Research ArticleCAR-T CELLS

Engineering γδT cells limits tonic signaling associated with chimeric antigen receptors

See allHide authors and affiliations

Science Signaling  10 Sep 2019:
Vol. 12, Issue 598, eaax1872
DOI: 10.1126/scisignal.aax1872

Taming tonic signaling

Adoptive transfer of T cells engineered to express a chimeric antigen receptor (CAR) is an effective therapy for select lymphomas. The potency of this therapy can be limited by antigen-independent (tonic) signaling, which promotes progressive CAR-T cell inactivation. Fisher et al. used mass cytometry to analyze CAR-T cells and found that the process of increasing αβ T cell numbers during CAR-T cell production (expansion) was sufficient to increase tonic signaling in αβ T cells. In contrast, expansion of γδT cells did not alter their basal activity. When these cells were engineered to express a chimeric costimulatory receptor, they specifically recognized transformed, but not healthy, myeloid cell targets. These data demonstrate a strategy for engineering specific antitumor responses free of the complications associated with tonic signaling.


Despite the benefits of chimeric antigen receptor (CAR)–T cell therapies against lymphoid malignancies, responses in solid tumors have been more limited and off-target toxicities have been more marked. Among the possible design limitations of CAR-T cells for cancer are unwanted tonic (antigen-independent) signaling and off-target activation. Efforts to overcome these hurdles have been blunted by a lack of mechanistic understanding. Here, we showed that single-cell analysis with time course mass cytometry provided a rapid means of assessing CAR-T cell activation. We compared signal transduction in expanded T cells to that in T cells transduced to express second-generation CARs and found that cell expansion enhanced the response to stimulation. However, expansion also induced tonic signaling and reduced network plasticity, which were associated with expression of the T cell exhaustion markers PD-1 and TIM-3. Because this was most evident in pathways downstream of CD3ζ, we performed similar analyses on γδT cells that expressed chimeric costimulatory receptors (CCRs) lacking CD3ζ but containing DAP10 stimulatory domains. These CCR-γδT cells did not exhibit tonic signaling but were efficiently activated and mounted cytotoxic responses in the presence of CCR-specific stimuli or cognate leukemic cells. Single-cell signaling analysis enabled detailed characterization of CAR-T and CCR-T cell activation to better understand their functional activities. Furthermore, we demonstrated that CCR-γδT cells may offer the potential to avoid on-target, off-tumor toxicity and allo-reactivity in the context of myeloid malignancies.


Chimeric antigen receptors (CARs) represent a major technological advance in cancer therapeutics. Expression of CARs redirects effector T cells to become reactive against cell surface antigens in a manner that is independent of the major histocompatibility complex (MHC). This synthetic form of immunotherapy enables effective killing of immunologically “quiet” tumors and has achieved dramatic and encouraging successes against B lymphoid malignancies in particular (15). CARs are engineered to include an ectodomain, typically a single-chain antibody (scFv), that determines tumor antigen specificity, and endodomains comprising signal transduction motifs from T cell immunoreceptors. The most popular endodomain configuration uses the CD3ζ chain in combination with one or more costimulatory endodomains such as CD28 or 4-1BB. When activated, the CAR provides the equivalent of a T cell receptor (TCR) signal through CD3ζ, in addition to the costimulation required for T cell activation. The specific combination of the ectodomain and the costimulatory endodomain(s) included in the CAR influences the phenotype and longevity of CAR-T cells, and relative efficacy has been linked to differing phosphorylation kinetics in downstream signaling molecules (6, 7).

Two major limitations of CARs have emerged, especially as the field progresses from early demonstrations of efficacy in B cell disease to the more challenging solid tumors. First, CARs based on single antibodies cannot discriminate between tumor and healthy cells that express the target antigen, raising the risk of on-target off-tumor toxicity (811). Second, antigen-independent CAR activation or “leakiness,” termed tonic signaling, can have deleterious effects on CAR-T cell phenotype and efficacy (1214). The ability to fine-tune CAR signaling is thus critical to minimize tonic signaling and overcome these limitations. However, there is a paucity of detailed mechanistic analysis of CAR-T signaling, either as cells generated and expanded ex vivo or following encounter with tumor. Whereas recent reviews have favored research into novel antigens and combination therapies (15), if the fundamental signaling mechanics and behavior of CARs are not understood, changing antigens by itself may provide little advantage. Bulk phosphoproteomic analyses of CAR-T cell signaling have tried to address this (6, 7, 16), but these lack the resolution to decipher molecular signaling relationships and their function within a cell. Techniques such as mass cytometry measure multiple molecular epitopes at single-cell resolution and thus have potential to elucidate CAR-T cell signaling dynamics. We therefore sought to compare signaling at single-cell resolution in T cells before and after and ex vivo expansion and CAR transduction, to interrogate the effects of CAR expression on the intracellular signaling network, and to learn how this information can be used to guide CAR design.

Here, we used mass cytometry to diagnose problems in CAR design, with a focus on the longstanding challenges of tonic signaling and off-target toxicity. To quantify changes in phosphoprotein abundance and codependence in canonical signaling networks in CAR-transduced cells, we adapted analytical approaches including the earth mover’s distance (EMD) (17, 18) and density resampled estimate of mutual information (DREMI) (19), which provided a less biased assessment of information transfer between pairs of signaling molecules and enable robust comparisons across conditions. Our analysis pinpointed that the CD3ζ module was the source of second-generation CAR tonic signaling in CARs with three different antigen recognition domains. We then demonstrated that an alternative CAR design, which lacked TCR signal transduction elements and contained an AND gate, avoided on-target off-tumor toxicity when expressed in Vδ2+ γδT cells. We found that despite using the same ectodomains as 28ζ CARs, chimeric costimulatory receptors (CCRs) expressed in γδT cells produce limited network perturbation and no tonic signaling. Single-cell analysis of signaling through synthetic receptors offers a rapid and powerful way to understand the limitations of CAR therapy.


Stimulation and expansion increase co-dependency in T cell signaling networks

Intracellular signaling networks involve protein phosphorylation cascades that govern the transmission of information, from the reception of environmental stimuli at the cell surface to changes in the nucleus. To understand this information flow, it is necessary to determine the influence of one network component on another in a data-driven manner. Using mass cytometry, we can quantitatively measure the abundance of multiple proteins in tens of thousands of individual cells in a single experiment. This scale provides statistical power to identify dependencies between the measured phosphoproteins. We consider a network edge to be an interaction between two molecules, whose strength can be quantified from data using statistical dependency between the measured protein abundances. However, conventional correlation metrics are biased toward densely sampled cell phenotypes and are thus insensitive to the entire dynamic range. To this end, the density rescaled visualization (DREVI) plot renormalizes the density of cells across the entire dynamic range of expression for any interacting protein pair, and the DREMI score quantifies the degree of influence between these proteins. The DREMI score for an edge (XY, with direction assigned a priori) thus indicates the degree of dependence of Y on X (Fig. 1A). DREMI is far more reliable than conventional correlation since it gives equal significance to the extremes of marker detection, which often encompass responding phenotypes (Fig. 1B).

Fig. 1 Expansion augments subsequent activation of CD8+ T cells.

(A) Example DREMI analysis of correlation between molecules X and Y. A high DREMI score indicates that Y expression is highly dependent on X. (B) Mass cytometry analysis of pSLP-76 and pERK abundance in fresh CD8+ T cells at resting state (US) or after 60 or 360 s of CD3 stimulus. Conventional biaxial scatter plots (upper) and corresponding DREVI plots (lower) are from three independent experiments. (C) Mass cytometry analysis of pSLP-76 correlation with CD3 and pERK correlation with pSLP-76 in fresh or expanded CD8+ T cells before and after 360-s stimulus with CD3 or CD3+CD28. DREVI plots (right) are representative of and DREMI scores (left) are means ± SEM pooled from at least three biological replicates. (D) Mass cytometry analysis of pAKT in fresh or expanded CD8+ T cells stimulated as indicated. Histograms (left) are representative of and EMD scores (left) are means ± SEM pooled from at least three biological replicates. Individual Spearman’s correlation values are indicated (B) and ***P = 0.0002 and ****P < 0.0001 by one-way analysis of variance (ANOVA) test with Sidak’s correction (C and D).

When CAR-T cells are manufactured, they are typically subjected to ex vivo expansion using a stimulus such as a combination of antibodies against CD3 and CD28. To investigate whether stimulation and expansion alter signaling responses and network states, we used mass cytometry together with DREMI analysis. We measured canonical T cell signaling phosphoproteins in the phosphatidylinositol 3-kinase (PI3K), mitogen-activated protein kinase (MAPK)/extracellular signal–regulated kinase (ERK), and p38/MAPK pathways (fig. S1) (2022) of freshly isolated T cells, T cells expanded for 8 days using CD3+CD28 stimulus, and T cells that had been expanded and transduced with a CD19-28ζ CAR. Cells from each group were stimulated by cross-linking αCD3, αCD28, αCD3+αCD28, or αCAR antibodies for 60, 180, or 360 s at 37°C before analysis. In mass cytometry data pooled from experimental replicates and different antibody stimuli in CD8+ cells, we observed that, in both freshly isolated and expanded cells, the stimulus led to increased codependence in canonical TCR signaling (CD3→pSLP-76 and pSLP-76→pERK edges), indicating that stimulus generated greater information transfer (Fig. 1C). Moreover, expanded CD8+ cells had increased DREMI scores compared to freshly isolated cells, both in resting state and after antibody stimulation. Baseline DREMI scores for expanded CD8+ cells often exceeded the peak post-stimulation DREMI scores of fresh cells (Fig. 1C).

In addition to the insights from DREMI analysis, a population-level estimation of the changes in phosphoprotein abundance would give a more comprehensive overview of signaling events. Because of the many conditions, markers, and amount of background phosphoprotein between donors, it is unwieldy and impractical to visualize expression differences globally using conventional histograms. We therefore explored succinct techniques for describing differences between stimulated cells and matched unstimulated controls. Mean or median signal intensity, although commonly used in flow cytometry, neglects the single-cell resolution of the dataset by collapsing values to averages. Instead, distribution-based methods such as the Kolmogorov-Smirnov test can be used to assess the differences in marker expression in different conditions. However, Kolmogorov-Smirnov test has been heavily criticized in the past for its sensitivity toward outliers (23, 24). To this end, EMD (17) provides an alternative for robustly determining signal. EMD describes change in signal strength based on difference in probability distribution, with a higher EMD denoting a larger change; it has previously been used with mass cytometry data to describe changes between groups controlling for unequal size between groups (18). The use of EMD to describe changes in signaling expression allows multiple biological replicates to be characterized with a high degree of consistency, without collapsing data to mean or median values. Given that stimulation tends to increase the expression of a marker, a strongly positive EMD score can be reduced by an increase in baseline marker expression, a decrease in expression after stimulation, or both. Similar to the DREMI analysis, we observed difference in EMD between fresh and expanded CD8+ cells, with greater EMD for pAKT (Fig. 1D), pSLP-76, and pERK in expanded cells (fig. S2A). A similar pattern was observed for CD4+ T cells (fig. S2B). Together, our results show that expanded T cells respond to stimulus with a greater magnitude of phosphorylation and more information flow through T cell signaling pathways.

Effects of CAR transduction on T cell signaling

Synthetic signaling molecules such as CARs have not evolved within the natural cellular context, and their introduction could have unexpected effects on the carefully regulated intracellular environment. In pursuit of precise and effective cancer immunotherapy, they allow an entire polyclonal T cell population to be redirected against a single-target epitope, bypassing MHC restricted epitope-TCR interactions, which allow only a minority of T cells to recognize the tumor. Second-generation CARs that provide both a CD3ζ stimulus and CD28 costimulation are prone to tonic signaling, which can promote functional exhaustion and inhibit CAR-T cell function (1214). We therefore investigated the signaling phenotype of expanded T cells transduced with second-generation CARs and compared it to that of identically expanded controls that were not transduced.

T cells from healthy donors were expanded for 8 days in the presence of interleukin-2 (IL-2) using an initial after stimulation with antibodies against CD3 and CD28. A set of these cells were transduced on day 3 with a retrovirus encoding CD19-28ζ, a second-generation CAR that has exhibited great clinical efficacy (25). Whereas there is wide variation in basal phosphoprotein abundance between donors, samples transduced with CD19-28ζ consistently demonstrated increased basal phosphorylation of various proteins, including SLP-76, ERK, RelA(p65), and MAPKAPK2 (Fig. 2A). To quantify this basal “leakiness,” we determined EMD scores between matched transduced and untransduced samples for pSLP-76, pERK, and pRelA, which were correlated with transduction efficiency, consistent with tonic signaling from the CAR. The strongest correlation was seen in pERK, whereas weak correlation was seen with pAKT and pMAPKAPK2 (Fig. 2B). We next compared the EMD scores of CD19-28ζ–transduced and CD19-28ζ–untransduced cells for each marker after stimulation with antibodies against CD3, CD28, or CD3+CD28 (pooled data from all antibody stimuli). The amount of change in pSLP-76, pERK, pRelA, and pMAPKAPK2 was reduced (lower EMD scores between unstimulated and stimulated) in transduced compared to untransduced cells, as might be expected from the increased basal phosphorylation. This differential responsiveness was quantified as the difference in slope between the linear regression and the equivalence (y = x) lines in a plot of transduced EMD versus untransduced EMD, and it is most significant in the canonical TCR signaling pathway containing pSLP-76, pERK, and pRelA. There was no difference in pAKT responsiveness, which suggests that the CD3ζ, but not the CD28, endodomain predominantly contributes to tonic signaling (Fig. 2C).

Fig. 2 CAR expression promotes tonic activation of signaling networks.

(A) Mass cytometry analysis of baseline phosphoprotein abundance in expanded CD8+ T cells transduced to express the CD19-28ζ CAR as indicated. Histograms are representative of three independent donors. (B) Mass cytometry analysis of baseline phosphoprotein abundance in CD4+ (green), CD8+ (blue), and Vδ2+ (yellow) T cells transduced to express the CD19-28ζ CAR. Correlation of EMD score with transduction efficiency is from analysis of at least three independent donors. (C and D) Mass cytometry analysis of phosphoprotein abundance in expanded T cells transduced with CD19-28ζ stimulated with antibodies against CD3 (blue), CD28 (green), CD3+CD28 (red), or CAR (violet). Correlation of EMD scores in untransduced and CAR-transduced cells (C) and EMD scores for pSLP-76, pERK, and pMAPKAPK2 (D) are means ± SEM of at least three independent donors. (E) Flow cytometry analysis of TIM-3 and PD-1 abundance on T cells expressing a CD19-28ζ, GD2-28ζ (huk666), or GD2-28ζ (14G2A) CAR. Data are means ± SEM pooled from at least four biological replicates (see also fig. S11). Pearson correlation (B) and analysis of covariance (ANCOVA) (C) P values are displayed; (D) *P < 0.05, **P < 0.01 by paired t test. ns, not significant.

We were interested in the strength of response downstream from a 28ζ CAR compared to that provided by stimulus of the native CD3 and CD28 receptors. We hypothesized that tonic CAR signaling could blunt T cell responsiveness due to increased basal phosphoprotein abundance. In fresh cells with different memory phenotypes after T cell stimulation, there is decreased co-dependency (lower DREMI) between signaling edges in mouse effector memory compared to naïve T cells and blunted responses to stimulation (25, 26), which correlates with effector memory cells having increased basal phosphoprotein expression (19). Counter to expectation, the EMD scores for pERK and pMAPKAPK2 after CAR cross-linking were significantly increased when compared to those generated by CD3+CD28 cross-linking in donor-matched untransduced expanded cells (Fig. 2D). These data support the notion that CD3 and CD28 signaling from a CAR is stronger than from native receptors and suggests that tonic signaling is associated with the very effective signal transduction associated with CARs.

CAR tonic signaling is, in part, mediated by clustering, attributable to scFv framework interactions (12). One proposed sequela of tonic signaling is T cell exhaustion, a factor that has been linked to poorer clinical efficacy in CAR-T cell products and is associated with the expression of markers such as PD-1 and TIM-3. Consistent with the recognized effects of continuous stimulus, we found that T cells expressing any of three second-generation (28ζ) CARs were more likely than untransduced cells to display PD-1 and TIM-3 (Fig. 2E). However, despite containing different scFvs targeting either GD2 (27, 28) or CD19, there were no differences in abundance of these markers on the different CAR-T populations, which suggests that the exhaustion potential of 28ζ CARs may be independent of scFv identity. Given that the CARs in these experiments target antigens not expressed on T cells, it is highly unlikely that the signaling is due to unexpected CAR ligation.

Sustained nonphysiological stimulus hardwires TCR signaling networks

T cells are exposed to a variety of signaling events during CAR-T engineering. To investigate whether ex vivo manipulation and CAR expression alters the signaling network itself in the process of human T cell engineering (fig. S3), we examined the variance in DREMI score across multiple stimuli. Because T cells from a single donor are all treated under the same conditions, for each stimulus that is applied, DREMI scores can be generated for edges of interest in each population (CD4+ or CD8+ for instance). The addition of stimulus affects DREMI scores to varying degrees, depending on the baseline phosphoprotein abundance and the amplitude of any change. For a defined cell population from one donor, variance in DREMI for a particular edge across a panel of stimuli indicates how much the signaling relationship can be influenced by stimulation in general—in other words, the plasticity of the network. Edges with low DREMI variance are not as susceptible to stimulus-induced change as edges with high variance, irrespective of the overall mean DREMI score.

Using a panel of stimulus conditions and unstimulated controls, we observed two broadly different behaviors categorized by signaling edge and T cell subset. CD4+ cells tended to undergo network state changes in two steps: first upon expansion and again after transduction. In contrast, CD8+ cell networks mainly changed after expansion, with little further change following transduction. This factor was particularly noticeable in the early TCR signaling edge CD3→pSLP-76 (Fig. 3, A and B). When examining the behavior of all of the measured signaling edges in αβT cells, two distinct patterns of network behavior were observed. Edges in the canonical TCR signaling pathway increased mean DREMI but retained a low variance or even reduced variance after stimulation or transduction (Fig. 3, C and E, and fig. S4). Edges that signaled through pAKT showed a different behavior, increasing in mean and variance after expansion and transduction (Fig. 3, C and D, and fig. S4). The differences between CD4+ and CD8+ αβT cells remained consistent for other signaling edges but were most pronounced in the canonical TCR pathway (Fig. 3H, fig. S5, and tables S1 to S3). The increase in DREMI after CAR transduction, but in the absence of further stimulation, was also observed in cells transduced with three second-generation CARs (Fig. 3I and fig. S6, A and B). The abundance of the exhaustion marker Tim-3 was greater in expanded CD8+ cells than in expanded CD4+ cells (fig. S7A), but these differences were lost when cells were transduced with second-generation CARs (fig. S7B). Thus, we found that second-generation CAR expression was associated with increased basal signaling, greater amounts of PD-1 and TIM-3, and larger DREMI scores. The increased basal signaling affected both CD4+ and CD8+ T cell subsets equally after CAR expression. At the same time, we found reduced DREMI variation in canonical TCR pathway edges, which suggests that tonic stimulation limited network plasticity. These phenotypes are also associated with more “exhausted” T cells and correlated with poor in vivo efficacy. We hypothesize that CAR technologies that lack tonic signaling may restore network plasticity and relieve T cell exhaustion.

Fig. 3 Expansion and CAR expression have distinct effects on CD4+ and CD8+ T cells.

(A and B) Mass cytometry analysis of the CD3 and pSLP-76 abundance in stimulated fresh (green), expanded (blue), or CD19-28ζ–transduced (red) CD4+ and CD8+ T cells, as indicated. DREMI scores after CD3, CD28, or CD3+CD28 stimulation with means ± SEM (A) are pooled from three independent donors. DREVI plots with DREMI scores after CD3 stimulation alone (B) are representative of all donors. (C to H) Analysis of DREMI mean and variance in the indicated cells across all stimuli (C) identified two distinct types of responses (pink and blue areas). pAKT signaling edges (D) exhibited greater variance than known TCR signaling pathways (E). Differences between CD4+ (F) and CD8+ (G) T cells were most apparent in known TCR signaling pathways. DREMI scores for known PI3K and TCR signaling interactions in fresh, expanded, and transduced CD4+ and CD8+ T cells (H) with means ± SEM are pooled from all experiments. (I) Mass cytometry analysis of pERK and pRelA in expanded and CAR-transduced cells without further stimulus. Data with means ± SEM are pooled from three independent donors. *P = <0.05 and ****P < 0.0001 by one-way ANOVA with Sidak’s correction (A) or two-way ANOVA with Sidak’s multiple comparison correction (I).

Boolean logic approach to remove tonic TCR pathway activity

αβTCRs recognize specific peptide fragments presented on the MHC, but γδT cells that express Vγ9Vδ2 TCR (Vδ2+ cells) recognize tumor cells in an MHC-independent manner. The Vγ9Vδ2 TCR interacts with stress markers that are present on malignant or infected cells but are absent on their untransformed/uninfected counterparts (2933) and signals in a similar manner to conventional αβTCRs through CD3ζ molecules. A CCR, which provides only costimulation and relies on the γδTCR to provide CD3ζ signals, avoids reactivity against healthy cells, which do not engage the γδTCR (34). We therefore reasoned that this CCR design would provide beneficial costimulation without causing tonic CD3ζ signaling. Moreover, because T cells typically require both a TCR/CD3ζ signal and a costimulatory signal for full activation, removal of CD3ζ from the synthetic receptor should introduce two requirements for full activation—γδTCR and CCR ligation—forming a Boolean “AND” logic gate, which may elicit less on-target off-tumor toxicity (Fig. 4A).

Fig. 4 Engineering Vδ2+ T cells to express CCRs promotes TCR signaling and effector functions.

(A) A schematic highlighting that CARs provide all activation signals from one receptor, whereas the CCR requires a CD3ζ signal to be provided from another source. In this case, the engagement of tumor target butyrophilin (BTN3A1) by the Vδ2 γδTCR and the CCR is required for activation. (B) Phospho-flow analysis of pAKT and pERK abundance in CD33-CD28 CCR–transduced Vδ2+ T cells stimulated as indicated up to 600 s. Data are means ± SEM pooled from three independent donors. (C) Flow cytometry analysis of TNFα production by CD33-CD28 CCR–transduced Vδ2+ T cells stimulated as indicated. Data are means ± SEM pooled from three independent donors. (D) 51Cr-release assay of cytotoxicity by untransduced or GD2-CD28 CCR–transduced Vδ2+ T cells cocultured for 4 hours with GD2+ neuroblastoma (LAN1) target cells. Data are means ± SEM pooled from four independent donors. (E) Mass cytometry analysis of the indicated phosphoproteins in expanded Vδ2+ T cells after stimulation with antibodies against CD3 (red), NKG2D (green), or CD3+NKG2D (blue). EMD scores are means ± SEM pooled from four independent donors. (F) Mass cytometry analysis of the indicated phosphoproteins in expanded, CD33-CD28 CCR–transduced Vδ2+ T cells after stimulation with antibodies against CD3 (red), CCR (green), or CD3+CCR (blue). EMD scores are means ± SEM pooled from three independent donors. *P < 0.05, **P < 0.01, ****P < 0.0001 by one-way ANOVA with Sidak’s correction (C and D) or two-way ANOVA with Tukey’s correction (B, E, and F).

To determine whether a CAR lacking CD3ζ could signal and induce important T cell effector functions such as cytokine production and cytotoxicity, we created three CCRs (GD2-CD28, CD33-CD28, and ErbB2-CD28) and expressed them in Vδ2+ cells. Stimulation of both CD3 and CD33-CCR significantly enhanced phosphorylation of signaling proteins compared to CD3 stimulus alone, with ErbB2-CD28 and GD2-CD28 showing similar results (Fig. 4B and fig. S8A). Stimulation of CD33-CD28 CCR did not increase the abundance of either pAKT or pERK. Stimulating cells through both the CCR and CD3 increased production of the cytokine tumor necrosis factor α (TNFα) (Fig. 4C) and the anti–GD2-CD28 CCR promoted cytolytic activity when cultured with GD2+ LAN1 neuroblastoma (Fig. 4D). There was no evidence of tonic signaling from CD28-CCRs (fig. S8B).

NKG2D and DAP10-CCR signaling in Vδ2+ γδT cells

NKG2D is an innate immune receptor expressed by a range of effector lymphocytes, including Vδ2+ γδT cells, which responds to danger-associated molecular markers such as MICA, MICB, and ULBP1 to ULBP6 (35). Its role in γδT cell activation remains a matter of debate; some studies suggest that it acts as a costimulatory molecule (36, 37), and others indicate that NKG2D alone is sufficient to activate γδT cells (34, 35). NKG2D does not contain any signaling endodomains but instead forms a complex with the adaptor protein DAP10 (38). Engagement of DAP10 promotes PI3K activation (35, 39) and SLP-76 phosphorylation (37). We observed that ex vivo expanded Vδ2+ γδT cells retain responsiveness to both CD3 and NKG2D stimulation. As expected, CD3 cross-linking increased phosphorylation of pZAP70, pSLP-76, pAKT, pMAPKAPK2, pERK, and pS6, whereas NKG2D cross-linking only increased pAKT, consistent with PI3K pathway activity (Fig. 4E). NKG2D stimulation did not increase signal strength when combined with CD3 stimulus. Thus, activation of CD3 and DAP10 signaling likely promote activation of distinct molecular pathways.

GD2-directed CCRs that incorporate the DAP10 endodomain provide effective Boolean AND gating in Vδ2+ γδT cells (40). To determine whether this approach could be extended to targeting CD33, which is commonly expressed by acute myeloid leukemia, we engineered CD33-CCR and evaluated its downstream signaling function. Cross-linking of CD33-DAP10-CCR had minimal effect on the proximal mediators of TCR signaling, ZAP70 and SLP-76, but induced strong responses in pMAPKAPK2, pAKT, and pERK (Fig. 4F). DAP10-CCR induced pAKT more robustly than NKG2D stimulus, led to faster pMAPKAPK2 response than CD3 stimulus alone, and produced statistically significantly larger effects on pMAPKAPK2 and pERK when combined with CD3 than either stimulus alone. CD33-DAP10+ γδT cells increased TNFα production in response to CCR stimulation, although little response was seen to CD3 stimulus (fig. S9). These data indicate that CD33-DAP10-CCR promotes functional responses, and further suggest that DAP10-dependent receptors may contribute to the production of TNFα in γδT cells.

DAP10 CCR limits tonic signaling and synergizes with CD3 stimulation

Consistent with our earlier results in engineered αβT cells (Fig. 2H), we found that introducing CD33-28ζ into expanded Vδ2+ γδT cells promoted tonic signaling arising from CD3ζ (Fig. 5A). EMD scores for pSLP-76 and pERK were greater in unstimulated CD33-28ζ–transduced Vδ2 T cells than in CD33-DAP10–transduced Vδ2 T cells. To look for more subtle evidence of tonic activity, we used DREMI analysis of a CD33-DAP10 CAR and evaluated all pAKT and pMAPKAPK2 signaling edges, as these were strongly influenced by DAP10-CCR cross-linking. Using DREMI, we were also able to examine any simultaneous increases in pAKT and pMAPKAPK2. No edges apart from CD3→pAKT demonstrated changes in mean DREMI across a range of donor- and time-matched stimulus conditions (Fig. 5B and fig. S10). In contrast, stimulation of the CCR augmented pMAPKAPK2-pAKT DREMI score (Fig. 5C). Thus, these data indicate that the DAP10-CCR limits tonic signaling but, when activated, is capable of rapidly triggering a downstream response (Fig. 4F).

Fig. 5 DAP10-CCR does not induce tonic signaling and synergizes with CD3 stimulation.

(A) Mass cytometry analysis of basal phosphoprotein abundance in expanded untransduced, CD33-DAP10 CCR–transduced, or CD33-28ζ CAR–transduced Vδ2+ T cells. Histograms (A, upper) are representative of and EMD scores in expanded CD33-DAP10 (purple) or CD33-28ζ (red) cells (A, lower) are means + SEM pooled from three independent donors. (B to E) Mass cytometry analysis of the indicated phosphoprotein abundance in expanded untransduced or CD33-DAP10 CCR–transduced Vδ2+ T cells after stimulation. DREMI scores with means ± SEM (B) are pooled from three independent donors. DREVI plots of pAKT versus pMAPKAPK2 (C) and pERK versus pSLP-76 (D) are representative of all donors. Network representation of EMD and DREMI scores in DAP10-CCR–transduced Vδ2+ cells (E) stimulated with CCR alone (middle) or CD3+CCR (right). Stimulus inputs (green), outputs colored according to EMD score, and connections indicated by line color and thickness determined by the mean DREMI score are from the analysis of all donors. *P < 0.05 and **P < 0.01 by one-way ANOVA with Sidak’s correction for multiple comparisons (B).

To detect the influence of more than one signaling molecule on the activation of a downstream target, it is possible to analyze the area under a curve (AUC) fitted to the points of maximum density of a DREVI plot. This may be more useful than DREMI scores at uncovering additive or synergistic effects, because AUC accounts for differences in protein abundance represented in the maximum density curve (i.e., higher Y-axis intercept) that can result from independent signaling inputs. For example, although the DAP10 CCR does not influence pSLP-76 (Fig. 4F) and pSLP-76–pERK DREMI score remains low after DAP10 CCR stimulation, the pSLP-76–pERK AUC increases when compared to unstimulated as a result of pSLP-76–independent ERK phosphorylation (Fig. 5D). When both CD3 and the DAP10 CCR are engaged, the AUC reaches its maximum as both signals contribute to pERK activation. To gain a network-level understanding of the combinatorial effects of DAP10-CCR and CD3 stimuli, we built upon the “network DREMI” visualization (19). By using the post-stimulation EMD score to color nodes in the network and DREMI to define the color and thickness of connectors, we were able to intuitively visualize and summarize changes in codependency in addition to the resulting signals (Fig. 5E).

A range of different CCRs enhance effector function without on-target off-tumor toxicity

We were interested in whether other CCRs expressed in Vδ2+ γδT cells displayed the AND gate properties demonstrated by CD33-DAP10. In Vδ2+ γδT cells, expression of a second-generation CD19 CAR, but not a CD19-DAP10 CCR, was associated with significant increase in exhaustion markers compared with untransduced controls (Fig. 6A). The avoidance of exhaustion was not specific to the CD19 scFv and DAP10 endodomain; the same pattern was observed with a GD2 scFv and CD28 CCR with longer-term culture (Fig. 6B) and with a CD33-DAP10 CAR (Fig. 6C).

Fig. 6 CCR expression in Vδ2+ T cells avoids on-target off-tumor toxicity.

(A to C) Flow cytometry analysis of TIM-3 and PD-1 abundance on untransduced, CD19-28ζ CAR–transduced, and CD19-DAP10 CCR–transduced Vδ2+ cells (A) or GD2-CD28 CCR–transduced and untransduced Vδ2+ cells (B) or untransduced and CD33-DAP10–transduced Vδ2+ cells 8 days after transduction (see also fig. S15). Data are means + SEM pooled from at least three independent donors. (D) Flow cytometry analysis of cell proliferation by CD33-DAP10 CCR ± or CD33-CD28 CCR ± Vδ2+ T cells cultured with irradiated MV4-11 AML cells. Data are means ± SEM pooled from three independent donors. (E) Flow cytometry analysis of IFN-γ and TNFα production by DAP10-CCR–transduced Vδ2+ T cells cocultured overnight with either MV4-11 or allogeneic monocytes with and without butyrophilin blockade. Data are means + SEM pooled from three independent donors. (F and G) 51Cr-release assay of cytotoxicity by untransduced, CD33-DAP10 CCR–transduced (purple), CD33-28ζ CAR–transduced (yellow), or CD33-CD28 CAR lacking CD3ζ domain–transduced (green) Vδ2+ T cells cocultured with allogeneic monocytes or MV4-11 AML cells and treated with zoledronic acid where indicated. Data are means ± SEM of 3 to 10 independent donors (F) and 3 to 9 independent donors (G). (H) Myeloid colony formation assay by healthy bone marrow cultured overnight with or without untransduced, CD33-DAP10 CCR–transduced, or CD33-CD28ζ CAR–transduced Vδ2+ T cells. Data are means + SEM from three independent donors. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 by one-way ANOVA with Sidak’s correction.

The GD2-DAP10 CCR enhances antigen- and tumor-specific immune responses (40). Because of the particular challenge of on-target off-tumor toxicity in targeting myeloid malignancies, we investigated whether this approach was more broadly applicable by promoting key effector functions such as proliferation, cytokine production, and cytotoxicity in the context of CCRs targeting CD33. Using a multiplexed assay of protein tyrosine kinase activity, we confirmed that the acute myeloid leukemia (AML) cell line (MV4-11) activated TCR signaling pathway in Vδ2+ γδT cells (fig. S11). After comparing Vδ2+ cells cocultured with MV4-11 with target-free Vδ2, the most specifically up-regulated kinases were Syk, Abl, and ZAP70, which have known roles in TCR signaling (41). We found that irradiated CD33+ MV4-11 AML cells stimulated greater proliferation in both CD33-DAP10 and CD33-CD28 CCR Vδ2 T cells when compared to target-free controls (Fig. 6D). After overnight coculture with MV4-11 AML cells, more CD33-DAP10 CCR+ Vδ2 cells produced TNFα and interferon-γ (IFN-γ) than CD33-DAP10-CCR cells (Fig. 6E). The enhanced effector function against MV4-11 was not blocked by the presence of an antibody against butyrophilin, which suggests that cytokine production was independent of TCR engagement (29). Experiments using a panel of blocking antibodies suggested that the combined recognition of γδTCR, NKG2D, and DNAM-1 was involved (fig. S12B), consistent with studies on the interaction between γδT cells and AML (42).

CD33-DAP10 Vδ2 showed minimal reactivity against allogeneic monocytes or myeloid progenitors, whereas CD33-28ζ Vδ2 were highly cytotoxic to myeloid cells in both short- and long-term coculture (fig. S12A and Fig. 6, E to H). Pretreatment with zoledronic acid enhanced the susceptibility of monocytes but not MV4-11 to killing by unmodified, CD33-28ζ and CD33-DAP10 Vδ2 T cells (Fig. 6F). Given the multiple pathways involved in γδT cell recognition of AML, these data may imply that CD33-DAP10 promotes but does not increase maximum cytotoxicity in these short-term assays. In studies of costimulation in trans, CCRs are unable to enhance cytotoxicity, but this does not necessarily limit their utility (43, 44). Of importance for the cellular therapy field, even in the allogeneic setting, we found that there is minimal off-tumor toxicity. This approach therefore both addresses the problem of on-target off-tumor toxicity and brings the field closer to developing a cellular therapy with limited allo-reactivity. Together, our data identify how the process of CAR-T cell development promotes tonic signaling and describes an alternate engineering strategy to limit its harmful effects.


The remarkable success of second-generation CAR-T cells in the treatment of lymphoid malignancies (15) has been frustrated in other settings. Targeting of myeloid malignancies has been complicated by a lack of AML-specific antigens (45) such that patients receiving CAR-T cell therapy for AML typically require rescue by hematopoietic stem cell transplantation. In solid tumors, the CAR-T experience has been punctuated by failure of T cell expansion and substantial off-tumor toxicity (8, 10).

Functional assays dominate the CAR-T cell literature, and although many studies assess whether similar CARs will target new antigens, very few have dissected the mechanism of existing designs. Quantitative, reproducible, and detailed signaling analysis is needed to inform future CAR designs. This will help address mechanistic problems such as target-autonomous CAR signaling, which have been linked to T cell exhaustion and poor efficacy (12, 46). Our single-cell analysis of CAR-T signaling allowed a detailed description of the molecular CAR-T response to activation within a population. We also examined the effects of changing the cellular “chassis” within which CARs are expressed. In γδT cells, we found that CCRs reduced the tonic signaling associated with CD3ζ signaling domains and exhibited features of an AND gate that was only capable of response after both CAR and TCR engagement. This example of data-driven design to avoid off-tumor toxicity may prove a paradigm in CAR construction.

The process of CAR-T engineering requires a number of artificial cellular activation events. Expansion using antibodies against CD3 and CD28 augmented basal T cell stimulation, and introduction of a 28ζ CAR promoted continuous and autonomous tonic CD3ζ signaling. Unlike previous studies using biochemical or bulk phosphoproteomic readouts (6, 12, 16), single-cell resolution analysis enabled DREMI to be used as a measure of codependence between phosphoproteins, allowing network analysis of the cellular response under multiple conditions. In fresh cells, brief stimulus increased phosphoprotein expression, as well as in DREMI score. Expanded cells, having already received a stimulus, had increased baseline DREMI scores in the canonical TCR pathway than fresh cells. Further stimulation with antibodies against CD3, or CD3 and CD28, was less able to alter DREMI, which reduced DREMI variance, an indicator of the plasticity in established signaling relationships. 28ζ CAR-T cells had the highest DREMI scores, indicating a large amount of signal flow that resists augmentation using further stimulus, as indicated by the low DREMI variance. PI3K pathway edges did not show this reduction in plasticity, retaining the potential to be altered by further stimulation (Fig. 3D) and consistent with tonic signaling being restricted to the CD3ζ endodomain (Fig. 2C). These findings support a model whereby continuous signal flow “erodes” highly co-dependent but rigid connections between signaling nodes. Thus, signaling edges in the canonical TCR pathway increase mean DREMI and reduce DREMI variance as stimulus progresses from brief to sustained to tonic, and restimulation of expanded or transduced cells cannot further increase DREMI (that is, transmit information), despite eliciting protein phosphorylation.

We observed additional differences in network hardwiring in another context. Expanded CD8+ cells, which exhibit less plastic canonical TCR edges than expanded CD4+ cells, also had increased TIM-3 abundance. Whether a direct link exists between T cell functional exhaustion and the “hardening” of network connections (as observed by reduced DREMI variance) remain to be seen. In support of this possibility, 28ζ CAR–expressing cells have the highest DREMI means with lowest variance in canonical TCR pathway edges and are also much more likely to have a TIM-3+/PD-1+ phenotype, associated with late activation and exhaustion (4749), than expanded untransduced controls (Fig. 2E). Because of tonic signaling, 28ζ CAR–transduced cells have increased baseline phosphoprotein expression in the TCR signaling pathway (pSLP-26, pERK, and pRelA, Fig. 2C) with better-connected nodes (higher DREMI) at baseline. Chronically increased MAPK kinase (MEK)/ERK pathway activity and chronic IκB kinase (IKK) activation [with resultant nuclear factor κB (NFκB) p65 RelA phosphorylation] are both associated with T cell exhaustion (50, 51). Together, these observations suggest that a rigid signaling network, with reduced capacity to remodel in response to stimulus, may lead to reduced phenotypic responses and a functionally “exhausted” phenotype. The functional in vivo correlates of this network rigidity are impractical to determine due to the large number of cells required for CyTOF analysis, and short-term cytotoxicity assays are known to be poor indicators of long-term CAR-T cell health. However, there is substantial preexisting evidence that tonic signaling, now demonstrated to be associated with reduced network plasticity and increased exhaustion marker expression, leads to reduced longevity and proliferative capacity of CAR-T cells (12, 52). The emerging constellation of characteristics displayed by tonically signaling CAR-T cells—increased exhaustion marker expression, reduced in vivo longevity, and now loss of network plasticity—is tantalizing and hints at network plasticity being a marker of cellular health after expansion and transduction. A broader screen of other CAR constructs and endodomain combinations would provide more information to validate this association.

We examined alternative CAR designs, in part, because excess and redundant CD3ζ signals from 28ζ CARs are an established phenomenon (12, 52); removal of one or two of the three CD3ζ ITAMs (immunoreceptor tyrosine-based activation motifs) leads to enhanced in vivo persistence of CAR-T cells without reducing efficacy (52). Our alternative approach of removing the synthetic CD3ζ signaling motif was made possible by using γδT as a cellular chassis. Often overlooked, tumor-infiltrating γδT cells are the strongest predictor of positive outcome in solid tumors (53). These cells readily accept both conventional CARs and CCRs (40, 5457) without loss of innate function or tumor tropism. The MHC-unrestricted promiscuity of the γδTCR gives them a key advantage over αβT cells, insomuch as the native γδTCR can be used to provide a CD3ζ signal. By not relying on a CAR to deliver a CD3ζ signal, we limited tonic signaling and introduced an important check on detection of whether a target is “healthy” or “tumor.” Our DAP10-CCR design was informed by the important role of NKG2D in γδT cell activation (37) and its susceptibility to tumor immune escape mechanisms (58, 59). Perhaps due to differences in intracellular packing or amounts of NKG2D/DAP10 complexes and DAP10-CCRs, the stimulation of the synthetic receptor generates more potent signals and activates p38 MAPK and MEK pathways that were not activated by NKG2D stimulus. Activation is only observed when the CCR is stimulated (Fig. 5), preserving the plasticity of signaling and having no effect on γδT cell exhaustion profiles (Fig. 6, A to C).

Cancer immunotherapeutic strategies that redirect the immune system against a tumor-associated antigen must strike a balance between efficacy and toxicity. In CAR-T trials targeting lymphoid malignancies, depletion of healthy CD19+ cells is a recognized but tolerable side effect. Ablation of healthy myeloid cells by T cells expressing CD33-28ζ CARs (Fig. 6, F to H) could be profoundly toxic if the CAR-T cells persisted in a patient. Inter-species differences between mouse and human γδT cell biology and immunological niche preclude modeling this in murine model without introducing problematic confounders such as weekly zoledronic acid treatments (60, 61), which sensitize healthy myeloid cells to γδT cell cytotoxicity. In vitro, we have demonstrated that CD33-DAP10 enhances the function of Vδ2+ γδT cells, conferring significant proliferative advantage over untransduced cells cultured with AML cells (Fig. 6D). CD33-DAP10 Vδ2 cells are also capable of killing AML cells without demonstrating toxicity against allogeneic monocytes or myeloid progenitors (Fig. 6, F to H).

In summary, we have shown that mass cytometry of native and synthetic T cell signaling in combination with DREMI and EMD analysis provides a rapid means of assessing the effects of CAR production on T cell biology. In addition, we described how CCR-expressing γδT cells may provide a means of improving the safety of cellular immunotherapy by avoiding on-target off-tumor toxicity. Furthermore, they may allow targeting of antigens, which to date have been off-limits due to toxicity concerns. More work is needed to determine whether these cells are able to be used as a viable cellular therapy chassis, but studies using unmodified γδT cell infusions have been promising (62, 63), and we look forward to seeing early-phase clinical studies of this technology.


Isolation and prestimulation handling of fresh PBMCs

Whole blood (20 ml) was diluted with 10 ml of phosphate-buffered saline (PBS) + 500 μl of 100 mM EDTA and layered on 20 ml of Percoll. Interface peripheral blood mononuclear cells (PBMCs) (20 min, 300g, room temperature) were washed in PBS and resuspended in 25 ml of T cell medium (X-VIVO 15, Lonza BioWhittaker, Maryland, USA) supplemented with penicillin (100 IU/ml)/streptomycin (100 μg/ml) (Sigma-Aldrich, Missouri, USA) and cultured overnight before use.

T cell expansion

PBMCs were isolated as described above. Cells were cultured in RPMI 1640 medium supplemented with l-glutamine (2 mM, Sigma-Aldrich), penicillin (100 IU/ml)/streptomycin (100 μg/ml) (Sigma-Aldrich), and 10% fetal calf serum (FCS) (v/v) (Gibco, Massachusetts, USA). T cell expansion was induced by addition of anti-CD3 (OKT3) and anti-CD28 (clone CD28.2). IL-2 (100 IU/ml; Aldesleukin, Novartis, Frimley, UK) was added to PBMC suspension after PBMC isolation (day 1) and was replenished every 2 to 3 days by removing half of the media from the well and replacing with fresh media containing IL-2 (200 IU/ml).

Vδ2+ T cell expansion

For specific Vδ2+ γδT cell expansion, PBMCs were isolated as described above. They were cultured in RPMI 1640 medium supplemented with l-glutamine (2 mM, Sigma-Aldrich), penicillin (100 IU/ml)/streptomycin (100 μg/ml) (Sigma-Aldrich), and 10% FCS (v/v) (Gibco, Massachusetts, USA). Vδ2+ γδT cell expansion was stimulated using 5 μM zoledronic acid (Actavis, New Jersey, USA) and IL-2 (100 IU/ml; Aldesleukin, Novartis, Frimley, UK), which was added to PBMC suspension after PBMC isolation (day 1). IL-2 was replenished every 2 to 3 days by removing half of the media from the well and replacing with fresh media containing IL-2 (200 IU/ml).

Construction of retroviral constructs

The gammaretroviral vector used in all constructs was SFG (64), pseudo-typed with an RD114 envelope. DNA fragments were amplified using the Phusion HT II polymerase according to the manufacturer’s instructions (Thermo Fisher Scientific, Massachusetts, USA). Polymerase chain reaction (PCR) was carried out in a PTC-200 DNA Engine (MJ Research, Massachusetts, USA). PCR products were extracted from 1% agarose gels using the Wizard SV Gel and PCR Clean-Up Kit (Promega, Wisconsin, USA). Sample concentrations were determined using a NanoDrop ND-1000 spectrophotometer (Thermo Fisher Scientific, Massachusetts, USA). The CAR ectodomain comprised one of a range of scFvs against a panel of targets, including human GD2 (clone HUK666 or 14G2A), human CD33 (clone 113), human ErbB2 (clone 4D5), or human CD19 (clone FMC63) and a spacer derived from the human immunoglobulin G4 (IgG4) CH2-CH3 portion, which has been described before in third-generation format (27).

The endodomains were generated using oligo-assembly PCR based on a codon-optimized sequence of the human CD3ζ, CD28, and DAP10 endodomains. In addition to the CAR construct, RQR8, which is a marker bearing a CD34 epitope (65), was included, separated from the CAR by a cleavable 2A peptide. This allows CAR-expressing cells to be detected by flow cytometry by staining using the anti-CD34 antibody clone QBend10. The CAR and CCR constructs used in this study are shown in fig. S13.

Production of viral particles by transfection

293T cells (1.5 × 106) per 100-mm2 dish (Nucleon Delta Surface, Thermo Fisher Scientific) were plated at day 1 in 293T medium [Dulbecco’s modified Eagle’s medium (DMEM), 10% FCS (v/v)]. γ-Retroviral particles were produced by cotransfection of 293T cells at day 2 using GeneJuice Transfection Reagent (Novagen/Millipore, Massachusetts, USA) in accordance with the manufacturer’s directions. Viral particle–containing supernatants were harvested at day 4; medium was replenished and harvested at day 5. Supernatants were pooled, filtered (0.45-μm filter, Millipore), and directly used for transductions or stored overnight at 4°C before use.

Transduction of T cells

Transduction of T cells was carried out in RetroNectin (Takara Bio, Tokyo, Japan)–coated 24-well plates, which were preloaded with viral supernatant. T cells (0.5 × 106) suspended in 0.5 ml of T cell medium + IL-2 (400 IU) were combined with 1.5 ml of viral supernatant and centrifuged for 40 min, 1000g at room temperature. Typically, 12 × 106 T cells per donor were plated for transduction.

Timing of transduction differed between γδT cells and αβT cells due to expansion dynamics. γδT cell expansion was stimulated with 5 μM zoledronic acid (Actavis, New Jersey, USA) and IL-2 (100 IU/ml) (Aldesleukin), and transduction was performed at day 5. At day 8 of culture (day 3 after transduction), cells were pooled, washed, and plated at 2 × 106 cells/ml in T cell medium + IL-2 (100 IU/ml) (24-well plates, Nucleon Delta Surface, Thermo Fisher Scientific, Massachusetts, USA). Transduction efficiency was determined by flow cytometry at day 10 (day 5 after transduction).

In the case of αβ T cells, transduction was performed 72 hours after stimulation with anti-CD3/CD28 antibodies in the presence of IL-2 (100 IU/ml) (Aldesleukin), with replating 3 days after transduction. Signaling analysis and cytotoxicity assays were performed 6 days after transduction in both cases.

Stimulation of T cells using receptor-antibody cross-linking

Before stimulation, T cells were cultured overnight in serum-free media (XVIVO-15 and X-VIVO 15, Lonza BioWhittaker, Maryland, USA) to reduce background phosphorylation abundance. Single samples containing 0.5 × 106 to 1 × 106 cells were used for each stimulation condition.

Primary antibodies against CD28, CD3, and NKG2D of Fc were loaded on ice for 10 min, and the cells were then washed in ice-cold PBS at 3°C. Secondary antibodies specific to the species of the primary antibody to be cross-linked (donkey anti-mouse or donkey anti-goat) were resuspended in XVIVO-15 and prewarmed to 37°C. Primary antibody–labeled T cells were added to the prewarmed tube containing secondary antibody and stimulated for 60, 180, or 360 s. Stimulation was halted by the addition of paraformaldehyde to a final concentration of 4%.

Control samples were made for baseline (t = 0 s) and for every time point (60, 180, and 360 s). Control samples were treated identically to stimulated samples, but the primary and secondary antibodies were excluded. A diagram of the stimulation protocol is shown in fig. S14.

Mass cytometry

All samples from a given donor and stimulation run were barcoded using the Cell-ID 20-Plex Pd Barcoding Kit (Fluidigm, San Francisco, California, USA) in accordance with the manufacturer’s protocol. Up to 20 samples were therefore stained simultaneously in the same tube in a total volume of 300 μl. A two-step staining procedure was used whereby surface markers including CD3, CD45RA, CD45RO, CD27, CCR7, TIM-3, CD4, CD8, NKG2D, CD16, and TCRVδ2 were stained, followed by intracellular markers including pERK1/2, pZAP70, pMAPKAPK2, pRelA(p65), pS6, pSLP-76, and pAKT. Cell fixation, permeabilization, and staining were performed as previously described (66). To ensure maximum comparability between samples, all data were acquired using internal metal isotope bead standards (EQ Beads, Fluidigm). A list of antibodies used and their conjugates is shown in table S4.

Mass cytometric analysis was performed using a Helios mass cytometer (Fluidigm). Approximately 100,000 events were acquired per sample—totaling 2 × 106 events for a full barcoded set.

CyTOF data postprocessing

Individual time series were normalized to the internal bead standards using the method described in (67), and bead events were removed from the resulting FCS files. In addition, as described in (66), abundance values reported by the mass cytometer were transformed using a scaled arcsinh, with a scaling factor of 5. Before arcsinh transformation with a cofactor of 5, values of zero in the FCS files had Gaussian noise with mean 0 and an SD of 0.5 added to them to enhance visualization using conventional flow cytometry packages and to later enable easier computation of density estimates. Given the minimal SD of the noise compared to actual “positive” signal intensities and the in-built noise and outlier handling capacity of DREMI [figure S6 of (19)], the addition of noise is not expected to significantly alter the DREMI scores. Furthermore, as the calculation of DREMI scores is done by binning across the dynamic range of values, addition of a small amount of noise to the bottom bins will not propagate across the rest of the distribution.

Debarcoded FCS files were postprocessed using a combination of FlowJo Vx and bespoke Python scripts. Gating was performed in FlowJo VX, and gated populations were exported before computational analysis.

Gating of cellular subsets

Cells were filtered through a series of gates to ensure compatibility between compared samples. Singlets were identified using DNA content, and live cells were detected by exclusion of Cell-ID Cisplatin (Fluidigm), which only stains dead cells. T cell subsets were identified by expression of CD3 ± CD4, CD8, or TCRVδ2. CAR expression was detected using anti-human Fc, which stains the stalk portion of the CAR. An example of the gating strategy used is shown in fig. S15A.

DREMI and DREVI analysis

DREMI and DREVI are information theory–based methods developed to quantify and visualize relationship between two molecular epitopes (19). Given two proteins epitopes X and Y, and assuming that we are interested in assessing the influence of X on Y, then DREVI visualizes the conditional dependence of Y on X. Specifically, DREVI computes the conditional probability density of Y given X, p(YX)=p(X,Y)p(X), where the joint distribution is computed using a heat diffusion–based kernel density estimation procedure (68). Once the conditional density is computed, the result is visualized as a heatmap.

DREMI quantifies the strength of relationship between two protein epitopes, using mutual information–based metric. Although a traditional mutual information metric relies on the joint distribution, it is more likely to be biased by dense regions, thereby missing out on interesting biology shown by extreme amounts of proteins. To circumvent this, DREMI computes the mutual information on the conditional density function (as computed for the DREVI visualization) as opposed to the joint density.

Thus, DREMI=Ic(X,Y)=ijp(yjxi)log(p(xi,yj)p(xi)p(yj))

Effectively, DREMI reweighs the regular mutual information so that all observed range of expression contributes uniformally. In this article, we used the simpledremi implementation of DREVI and DREMI for analysis (69).

Calculation of EMD

EMD was computed between stimulated samples and time- and donor-matched unstimulated controls. EMD was calculated between T cell populations that had undergone the same postprocessing in terms of gating of specific T cell populations. The python module wasserstein_distance, which is a component of scipy.stats, was used to calculate EMD between samples.

Flow cytometry

Flow cytometric analysis was performed using an LSRII flow cytometer (BD, New Jersey, USA). Data were collected using BD FACSDiva V8.0.1 and analyzed using FlowJo VX. Compensation was calculated on the basis of OneComp eBeads (Thermo Fisher Scientific, Massachusetts, USA) stained with single-color antibodies. When flow cytometry was used as an alternative to mass cytometry for rapid stimulation experiments, cells stimulated and fixed as described above were stained in a two-step protocol, whereby surface markers (CD3, TCRVδ2, and human Fc) were stained first, after which the cells were permeabilized and intracellular markers (pZAP70, pERK, pRelA, and pAKT) were subsequently stained. A list of antibodies used is shown in table S5.

51Cr-release killing assay

To assess the killing capacity of T cells, 5000 51Cr-labeled target cells were cocultured with 50,000, 25,000, 12,500, or 6250 effector cells in 200 μl of T cell medium + IL-2 (100 IU/ml). Cells were incubated in 96-well V-bottomed plates (Greiner Bio-one, Kremsmünster, Austria) for 4 hours. Fifty microliters of supernatant was transferred into Isoplate-96 HB plates (PerkinElmer, Massachusetts, USA). After addition of 150 μl of scintillation cocktail (OptiPhase SuperMix, PerkinElmer), samples were incubated overnight at room temperature. Counts were acquired using a 1450 MicroBeta TriLux scintillation counter (PerkinElmer). Percentage killing was calculated using the following formula:%killing=(Condition Cr51 releasebackgroundMaximal Cr51 releasebackground)*100

Activation of T cells using antibody-coated beads

Anti-Biotin MACS iBeads (Miltenyi Biotec) were labeled with 10 μg of anti-CD3–biotin (OKT3, BioLegend), 10 μg of anti-IgG (Fc)–biotin (Novex/Life Technologies), or 10 μg of anti-CD3–biotin and 10 μg of anti-IgG (Fc)–biotin. After incubation (10 min, 4°C), beads were washed twice with PBS and resuspended in T cell medium. Bead suspensions (100 μl) were plated in 96-well U-bottom plates (Thermo Fisher Scientific, Massachusetts, USA). Effector T cell preparations were added (0.2 × 106 cells per well in 100 μl of T cell medium) and incubated with beads for 24 hours. No exogenous IL-2 was added to the medium in either case. After 23 hours, cells were harvested for analysis of cytokine secretion by intracellular cytokine staining. The gating strategy was identical to that in fig. S15, with the exception that “% cytokine + ve” was used as the readout.

Colony formation assay

Cells were isolated from healthy donor bone marrow using Ficoll density gradient separation and cocultured overnight with Vδ2+ γδT cells expressing either CD33-28ζ, CD33-DAP10, or untransduced controls at a 1:1 effector:target ratio. After overnight coculture, remaining cells were transferred to semi-solid medium (MethoCult H4534 Classic Without EPO, StemmCell Technologies, Vancouver, Canada) and incubated at standard tissue culture conditions for 7 days. After incubation, all myeloid colonies were counted manually by two independent investigators blinded to the culture conditions.

Secondary expansion in response to target cells

Vδ2+ γδT cells transduced to express CD33-DAP10 or CD33-CD28 CCR were cocultured for 7 days with irradiated MV-411 at a 1:1 effector:target ratio in RPMI 1640 medium supplemented with l-glutamine (2 mM), penicillin (100 IU/ml)/streptomycin (100 μg/ml), 10% FCS (v/v), and IL-2 (100 IU/ml). Absolute numbers of CD3+/Vδ2+/CCR+ and CD3+/Vδ2+/CCR cells in the coculture were determined by flow cytometry using Precision Count Beads (BioLegend), and these values used to derive fold change of each population during the culture period.

Multiplex protein tyrosine kinase assay

Expanded Vδ2+ γδT cells were cocultured with live MV4-11, which had been precoated in magnetic anti-CD33 microbeads. After 30 min of coculture, MV4-11 was removed using magnetic column separation (Miltenyi, Bergisch Gladbach, Germany). The flow-through, containing effector cells only, was lysed and protein tyrosine kinase activity in the lysate was assessed using the PamChip PTK assay using a PamStation 12 (PamGene, Wolvenhoek, The Netherlands). Data were processed using BioNavigator6 (PamGene, Wolvenhoek, The Netherlands). A (normalized) kinase statistic Sk for the change in phosphorylation between samples in group x (target-free Vδ2) and group y (Vδ2 cocultured with MV-411) can be calculated as the mean signal-to-noise ratio of the individual peptides in the group: the mean signal in x minus the mean signal in y divided by the SD. If there is, on average, a larger change of the peptides in the same direction (i.e., all upwards or all downwards), a larger absolute value of Sk would result.


Fig. S1. Map of signaling networks analyzed in distinct cell populations.

Fig. S2. Response of αβ T cells to stimulus.

Fig. S3. Network behavior of activated αβ T or CD19-28ζ CAR-T cells.

Fig. S4. Representative DREVI plots for expanded CD8+ T cells.

Fig. S5. DREMI scores for expanded transduced and nontransduced T cells.

Fig. S6. Representative DREVI plots for expanded transduced and nontransduced T cells.

Fig. S7. TIM-3 abundance in expanded transduced and nontransduced αβ T cells.

Fig. S8. Phosphoprotein abundance in CD28-CCRs in Vδ2+ γδT cells.

Fig. S9. TNFα production in response to CCR stimulus.

Fig. S10. DREMI scores in expanded or CD33-DAP10 CCR+ γδT cells.

Fig. S11. SYK, ABL, and ZAP70 activity are increased in Vδ2+ γδT cells cultured with AML cells.

Fig. S12. CD33-DAP10 CCR+ Vδ2+ γδT cells produce cytokines in response to AML cells.

Fig. S13. Schematics of CAR and CCR constructs used in this study.

Fig. S14. Cartoon illustrating stimulation of CAR-transduced αβ T cells by antibody cross-linking.

Fig. S15. Flow and mass cytometry gating strategies.

Table S1. Statistical analysis of data displayed in Fig. 3H (CD4+).

Table S2. Statistical analysis of data displayed in Fig. 3H (CD8+).

Table S3. Statistical analysis of data displayed in fig. S5.

Table S4. Antibodies used in mass cytometry.

Table S5. Antibodies used in flow cytometry.


Acknowledgments: We thank staff at the UCL Cancer Institute Cytometry Core Facility for their technical assistance. We thank M. Pule (UCL Cancer Institute) for several of the constructs used in this study. We also thank T. Nawy (MSKCC) for his assistance in the production of the manuscript. Funding: Supported by a Wellcome Trust Fellowship (110022/Z/15/Z to J.F.), Olivia Hodson Cancer Fund via GOSHCC (to D.W.D.), GOSHCC infrastructure award, NIHR GOSH Biomedical Research Centre (to J.A.), GOSHCC leadership grant (to J.A.), Research in Childhood Cancer (RICC; to J.A.), Great Ormond Street Hospital Charity (leadership award and grants W1134, VS0118, W1029, and W1076 to J.A.), Action Medical Research GN2400 (to M.B.), Dr. Mildred Scheel Foundation for Cancer Research, and the German Cancer Aid (to P.A.). D.P. received funding from the following grants: NIH (DP1-HD084071 and R01CA164729), Cancer Center Support (P30 CA008748), and Gerry Center for Metastasis and Tumor Ecosystems. Author contributions: J.F. designed the experiments and wrote the manuscript. J.F., D.W.D., M.B., M.O.H., P.A., L.P., and R.B. performed experiments generating data for this paper. W.D. provided technical assistance in the collection and pre-processing of CyTOF data. R.S. assisted in the design of analysis strategies, computational methods, and data analysis. S.I. provided technical assistance regarding colony formation assays. J.A. supervised synthetic biology and immunological aspects of the work. D.P. supervised computational analysis and systems biology. Competing interests: J.A. holds company stock in Autolus Ltd. and is a paid consultant for TC Biopharm. J.F. has also undertaken paid consultancy work for TC Biopharm. J.A. and J.F. are both inventors on a patent pertaining to CCRs in γδT cells, which has been licensed to TC Biopharm (WO/2016/174461). M.B. performed the work contributing to this paper when employed at University College London but is now employed by TC Biopharm. All other authors declare that they have no competing interests. Data and materials availability: Data will be made available through the OpenScience Framework (, 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