I with rotenone. J. Neurosci. 27, 7310–7317. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Zhao Y, Moon E, Carpenito C, Paulos CM, Liu X, Brennan AL, Chew A, Carroll RG, Scholler J, Levine BL, et al. (2010). Multiple injections of electroporated autologous T cells expressing a chimeric antigen receptor mediate regression of human disseminated tumor. Cancer Res. 70, 9053–9061. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Document S1. Supplemental Experimental Procedures and Figures S1–S4.
NIHMS2129298-supplement-Document_S1__Supplemental_Experimental_Procedures_and_Figures_S1_S4_.pdf (1.3MB, pdf)

RESOURCES