Research ArticleCAR–T CELL SIGNALING

Tuning ITAM multiplicity on T cell receptors can control potency and selectivity to ligand density

See allHide authors and affiliations

Science Signaling  22 May 2018:
Vol. 11, Issue 531, eaan1088
DOI: 10.1126/scisignal.aan1088

ITAM abundance increases receptor potency

T cells transduce extracellular signals by phosphorylating immunoreceptor tyrosine-based activation motifs (ITAMs) within the T cell receptor complex. Why this complex contains more ITAMs than any other immune receptor remains unclear. James engineered drug-responsive synthetic receptors that varied in ITAM abundance to determine how ITAM frequency influenced T cell signaling at the single-cell level. The experimental data fit computational models predicting that increasing ITAM number influenced the efficiency of signal transduction, but not the signal amplitude. Application of these results to human CAR–T cells identified strategies to improve the specificity of these cancer therapeutics.

Abstract

The T cell antigen receptor (TCR) recognizes peptides from pathogenic proteins bound in the major histocompatibility complex (MHC). To convert this binding event into downstream signaling, the TCR complex contains immunoreceptor tyrosine-based activation motifs (ITAMs) that act as docking sites for the cytoplasmic tyrosine kinase ZAP-70. Unique among antigen receptors, the TCR complex uses 10 ITAMs to transduce peptide-MHC binding to the cell interior. Using synthetic, drug-inducible receptor-ligand pairs, it was found that greater ITAM multiplicity primarily enhanced the efficiency with which ligand binding was converted into an intracellular signal. This manifested as an increase in the fraction of cells that became activated in response to antigen, and a more synchronous initiation of TCR-proximal signaling, rather than direct amplification of the intracellular signals. Exploiting these findings, the potency and selectivity of chimeric antigen receptors targeted against cancer were substantially enhanced by modulating the number of encoded ITAMs.

INTRODUCTION

T cells continuously interrogate the intracellular state of host cells through T cell antigen receptor (TCR) recognition of antigenic peptides bound within the major histocompatibility protein complex (pMHC). Recognition of “foreign” peptides presented at the surface of host cells activates TCR signaling and T cell functional responses. On productive ligand binding, TCR signaling is initiated by lymphocyte cell-specific protein tyrosine kinase (LCK)–mediated tyrosine phosphorylation of YXXL motifs within the intracellular tails of the TCR complex (1). These motifs, known as immunoreceptor tyrosine-based activation motifs (ITAMs), then act as docking sites for ZAP-70. Provided that the TCR remains bound by ligand, ZAP-70 becomes activated and continues to phosphorylate proteins such as linker for activation of T cells (LAT), which is a signaling scaffold that nucleates many canonical downstream pathways.

The TCR complex is an oligomeric association of the TCRαβ heterodimer, which is responsible for ligand binding, and the CD3γε, CD3δε, and CD3ζζ dimers required for intracellular signaling. Each CD3ζ contains three ITAMs, whereas the remaining CD3 chains have one ITAM each, for a combined total of 10 ITAMs per TCR complex. A long-standing question in T cell biology is why the TCR complex has so many of these binding sites, when almost all other immune receptors function effectively with no more than two. The best evidence suggests that TCR ITAMs are functionally redundant, and no single ITAM is required for normal T cell responses (2). However, TCR signaling and activation-induced cell proliferation appear to “scale” linearly with ITAM count in peripheral T cells. By contrast, cytokine production is almost invariant to changes in ITAM number (3, 4). Similarly, decreasing ITAM multiplicity in vivo inhibits T cell development, where fewer ITAMs diminish positive selection and impair thymocyte lineage commitment (2). Expression of seven or more CD3 ITAMs is necessary to establish a normal pattern of endogenous T cell development (3). However, TCR-ligand affinity may also influence the outcome of this kind of ITAM multiplicity experiment. Although the absence of CD3ζ ITAMs does not impair thymic positive selection when T cells express a high-affinity P14 transgenic TCR (5), the same is not true when T cells express low-affinity receptors for male self-antigen (3). Defects in negative selection that distort the peripheral T cell population have also been observed (6). Because of the noted defects in T cell development, experiments in in vivo models make it difficult to directly isolate the effect of receptor ITAM multiplicity from T cell signaling per se. This caveat perhaps explains why no consensus has been reached from these types of experiments on why TCRs require such high ITAM multiplicity. Although biophysical evidence suggests that increased ITAM number affects TCR phosphorylation and potency through an entropic mechanism (7), the effects of ITAM multiplicity on downstream signaling output remain unclear.

The evaluation of only the average signaling response of a population that expresses a given number of TCR ITAMs has made it difficult to determine whether modulating the number of ITAMs affects the signaling output homogeneously in all activated T cells or the proportion of cells that respond. This information is essential to understand the underlying functional requirement for high ITAM multiplicity in the TCR. To address the effect of receptor ITAM multiplicity without perturbing the underlying TCR signaling network, we developed synthetic receptors of varying ITAM multiplicities that responded to drug-mediated ligation and measured the output response at the single-cell level. We found that increasing the number of ITAMs robustly enhanced the probability of signal transduction by synthetic TCRs, which drove an increased fraction of activated T cells, rather than amplification of downstream responses. These conclusions were exploited to improve the efficiency of a chimeric antigen receptor (CAR)–T cell response.

RESULTS

A synthetic, drug-inducible TCR engineers control over T cell signaling

To isolate the effect of varying ITAM multiplicity from changes in receptor expression or affinity and any potentially confounding effects of network adaptation, we designed synthetic receptors that could be expressed in T cells in the presence of the endogenous TCR complex (Fig. 1A). In this system, the basal TCR signaling network was not disrupted by ITAM variance. These synthetic receptors used an extracellular ligand-binding domain based on the FK506-binding protein (FKBP) that provided an entirely orthogonal T cell input. To generate synthetic receptors containing zero to three ITAMs, we used the extracellular protein domains of CD86, a known monomer (8), as a scaffold to link the FKBP domain to the intracellular signaling region of the CD3ζ chain (Fig. 1A). Point mutations within the ITAMs of the CD3ζ chain were introduced to generate equivalent receptors with zero, one, or two functional ITAMs. To construct a synthetic receptor that expressed the full complement of 10 ITAMs, we repurposed the endogenously expressed TCR complex by introducing a modified version of the TCRα chain that encoded the FKBP domain extracellularly into T cells (Fig. 1A).

Fig. 1 Synthetic receptor signaling potency is enhanced at greater ITAM multiplicity.

(A) The FKBP-responsive synthetic receptors used in this study encode 0 to 3 ITAMs or 10 ITAMs. (B) Synthetic receptors are activated in T cells in the presence of endogenous TCR. Conjugation with cells presenting FRB, the cognate FKBP ligand, drives receptor engagement, but only in the presence of the rapalog drug. This can lead to LCK-mediated ITAM phosphorylation and the subsequent recruitment of the cytoplasmic kinase ZAP-70, which is necessary for phosphorylation of proteins, such as LAT, that initiate downstream signaling. (C and D) Jurkat cells expressing synthetic receptors encoding the indicated number of ITAMs and an NFAT-GFP reporter were activated with FRB-expressing Raji cells in the presence of 0.5 μM (C) or 2.5 μM (D) rapalog. Jurkat cell activation as measured by GFP expression was determined by flow cytometry analysis of cells gated for equivalent receptor surface expression. Data are representative of three independent experiments. au, arbitrary units. (E and F) The fraction of T cells activated (E) and the mean GFP intensity of this activated fraction (F) at the indicated rapalog concentration were calculated from the data in (C) and (D). Data are means ± SEM. (G) Representative density plots of GFP expression at each synthetic receptor ITAM multiplicity (white box) for the experiments shown in (D) and (I). Vertical lines denote the range of receptor expression analyzed, and the horizontal dashed line shows the threshold of activation. (H and I) Correlation of Jurkat cell activation [from (C) and (D)] with synthetic receptor expression for all variants in the presence of 0.5 μM (H) or 2.5 μM (I) rapalog. Data are means ± SEM. (J and K) The receptor density required to cause EC50 in the presence of 0.5 μM (J) or 2.5 μM (K) rapalog was calculated from data sets (H and I). Data are means ± 95% confidence interval.

A key feature of the FKBP domain is that its interaction with its binding partner, the FKBP12-rapamycin–binding (FRB) domain of mammalian target of rapamycin (mTOR) (9) is entirely dependent on the presence of a small molecule (10). Normally, this is the drug rapamycin, which can interfere with T cell activation (11). However, expressing FRB with the T2098L mutation enables the use of a rapamycin analog (a “rapalog,” known as AP21967 or A/C heterodimerizer) that has negligible binding to the equivalent domain of mTOR endogenously found in T cells. We used Raji cells (a B cell line) that expressed FRB at the cell surface as the ligand-presenting cells to drive synthetic receptor triggering by rapalog and subsequent downstream T cell activation (Fig. 1B). Synthetic receptors based on FKBP efficiently recapitulate TCR-mediated triggering (12). However, an important advantage of this drug-inducible receptor system was that it can provide fine temporal control over the initiation of signaling within the physiological context of apposing cell membranes, something that has previously not been possible with other methods for receptor activation.

ITAM number enhances receptor potency by increasing its efficacy

Increased ITAM multiplicity could affect T cell signaling output by two nonexclusive means. Multiple ITAMs may amplify the signal input and increase the absolute functional readout of T cell activation but would have little effect on the fraction of T cells that responded (fig. S1A). An alternative explanation is that having more ITAMs on a receptor enhances its potency, increasing the fraction of bound receptors that are capable of transducing a signal into the cell. This would not increase the absolute cellular output, but rather the number of cells that become activated (fig. S1B). Crucially, it would be very difficult to distinguish between these mechanistic explanations for ITAM multiplicity when only the average output response of the cell population is measured.

To provide the data required to distinguish between these two models, Jurkat cells (a human CD4+ T cell line) expressing a nuclear factor of activated T cells (NFAT) green fluorescent protein (GFP) reporter and the synthetic receptors with 0 to 3 or 10 ITAMs were stimulated with rapalog in the presence of FRB-expressing Raji cells. We used two different concentrations of the rapalog to explore the effect of ITAM multiplicity at different signaling input “strengths.” After stimulation of the cells for 16 hours, we measured GFP intensity as a readout of downstream NFAT activity output at the single-cell level (Fig. 1, C and D) and unmixed the total GFP output histograms to recover the distribution of activated cells (fig. S1C). We found that the number of ITAMs had a substantial effect on the fraction of cells that responded to stimulation (Fig. 1E) but did not greatly affect the overall magnitude of the output response, especially when more than one ITAM was present (Fig. 1F). This held true at both “low” (Fig. 1C) and “high” (Fig. 1D) receptor inputs mediated by the different rapalog concentrations.

As an alternative downstream functional output, we measured activation-induced cell surface CD69 expression, which is driven by the transcription factor activating protein 1 (AP-1) (13). We observed the same effect of increased ITAM multiplicity, with a substantial increase in the fraction of activated cells with essentially no change in the absolute abundance of CD69 (fig. S1, D to G). In agreement with others (14), we also observed that T cells increased their expression of CD69 when presented with increasing ligand density (fig. S1G), suggesting that CD69 expression is not an entirely digital response. These data also showed that rapalog stimulation was sufficient to activate essentially the entire T cell population, implying that the signaling threshold for CD69 expression was likely lower than that for NFAT activity where complete activation was not always observed. We also measured the effect of ITAM multiplicity on secretion of the cytokine interleukin-2 (IL-2) using an equivalent assay and found that IL-2 production correlated well with the fraction of activated cells (fig. S2).

By pooling T cells expressing synthetic receptors driven by promoters of different efficiencies, we could investigate the effect of a range of receptor surface expression within a single experiment. This enabled us to quantitatively determine the relationship between the receptor density and output to the signaling network while also varying the number of ITAMs (Fig. 1G and fig. S3). For the subsequent analysis, the absolute receptor abundance was quantified by flow cytometry so that we could directly compare NFAT activity at equivalent inputs for each ITAM variant. We plotted the fraction of responding cells as a function of receptor expression at two different rapalog concentrations. For both rapalog concentrations, it was evident that increasing ITAM multiplicity decreased the number of receptors required to generate an equivalent fraction of activated cells (Fig. 1, H and I). By fitting these data sets, we could quantify this effect on receptor potency directly by calculating the half-maximal cell activation (EC50) values from the number of synthetic receptors per cell required for a half-maximal response (Fig. 1, J and K). Given that all the synthetic receptors bear an identical ligand-binding domain and so have equivalent affinity, the increased potency that we observed with increasing ITAM multiplicity must be the result of increased signaling efficacy. This suggested that the primary effect of increasing the number of ITAMs was to improve the likelihood of productive signal transduction after ligand binding and did not substantially amplify the input signal per se. It was nonetheless possible for synthetic receptors with a reduced number of ITAMs to elicit the same response as the complete TCR construct, but this required a greater number of receptors to be engaged to drive an equivalent response. These data suggest that low potency may be overcome by increasing the total ligand input, at least for the outputs that we measured.

Increased receptor ITAM number synchronizes the initiation of T cell signaling

Our results demonstrated that the number of ITAMs within a synthetic receptor affected the efficiency with which the receptor transduced ligand binding to the intracellular signaling network. This suggests that ITAM multiplicity should have a substantial effect on the rate at which signaling is initiated from synthetic receptors, with a higher number of ITAMs predicted to drive more synchronous signaling. Because this effect should be evident in the kinetics of the early events of receptor signaling in the T cells, we measured the intracellular Ca2+ flux rapidly induced after T cell activation. This signal can be measured at single-cell resolution by flow cytometry using the Indo-1 ratiometric Ca2+ indicator (15).

Jurkat cells expressing equivalent amounts of the synthetic receptors (fig. S4A) were first conjugated with FRB-expressing Raji cells in the absence of the rapalog. These cell conjugates could then be gated on by flow cytometry (fig. S4B), and the intracellular Ca2+ concentration in these conjugates was measured for an initial period to define a baseline before the addition of the rapalog (Fig. 2A). From these data, we quantified the activated fraction of cells over time (Fig. 2, B and C) and the mean Ca2+ flux in the activated population (Fig. 2, D and E). These data show that the number of ITAMs present on the receptor had a potent effect on the rate at which T cells responded to stimulation, at both the low (Fig. 2, B and D) and high (Fig. 2, C and E) rapalog concentrations. We found a strong correlation between the number of ITAMs per receptor and the synchrony of the response to activation. This was best illustrated by taking the differential of the response over time to derive the rate of activation, which showed that increased ITAM multiplicity enhanced the “normal” temporal distribution of receptor triggering (Fig. 2, B to E). This effect can be solely ascribed to the number of ITAMs present on the receptor because both receptor expression and ligand affinity (fig. S4A) were consistent between all experiments.

Fig. 2 Increasing ITAM multiplicity drives synchronous T cell Ca2+ flux.

(A) Jurkat cells expressing synthetic receptors encoding the indicated number of ITAMs were loaded with the Ca2+ indicator Indo-1 before conjugation with FRB-expressing Raji cells. After activation with 2.5 μM rapalog, Ca2+ flux was determined by flow cytometry in conjugated cells. Data are representative of three independent experiments, and the density is denoted by color scale. (B and C) Left: The fraction of T cells that exhibited Ca2+ signaling after exposure to 0.5 μM (B) or 2.5 μM (C) rapalog was calculated from the data in (A). Data are means ± SEM. Right: The derived rate of activation (after smoothing) was calculated from the mean fraction of activated cells. (D and E) Left: The amplitude of Ca2+ flux as measured by the mean Indo-1 ratio of activated Jurkat cells after exposure to 0.5 μM (D) or 2.5 μM (E) rapalog was calculated from the data in (A). Data are means ± SEM. Right: The derived rate of Ca2+ flux (after smoothing) was calculated from the amplitude of activation.

Modeling our data set suggested that the increased synchronicity of the observed Ca2+ response could be explained by a combination of an invariant individual cellular response and variable temporal activation (fig. S4C). This suggests that the absolute Ca2+ flux of individual cells after synthetic receptor triggering could be constant despite altered ITAM multiplicity. Nonetheless, we cannot exclude the possibility of ITAM-mediated amplification of receptor signaling for Ca2+ flux during T cell activation, perhaps driven by the positive feedback loop of Ca2+-induced Ca2+ entry into T cells through calcium release–activated channels.

ITAM multiplicity affects signaling through the mitogen-activated protein kinase pathway

Our results demonstrated that increasing ITAM multiplicity on synthetic receptors caused an increased fraction of cells to be activated more synchronously. To investigate whether these responses could be observed as receptor activation was propagated through distal TCR signaling networks, we measured the extent of phosphorylation of extracellular signal–regulated kinase (ERK) at the single-cell level using intracellular flow cytometry. As described earlier, synthetic receptor–expressing Jurkat cells were conjugated with FRB-presenting Raji cells in the absence of the rapalog drug and the addition of rapalog initiated synchronous activation of receptor signaling. We found that after 5 min of signaling, ERK phosphorylation was readily detectable for the rapalog concentrations tested (Fig. 3, A and B). These results indicated that the effect of higher receptor ITAM multiplicity was to increase the fraction of activated cells (Fig. 3C), with almost no effect on the mean phosphorylation of ERK (Fig. 3D). This result agrees with data indicating that ERK phosphorylation is a “digital” response to graded T cell activation (16). Because of the rapid kinetics of ERK phosphorylation, we also measured ERK phosphorylation at earlier time points. In agreement with our Ca2+ flux results, we found that increased ITAM multiplicity drove a more rapid response, which was evident especially at 2 min after rapalog addition (Fig. 3E and fig. S5), and the mean ERK phosphorylation became essentially equivalent once the dynamics reached equilibrium after ~3 min (Fig. 3F). Overall, our data confirmed our previous findings that greater ITAM multiplicity increased both the synchrony of signaling and the fraction of cells activated.

Fig. 3 Increased ITAM multiplicity stimulates faster ERK phosphorylation.

(A and B) Jurkat cells expressing synthetic receptors encoding the indicated number of ITAMs at equivalent abundance were conjugated with FRB-expressing Raji cells. After 5 min of activation with 0.5 μM (A) or 2.5 μM (B) rapalog, pERK1/2 in conjugated cells was detected by flow cytometry. Data are representative of three independent experiments. (C and D) The fraction of Jurkat cells exhibiting ERK activation (C) and the mean intensity of pERK of this fraction (D) at the indicated rapalog concentrations were calculated from the data in (A) and (B). Data are means ± SEM. (E and F) Phosphorylation of ERK1/2 was analyzed by flow cytometry in the same cells activated for 0 to 5 min with 2.5 μM rapalog in the presence of ligand-expressing Raji cells. The fraction of Jurkat cells exhibiting ERK activation over time (E) and the mean intensity of pERK in this activated fraction (F) were calculated. Data are means ± SEM of three independent experiments.

ZAP-70 kinase activation primarily depends on autophosphorylation

The preceding experiments showed that the number of ITAMs present on synthetic receptors predominantly affected the efficiency of signal transduction after ligand binding. How could high ITAM multiplicity improve receptor potency, if not through downstream signal amplification? One possibility could be that multiple binding sites within a single antigen receptor are required to bring multiple kinases into close physical proximity. This could be important for ZAP-70 because full activation of its kinase activity requires the phosphorylation of Tyr493 (Y493) in its kinase domain activation loop (17). For high ITAM multiplicity to cause efficient signal transduction in this manner, ZAP-70 activation would need to occur through trans-autophosphorylation (Fig. 4A). However, there is limited experimental evidence for an autophosphorylation mechanism for ZAP-70 activation (18, 19). The best evidence suggests that the ZAP-70 activation loop sequence should be a poor substrate for its own kinase domain (20) and that ZAP-70 is primarily activated by LCK (Fig. 4A).

Fig. 4 ZAP-70 activation is primarily driven by autophosphorylation.

(A) ZAP-70Y493 phosphorylation could be driven by LCK-mediated kinase activity (left) or through ZAP-70 trans-autophosphorylation (right). (B) HEK-293T cells expressing the minimal components required for TCR signaling [TCR, LCK, ZAP-70, LAT, CD45, and C-terminal Src kinase (CSK)/CSK-binding protein C (CBP)] were purified by sorting and synchronously activated by inhibiting the phosphatase CD45 with pervanadate (PerVi). At defined times, aliquots of the activated cells were removed and snap-frozen for phospo–Western blotting analysis. (C and D) ZAP-70 phosphorylation at Y493 (C) or LAT phosphorylation at Y132 (D) in cells expressing either the WT or kinase-deficient (“Dead”) versions of LCK and ZAP-70. Data are means ± SEM of three independent experiments.

To determine whether ZAP-70 autophosphorylation was important for its own activation, we used a previously characterized system to recapitulate proximal TCR signaling events in human embryonic kidney (HEK) 293T cells (12). This approach relies on the expression of a minimal set of proteins necessary for proximal TCR signaling and enables us to quantitatively investigate ZAP-70 activation within a cellular system that expresses none of the T cell proteins that could potentially confound the interpretation of our results. In this system, phosphatase inhibition was used to synchronously initiate proximal TCR signaling (Fig. 4B and fig. S6A), and the intensity of ZAP-70Y493 phosphorylation was quantified over time by Western blotting analysis. By coexpressing the wild-type (WT) forms of both LCK and ZAP-70, we could readily observe that the phosphorylation of ZAP-70Y493 increased over time (Fig. 4C). To confirm that ZAP-70 phosphorylation in the reconstituted signaling cascade was dependent on LCK activity, similar to TCR signaling (12), we used a catalytically inactive mutant of LCK (LCKK273R) and found that ZAP-70 phosphorylation was undetectable. A substantial reduction in ZAP-70Y493 phosphorylation was also observed after repeating this assay with a kinase-deficient version of ZAP-70 (ZAP-70K369R), indicating that ZAP-70 activation has a strong dependence on its own kinase activity. In agreement with other studies, these data suggest that trans-autophosphorylation can play a crucial role in ZAP-70 activation (21). This requirement was not absolute, however, because there was still a small fraction of total Y493 phosphorylation that was mediated by LCK when ZAP-70 was catalytically inactive. This may be expected if LCK acts to “prime” the ITAM-bound ZAP-70 to initiate the autophosphorylation, although the bulk of ZAP-70Y493 phosphorylation appeared to derive from trans-autophosphorylation. ZAP-70 kinase activity was also required for LATY132 phosphorylation after stimulation of reconstituted proximal TCR signaling (Fig. 4D). Because LATY132 phosphorylation recruits phospholipase–Cγ1 (PLC-γ1) (22, 23), which is necessary for downstream NFAT activity (Fig. 1) and Ca2+ flux (Fig. 2), this result implies that the effects of ITAM multiplicity that we observed may be due to ZAP-70 autophosphorylation.

To determine whether the requirement for ZAP-70 autophosphorylation could also explain the kinetics of our Ca2+ flux (Fig. 2) and pERK data (Fig. 3), we generated network models of the known steps in proximal TCR triggering using BioNetGen software (software S1) (24). We scaled these models with receptor ITAM density and used LAT phosphorylation as a “readout” of receptor activation (fig. S6, B to E). A model that relied on LCK-mediated ZAP-70 activation alone could not readily replicate our experimental data (fig. S6, B and C). However, a model in which ZAP-70 autophosphorylation within a single receptor complex was greatly favored over LCK-mediated phosphorylation appeared to fit the substantial increase in signal output from one to two ITAMs (fig. S6, D and E).

CAR activation and discrimination can be controlled by modulating the number of ITAMs

CARs are a new class of cancer therapy that use genetically engineered T cells to attack malignant cells. These constructs splice the high-affinity binding of antibodies onto the ITAMs of the TCR, including costimulatory motifs to maintain their expression in the host (fig. S7A). Although much work has gone into improving the latter sequences, there has been very limited work on optimizing the signaling potency mediated by the ITAMs (25). Current CARs use the three ITAMs from the CD3ζ chain of the TCR, but we hypothesized that the efficiency of CAR activation could be improved by simply increasing the ITAM density of the construct. To do this, we duplicated the CD3ζ sequence to generate a six-ITAM CAR variant (fig. S7A). We also made equivalent versions that had zero to two ITAMs to quantify the relationship between the number of ITAMs and the CAR-mediated response. As a proof of principle, we used a CAR that is reactive to CD19, a B cell–specific surface protein that has been used to target CAR-expressing T cells against lymphoblastic leukemias (26, 27). We transduced human primary CD4+ T cells (fig. S7B) to express anti-CD19 CAR variants with zero to six ITAMs. As surrogate target cells, we used K562 myeloma cells that were transduced to express different amounts of the CD19 antigen that span the normal range of CD19 expression (fig. S7C). The CAR–T cells were conjugated with the target cells for 24 hours to drive CD19-induced T cell activation, which was determined by quantifying the expression of CD137 (fig. S7D), a robust marker of activation that correlates well with cell proliferation (28).

The “standard” CAR construct harboring three ITAMs activated the transduced T cells when they were mixed with target cells presenting antigen at all three densities (Fig. 5A). We again found that the effect of increased ITAM number primarily caused an increase in the fraction of activated cells (Fig. 5B) and not the absolute amount of the cell output (Fig. 5C). This effect held true for all three amounts of CD19 expressed on the target cells (Fig. 5, B and C), although activation was enhanced with higher target density, which suggests that CAR-mediated signaling was not saturated. Increasing the number of ITAMs from three to six increased the efficiency of T cell activation by ~15% for all target cells (Fig. 5, A and D). Similar results were found using CD69 as the readout for T cell activation, although the activation period may have enabled subsequent down-regulation of CD69, which slightly confounded the analysis (fig. S7, E to G). An unanticipated result from this data set was that by decreasing the number of ITAMs in the receptor construct from three to two ITAMs, the CAR–T cells were now substantially more selective for target cells expressing CD19 at high density (Fig. 5D), something that CARs are not currently optimized for. This agreed well with earlier results from synthetic receptors showing that equivalent T cell activation could be achieved with fewer ITAMs when more receptors were engaged, which could also be the case in the response to target cells expressing high amounts of CD19. To expand on this finding, we manipulated the available target density on the CD19-expressing target cells with the addition of an antibody against CD19 and tested the specificity of the response by a CAR with three ITAMs. We found that the CD19-blocking antibody artificially shifted the effective antigen densities required for CAR–T cell responses and increased selectivity for target cells that overexpressed CD19 (Fig. 5E). The addition of the blocking antibody “protected” those target cells that expressed physiological amounts of CD19 without inhibiting the response to cells that overexpressed CD19 (Fig. 5F). These data suggest that understanding how ITAM multiplicity affects receptor potency may inform new strategies to improve CAR–T cell specificity.

Fig. 5 CAR potency can be modulated by ITAM multiplicity.

(A) Human CD4+ T cells that expressed CD19-reactive CARs with the indicated number of ITAMs were mixed with K562 target cells that expressed distinct amounts of CD19. T cell activation as measured by cell surface expression of CD137 (4-1BB) was determined 24 hours later by flow cytometry. Data are representative of three independent experiments. (B and C) The fraction of activated T cells (B) and the mean intensity of CD137 of this activated fraction (C) were calculated from the data in (A). Data are means ± SEM. (D) The fraction of activated T cells for each CAR expressing the indicated number of ITAMs was calculated from the data in (B). Data are means ± SEM. (E) T cells expressing the WT CAR (three ITAMs) were activated with the indicated target cells with or without a blocking antibody (Ab) against CD19. Data are representative of three independent experiments. (F) The fraction of CAR-expressing T cells activated by target cells in the presence or absence of the CD19-blocking antibody was calculated from (E). Data are means ± SEM.

DISCUSSION

We have developed a synthetic receptor system with which to quantitatively investigate how the number of ITAMs encoded by a cell surface receptor modulates its ability to drive intracellular signaling in T cells. We found that greater receptor ITAM multiplicity enhanced the potency of the receptor by increasing the transduction efficiency stimulated by ligand binding. This effect manifested as receptors with a greater number of ITAMs stimulating an increase in the fraction of T cells that were activated and more synchronous initiation of downstream signaling across experiments testing multiple downstream signaling outputs. Through reconstitution experiments, we showed that there was a substantial requirement for the autophosphorylation of ZAP-70 to stimulate its kinase activity, which could provide a mechanistic explanation for the increased efficacy observed with high ITAM multiplicity. We then applied our results to demonstrate that we could improve the therapeutic potential of CARs. Similar to our results on synthetic ITAM-containing receptors, we found that increased ITAM count on an anti-CD19 CAR improved the activation of primary human T cells. This result is applicable to all current CARs being investigated and may be critical to augmenting responses to target antigens expressed at low amounts. Conversely, our results suggest that decreasing the number of ITAMs or using different ITAM sequences with weaker binding affinities (29) could be an interesting new avenue to provide increased CAR–T cell selectivity.

These findings contrast with the prevailing view that ITAM multiplicity is principally for intracellular signal amplification, which should lead to greater absolute signaling output. We believe that our experiments at single-cell resolution demonstrate that this is not the primary effect of a high number of ITAMs. Instead, we found that greater ITAM multiplicity led to a decrease in the number of triggered receptors required to cause a half-maximal output response. Equivalent signaling output was possible with fewer ITAMs, but this required more receptors to be engaged. By only measuring the mean response of a population of T cells, as essentially all other experiments on ITAM number have been performed, it would not be possible to separate these two explanations. Our results suggest that the ITAM multiplicity of the receptor influences the efficiency of converting ligand binding into an intracellular signal. Although it could be argued that simply having more bound ZAP-70 is amplification per se, we believe that this is not an appropriate use of the term because of the change in input modality, with amplification normally only defined as an increase in a signal amplitude.

A key advantage of our experiments using the synthetic receptors is that a gain-of-function approach to investigate ITAM multiplicity obviates the need to directly disrupt the underlying signaling network, which can often lead to unintended consequences such as systemic adaptation. There are of course caveats to our approach too. By expressing the synthetic receptors in an endogenous T cell line, we are inherently increasing the total density of ITAMs at the cell surface, which could alter basal signaling. We were therefore careful not to overexpress the receptor, which increased the total ITAM density by no more than ~10% at the highest receptor expression used. The synthetic receptor also relied on the FKBP/FRB interaction, which has a higher affinity than that of the normal TCR/pMHC equivalent, although the kinetics of the FKBP unbinding under tension have never been measured and so may not be so different to that for pMHC dissociation. The presence of the endogenous TCR in our synthetic receptor–expressing T cells could potentially lead to signal augmentation by co-opting these receptors. However, we observed no evidence for the down-regulation of WT TCR in our assays even with potent stimulation, suggesting that only the engaged receptors were providing signal input to the cells.

Although the synthetic receptors used in this study may not replicate all of the features of the endogenous TCR, we believe that the work presented here can explain the requirement for 10 ITAMs being present within the TCR complex. Why does the TCR have this requirement? The unique aspect of TCR-mediated signaling is that it must discriminate potential ligands with very high sensitivity, yet it relies on only a few ligands to drive cell activation (30, 31). To achieve this, the TCR must be able to convert cognate pMHC binding events as effectively as possible into intracellular signals; this efficacy in signal transmission across the plasma membrane is likely derived from the high ITAM multiplicity, something that is not required for most other immune receptors. Furthermore, having the ITAMs distributed over multiple chains, rather than encoded by a single large sequence, might increase the efficiency of receptor triggering by increasing the effective local concentration of ZAP-70 binding sites. It is also possible that ligand-induced receptor clustering (32) or preformed TCR “nanoclusters” in the plasma membrane (33) could further enhance the efficacy of receptor triggering in response to low pMHC abundance by increasing the effective number of ITAMs that could drive ZAP-70 autophosphorylation and proximal signaling.

MATERIALS AND METHODS

Vector construction

Modification of pHR-SIN lentiviral backbone vectors was performed using standard cloning techniques (12). When required, fluorophores were fused in-frame at the C terminus of the gene to be tagged. All oligonucleotides used in this study are given in table S1. To drive a range of gene expression, the spleen focus-forming virus (SFFV) promoter from pHR vector was replaced by EEF1A1 (pHREF) or cytomegalovirus (CMV) (pHCM) promoters. The TCRαβ sequences from clone 1G4 were linked through a 2A sequence into one contiguous reading frame and inserted into pHREF vector to generate pHREF-1G4. To generate pHR-CBP/CSK-mIFP, the genes encoding CBP and CSK replaced the TCR genes in pHREF-1G4, and mIFP was fused to the C terminus of CSK. To generate pHCM-CD45RO–monomeric enhanced GFP (mEGFP), the gene for CD45RO (including the sequence encoding an N-terminal His tag) was inserted into pHCM modified to express mEGFP. To drive high expression, the SV40 introns (16S) were inserted before the 5′ end of this construct. To generate pHR-FKBP-CD86-CD3ζ from FKBPExζInt (12), the sequence encoding the signal peptide of CD86 was replaced with one from Gaussia luciferase and included an hemagglutinin (HA) epitope at the mature N terminus. In addition, a Spe I restriction site was inserted just after the sequence encoding CD86 by polymerase chain reaction (PCR) mutagenesis to facilitate the mutation of CD3ζ ITAMs. To generate pHR-FKBP-TCRα, the sequence encoding the FKBP domain was fused to the mature N-terminal sequence of the Jurkat TCRα chain by extension PCR. To generate pHREF-CAR-mRuby2 (αCD19), the sequence encoding the anti-CD19 antibody fragment from clone FMC63, along with those encoding the CD8 stalk and transmembrane regions and the 4-1BB intracellular domain, were synthesized and inserted into pHR-FKBP-CD86-CD3ζ. To make the six-ITAM CAR variant, a second copy of the CD3ζ ITAMs (synonymously mutated) were synthesized as a GeneArt String and inserted into the vector.

Cell culture

HEK-293T cells and all derivatives were grown in Dulbecco’s modified Eagle’s medium (Sigma-Aldrich) supplemented with 10% fetal bovine serum (FBS) (Life Technologies) and antibiotics. WT Jurkat cells and Raji cells (including all derivatives) were grown in RPMI 1640 (Sigma-Aldrich), supplemented with 10% FBS, 10 mM Hepes, and antibiotics. K562 cells (including all derivatives) were grown in Iscove’s modified Dulbecco’s medium (Life Technologies) supplemented with 10% FBS. A clonal Jurkat derivative (J.NFAT) that expressed GFP upon TCR stimulation using NFAT-response elements and constitutively expressing the fluorophore iRFP713 for identification was generated. To construct K562 target cells expressing different amounts of the CD19 protein, the CD19 gene (GeneArt, Thermo Fisher Scientific) was synthesized and inserted into lentiviral vectors with different promoters. K562 cells were transduced with these constructs and sorted on the basis of cell surface CD19 expression. Human primary CD4+ T cells were purified by negative selection from whole blood samples (Cambridge Blood and Stem Cell Biobank) using the EasySep Direct Human CD4 Isolation Kit (STEMCELL Technologies) and were cultured in ImmunoCult-XF T Cell Expansion Medium (STEMCELL Technologies).

Lentivirus production

HEK cells were transiently transfected with a lentiviral backbone, pCMVΔ8.91(34), and pMD2.G (gift from D. Trono) at a 2:2:1 ratio using GeneJuice (EMD Millipore) according to the manufacturer’s instructions. After 48 to 72 hours, cell culture medium was harvested, centrifuged to remove debris, and used to innoculate ~1 × 106 target cells. CAR constructs were concentrated ~40× using Lenti-X (Takara Clontech). After 16 hours, fresh medium was added, and the cells were rested for at least 3 days until expression of the target protein could be detected.

Flow cytometry

Cell samples were stained with primary antibodies (~10 μg/ml) in flow wash buffer [2.5% (v/v) FBS and 0.1% (w/v) NaN3 in phosphate-buffered saline (PBS) (pH 7.4)] and incubated on ice for at least 30 min. For Raji cells, endogenous Fc receptors were preblocked with Human TruStain FcX (BioLegend) for 10 min at room temperature. The following antibodies were used: anti-HA (6E2) from Cell Signaling Technology and anti-human CD69 (FN50), CD19 (HIB19), and CD137 (4B4-1) from BioLegend. Samples were washed and fixed in flow fix buffer [1.6% (v/v) formaldehyde, 2% (w/v) glucose, and 0.1% (w/v) NaN3 in PBS (pH 7.4)] or labeled where necessary with an Alexa Fluor 647–conjugated anti-mouse secondary antibody (20 μg/ml) before being washed and fixed. Cell staining was assessed using a LSRFortessa flow cytometer (BD Biosciences), and the data were analyzed with FlowJo software (FlowJo). A complete list of antibodies is given in table S2.

Chemically inducible synthetic receptor system

Construction of the vectors for the FKBP-CD86-CD3ζ receptor (previously described as FKBPExζInt), its ligand, FRBEx, and the Raji cell line expressing this ligand have been previously described (12), although the synthetic receptor now included an HA epitope at the mature N terminus for the quantification of expression. The first tyrosine in each ITAM was mutated to a phenylalanine by PCR mutagenesis to sequentially decrease the number of ITAMs. To engineer a receptor complex with 10 ITAMs, we fused FKBP to the mature N terminus of the TCRα sequence from Jurkat cells (along with an HA epitope), which was incorporated into the endogenous Jurkat TCR complex through competition with the WT equivalent TCRα chain. J.NFAT cells were transduced with constructs encoding the synthetic receptors, and expression was measured by detecting the HA epitope, which was quantified by comparison to standardized Alexa Fluor 647 MESF calibration beads (Bangs Laboratories) according to the manufacturer’s instructions.

Calcium flux and pERK analysis

Synthetic receptor–expressing Jurkat cells were loaded with Indo-1 (TEFLabs), a fluorescent Ca2+ indicator, before conjugation with Raji-FRBEx cells expressing the cognate binding domain. The fluorescence of cell conjugates was monitored by flow cytometry during baseline acquisition and after the addition of the rapalog drug (AP21967, Takara Clontech). Ca2+ flux in T cell conjugates was determined by calculating the Indo-1 ratio of the fluorescence emission detected at 405 nm/485 nm, and the data were further analyzed in MATLAB. For each sample, the data were binned into 1-s windows. The histogram of the Indo-1 ratio at each time point was calculated before normalization using the 2-norm of the vector. The resulting array was either directly plotted or used for further analysis. A two-peak Gaussian model was fitted to each time bin to identify the background signal distribution, and a value 2σ from the mean was used as the threshold. The median of the Indo-1 ratio above this threshold was calculated for each time point. Similarly, to measure pERK abundance, synthetic receptor–expressing T cells were conjugated with Raji-FRBEx cells and then stimulated with rapalog for a defined period at 37°C. Cells were then fixed with 4% formaldehyde, permeabilized with methanol, and subsequently stained with a directly conjugated fluorescent antibody specific for pERK1/2 (T202/Y204) (Cell Signaling Technology).

Downstream cellular activation

J.NFAT cells expressing the appropriate synthetic receptor were conjugated to Raji-FRBEx cells for 30 min in the presence of rapalog drug at the appropriate concentration. Conjugates were then cultured for 16 hours to enable activation-induced expression of GFP from the NFAT-responsive promoter. The culture medium of the activated cells was also collected as this point to assay for IL-2 secretion, which was quantified with the IL-2 Human ELISA Kit (Thermo Fisher Scientific) according to the manufacturer’s instructions. Cells were stained with a fluorescently conjugated antibody against the HA epitope before analysis by flow cytometry for receptor and GFP expression. The data were gated for equivalent receptor expression in FlowJo and exported to MATLAB. The control sample without drug addition was used to define the background intensity distribution and “unmix” the positive population from the sample histogram. In MATLAB, the sample distribution of the activated T cells was calculated by subtracting the background population from the sample histogram. The arithmetic mean of the log-transformed data was calculated for the activated distribution. For the derivation of receptor potency, the expression of the synthetic receptor was used to bin the data, and the fraction of activated cells in each bin was calculated. Data sets were fitted using a three-parameter logistic regression with the asymptote constrained to that of the three-ITAM data, from which EC50 values were derived.

Cellular reconstitution of TCR phosphorylation kinetics

HEK-TCR cells, which express the entire TCR complex (1G4 clone), were transiently transfected [to express LCK, ZAP-70, LAT, CD45, and CSK/CBP as previously described (12)] and purified by sorting. Sorted HEK cells were then activated at 21°C with pervanadate and aliquots taken at defined times by rapid snap-freezing. Frozen cell lysates were analyzed by Western blotting with the following antibodies: anti-human ZAP-70 (D1C10E), anti-human ZAP-70 pY493 (rabbit polyclonal), anti-human LAT (rabbit polyclonal) from Cell Signaling Technology, and anti-human LAT pY132 (rabbit polyclonal) from Thermo Fisher Scientific. Blots were developed with DyLight 800 Goat anti-rabbit immunoglobulin G (Cell Signaling Technology) and imaged using an Odyssey Imaging System. A custom Fiji ImageJ plugin was used to extract the integrated intensities of the appropriate bands, which were corrected for local background fluorescence. To convert the raw intensities derived from the blots into absolute numbers so that different experiments could be directly compared, a subset of samples at equivalent times were resolved on a single gel, and the integrated band intensities for both the phosphorylated (IP) and native proteins (IN) were calculated. The fraction of phosphorylated protein (IP/IN) at the time point was then used to scale the raw data to that value. Data sets were fitted in MATLAB using a five-parameter logistic regression.

Modeling effect of ITAM number on receptor activation

BioNetGen software (24) was used to model proximal TCR signaling incorporating the known initiating steps in TCR activation that lead to LAT phosphorylation. The two tyrosine residues of an ITAM were modeled as a single phosphorylation event to minimize complexity. To define all the TCR-bound ZAP-70 variants with different numbers of ITAM-binding sites, a bash script was used to iterate over all potential combinations. Because BioNetGen does not account for multiple enzymes being bound to a single species (in this case, multiple ZAP-70 molecules bound to the TCR), it was necessary to include rules that explicitly accounted for this when describing LAT phosphorylation by ZAP-70. Up to six ZAP-70–binding sites could be computed (more than ~20 hours), but more than this became unfeasible. For the model without ZAP-70 autophosphorylation, the rate parameters were simply adjusted to linearize ZAP-70 activation. Mass action kinetics, rather than Michaelis-Menten kinetics, were used to simplify the reaction network and reduce computational complexity.

CAR activation

Human primary CD4+ T cells were transduced with a construct encoding the anti-CD19 CAR, including mutations of the ITAMs of the CD3ζ sequence as described earlier for the synthetic receptor earlier. The cells were allowed to proliferate for more than ~10 days until there was a clear decrease in cell volume, as judged by assessing scatter on a flow cytometer. The T cells were conjugated with CD19-expressing K562 target cells for 24 hours, stained with fluorescently labeled antibodies against CD137 and CD3, and then analyzed by flow cytometry. When required, the target cells were preincubated with an anti-CD19 antibody (FMC63, Absolute Antibody) to block the CAR epitope.

SUPPLEMENTARY MATERIALS

www.sciencesignaling.org/cgi/content/full/11/531/eaan1088/DC1

Fig. S1. Approach to evaluating the effects of synthetic receptor ITAM multiplicity.

Fig. S2. Effect of ITAM multiplicity on IL-2 secretion.

Fig. S3. Rapalog stimulates synthetic ITAM-containing receptors.

Fig. S4. Approach to evaluating synthetic receptor–induced Ca2+ flux.

Fig. S5. Kinetics of ERK phosphorylation after stimulation of synthetic receptors.

Fig. S6. Modeling the mechanism of ZAP-70 activation by autophosphorylation.

Fig. S7. Approach to evaluating the effect of ITAM multiplicity in anti-CD19 CAR–T cells.

Table S1. Oligonucleotide sequences.

Table S2. List of select reagents and resources.

Software S1. Modeling TCR-proximal signaling with BioNetGen.

REFERENCES AND NOTES

Acknowledgments: We thank the Trono (École polytechnique fédérale de Lausanne) and Cerundolo (University of Oxford) laboratories for sharing plasmids. We thank the staff at the Cambridge Blood and Stem Cell Biobank for blood collection. Funding: This work was supported by the Wellcome Trust and the Royal Society (#099966/Z/12/Z to J.R.J.). Author contributions: J.R.J. designed the project, performed the experiments, analyzed the data, and wrote the manuscript. Competing interests: The author declares that he has no competing interests. Data and materials availability: The pHREF-1G4 and pHCM-CD45RO plasmids require a material transfer agreement from the University of Cambridge, UK. BioNetGen code is available from the Faeder Lab at the University of Pittsburgh (www.csb.pitt.edu/Faculty/Faeder/). All other data needed to evaluate the conclusions in the paper are present in the paper or the Supplementary Materials.
View Abstract

Navigate This Article