A Small-Molecule c-Rel Inhibitor Reduces Alloactivation of T Cells without Compromising Antitumor Activity

Allogeneic hematopoietic stem cell transplantation (allo-HSCT) represents a potent therapy for malignant and nonmalignant hematologic diseases (1). In the care of patients with malignant diseases, it was initially developed to follow high-dose chemotherapy/radiotherapy to rescue from therapy-related bone marrow (BM) failure; however, the emphasis has now shifted toward allo-HSCT as a strategy to facilitate graft-versus-tumor (GVT) activity (2). GVHD remains a major complication of allo-HSCT, resulting in significant morbidity and mortality (3, 4), and strategies to suppress GVHD are often associated with broad suppression of the immune system, leading to immune deficiency and compromised antitumor activity (5). Molecular pathways that have been targeted in preclinical studies to accomplish separation of GVHD from GVT activity include the mTOR, histone deacetylases, and NF-κB (via proteasome inhibition; ref. 6). The NF-κB family member c-Rel is a transcription factor that regulates lymphocyte survival and proliferation following antigen receptor triggering and plays a dominant role in inflammation, autoimmunity, and alloimmunity (7–10). c-Rel regulates expression of many inflammatory cytokines and is expressed in T and B cells as well as monocytes/macrophages and dendritic cells (11). Immune defects secondary to c-Rel deficiency have been attributed to impaired activation of lymphocytes, particularly T cells. In T cells, the main target gene of c-Rel is interleukin (IL)-2, a cytokine required for normal T-cell proliferation and differentiation. Previous reports demonstrated that c-Rel–deficient T cells showed reduced Th1 but normal Th2 responses (12, 13) and are compromised in the generation of anti-inflammatory natural regulatory T cells (Treg; refs. 14–18) and proinflammatory Th17 cells (18, 19). Given the limitations of existing immunosuppression-based therapies for GVHD and the pivotal role of c-Rel in T-cell proliferation and function, we used strategies targeting the c-Rel pathway, including small molecule–based inhibition of c-Rel activity (20, 21), to modulate T-cell responses in the context of GVHD and malignant diseases.

We first analyzed the biologic significance of c-Rel for hematopoietic reconstitution and T-cell activation after allo-HSCT. We performed studies assessing c-Rel expression in T cells in the setting of radiation-induced injury as well as GVHD (Fig. 1A and Supplementary Fig. S1). After irradiation and especially during GVHD, IL-2, CD25 (IL-2Rα), and c-Rel were upregulated, consistent with lymphocyte activation. c-Rel expression was analyzed by both intracellular and intranuclear staining (22), and the results were closely correlated (Fig. 1B and C). Hematopoietic stem cell functions, including the development of lymphoid and myeloid lineages, are not known to be dependent on c-Rel activity. To confirm this, we performed an allo-HSCT using c-Rel^−/− donor BM in an MHC-disparate allo-HSCT model. c-Rel^−/− BM engrafted and reconstituted recipients as efficiently as wild-type (WT) BM (Fig. 1D). We also sought to determine whether c-Rel activity in cells other than donor T cells played a role for GVHD development by using different combinations of c-Rel^−/− BM and T cells as the donor source (Fig. 1E), as well as using c-Rel^−/− mice as recipients (Fig. 1F). c-Rel deficiency in donor BM or in the recipient did not affect survival or GVHD scores.

A recent study found evidence for a role of c-Rel in donor T cells during acute GVHD in mice (10). We were able to confirm these observations by using c-Rel^−/− donor T cells in mouse models of MHC-disparate as well as a more clinically relevant minor antigen-mismatched allo-HSCT, which in both cases resulted in significant amelioration of GVHD as determined by survival, weight loss, histopathology of GVHD target organs, and clinical GVHD scores (Supplementary Fig. S2 and data not shown). Moreover, we evaluated the profiles of donor T cells in the spleens of recipient mice on day 7, revealing that the effector:naïve T-cell (T_E/T_N) ratios were decreased for both CD4⁺ and CD8⁺ T cells, and we observed lower CD8⁺ T_E/Tregs ratios in mice receiving c-Rel^−/− T cells compared with recipients of WT T cells (Fig. 2A). We also found increased thymic cellularity on day 14 in recipients of c-Rel^−/− T cells compared with WT T cells, indicating reduced thymic GVHD (Fig. 2B). Furthermore, c-Rel^−/− T cells expressed significantly lower levels of the intestinal homing marker LPAM-1 on T_E cells (Fig. 2C), consistent with the findings that recipients of c-Rel^−/− T cells had much lower numbers of infiltrating donor T cells in mesenteric lymph nodes (mLN) and the small intestine (Fig. 2D and E). Importantly, the absolute numbers of Tregs infiltrating the small intestines showed no difference between recipients of WT and c-Rel^−/− T cells (Fig. 2F), reinforcing that the net outcome of the T-cell balance is shifted toward T-cell suppression with significantly lower CD8⁺ T_E/Tregs ratios in the small intestines of recipients of Rel^−/− T cells (Fig. 2E).

Unlike in vitro (7), we found that a greater percentage of c-Rel^−/− donor T cells produced IL-2 compared with WT T cells on day 7 after allo-HSCT, resulting in IL-2 serum concentrations above the dissociation constant (K_d; that means more than 50% of IL-2 receptors will be occupied) in recipients of c-Rel^−/− T cells and in significantly lower levels in recipients of WT T cells (Fig. 3A and B). STAT5 is a negative regulator of IL-2 production (23, 24). When we analyzed STAT5 phosphorylation (pSTAT5) in c-Rel^−/− and WT T cells, we found that splenic c-Rel^−/− T cells harvested 96 hours after allo-HSCT exhibited impaired IL-2–mediated induction of pSTAT5 (Fig. 3C). This lower responsiveness to IL-2 is consistent with decreased expression of CD25 by c-Rel^−/− T_E cells (data not shown) and suggests decreased negative feedback on IL-2 production, resulting in a relative increase in IL-2 secretion compared with WT T cells. Furthermore, consistent with a previous report (10), we observed a dramatic increase in the numbers of donor-derived Tregs in recipients of c-Rel^−/− T cells starting on day 7 after HSCT (Fig. 3D), whereas host Tregs were hardly detectable. Expression of CD25 on those Tregs was higher than that of WT Tregs (Fig. 3E). These data suggest that competition for IL-2 between T_E and Tregs results in promotion of alloactivation in recipients of WT T cells, and in expansion of Tregs in recipients of c-Rel–deficient T cells (Fig. 3F). Neuropilin-1 (NRP-1) was recently identified as a marker for natural Tregs (25, 26). As expected, the vast majority of c-Rel–deficient Tregs in our experiments were Nrp-1⁻–induced Tregs (data not shown). To further investigate the potential role of Tregs in the amelioration of GVHD, we performed in vivo depletion of Tregs using FoxP3-DTR transgenic mice that express the diphtheria toxin receptor (DTR) under control of the Foxp3 promoter (27) as donor T-cell source. Elimination of Tregs in vivo by diphtheria toxin administration on day 13 worsened the survival of recipients of c-Rel–deficient T cells (c-Rel deficiency was achieved by pretreatment of donor T cells with a small-molecule inhibitor compound; see below for more details), suggesting that donor Tregs indeed contributed to the amelioration of GVHD (Fig. 3G and Supplementary Fig. S3). Depletion of donor Tregs in recipients of T cells with normal c-Rel activity did not affect GVHD severity, indicating that in contrast with animals receiving c-Rel–deficient T cells, the Treg number in animals receiving WT T cells was too low during early GVHD to be biologically significant. Importantly, Tregs from c-Rel^−/− mice are functional (28, 29) and could reduce GVHD in our model (Fig. 3H).

Some natural compounds and their synthetic derivatives inhibit Rel or NF-κB via specific interaction with the cysteine residues critical for binding to the specific κB-DNA sequence (20). For example, dehydroxymethylepoxyquinomicin (DHMEQ) was shown to inhibit NF-κB binding activity; however, the complex structures and poor pharmacokinetics due to reactive sites found in natural products made them unsuitable for further development. To develop direct Rel inhibitors, we established an assay [fluorescence polarization (FP) assay; for more details, see Methods] for high-throughput screening. Screening of a library of 15,000 compounds yielded 20 hits. We conducted structure–activity relationship studies of these initial hits using electrophoretic mobility shift assay (EMSA) and finally identified the hydrophobic small-molecule pyrimidinetrione and its derivatives as potent and highly specific inhibitors of c-Rel activity, having a 20- to 200-fold higher inhibitory effect on c-Rel and NF-κB than on other transcription factors such as OCT1 and AP1 (Fig. 4A and data not shown). These compounds bind c-Rel directly and change the conformation of the protein, inhibiting DNA binding and transcriptional activity. We developed several generations of pyrimidinetrione derivatives to further optimize the inhibitory effect and found that treatment of activated T cells with one of those compounds (IT-603) resulted in most efficient inhibition of c-Rel activity as well as IL-2 expression in vitro (Fig. 4B–D) without altering cell viability after 24 hours of incubation (Fig. 4E). IT-603 was used for all subsequent experiments involving chemically induced c-Rel inhibition. Basic characteristics of IT-603 are summarized in Table 1.

We next tested the capacity of IT-603–pretreated T cells to induce GVHD. By day 4 after allo-HSCT, T cells pretreated with IT-603 expanded less compared with T cells treated with empty vehicle [dimethyl sulfoxide (DMSO); Fig. 4F], and even though the inhibitory effect of the compound was temporary and c-Rel activity returned to normal by day 4 after allo-HSCT (data not shown), those T cells caused significantly reduced GVHD (Fig. 4G). We also observed higher IL-2 secretion from donor CD4⁺ T cells on day 4 in this experiment (Supplementary Fig. S4). Of note, we recently developed a DMSO-free lipid-based formulation of IT-603 using the U.S. Food and Drug Administration (FDA)–approved nonionic surfactant Cremophor. Intraperitoneal injection of this formulation was well tolerated and allowed us to study pharmacokinetics (Table 1 and Fig. 4H), in a first step toward development of a c-Rel inhibitor drug.

A recent study found evidence indicating that in recipients of c-Rel^−/− T cells, GVT activity against A20-TGL mouse lymphoma cells can be intact (10). We confirmed these findings and were, moreover, able to demonstrate that this effect can be sustained with T-cell doses as low as 25% of the standard dose (data not shown), and when targeting less immunogenic A20 tumor cells as well as a solid tumor RENCA (renal cell carcinoma; data not shown).

In addition, c-Rel^−/− T cells displayed strong antitumor activity against EL4-TGL T-cell lymphoma cells even in the absence of T-cell alloactivation (Fig. 5A). In this syngeneic model, the level of CD25 on c-Rel^−/− CD8⁺ T cells on day 7 after transplantation was similar to that of WT T cells (Fig. 5B), in striking contrast with the allo-HSCT setting, in which CD25 levels in c-Rel^−/− T cells were significantly decreased (Fig. 5C). We performed a similar experiment using B16-TGL melanoma tumor cells and melanoma-specific syngeneic donor T cells from Pmel1^+/+ mice after incubating those T cells with the c-Rel inhibitor compound. Pmel1^+/+ T cells retained normal GVT activity under c-Rel–deficient conditions (Fig. 5D), indicating that inhibition of c-Rel activity is not sufficient to prevent antigen-specific T-cell receptor (TCR) activation. To further dissect the role of c-Rel in antigen-specific activation versus alloactivation of T cells, we analyzed the expression levels of c-Rel in Pmel1^+/+ T cells stimulated by B16 tumor cells and compared them with those of WT C57BL/6 T cells stimulated with MHC-mismatched splenocytes (Fig. 5E and F). We found significantly lower levels of c-Rel in Pmel1^+/+ T cells specifically stimulated with target antigen, suggesting that c-Rel activity is less required for antigen-specific T-cell activation. Moreover, T-cell stimulation in the presence of the c-Rel inhibitor compound resulted in reduced c-Rel activity only when allogeneic stimulators were used (Fig. 5E).

Importantly, WT T cells pretreated with a c-Rel antagonist mediated GVT activity in several tumor models (Fig. 5D and G–I), indicating that viable and functional T cells are present. It is therefore highly likely that the diminished capacity of these T cells to induce GVHD is due to the c-Rel inhibitory mechanism of the small-molecule compound and not simply due to a nonspecific effect of the in vitro manipulation procedure. However, we did observe that GVT activity of WT T cells treated with inhibitor compound was slightly less than the GVT activity of c-Rel^−/− T cells, and we found moderate reduction in antigen-specific T-cell activation (but not alloactivation) in vitro in the presence of vehicle alone (Supplementary Fig. S5). As a result of our ongoing efforts to establish less-toxic compound formulations, we recently succeeded in developing a regimen for systemic administration of the inhibitor compound as GVHD therapy, and we were able to establish in vivo efficacy (Fig. 5J). Although the administration of empty vehicle was not associated with any signs of toxicity, daily intraperitoneal administration of c-Rel inhibitor solution at a dose of 24 mg/kg for more than 2 weeks resulted in mild diarrhea and ruffled fur. Importantly, c-Rel deficiency (as a result of c-Rel inhibitor administration or as a result of transplantation of c-Rel^−/− BM) did not have a negative impact on hematopoiesis and immune reconstitution (Fig. 1D and data not shown).

Our inhibitor compound effectively inhibited c-Rel activity of human T cells (Fig. 5K). Moreover, in vitro cytotoxicity analysis of human cytomegalovirus (CMV)–specific T cells (Fig. 5L; ref. 30) as well as Wilms tumor 1 (WT1) and Epstein–Barr virus (EBV)–specific T cells (Fig. 5M and data not shown) demonstrated that inhibition of c-Rel activity did not impair antigen-specific TCR-mediated killing. Alloreactivity of human T cells was on the other hand dramatically reduced when human peripheral blood mononuclear cells (PBMC) were cultured with HLA-mismatched stimulators in the presence of the inhibitor compound (Fig. 5M). These data reinforce the notion of separation of GVHD from GVT activity through inhibition of c-Rel activity even in human T cells.

The NF-κB/Rel transcription factor family is composed of five members of interacting proteins: c-Rel, p50, p65, p52, and RelB. c-Rel is involved in the pathway downstream of antigen-stimulated canonical NF-κB signal transduction and is crucial for T-cell proliferation and differentiation (7, 8). Studies in mouse models revealed unique roles for c-Rel in the pathophysiology of allergic reactions, autoimmunity, and allogeneic transplantation (10, 11, 31, 32). Our study revealed that deficiency of c-Rel in donor T cells ameliorated GVHD due to impaired alloactivation and proliferation of T_E, decreased homing of donor c-Rel^−/− T_E to the small intestine, and increased Treg-mediated suppression.

Surprisingly, levels of IL-2, one of the key target genes of c-Rel, were increased in recipients of c-Rel^−/− T cells on day 7 after allo-HSCT. We identified a feedback mechanism that resulted in reduced downregulation of IL-2 production by c-Rel^−/− T cells during periods of T-cell activation in the setting of acute GVHD. In addition to the relative increase of IL-2 at a critical time period during early GVHD, we also found that serum levels of Th2-type cytokines, such as IL-4, IL-5, and IL-13, were significantly increased on day 7 after allo-HSCT, and that GATA-3 expression of c-Rel^−/− T cells was increased, whereas expression of T-bet was decreased (Supplementary Fig. S6), suggestive of Th2 polarization in recipients of c-Rel^−/− T cells (12, 13). The availability of optimal IL-2 levels in the absence of competition by effector T cells resulted in expansion of c-Rel–deficient natural Tregs and, more importantly, induction of Tregs. These findings can certainly be considered counterintuitive, given the role of c-Rel for IL-2 as well as Treg generation under physiologic conditions. However, our experiments reveal a differential outcome of the IL-2–STAT5 pathway in WT compared with c-Rel–deficient T cells during GVHD. In the setting of GVHD, pSTAT5 in c-Rel–deficient donor T cells is decreased, which over time leads to increased IL-2 levels (due to decreased negative feedback) and renders T_E at the same time noncompetitive (due to decreased CD25 expression), while allowing Tregs, a T-cell population that expresses CD25 by definition, to outcompete T_E. Our findings as well as a recently described similar concept of competition for IL-2 between T_E and Tregs (23, 24) indicate that this mechanism may indeed have significant implications for alloactivation and GVHD, and it will be an important line of future research to evaluate the kinetics of T_E and Tregs in response to interventions modulating the IL-2 pathway (33). Reported roles of c-Rel in T-cell responses and transplantation immunology in comparison with our findings during GVHD are summarized in Table 2.

Inhibition of c-Rel activity did not impair GVT activity, as demonstrated by various tumor models in the setting of allogeneic and syngeneic HSCT. This phenomenon was most evident when using c-Rel^−/− T cells as opposed to inhibitor compound–treated T cells, which can be attributed to expected potential limitations of any chemical strategy to inhibit protein function (such as limited duration of the effect, potential off-target effects, and nonspecific toxicities). Furthermore, antigen-specific T-cell activation was associated with significantly less c-Rel expression than allostimulation. The redundancy of c-Rel activity for antigen-specific TCR activation provides an underlying molecular mechanism for the observed separation of GVHD from GVT activity: upon only antigen-specific activation (but not alloactivation), c-Rel–deficient T cells can differentiate into activated effector T cells that exhibit normal cytotoxicity responses. In addition, we found that in the setting of c-Rel deficiency, Treg-mediated suppression of GVHD represents an important additive factor without equally affecting GVT activity. Although it is possible that this phenomenon is more pronounced in experimental models in which the kinetics of tumor eradication can be much different from real-life scenarios, it is important to keep in mind that there is accumulating evidence that Tregs can preferentially affect GVHD while preserving protective immunity (34, 35). Our findings in the setting of an MHC-matched allo-HSCT indicate that c-Rel activity is also involved in T-cell activation mediated by minor transplantation antigens. This and other important questions about the differential requirements for c-Rel activity for TCR triggering will need to be addressed in more depth in future studies; however, our current evidence that inhibition of c-Rel activity separates GVHD from GVT activity at the level of TCR signaling has important clinical implications. Administration of donor lymphocyte infusions containing high levels of allogeneic CTL precursors (allo-CTLp) in the early posttransplant period correlates with a high risk of GVHD, whereas infusion of antigen-specific T cells with low levels of allo-CTLp never causes GVHD. Using limited dilution analysis, we were able to show that pretreatment of PBMCs with our inhibitor compound prevents activation of alloreactive T cells and results in a 20-fold decrease of allo-CTLp to levels comparable with allo-CTLp levels detected in EBV CTLs that are safely used for adoptive therapy for EBV lymphoproliferative disorders (36).

Two recent studies have validated possible uses of a proteasome inhibitor and an IκB kinase inhibitor to target NF-κB (37, 38). These pan–NF-κB inhibitors have broad effects on the downstream signaling pathways and have significant potential to cause serious adverse effects such as increased radiation-induced epithelial damage (39). However, because c-Rel activity is restricted to a small number of hematopoietic lineages, a c-Rel–specific inhibitor will likely have a better safety profile than a pan–NF-κB inhibitor. Indeed, we have evidence that systemic administration of the c-Rel inhibitor compound discovered by us is feasible, safe, and effective. Furthermore, Rel/NF-κB factors are also known for their roles as proto-oncogenes by contributing to tumor growth, survival, drug resistance, and metastasis of lymphoid malignancies and breast, head, and neck cancers (9, 40). We found in a preliminary experiment that intraperitoneal c-Rel inhibitor compound administration displayed antineoplastic activity in a xenograft model of human diffuse large B-cell lymphoma (data not shown).

In conclusion, our data provide evidence, for the first time, for a differential role of c-Rel for alloactivation versus antigen-specific T-cell activation: although c-Rel activity is critically important for T-cell activation during GVHD, it is dispensable for antigen-specific TCR activation. As a result, inhibition of c-Rel activity reduces the severity of GVHD without compromising antitumor activity of T cells. Our findings validate c-Rel as a highly promising therapeutic target, and we demonstrate biologic benefits of inhibition of c-Rel activity in both mouse and human T cells with a highly specific small-molecule compound. Drug development studies are currently under way in an effort to translate this technology from bench to bedside.

We obtained female C57BL/6 (B6, H-2^b), LP/J (H-2^b), B10.BR (H-2^k), and BALB/c (H-2^d) mice from The Jackson Laboratory. B6 mice carrying the c-Rel gene null mutation (c-Rel^−/−) were originally generated by inserting the neomycin cassette into the fifth exon of the c-Rel gene (7). c-Rel^−/− B6 mice and Foxp3-DTR transgenic B6 mice that expressed the DTR under control of the Foxp3 promoter (27) as well as Pmel1^+/+ transgenic B6 mice (41) were maintained at Memorial Sloan-Kettering Cancer Center (MSKCC; New York, NY) in accordance with the Institutional Animal Care and Use Committee Standards. Mice used for experiments were 6 to 9 weeks old. Mouse HSCT experiments were performed as previously described (42), with 850 cGy split-dosed lethal irradiation of BALB/c recipients transplanted with BM (5 × 10⁶), T cell depleted (TCD) with anti–Thy-1.2 and low-TOX-M rabbit complement (Cedarlane Laboratories), or with 1,100 cGy split-dosed lethal irradiation of B6 or LP recipients transplanted with TCD BM (5 × 10⁶) as well. Donor T cells were prepared by harvesting donor splenocytes and enriching T cells by Miltenyi MACS purification of CD5 (routinely >90% purity). For the Treg transfer experiment, highly enriched (routinely >90% purity) CD4⁺CD25⁺ T cells were obtained by positive selection with the Miltenyi MACS magnetic sorting system. In GVT experiments, animals received tumor cells intravenously in a separate injection on day 0.

Pyrimidinetrione derivatives were previously identified as small-molecule c-Rel inhibitor compounds (20, 21). c-Rel inhibitory activity of the compounds was confirmed by FP as well as EMSA using the DNA-binding property of the c-Rel protein. Detailed methods for these analyses were described previously (7, 20, 43). Unless otherwise indicated, we used IT-603 [(5Z)-5-[(5-bromo-2-hydroxy-3-methoxyphenyl)methylidene]-2-sulfanylideneimidazolidin-4-one; molecular weight = 329.1 g/mol; obtained from ChemDiv] by incubating cells with the compound for 24 hours at a concentration of 20 μmol/L. For high-throughput screening, we used an FP assay that uses the DNA-binding property of the c-Rel protein. Specifically, c-Rel binds with high affinity to the CD28 responsive element (CD28RE) in the IL-2 promoter. The differential signals of free CD28RE versus bound CD28–Rel complex were used to screen for compounds that disrupted CD28–Rel interaction.

Mice were monitored daily for survival and weekly for GVHD clinical scores (44). Small intestine, large intestine, liver, and skin samples were evaluated histologically for evidence of GVHD and scored as previously described (45). In GVT experiments, we determined the bioluminescent signal intensity (BLI) of tumor-bearing mice twice weekly as described previously (46). We superimposed pseudocolor images showing the whole-body distribution of BLI on grayscale photographs and determined total flux (photons s⁻¹) for individual mice. We determined the cause of death (tumor vs. GVHD) by necropsy and histopathology as previously described (45). All cell lines used in our experiments were originally obtained from the American Type Culture Collection (ATCC) in 2006. All cells were maintained and propagated according to the recommendations of ATCC and were validated as Mycoplasma-negative. All cell lines underwent authentication testing at ATCC; in addition, the histologic origin of tumors derived from cancer cell lines in our animal experiments was confirmed by histopathology.

Blood was collected into microcentrifuge tubes, allowed to clot, and centrifuged, and the supernatant was collected. Multiplex ELISA was conducted as per the manufacturer's instructions (Millipore). Results were acquired with a Luminex 200 instrument and analyzed with xPONENT software (Luminex Corporation).

All antibodies other than the anti–c-Rel antibody (Santa Cruz Biotechnology) and the anti-CD44 antibody (BioLegend) were obtained from BD Biosciences–Pharmingen. For cell analysis of surface markers, cells were stained for 20 minutes at 4°C in PBS with 0.5% bovine serum albumin (PBS/BSA) after Fc block, washed, and resuspended in 4′,6-diamidino-2-phenylindole (DAPI) in PBS/BSA. IL-2 secretion analyses were performed using the IL-2 Secretion Assay Kit as per the manufacturer's instructions (Miltenyi Biotec). c-Rel expression was analyzed by intracellular staining after cells were stimulated for 5 hours with Cell Stimulation Cocktail (eBioscience) unless otherwise indicated. Cell surface staining was followed by intracellular staining with the eBioscience kit as per the manufacturer's instructions. Dead cells were excluded with the LIVE/DEAD Fixable Dead Cell Stain kit (Invitrogen). Intracellular pSTAT5 was detected as described previously (24). Briefly, mice were sacrificed and splenocytes were instantaneously exposed to different concentrations of mouse IL-2 for 10 minutes at 37°C, followed by fixation with 1.6% paraformaldehyde (PFA) and permeabilization with 90% methanol. During this process, membrane-bound IL-2 was stripped from cell surfaces by a 2-minute incubation with 0.1 mol/L glycine buffer equilibrated at pH 4.0, followed by a 5-minute wash in RPMI before the exposure to exogenous IL-2. All flow cytometry was performed on an LSR II flow cytometer (BD Biosciences) and analyzed with FlowJo (TreeStar Software).

Plasma samples were analyzed at 30 minutes, 1, 2, 4, 6, and 16 hours after intraperitoneal administration (12 mg/kg) of c-Rel inhibitor compound IT-603. To assess the level of IT-603 in blood, samples were analyzed by liquid chromatography/tandem mass spectrometry (LC/MS-MS) as previously described (47, 48). Calibration curves were determined for IT-603 to permit conversion of peak areas to the drug amounts against external reference standards. The MS-MS detector (Model ABI/Sciex API 4000; Applied Biosystems) permitted verification of peak identity as well as a quantitative assessment of the compounds in the samples.

In vitro cytotoxicity analyses of human CMV-specific T cells as well as human WT1 and human EBV-specific T cells were performed as previously described (30). In brief, CMV-specific T cells were generated from PBMCs of a healthy donor by repeated in vitro stimulations with an autologous EBV-transformed B-cell line (EBV BLCL) loaded with a pool of CMVpp65-derived pentadecapeptides overlapping by 11 amino acids and including the entire sequence of the protein. T cells were tested on day 34 of culture in a standard ⁵¹Cr release assay against a panel of targets including autologous dendritic cells without CMVpp65 peptide, autologous dendritic cells loaded with a CMVpp65 peptide, and fully HLA-mismatched allogeneic BLCL. The effector:target ratio was 25:1. We compared unmodified T cells with control- and compound-treated T cells (pretreatment for 24 hours in media containing IT-603 compound at 20 μmol/L).

The frequencies of EBV-specific T cells and alloreactive T cells were measured in the PBMCs from a normal EBV seropositive donor by limited dilution analysis as previously described (36). Briefly, the aliquots of PBMCs were stimulated on day 0 with either autologous EBV BLCL or allogeneic HLA-mismatched BLCL after coincubation with the inhibitor compound IT-603 or with the control solution (DMSO). The stimulated PBMCs were plated at serial dilutions and cultured in the presence of IL-2 for 14 days. The CTL precursor frequencies were measured in a ⁵¹Cr assay against either autologous BLCL or allogeneic HLA-mismatched BLCL.

The HLA type of the PBMC donor was: A0201 A3601 B 44WRJ B1510 C0501 C04JERF DRB1 0401 DRB1 0301 DQB1 0301 DQB1 0201.

The HLA type of the HLA-mismatched BLCL was: A2601 B 0801 C0701 DRB1 1501 DQB1 0602 (homozygous by all alleles).

Data are presented as mean ± SEM. Survival data were analyzed with the Mantel–Cox log-rank test. For nonsurvival pointwise analyses, an unpaired t test was used for comparisons between two experimental groups, or the nonparametric Mann–Whitney U test was used for non-Gaussian distributions, and ANOVA was used for comparisons of more than two groups. All statistical analyses were performed using GraphPad Prism 5. A P value of less than 0.05 was considered statistically significant.

H.-C. Liou is employed as the CEO of ImmuneTarget, Inc. J.L. Zakrzewski is a consultant/advisory board member of ImmuneTarget, Inc. No potential conflicts of interest were disclosed by the other authors.

This content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH.

Conception and design: Y. Shono, S. Ouk, H.-C. Liou, J.J. Tsai, E. Doubrovina, M.R.M. van den Brink, J.L. Zakrzewski

Development of methodology: Y. Shono, G. Altan-Bonnet, E. Doubrovina, D. Pankov, J.L. Zakrzewski

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): Y. Shono, A.Z. Tuckett, J.E. Oyler, N.V. Singer, E. Doubrovina, C.V. Undhad, G.F. Murphy, C. Lezcano, R.J. O'Reilly, M.R.M. van den Brink, J.L. Zakrzewski

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): Y. Shono, A.Z. Tuckett, G. Altan-Bonnet, J.J. Tsai, J.E. Oyler, E. Doubrovina, D. Pankov, C. Liu, M.R.M. van den Brink, J.L. Zakrzewski

Writing, review, and/or revision of the manuscript: Y. Shono, A.Z. Tuckett, S. Ouk, G. Altan-Bonnet, J.J. Tsai, J.E. Oyler, E. Doubrovina, C. Lezcano, M.R.M. van den Brink, J.L. Zakrzewski

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): Y. Shono, O.M. Smith, M.L. West, N.V. Singer, C.V. Undhad, R.J. O'Reilly, M.R.M. van den Brink, J.L. Zakrzewski

Study supervision: H.-C. Liou, M.R.M. van den Brink, J.L. Zakrzewski

The authors thank the Laboratory of Comparative Pathology, the Analytical Pharmacology Core Facility, and the Molecular Cytology Core Facility (supported by Cancer Support Grant NCI P30-CA008748) of MSKCC for their invaluable help. T. Merghoub and A.Y. Rudensky (MSKCC) provided PMEL1 and FoxP3-DTR mice. The authors also thank E. Velardi, J.A. Dudakov, A.M. Hanash, R.R. Jenq, L.F. Young, A. Ghosh, A.M. Holland, and T. Chinen for helpful discussion.

This work was supported by NIH award numbers R01-HL069929 (to M.R.M. van den Brink), R01-AI100288 (to M.R.M. van den Brink), R01-AI080455 (to M.R.M. van den Brink), R01-AI101406 (to M.R.M. van den Brink), and 1K08CA160659-01 (to J.L. Zakrzewski). Support was also received from the Leukemia Research Foundation (to J.L. Zakrzewski); The Radiation Effects Research Foundation (RERF-NIAID; to M.R.M. van den Brink); The Uehara Memorial Foundation (to Y. Shono); a Lymphoma Research Foundation Post-Doctoral Fellowship Research Grant (to Y. Shono); the MSKCC Center for Molecular Imaging and Nanotechnology (CMINT); The Experimental Therapeutics Center of MSKCC, funded by Mr. William H. Goodwin and Mrs. Alice Goodwin; The Lymphoma Foundation; Alex's Lemonade Stand; The Geoffrey Beene Cancer Research Center at MSKCC, and The Peter Solomon Fund.