|Home | About | Journals | Submit | Contact Us | Français|
It is likely that the enhancement of signaling after antigenic stimulation, particularly in the tumor microenvironment, would improve the function of adoptively transferred T cells. Linker for activation of T cells (LAT) plays a central role in T cell activation. We hypothesized that the ubiquitylation-resistant form of LAT in cells would enhance T cell signaling and thus augment antitumor activity. To test this, human CD4+ or CD8+ T cells were electroporated with small interfering RNA (siRNA) to repress endogenous LAT and ubiquitylation-resistant LAT 2KR or wild-type LAT mRNA was introduced for reexpression. Significantly enhanced phosphorylation of LAT and phospholipase C-γ (PLCγ) was observed, and augmented calcium signaling after T cell receptor (TCR) triggering was observed in LAT 2KR-expressing T cells. TCR-induced calcium signaling was abrogated in LAT knockdown cells, but the baseline was higher than that of control siRNA-electroporated cells, suggesting a fundamental requirement of LAT to maintain calcium homeostasis. Redirected LAT 2KR T cells expressing a chimeric antigen receptor or an MHC class I-restricted TCR showed augmented function as assessed by enhanced cytokine secretion and cytotoxicity. These results indicate that interruption of LAT ubiquitylation is a promising strategy to augment effector T cell function for adoptive cell therapy.
The development of redirected T cells encoding transgenic T cell receptors (TCRs) or chimeric antigen receptors (CARs) is a promising approach to overcome tolerance in patients with cancer and chronic infections (Greenberg and Riddell, 1999; Schumacher, 2002; Sadelain et al., 2003; Johnson et al., 2009; Bonini et al., 2011; Kohn et al., 2011). However, the clinical application of adoptive transfer has with some exceptions led to suboptimal efficacy for several reasons. To eliminate tumor, infused T cells must proliferate extensively in vivo, traffic to the tumor, and retain function in the immunosuppressive tumor microenvironment. Consequently, methods to enhance in vivo proliferation and function of transferred T cells may improve antitumor efficacy. In particular, decreasing the threshold for antigen signaling, increasing cytolytic activity, increasing resistance to tumor-derived immunosuppressive mechanisms, and enforcing differentiation to helper T cell type 1 (Th1) or (Th17) lineages have promise (Paulos et al., 2010). Approaches suggested to improve antigen signaling have included the incorporation of costimulatory molecules (Krause et al., 1998; Kalos et al., 2011), ablation of Cbl-b (Chiang et al., 2007), and induction of diacylglycerol kinase-ζ deficiency (Riese et al., 2011).
Linker for activation of T cells (LAT) is a type III transmembrane molecule that is critical for the coordination of signals triggered on T cell activation (Zhang et al., 1998, 2000; Simeoni et al., 2008). LAT contains a short extracellular domain, a single transmembrane-spanning region, and a long intracellular region containing nine tyrosines, of which five can be phosphorylated (Paz et al., 2001; Zhu et al., 2003). TCR signaling is initiated by phosphorylated Zap-70, which in turn phosphorylates LAT at conserved tyrosine residues (Horejsi et al., 2004; Simeoni et al., 2008). Phosphorylated LAT recruits a number of signaling molecules resulting in calcium influx, nuclear factor of activated T cells (NFAT) activation, and ultimately expression of cytokine genes (Malissen et al., 2005). Studies in mice with knockout or modified knock-in LAT molecules have revealed that LAT has a function in both the differentiation of T cells along the Th1 and Th2 pathways, as well as in the control of lymphocyte proliferation (Mingueneau et al., 2009; Roncagalli et al., 2010; Shen et al., 2010). The function of LAT in postthymic mature T cells is relatively unknown, and the function in human T cells is not well studied, as most studies to date have been done in mouse T cells. In one study we reported that expression of a ubiquitylation-resistant form of LAT in Jurkat cells was more stable than that of wild-type LAT and induced the enhancement of downstream signaling molecules and calcium flux (Balagopalan et al., 2011). On the basis of these findings, we hypothesized that expression of ubiquitylation-resistant LAT in primary postthymic mature human T cells might enhance T cell signal transduction and function. We show that knockdown of wild-type LAT and expression of ubiquitylation-resistant LAT has the potential to generate engineered T cells with increased therapeutic efficacy.
Primary human CD4+ T and CD8+ T cells were isolated from healthy volunteer donors after leukapheresis by negative selection, using RosetteSep kits (STEMCELL Technologies, Vancouver, BC, Canada). All specimens were collected under a university institutional review board-approved protocol and written informed consent was obtained from each donor. CD4+ and CD8+ T cells were expanded by CD3/CD28 bead stimulation (immunomagnetic beads to which CD3-specific and CD28-specific monoclonal antibodies had been conjugated; provided by the Clinical Cell and Vaccine Production Facility of the University of Pennsylvania, Philadelphia, PA) at a 1:1 cell-to-bead ratio in culture medium (RPMI 1640 [Invitrogen, Carlsbad, CA], 10% fetal calf serum [FCS; Gemini Bio-Products, West Sacramento, CA], penicillin [100U/ml; Invitrogen], and streptomycin [100μg/ml; Invitrogen]), with CD8+ T cells supplemented with human interleukin (IL)-2 (Chiron, Emeryville, CA) every other day to 300IU/ml. After 10 days, these in vitro-expanded T cells were collected and cryopreserved. Cells were thawed and incubated for 24hr before electroporation with in vitro-transcribed (IVT) RNAs, using a BTX CM830 system (Harvard Apparatus BTX, Holliston, MA) as previously described (Zhao et al., 2010) and immediately placed into prewarmed culture medium.
The small interfering RNA (siRNA) targeting human LAT at CCAACAGUGUGGCGAGCUA (nucleotides 311 to 329) and the nontargeting control siRNA were purchased from Integrated DNA Technologies (Coralville, IA). Six constructs were cloned into the RNA transcription vector pGEM: (1) yellow fluorescent protein (YFP)–LAT wild-type (LAT WT) fusion protein, (2) YFP–ubiquitylation-resistant LAT fusion protein, in which K52 and K204 were replaced with arginines (LAT 2KR) (Balagopalan et al., 2011), (3) YFP–phosphorylation-resistant LAT fusion, in which Y132, Y171, Y191, and Y226 were replaced with phenylalanines (LAT 4YF) (Supplementary Fig. S1; supplementary data are available online at www.liebertonline.com/hum), (4 and 5) the α and β chains, respectively, of 1G4 HLA-A2-restricted NY-ESO-1 TCR (Zhao et al., 2005), and (6) chimeric antigen receptor (CAR) of anti-CD19-ζ (Barrett et al., 2011). In addition, the codons of LAT WT, LAT 2KR, and LAT 4YF were changed within the siRNA target (nucleotides 311 to 329), so that the amino acid sequence remained conserved but was resistant to siRNA degradation. IVT RNA was transcribed from SpeI-linearized plasmids, using a T7 mScript standard mRNA production system (CELLSCRIPT, Madison, WI) according to the manufacturer's protocol.
For knockdown of endogenous LAT, 1nmol (12.7μg) of LAT-targeting siRNA or control siRNA was used with or without various amounts of YFP–LAT fusions and coelectroporated into CD4+ T or CD8+ T cells (5×106/100μl). To analyze the cytotoxicity of these cells, 5μg of IVT RNA of each chain of the NY-ESO-1 TCR or 10μg of IVT RNA of anti-CD19-ζ CAR was added into the mixtures.
Pellets of cells were lysed for 30min in radioimmunoprecipitation (RIPA) buffer (Cell Signaling Technology, Danvers, MA) with a protease–phosphatase inhibitor cocktail (Cell Signaling Technology). Lysates were cleared by centrifugation at 16,000×g for 10min. Samples were mixed with NuPAGE lithium dodecyl sulfate (LDS) sample buffer with NuPaGE Sample Reducing Agent (Invitrogen), boiled for 5min, and loaded on a 4–12% Bis-Tris polyacrylamide gel (Invitrogen). Separated proteins were transferred to polyvinylidene difluoride (PVDF; Millipore, Billerica, MA). Membranes were blocked with 5% nonfat milk or 5% bovine serum albumin (BSA) in Tris-buffered saline with Tween 20 (TBST) and each antibody was probed in 5% BSA–TBST. Antibodies for LAT, phospho-LAT (Y191), phosphorylated phospholipase C-γ (PLCγ) (Y783), and β-actin were purchased from Cell Signaling Technology. Identified proteins were detected with stabilized horseradish peroxidase (HRP)-conjugated goat anti-rabbit antibody (Cell Signaling Technology) in blocking buffer. Membranes were visualized with SuperSignal West Femto chemiluminescent substrate (Thermo Fisher Scientific, Barrington, IL). The bands were scanned and quantified with ImageJ (National Institutes of Health, Bethesda, MD).
To measure the cytoplasmic calcium concentration, a slightly modified version of a previously published method (June and Moore, 2004) was used. We electroporated 1nmol of control siRNA, LAT siRNA alone, or LAT siRNA with 5 or 20μg of IVT RNA of LAT–YFP into in vitro-expanded CD8+T cells. After 24hr of incubation, 2×106 cells were resuspended to 1ml with culture medium containing Indo-1-AM (2μg/ml; eBioscience, San Diego, CA) and incubated at 37°C for 30min. The cells were washed and stimulated with soluble anti-CD3 (OKT3 clone, 24μg/ml; eBioscience). The ratio of the two emission wavelengths (405/20 and 530/30 with a 450-nm long-pass filter) was recorded over time with an LSR II flow cytometer (BD Biosciences, Sparks, MD) equipped with a 325-nm laser. To confirm proper Indo-1 loading and the functional potential of cells, ionomycin (15μg/ml; Sigma-Aldrich, St. Louis, MO) was used. The kinetics of the data were analyzed with FlowJo software (Tree Star, Ashland, OR).
The capture antibodies and biotinylated detection antibodies for interferon (IFN)-γ, IL-2, tumor necrosis factor (TNF)-α, and IL-4 were purchased from Mabtech (Stockholm, Sweden). LAT-substituted CD4+ T cells were rested for 24hr after electroporation and then quantified with a Multisizer 3 COULTER COUNTER (Beckman Coulter, Brea, CA), cell concentration was adjusted, and cells were stimulated with anti-CD3/CD28 beads at a 1:1 ratio on capture antibody-coated 96-well filtration plates (Millipore) for 16hr. After three washes, detection antibodies were added according to the manufacturer's protocol. Two hours later, spots were detected with an avidin–biotin–peroxidase complex (Mabtech, Nacka Strand, Sweden) and 1-Step NBT/BCIP (Thermo Fisher Scientific) substrate (NBT nitroblue tetrazolium chloride; BCIP, 5-bromo-4-chloro-3′-indolyl phosphate p-toluidine salt). Spot-forming cells (SFCs) were objectively quantified with an AID iSpot FluoroSpot reader system (AID, Strassberg, Germany).
For intracellular cytokine staining, LAT-substituted CD4+ T cells were stimulated for 4hr with immobilized anti-CD3 (5μg/ml) or phorbol myristate acetate (PMA, 1μg/ml; Sigma-Aldrich) and ionomycin (1μM; Sigma-Aldrich), adding GolgiStop (BD Biosciences) at 1hr poststimulation. These cells were treated with a FIX & PERM kit (Invitrogen) and stained with eFluor 450-conjugated anti-IFN-γ (eBioscience) and phycoerythrin–cyanine 7 (PE–Cy7)-conjugated anti-IL-4 (eBioscience) according to the manufacturer's protocol. Cytometric data were collected with the LSR II and analyzed with FlowJo software.
After 24hr of incubation, LAT-substituted CD4+ T cells were stained with PKH26 (Sigma-Aldrich) according to the manufacturer's protocol and cultured on an anti-CD3 antibody-coated plate for 72hr. PKH26 expression levels of these cells were measured with a FACSCalibur (BD Biosciences) and mean T(X) between 0 and 72hr for each cell population was calculated, using the population comparison mode in the FlowJo program as previously published (Roederer et al., 2001; Munson, 2010). T(X) is a statistic which provides an indication of the probability with which two distributions between 0 and 72 hr are different. Higher values of T(X) reflect more proliferation.
Flow-based cytotoxicity assays were performed as previously described (Cao et al., 2010; Zhao et al., 2010). NY-ESO-1- and CD19-coexpressing target cells were generated by lentiviral transduction (Carpenito et al., 2009; Milone et al., 2009). To express the green fluorescent protein (GFP)–NY-ESO-1 fusion protein in Nalm-6 cells (human B cell leukemia, CD19+HLA-A2+; DSMZ, Braunschweig, Germany), the parental Nalm-6 line was transduced with GFP-NY-ESO-1 to obtain Nalm6-ESO. As a negative control, K562 cells (chronic myeloid leukemia, NY-ESO-1−CD19−; American Type Culture Collection [ATCC], Manassas, VA) were transduced to express mesothelin (K562-Meso).
Nalm6-ESO and K562-Meso cells were mixed at a 1:1 ratio and 100-μl volumes (5×104 cells) were aliquoted per well into 96-well round-bottom plates and then mixed with 100μl of effector cells at several effector-to-target (E:T) ratios. After 4hr of incubation, these cells were stained with anti-CD3–allophycocyanin (APC) (eBioscience) and 7-aminoactinomycin D (7AAD; BD Biosciences), and fluorescence was measured with the FACSCalibur (BD Biosciences). Specific lysis was calculated as the ratio of live Nalm6-ESO (CD3−7AAD−GFP+) cells and live K562-Meso (CD3−7AAD−GFP−) cells with FlowJo software. Result were normalized with six wells of target cells without effector cells.
The CD107a mobilization assay was performed as described (Cao et al., 2010). One hundred microliters of effector and target cells was plated in 96-well round-bottom plates at a ratio of 1:1, 2:1, or 4:1 (1, 2, or 4×105 effectors+105 targets). Ten microliters of PE–Cy5-labeled anti-CD107a antibody (eBioscience) was added to each cell mixture and incubated at 37°C for 1hr before adding GolgiStop. After an additional 2.5hr of incubation, these cells were stained with anti-CD8-APC–Cy7 (BD Biosciences) and levels of surface CD107 were determined by flow cytometry, using the LSR II.
For statistical evaluations, Student t tests were performed between two groups of homoscedastic samples, and the samples with unequal variance were judged by Cochran–Cox.
In one study we reported that expression of a ubiquitylation-resistant LAT mutant (LAT 2KR) in Jurkat cells was more stable than that of wild-type LAT (LAT WT) and induced the enhancement of downstream signaling, including phosphorylation and calcium flux (Balagopalan et al., 2011). In addition, the previous works showed increased CD69 upregulation in CD4+ T cells. Here we have extended the previous observations to include CD8+ T cells and have now investigated the effects of LAT modification on functional assays, including cytokine production, calcium signaling, phosphorylation of signaling molecules, and importantly, cytotoxicity.
To study the effects of LAT in primary human T cells, we prepared in vitro-transcribed (IVT) mRNA for LAT 2KR, LAT 4YF, and wild-type LAT fused to YFP (Supplementary Fig. S1). Freshly purified CD4+ T cells from a healthy donor were electroporated with LAT 2KR–YFP (ubiquitylation-resistant mutant), LAT 4YF–YFP (phosphorylation-resistant mutant), and LAT WT–YFP. LAT siRNA was concurrently electroporated to selectively knock down the endogenous LAT, because the target site of this siRNA was mutated in LAT WT–YFP, LAT 2KR–YFP, and LAT 4YF–YFP. LAT siRNA and LAT–YFP IVT mRNA-electroporated cells were lysed and analyzed by Western blot (Fig. 1A). More than 65% of endogenous LAT (36kDa) was knocked down by LAT siRNA electroporation between 24 and 48hr after treatment. Expression of the LAT fusion proteins was dose dependent, as 20μg of RNA-electroporated cells expressed LAT–YFP (64kDa) at higher levels than 10μg of RNA-electroporated cells, and these differences in expression were sustained for at least 48hr. LAT expression was confirmed by measurement of YFP expression by flow cytometry (Fig. 1B and C, and Supplementary Fig. S2). expression of LAT 2KR and LAT WT–YFP fusion proteins was comparable. At 24hr after electroporation, the YFP-positive frequencies were more than 90% in all groups of LAT–YFP-electroporated cells, and these frequencies were continued by 48hr. Thereafter, expression gradually decreased and was undetectable on day 6. There was no significant difference between LAT 2KR expression and LAT WT expression for 48hr; however, the YFP expression of LAT 2KR-electroporated cells was higher than that of LAT WT from 72 to 96hr after electroporation (Supplementary Fig. S2), confirming previous data in Jurkat cells that the LAT 2KR protein was more stable (Balagopalan et al., 2011). Endogenous LAT expression had almost recovered to baseline by 72hr after LAT siRNA electroporation.
Conditions were also established to express the altered LAT proteins in in vitro-expanded CD8+ T cells, prepared as described in Materials and Methods. In cultured CD8+ T cells, endogenous LAT was significantly knocked down (p<0.01) and YFP-fused LAT was expressed (p<0.01) (Fig. 1D and E). There was no significant difference between the expression of LAT 2KR and LAT WT. On the basis of these results, endogenous LAT could be efficiently substituted with LAT 2KR for at least 48hr after electroporation when 20μg of RNA of LAT–YFP was used. These results were confirmed in CD8+ cells from three different donors. In subsequent experiments, the above-described conditions were used to assess the effects of the various LAT mutants on T cell signaling and function.
The T cell signaling of LAT-substituted cells was assessed in CD8+ T cells electroporated with YFP-fused LAT 2KR or WT IVT mRNA together with LAT siRNA. The cells were stimulated with anti-CD3/CD28 magnetic beads 24hr after electroporation, and were lysed 10 or 30sec after stimulation. The bands of phospho-PLCγ783, phospho-LAT191, and total LAT are shown in Fig. 2A. The phosphorylation LAT was elevated in both LAT 2KR- and LAT WT-expressing cells at 10sec, and the relative phosphorylation of LAT 2KR was higher than that of LAT WT. Similarly, the phosphorylation of PLCγ in LAT 2KR-expressing cells was ~4-fold higher than in LAT WT-expressing cells, and the phosphorylation was sustained in LAT-2KR-expressing cells in contrast to the transient signal in LAT WT cells. The kinetics of phosphorylation were further assessed by comparison of the tyrosine phosphorylation band intensities to that of LAT–YFP (Fig. 2B). Endogenous LAT was confirmed to be knocked down.
The augmented and sustained phosphorylation of PLCγ suggested that calcium signaling might be augmented in cells expressing LAT 2KR. To test this notion, we next measured calcium flux in single cells after stimulation with soluble anti-CD3 antibody, and the changes in mean intracellular free calcium concentration are shown in Fig. 2C and D. Cells expressing LAT 2KR had an augmented peak calcium signal compared with cells expressing LAT WT. The effect was dose dependent, in that cells electroporated with 20μg of 2KR had more pronounced calcium signaling compared with cells electroporated with 5μg of 2KR (Fig. 2D, middle and bottom).
Unexpectedly, we observed that cells with knocked-down endogenous LAT had an elevation of baseline calcium and an attenuated calcium signal after anti-CD3 stimulation (Fig. 2D, top). Similarly, cells with knocked-down endogenous LAT and reconstituted with the phosphorylation-defective LAT 4YF also had constitutive elevation of calcium and a blunted calcium flux after TCR stimulation. In contrast, control siRNA-electroporated cells had the same resting calcium concentration as mock-electroporated T cells. Thus primary T cells expressing the LAT 2KR protein have increased tyrosine phosphorylation and calcium signaling compared with T cells expressing LAT WT.
Studies have shown that mice expressing a tyrosine phosphorylation-defective form of LAT develop lymphoproliferative disorders involving polyclonal T cells that produced high amounts of Th2 cytokines (Malissen et al., 2005). Therefore, we evaluated the effect of the LAT 2KR construct on the cytokine-producing profiles of CD4+ T cells by single-cell ELISPOT assay (Fig. 3A and Supplementary Fig. S3). After stimulation with anti-CD3 and anti-CD28, the frequency of IFN-γ-secreting cells in LAT 2KR-substituted cells was significantly increased compared with that of cells expressing LAT WT (p<0.01). Similarly, the frequency of IL-2 and TNF-α spot-forming cells was also increased in T cells expressing LAT 2KR. Consistent with a Th1 bias, IL-4-secreting cells were decreased in LAT 2KR-substituted CD4+cells (p<0.001). In contrast, in LAT knockdown cells the frequency of IFN-γ secretion was less than that of control siRNA-electroporated cells. IL-2 and TNF-α spot-forming cells were also attenuated in LAT knockdown cells. As in LAT-deficient mice (Mingueneau et al., 2009), we found that the frequency of IL-4-secreting cells was increased in LAT knocked-down T cells.
We used flow cytometry to measure IFN-γ- and IL-4-producing cells from three different donors by intracellular cytokine staining, as shown in Fig. 3B; the ELISPOT assay in Fig. 3A was performed on cells from donor 1. In all donors, more IFN-γ-positive cells were detected after anti-CD3 stimulation in LAT 2KR-substituted cells than in LAT WT-substituted cells. And in donor 1, fewer IL-4-positive cells were observed in LAT 2KR-expressing cells, whereas in donors 2 and 3 the frequency of IL-4-secreting cells was unchanged. In contrast, after pharmacological stimulation with PMA and ionomycin there was no difference in the frequency of cells secreting IFN-γ (Supplementary Fig. S4), consistent with a requirement for TCR stimulation for the enhanced function of cells expressing LAT 2KR. Last, cells expressing the tyrosine phosphorylation-deficient LAT 4YF molecule had impaired Th1 cytokine secretion, similar to LAT knockdown cells (Fig. 3A).
In the above-described experiments, the immediate effects of LAT knockdown or overexpression were assessed. We next determined whether manipulation of LAT would have an impact on T cell proliferation and function. Freshly isolated CD4+ T cells were electroporated, rested for 24hr, and stained with PKH26. These cells were stimulated with immobilized OKT3 for 72hr and the dilution of PKH26 was measured (Fig. 4A). There were no significant differences in the proliferation of T cells expressing 2KR, 4YF, or WT LAT (Fig. 4B). However, when these cells were restimulated with PMA and ionomycin to evaluate Th1 and Th2 polarization, there was a robust increase in IFN-γ and a decrease in IL-4 in cells expressing LAT 2KR compared with LAT WT or LAT 4YF cells (Fig. 4C). These results suggest that expression of ubiquitylation-resistant LAT biases cells toward Th1 differentiation. Importantly, cells expressing LAT 2KR do not have spontaneous proliferation or constitutive cytokine secretion (Fig. 4C).
The observation that ubiquitylation-resistant LAT augments signaling and cytokine secretion in mature T cells suggested that this strategy might enhance the potency of redirected T cells. To test this hypothesis we used the 1G4 TCR that we have previously shown to redirect T cells to recognize NY-ESO-1157–165 in the context of HLA-A2 (Zhao et al., 2006). mRNA encoding the α and β chains of the 1G4 TCR was coelectroporated into in vitro-expanded resting CD8+ T cells along with LAT siRNA and LAT–YFP. The anti-NY-ESO-1-specific TCR was expressed in more than 90% of cells and there was no significant difference in expression due to the introduction of the various LAT mutants (Supplementary Fig. S5A). The cytotoxic activities of these cells against Nalm6-ESO-1 targets expressing HLA-A2 and NY-ESO-1 were analyzed by a flow cytometry-based cytotoxic assay (Fig. 5A). Importantly, the cytotoxicity of LAT 2KR-substituted cells was significantly augmented compared with LAT WT-substituted cells and with T cells expressing endogenous LAT. In contrast, the cytotoxicity of LAT knockdown cells and LAT 4YF-substituted cells was inhibited compared with redirected T cells expressing endogenous LAT.
To further characterize the function of the NY-ESO-1 TCR we assessed the mobilization of CD107a (Fig. 5B). CD8+ T cells expressing the anti-NY-ESO-1 TCR and the engineered LAT molecules were coincubated with 624.38 melanoma cells (HLA-A2+NY-ESO-1+) or with 526 melanoma cells (HLA-A2+NY-ESO-1−).LAT 2KR T cells had augmented CD107a translocation in comparison with all other cell populations. The effector activity was specific for NY-ESO-1, and CD107a was induced in a tumor dose-dependent manner. The cells with endogenous LAT knockdown had a modest inhibition of CD107a translocation compared with redirected T cells reconstituted with LAT WT or with T cells electroporated only with the TCR mRNA. These results are consistent with the previous cytotoxicity assays.
We next explored whether it was necessary to knock down endogenous LAT in order augment the cytotoxic function of T cells. IVT mRNA directing the expression of the NY-ESO-1 TCRα/β and LAT–YFP were electroporated into CD8+ T cells without LAT siRNA. The introduction of TCR mRNA did not impair expression of the LAT fusion proteins (Supplementary Fig. S5B). Immunoblot analysis confirmed that the LAT fusion proteins could be expressed in T cells at levels exceeding endogenous LAT (Supplementary Fig. S5C), and flow cytometry indicated that more than 90% of the cells expressed the fusion proteins (Supplementary Fig. S5D). LAT 2KR-overexpressing T cells showed significantly increased cytotoxicity (Fig. 5B). This effect was specific for the ubiquitylation-resistant form of LAT, as LAT WT- or LAT 4YF-overexpressing cells did not exhibit any enhancement compared with control cells that were expressing the specific TCR without LAT–YFP. The finding that the knockdown of endogenous LAT is not required for the therapeutic strategy using LAT 2KR has important practical implications.
Last, we investigated whether T cells engineered to express LAT 2KR and a chimeric antigen receptor (CAR) would have enhanced functional activity. For these experiments we used the anti-CD19-ζ CAR (Milone et al., 2009). To simplify the interpretation of this approach we used a CD3ζ-only CAR, because the signaling via costimulatory molecules would complicate the evaluation of the various LAT constructs. CD8+ T cells were electroporated with 10μg of anti-CD19-ζ CAR mRNA together with LAT siRNA and LAT–YFP constructs. The CAR was expressed at the cell surface in more than 75% of cells and there were no significant differences in the efficiency of expression between cells expressing LAT 2KR, LAT WT, or LAT 4YF (Supplementary Fig. S5E). When tested on the Nalm-6 tumor line, the CD19-specific cytotoxicity of T cells expressing LAT 2KR was significantly augmented (Fig. 5D). In contrast, the cytotoxicity of LAT knockdown cells and LAT 4YF-substituted cells was attenuated. Thus, T cells with engineered LAT have the potential to augment the therapeutic efficacy of CAR T cells.
Here we report the potential of modified LAT to improve the effector functions of primary postthymic human T cells. In one study Balagopalan and colleagues (2011) showed that expression of ubiquitylation-resistant LAT 2KR increased signaling in human T leukemia cells by two mechanisms: more LAT was expressed and LAT 2KR was more potent in a molecule-to-molecule comparison. We find here that electroporation of mRNA results in efficient expression of LAT 2KR and that this results in increased signaling and function of CD4+ and CD8+ human T cells. We found that in primary T cells, the phosphorylation of wild-type LAT was rapid in onset and transient, whereas phosphorylation of 2KR LAT was augmented and sustained. Because phosphorylation and ubiquitylation reactions occur within seconds (Fig. 2A; and Pierce et al., 2009), it is likely that once phosphorylated, wild-type LAT becomes rapidly ubiquitylated. Ubiquitylation likely leads to a change in LAT localization, dephosphorylation, and finally degradation. In contrast, phosphorylated 2KR LAT persists because it is resistant to ubiquitylation.
Stimulation of CD4+ T cells that express LAT 2KR induced a pattern of cytokine secretion consistent with Th1 differentiation, whereas cytokine secretion in LAT knockdown cells was consistent with a Th2 pattern of differentiation. Signaling through the TCR can regulate Th1 and Th2 differentiation (Constant et al., 1995; Nakayama and Yamashita, 2010). Previous studies have shown that LAT has a role in T cell development in the balance of Th1 and Th2 differentiation (Malissen et al., 2005).
We also found that expression of LAT 2KR increased the activation of CD8+ T cells after tumor antigen presentation. It was notable that cytotoxicity was enhanced in LAT 2KR cells, and that this augmentation was observed not only after antigen presentation to MHC class I-restricted TCRs but also after antigen stimulation of T cells expressing chimeric antigen receptors. The enhanced cytotoxicity was specific for T cells expressing the ubiquitin-resistant form of LAT.
An unexpected finding in our studies was that LAT knockdown T cells had an elevated basal level of intracellular calcium concentration. This suggests that one function of LAT is to negatively regulate calcium homeostasis in quiescent cells. Others have previously reported that LAT can have negative regulatory function in mouse T cells (Mingueneau et al., 2009). In lymph nodes, there is evidence of continuous “tonic” signaling through the TCR, and that this is required to maintain antigen responsiveness (Hochweller et al., 2010). Consistent with a role for LAT in basal T cell signaling, imaging studies have shown that membrane domains of both LAT and TCRζ significantly overlap under both activating and nonactivating conditions (Sherman et al., 2011). Our studies suggest that LAT may play a role in regulating the tonic resting state of the lymphocyte (Hochweller et al., 2010) as well as the activated state during antigen stimulation (Lillemeier et al., 2010).
In these initial studies we have used transient expression of modified LAT to evaluate the potential of LAT 2KR for potential clinical applications. This would provide safety but may limit the potency in the clinic. Further studies will be required to evaluate the safety and efficacy of engineered T cells that are permanently modified to express LAT 2KR. Thus a limitation of the present studies is that the safety of this approach with constitutive expression of ubiquitin-resistant LAT remains unknown. Another limitation of this approach is that it is possible that augmented LAT signaling could result in the induction of activation-induced cell death (AICD). Our results show that the enhanced signaling mediated by LAT does not decrease proliferation during the first 3 days of stimulation; however, it is possible that during longer term stimulation this could result in the induction of AICD and impaired proliferation of LAT-engineered T cells in vivo.
Ubiquitylation-resistant LAT is a novel approach to enhance T cell signaling and has the potential to augment effector T cell function for adoptive cell therapy. We found that the function of T cells expressing an MHC class I-restricted TCR was enhanced by ubiquitin-resistant LAT. It is possible that this approach might improve the effector functions of T cells expressing low-affinity TCRs that are typical of many tumor-associated antigens (Kessels et al., 2001; Zehn et al., 2009). Last, the combination of an antigen-specific CAR with ubiquitylation-resistant LAT could be considered an example of a “next-generation CAR” that has promise to further enhance the effector functions of engineered T cells for cancer and chronic viral infections.
The authors thank Carmine Carpenito, Michael Kalos, and Caleph Wilson (University of Pennsylvania) and Shinichiro Motohashi (Department of Immunology and Thoracic Surgery, Graduate School of Medicine, Chiba University, Chiba, Japan) for advice; and members of the Human Immunology Core (Abramson Cancer Center) for providing lymphocytes. This work was funded by the National Institutes of Health (NIH) Common Fund Nanomedicine program (PN2 EY016586), and this research was supported, in part, by the Intramural Research Program of the NIH, NCI, CCR (L.B. and L.E.S.).
Dr. Balagopalan and Dr. Samelson have filed a patent on behalf of the U.S. Government. All other authors have no conflict of interest.