Acute graft-versus-host disease (aGVHD) is one major serious complication that is induced by alloreactive donor T cells recognizing host Ags and limits the success of allogeneic hematopoietic stem cell transplantation. In the current studies, we identified a critical role of Kras in regulating alloreactive T cell function during aGVHD. Kras deletion in donor T cells dramatically reduced aGVHD mortality and severity in an MHC-mismatched allogeneic hematopoietic stem cell transplantation mouse model but largely maintained the antitumor capacity. Kras-deficient CD4 and CD8 T cells exhibited impaired TCR-induced activation of the ERK pathway. Kras deficiency altered TCR-induced gene expression profiles, including the reduced expression of various inflammatory cytokines and chemokines. Moreover, Kras deficiency inhibited IL-6–mediated Th17 cell differentiation and impaired IL-6–induced ERK activation and gene expression in CD4 T cells. These findings support Kras as a novel and effective therapeutic target for aGVHD.
This article is featured in Top Reads, p.3235
Allogeneic hematopoietic stem cell transplantation (allo-HSCT) is a potentially curative therapy for patients with hematologic malignancies. However, acute graft-versus-host disease (aGVHD) is one of the most serious complications that cause morbidity and mortality following allo-HSCT (1–3). aGVHD is characterized by systemic inflammation and tissue destruction involving multiple organs, such as the gut, liver, lung, bone marrow (BM), thymus, and skin (1–3). aGVHD is caused by the activation of alloreactive donor T cells that recognize host Ags and attack host tissues through cytotoxicity and inflammatory cytokine production (4). Both CD4 and CD8 donor T cells play important roles in aGVHD pathogenesis (4). To prevent aGVHD, depletion and functional inactivation of donor T cells are the most effective approaches. However, it is important to note that donor T cells are also responsible for the graft-versus-tumor (GVT) effect (3). The first-line therapy for aGVHD is steroid-based global immunosuppression that suppresses T cells nonspecifically, inhibiting aGVHD, but also carries the risk of compromising the GVT effect (5). Thus, novel therapies that reduce aGVHD but maintain the GVT effect are urgently needed following allo-HSCT.
Ras protein is a small GTPase that functions as a central node conveying signals from cell surface receptors to the downstream effector pathways such as the Raf/MEK/ERK cascade (6). Ras is activated by guanine nucleotide exchange factors that catalyze the exchange of GDP for GTP (7). GTP-bound Ras directly interacts with and activates the serine/threonine kinase Raf that in turn activates the threonine/tyrosine kinases MEK1/2 (8–13). Activated MEK1/2 activate the serine/threonine kinases ERK1/2, ultimately leading to upregulation of the transcription factor AP-1 component c-Fos and promoting a wide variety of cellular events (12–15). The Ras/ERK pathway is an important signaling pathway emanating from the pre-TCR and TCR and activates the downstream effectors to control T cell early development and late maturation, respectively (16, 17). There are three highly homologous mammalian Ras members (Kras, Nras, and Hras) that share 85% identity but possess highly variable C-terminal regions (6, 18, 19). All three Ras isoforms are expressed in T cells (20). Constitutively active Ras compensates for the lack of pre-TCR to drive Rag1-deficient double-negative cell expansion and transition to double-positive cells (21). However, Nras or Hras single deficiency or hematopoietic-specific Kras single deficiency does not affect early T cell development (20, 22). Thus, the three Ras members have redundant functions in early T cell development. However, studies have shown that the individual Ras members appear to have different biological function. Nras- or Hras-deficient mice are largely normal, whereas Kras-deficient mice are embryonically lethal (23–25). All three Ras isoforms are activated by TCR engagement (20) and play specific and distinct roles in TCR-mediated biological function. Nras deficiency reduces CD8 T cell numbers, impairs CD8 T cell memory (26, 27), and inhibits IFN-γ production and Th1 response of CD4 T cells (20). Hras deficiency impairs IFN-γ production and Th1 response of CD4 T cells (20). However, because of embryonic lethality of Kras-deficient mice (25), the role of Kras in TCR-mediated T cell function is not known.
In the current studies, we studied the role of Kras in T cell function and found that deletion of Kras in donor T cells markedly reduced aGVHD severity and mortality but preserved the GVT effect in an MHC-mismatched mouse model of allo-HSCT. Kras deficiency reduced alloantigen/APC-induced T cell proliferation and TCR-induced ERK activation and altered TCR-induced gene expression profiles, including the decreased expression of various inflammatory cytokines and chemokines. Kras deficiency also reduced IL-6–mediated Th17 cell differentiation, ERK activation, and gene expression. Our findings demonstrate that targeting Kras is a promising strategy to control aGVHD and maintain the GVT effect following allo-HSCT.
Materials and Methods
Krasfl/fl mice were crossed with VavCre transgenic mice and the mouse line was maintained on C57BL/6 genetic background (>N10) (22). Experimental VavCreKrasfl/fl and control VavCreKrasfl/+ mice were 8–12 wk old. BALB/c (H-2d) and Rag1-deficient mice were from The Jackson Laboratory. The Medical College of Wisconsin’s Institutional Animal Care and Use Committee approved the animal protocols.
Allo-HSCT was performed as described previously (28). CD4 and CD8 T cells were isolated from donor spleens and purified by positive selection with the MACS Cell Separation system (Miltenyi Biotec, Auburn, CA) twice. A total of 1.2 × 106 CD4 T cells and 0.67 × 106 CD8 T cells from VavCreKrasfl/fl or control C57BL/6 donors and 5 × 106 BM cells from Rag1-deficient mice were transferred into lethally irradiated (900 rad) BALB/c recipients through tail vein injection. Recipients were then monitored every day for survival. The degree of aGVHD was assessed by a scoring system that included five clinical parameters: weight loss, posture, activity, fur texture, and skin integrity (29). Mice were graded from 0 to 2 for each criterion, and a graft-versus-host disease (GVHD) score was generated by summation of the five criteria scores.
Four weeks after allo-HSCT, the colons and lungs were collected and fixed in 10% formalin, embedded in paraffin, cut into 5-μm sections and stained with H&E. Colon and lung sections were graded by the pathologists in a blinded fashion (29, 30). In colon sections, epithelial apoptosis, crypt regeneration and erosion, mucosal ulceration, and lamina propria inflammatory cell infiltration were examined. In lung sections, interstitial/alveolar inflammation and periluminal bronchial and vascular lymphocyte infiltration were examined.
The B cell leukemia mouse model was established as previously described (28). Briefly, lethally irradiated (900 rad) BALB/c mice were i.v. injected with 0.6 × 106 B lymphoma cells of BALB/c genetic background with a luciferase reporter gene (A20luc). Five hours later, the mice were transferred with 5 × 106 Rag1-deficient BM cells alone or with Rag1-deficient BM cells plus 1.2 × 106 CD4 T cells and 0.67 × 106 CD8 T cells from VavCreKrasfl/fl or control donors. After cell transfer, the overall survival of the recipients was monitored. In addition, 10, 14, 21, and 30 d after cell transfer, mice were i.p. injected with luciferin (150 mg/kg body weight). After 10 min, mice were anesthetized, and A20luc B cells were tracked using the IVIS (Xenogen) to assess bioluminescence. Imaging data were analyzed with Living Image software (PerkinElmer).
Lamina propria, lung, and liver lymphocytes were isolated as described previously (31). Briefly, lamina propria lymphocytes from the intestine were isolated using the Lamina Propria Dissociation Kit (Miltenyi Biotec) according to the manufacturer’s instructions, followed by Percoll gradient centrifugation. Lungs were digested with collagenase D, and lymphocytes were isolated by Percoll gradient centrifugation. Liver lymphocytes were isolated by Percoll gradient centrifugation. Single-cell suspensions were treated with Gey solution and resuspended in PBS added with 2% BSA. Abs used for flow cytometric analysis were as follows: Percp-Cy5.5-conjugated anti-CD4 and anti-CD62L, allophycocyanin-Cy7-conjugated anti-CD8 and anti-CD44, PE-conjugated anti-H2Kb, anti-CD62L, RORγt and anti-Foxp3, PE-Cy7-conjugated anti–IFN-γ, anti-CD69, anti-CD8, anti-CD25 and anti-CD44, Alexa 647–conjugated anti-TNF-α, FITC-conjugated anti-CD44, anti-CD4 and anti-H2kb, allophycocyanin-conjugated anti–phospho-Erk, anti-H2Kd, and anti–IL-17. Abs were purchased from BD Biosciences or eBioscience. In some experiments, cells were stained with an Aqua dead cell exclusion dye. The Foxp3 Transcription Factor Staining Buffer Set was from eBioscience. Samples were applied to a BD LSR II flow cytometer (Becton Dickinson), and data were collected and analyzed using FACSDiva software (Becton Dickinson).
A total of 1 × 105 CD4 or CD8 T cells (responder cells) purified from VavCreKrasfl/fl and control mice were cocultured with 5 × 105 irradiated (2500 rad) splenocytes (APCs) from BALB/c mice in triplicate in complete RPMI 1640 medium containing 10% FBS in round-bottom 96-well plates for 5 or 3 d, respectively, and then pulsed with [3H]thymidine (1 μCi/well) for 16–18 h. As controls, 2 × 104 CD4 or CD8 T cells were stimulated with soluble anti-CD3 (2 μg/ml) plus anti-CD28 (2 μg/ml) or PMA plus ionomycin for 2 d and then pulsed with [3H]thymidine (1 μCi/well) for 16–18 h. Cells were harvested with a MACH 3 Harvester (Tomtec, Hamden, CT) and [3H]thymidine incorporation was determined with a Wallac MicroBeta TriLux scintillation system (PerkinElmer, Waltham, MA).
Western blot analysis
Isolated CD4+ or CD8+ T cells were resuspended in RPMI 1640 medium with 1% BSA, and stimulated with anti-CD3 (10 μg/ml; clone 500A2; eBioscience) at 37°C for the indicated times. Cell lysates were subjected to Western blot analysis with the indicated Abs. Rabbit polyclonal anti-ERK1/2 (sc-093) and mouse monoclonal anti–phospho-ERK1/2 (pThr202/pTyr204, sc-7383) and anti-ERK2 (sc-1647) Abs were purchased from Santa Cruz Biotechnology. Rabbit polyclonal anti–phospho-Akt (phospho-Thr308; 9275), anti–phospho-JNK (pThr183/pTyr185; 4668), anti–phospho-MEK1/2 (pThr180/pTyr182; 2338), anti–phospho-Stat3 (pTyr705; 9145p), and mouse monoclonal anti–phospho-p38 (pThr180/pTyr182; 9216) Abs were purchased from Cell Signaling Technology. Mouse monoclonal anti–phospho-Raf1 (Ser338; 05-338) and anti-Ras (05-516) Abs were purchased from EMD Millipore. Mouse monoclonal anti–phospho-Stat5 (pTyr694; 30979345) Ab was purchased from Zymed Laboratories. Mouse monoclonal anti-Stat3 (S21320/L3) Abs were purchased from Transduction Laboratories. Antisera against Stat5A has been described previously (32).
Gel mobility shift assay
CD4+ or CD8+ T cells were stimulated with anti-CD3 (10 μg/ml; clone 500A2; eBioscience) or IL-6 (100 ng/ml for CD4+ T cells) at 37°C for the indicated times and lysed in the lysis buffer (20 mM HEPES [pH 7.9], 350 mM NaCl, 1 mM MgCl2, 0.5 mM EDTA, 0.5 mM DTT, 20% glycerol, 1% nonylphenoxypolyethoxyethanol). Cell lysates were incubated with 32P-labeled NF-κB or AP-1 probe (Santa Cruz Biotechnology) for 15 min at room temperature and then resolved on a 4% polyacrylamide gel at 4°C.
Calcium flux analysis
The splenocytes (2 × 106) were loaded with indo-1AM (10 μg/ml) at room temperature for 30 min. Then cells were incubated with FITC–anti-CD4, PE–anti-CD8 and biotin–anti-CD3 (20 μg/ml) for 15 min. Streptavidin (8 μg/ml; Thermo Fisher Scientific) was added to cross-link the TCR. Calcium concentrations were determined in CD4+ or CD8+ T cells by flow cytometry.
In vitro differentiation of CD4 T cells
Naive CD4 T cells isolated from VavCreKrasfl/fl and control mice were activated by plate-bound anti-CD3 (3 μg/ml) and anti-CD28 (3 μg/ml) Abs for 4 d. The combinations of cytokines and Abs for inducing the differentiation of different CD4 Th cell subsets were as follows: for Th17 differentiation, IL-6 (20 ng/ml), TGF-β1 (2 ng/ml), IL-23 (50 ng/ml), and anti–IFN-γ (10 μg/ml) and anti–IL-4 (10 μg/ml) Abs; for T regulatory cell (Treg) differentiation, IL-2 (10 μ/ml) and TGF-β1 (5 ng/ml); for Th1 differentiation, IL-2 (10 μ/ml), IL-12 (2 ng/ml), and anti–IL-4 Abs (10 μg/ml); for Th2 differentiation, IL-2 (10 μ/ml), IL-4 (5 ng/ml), and anti–IFN-γ (10 μg/ml) and anti–IL-12 (10 μg/ml) Abs. For Th1 and Th2 cell differentiation, irriadiated (2500 rad) splenocytes were also used as APCs. Subsequently, the cells were restimulated with PMA plus ionomycin in the presence of monensin for 5 h followed by intracellular staining and flow cytometry analysis.
RNA-sequencing and data analysis
Following allo-HSCT, recipients were sacrificed on day 7, and donor CD4 T cells and CD8 T cells were purified from the spleens by flow cytometry. Cells were then restimulated with soluble anti-CD3 (10 μg/ml) for 6 h at 37°C. The CD4 T cells were also sorted from VavCreKrasfl/fl and control mice and then stimulated with IL-6 (200 ng/ml) for 6 h. Total RNA was prepared using TRIzol LS Reagent (LifeTechnologies), and mRNA was purified from 10 to 50 ng of total RNA by using the NEBNext Poly(A) mRNA Magnetic Isolate Module (New England Biolabs) and converted into libraries using the NEBNext Ultra RNA Library Prep Kit for Illumina (New England Biolabs). Libraries were quantified by a Qubit Fluorometer (Thermo Fisher Scientific) and a KAPA Library Quantification Kit (Kapa Biosystems), and their average size was estimated by a D1000 ScreenTape system (Agilent Technologies) with region selection from 150 to 900 bp. Equal moles of libraries were mixed, and a total of 1.7 pmoles of libraries were sequenced on Illumina NextSeq 500 with a NextSeq 500/550 (v2) Kit.
Raw RNA-sequencing (RNA-seq) reads were demultiplexed using the Illumina BaseSpace Sequence Hub and then aligned to the mouse reference genome Mus musculus/mm10 (RefSeq) using the aligner STAR. Read counts were summarized for each gene using featureCounts (PubMed Unique Identifier: 24227677), and libraries were normalized using DESeq2 (33). Differentially expressed genes with a false discovery rate ≤0.05 were discovered by comparing the Kras-deficient and control samples using Wald tests within DESeq2. Shrunken log2 fold change values were used to conduct preranked gene set enrichment analysis (GSEA) using fgsea (https://bioconductor.org/packages/release/bioc/html/fgsea.html). The analysis was restricted to the following from the Gene Ontology: ERK1 and ERK2 cascade, JNK cascade, and MAPK cascade. Potential NF-κB and AP-1 target genes were also derived from the RegNetwork database and tested with those previously listed in the Gene Ontology (PubMed Unique Identifier: 26424082). The gene sets for IL-6–inducible and –suppressing genes were created with the top 300 genes that were upregulated or downregulated upon IL-6 stimulation for 6 h in wild-type CD4 T cells (Supplemental Fig. 1A). For both differential gene expression and GSEA analyses, p values were adjusted for multiple testing using the Benjamini-Hochberg method with an adjusted p value significance threshold of 0.05.
Analysis of animal survival and aGVHD score data was performed using Prism software (GraphPad Software, La Jolla, CA). Animal survival and aGVHD score comparisons were performed using log-rank test and two-way ANOVA test, respectively. All the other statistical analysis was performed with the two-tailed unpaired Student t test. The p values <0.05 were considered significant.
Kras-deficient donor T cells largely reduce aGVHD
Kras deficiency results in embryonic lethality (25). We crossed Kras “floxed” mice with VavCre transgenic mice, in which Cre expression mediates deletion of the floxed gene throughout the hematopoietic compartment (34). VavCreKrasfl/fl mice display a complete deletion of Kras in BM cells (22). Although Kras is expressed in both CD4 and CD8 T cells, its deficiency has no effect on T cell development (22). In this study, we studied the potential role of Kras in T cell function. We examined the response of Kras-deficient T cells to alloantigens in C57BL/6 into BALB/c aGVHD model. Rag1-deficient BM cells alone or plus T cells from VavCreKrasfl/fl or control mice were i.v. transferred into lethally irradiated BALB/c mice. Mice that received Rag1-deficient BM alone were free of aGVHD and survived well, whereas mice transferred with Rag1-deficient BM plus control T cells all died rapidly (Fig. 1A). Of note, 70% of recipients transferred with Rag1-deficient BM plus Kras-deficient T cells survived (Fig. 1A). Consistent with the survival data, the assessment of clinical aGVHD scores showed a marked reduction of aGVHD severity in the recipients that received Kras-deficient relative to control T cells (Fig. 1B). Pathologic evaluation of aGVHD target organs found markedly reduced damages, such as epithelial cell apoptosis and crypt destruction, erosion and ulceration in the colons, and interstitial inflammation and luminal lymphocyte infiltration in the lungs, in the recipients that received Kras-deficient T cells relative to those of the recipients that received wild-type T cells (Fig. 1C, 1D). These data demonstrate that Kras-deficient T cells markedly reduce aGVHD.
Kras-deficient donor T cells preserve the GVT effect
Kras deficiency in donor T cells inhibits aGVHD; however, it is important to know whether Kras-deficient T cells maintain GVT abilities. Lethally irradiated BALB/c recipients were inoculated with A20luc B-lineage lymphoma cells. Afterwards, Rag1-deficient BM cells alone or plus Kras-deficient or control T cells were i.v. transferred into recipients. The recipients that received A20luc cells and Rag1-deficient BM without T cells died rapidly after cell transfer because of fast lymphoma development detected by bioluminescence imaging (Fig. 2). In contrast, the recipients that received A20luc cells and Rag1-deficient BM plus Kras-deficient or control T cells had no detectable tumor (Fig. 2A). However, the majority of the recipients that received control donor T cells died within 50 d after cell transfer because of aGVHD (Fig. 2B). Importantly, almost all of the recipients that received Kras-deficient donor T cells survived well with limited signs of aGVHD (Fig. 2). Taken together, Kras-deficient T cells can efficiently control lymphoma growth and markedly reduce aGVHD.
Kras deficiency reduces TCR-induced expression of a subset of inflammatory cytokines and chemokines
Kras is expressed in both CD4 and CD8 T cells, but its deficiency has no effect on T cell development (22). We found that Kras was activated following TCR ligation in both primary CD4 and CD8 T cells (Fig. 3A). To gain insight into the specific downstream target genes that were involved in the reduction of aGVHD, TCR-induced transcriptional profiling of wild-type and Kras-deficient donor CD4 and CD8 T cells from the spleens of the irradiated BALB/c recipients was examined through high-throughput RNA-seq analysis. Differential gene expression analysis results demonstrated that Kras-deficient CD4 T cells displayed a change of the gene expression profile compared with corresponding wild-type T cells (Fig. 3B). Specifically, 146 genes were upregulated and 51 genes were downregulated in Kras-deficient relative to control CD4+ T cells (adjusted p value <0.05) (Fig. 3B). Genes with reduced expression within the Kras-deficient CD4 T cells were associated with inflammatory cytokine and chemokine genes, such as Tnf, Ltb, Il1b, Il21, Xcl1, and Ccl6 (Fig. 3C). In addition, the downregulated genes in Kras-deficient CD4 T cells included many cytokine and chemokine/chemokine receptor genes closely associated with GVHD (Fig. 3C). Of note, the expression of cytotoxic effector molecules, such as GzmA, GzmB, GzmK, Fas, and Fasl, was not decreased in Kras-deficient relative to control CD4 T cells (Fig. 3C). GSEA found that the expression of genes associated with the activation of the Erk1/2 cascade, MAPK cascade, JNK cascade, and AP-1 targets but not NF-κB targets was significantly enriched in control relative to Kras-deficient CD4 T cells (Fig. 3D, Supplemental Fig. 1B). In addition, 717 or 1169 genes were reduced or increased in Kras-deficient CD8 T cells compared with control cells, respectively (Fig. 3E). The expression of various inflammatory cytokine, chemokine, and chemokine receptor genes, including Cxcl10, Ccr6, Lta, Ccl3, Ccl4, Tnfsf14, Ltb, etc., was reduced in Kras-deficient relative to control CD8 T cells (Fig. 3F). Again, the expression of cytotoxic effector molecules, such as GzmA, GzmB, GzmK and Fasl, was not decreased in Kras-deficient relative to control CD8 T cells (Fig. 3F). GSEA also found that the expression of genes associated with the Erk1/2 cascade, MAPK cascade, JNK cascade, and AP-1 targets but not NF-κB targets was reduced in Kras-deficient CD8 T cells compared with corresponding control T cells (Fig. 3G, Supplemental Fig. 1B). The expression of c-fos itself, the target of AP-1, was reduced in Kras-deficient relative to control CD8 T cells (log2 fold change = −0.72; adjusted p value <0.01). Therefore, Kras deficiency alters TCR-induced gene expression profiles, including reduced expression of a subset of inflammatory cytokines and chemokines but not cytotoxic effector molecules in both CD4 and CD8 T cells. Such alteration might be responsible for markedly reduced aGVHD but preserved GVT capabilities.
Kras deficiency impairs alloantigen/APC-induced proliferation of T cells
We next examined the response of Kras-deficient T cells to alloantigens in the MLR. CD4 and CD8 T cells isolated from VavCreKrasfl/fl or control mice were stimulated with splenocytes from BALB/c mice. The proliferation of both Kras-deficient CD4 and CD8 T cells was markedly reduced compared with that of the corresponding control T cells (Fig. 4A). Of note, Kras-deficient CD4 and CD8 T cells proliferated normally in response to PMA plus ionomycin stimulation compared with the corresponding control T cells (Fig. 4A). In addition, Kras-deficient CD4 and CD8 T cells proliferated normally in response to anti-CD3 plus anti-CD28 stimulation (Fig. 4A). Therefore, Kras-deficient CD4 and CD8 T cells display impaired alloantigen-driven proliferation by APC stimulation under physiological condition but exhibit largely normal response to strong stimulation by anti-CD3/CD28 Abs or PMA/ionomycin.
Kras deficiency impairs inflammatory cytokine production by T cells
To confirm that Kras deficiency alters TCR-induced expression of a subset of inflammatory cytokines in T cells (Fig. 3), we examined the effect of Kras deficiency on cytokine production of in both CD4 and CD8 T cells. CD4 and CD8 T cells from VavCreKrasfl/fl or control mice, along with Rag1-deficient BM cells, were transferred into lethally irradiated BALB/c recipients. At day 7 after cell transfer, Kras-deficient, relative to control donor CD4 T cells from the spleens of the recipients had a marked reduction in IFN-γ and IL-17a production following in vitro TCR engagement by anti-CD3 plus anti-CD28 stimulation (Fig. 4B, 4C). Kras-deficient donor CD4+ T cells also exhibited reduced TNF-α production, although the reduction did not reach statistical significance (Fig. 4D). Compared with control cells, Kras-deficient donor CD8 T cells displayed a marked reduction in IFN-γ production following in vitro TCR engagement (Fig. 4E). Therefore, Kras-deficient CD4 and CD8 T cells have decreased abilities to produce inflammatory cytokines.
In addition, compared with corresponding control T cells, Kras-deficient donor CD4 and CD8 T cells in the lungs of the recipients 14 d after transfer exhibited marked reduction in IFN-γ and TNF-α production (Fig. 5A). Kras-deficient donor CD4 T cells in the livers of the recipients also exhibited reduction tendency in IFN-γ and TNF-α production (Fig. 5B). Lastly, Kras-deficient donor CD4+ T cells in the guts of the recipients 28 d after cell transfer exhibited significant reduction in the production of IFN-γ (Fig. 5C). Therefore, Kras deficiency impairs the abilities of CD4 and CD8 T cells to produce inflammatory cytokines in vivo during aGVHD.
Kras deficiency impairs TCR-induced ERK activation
All three Ras isoforms, Kras, Nras, and Hras, are activated by TCR engagement and mainly control the activation of the ERK pathway (6, 20). We further examined whether Kras single deficiency could alter TCR signal transduction pathways. We first examined the effect of Kras deficiency on TCR signaling in naive T cells. Anti-CD3–induced Ca2+ flux and phosphorylation of p38 and JNK were comparable between Kras-deficient CD4 or CD8 T cells and corresponding control cells (Fig. 6A–C). In contrast, anti-CD3–induced ERK1/2 activation was impaired in Kras-deficient relative to control CD8 T cells, although the impairment was less obvious in Kras-deficient CD4 T cells (Fig. 6D, 6E). Anti-CD3–induced activation of Raf-1 and MEK1/2, the upstream kinases of ERK1/2, was markedly reduced in Kras-deficient CD8 but not CD4 T cells (Fig. 6D, 6E). Consistently, anti-CD3–induced activation of AP-1, the downstream transcription factor target of the Ras/ERK pathway, was normal in mutant CD4 T cells but markedly reduced in mutant CD8 T cells (Fig. 6F). Of note, anti-CD3–induced activation of NF-κB was normal in both Kras-deficient relative to control CD4 and CD8 T cells (Fig. 6F). Interestingly, the impairment of TCR-induced ERK activation in CD4 T cells was detected during aGVHD. At day 14 after cell transfer, donor splenic CD4 T cells isolated from the lethally irradiated BALB/c recipients that received Kras-deficient or control T cells along with Rag1-deficient BM cells were stimulated with anti-CD3. Kras-deficient relative to control CD4 donor T cells exhibited a marked reduction in TCR-induced ERK activation (Fig. 6G, 6H). Of note, anti-CD8 Ab staining did not work well after cell fixation for intracellular staining of phospho-ERK. Although decreased ERK activation in CD8 T cells in vivo could not be directly detected by intracellular staining, the impairment of ERK pathway activation was clearly demonstrated in CD8 T cells during GVHD reaction by the GSEA analysis (Fig. 3G, Supplemental Fig. 1B) and in vitro following TCR ligation (Fig. 6D–F). Thus, although all three Ras isoforms are expressed in CD4 and CD8 T cells, Kras single deficiency impairs TCR-induced activation of the Ras/ERK pathway in both CD4 and CD8 T cells.
Kras deficiency severely impairs IL-6–dependent Th17 differentiation
In addition to the TCR, cytokine receptors direct CD4 T cell function/differentiation and activate the Ras pathway (35, 36). We further examined whether Kras deficiency impaired cytokine-mediated CD4 T cell function/differentiation. Isolated naive Kras-deficient and control CD4+ T cells were cultured in vitro under the conditions for different subset T cell differentiation. The IL-12–driven differentiation of IFN-γ–producing Th1 cells and the IL-2–driven differentiation of CD25+Foxp3+ Treg cells were normal, whereas the IL-4–driven differentiation of IL-4–producing Th2 cells was slightly increased in Kras-deficient relative to control CD4+ T cells (Fig. 7A–C). We also examined the reconstitution of donor Treg cells in the recipients following allo-HSCT. The numbers of donor Tregs in the spleens were comparable between the recipients that received Kras-deficient or control donor T cells (Supplemental Fig. 2A, 2B). In addition, the in vitro suppressive assay showed that Tregs from Kras-deficient or control mice displayed similar suppressive function (Supplemental Fig. 2C). Thus, Kras deficiency has no impact on Treg reconstitution and function. In contrast, the differentiation of IL-17A–producing Th17 cells was markedly reduced in Kras-deficient relative to control CD4+ T cells (Fig. 7D). Consistently, the protein level of RORγt, the key transcription factor of Th17 cell differentiation, was markedly reduced in Kras-deficient relative to control CD4+ T cells (Supplemental Fig. 3). Thus, Kras deficiency specifically impairs Th17 cell differentiation.
Kras deficiency impairs IL-6–induced ERK activation and gene expression
IL-6 is one critical driver of Th17 cell differentiation, and IL-6 activates Kras (37, 38). We thus examined the effect of Kras deficiency on IL-6 signaling in CD4 T cells. IL-6 slightly induced ERK1/2 phosphorylation in control but not Kras-deficient CD4 T cells (Supplemental Fig. 4A). Furthermore, IL-6–induced activation of AP-1, the downstream ERK target, was markedly reduced in Kras-deficient relative to control CD4 cells (Fig. 7E). In contrast, Kras deficiency did not affect IL-6–induced Stat3 activation (Supplemental Fig. 4B). In addition, we examined the impact of Kras deficiency on IL-2 signaling, which induces Treg but inhibits Th17 differentiation. IL-2–induced activation of Stat5 and ERK1/2 in Kras-deficient CD4 T cells was comparable to that in control CD4 T cells (Supplemental Fig. 4C). Thus, Kras deficiency specifically impairs IL-6–induced ERK activation.
Moreover, we examined the effect of Kras deficiency on IL-6 signaling at the transcriptome levels. Upon IL-6 stimulation, the gene expression profile in Kras-deficient CD4+ T cells was largely different from that in control T cells (Fig. 7F). Specifically, 956 genes were upregulated and 148 genes were downregulated in Kras-deficient relative to control CD4+ T cells (adjusted p value <0.05; Fig. 7F). GSEA showed that the IL-6–induced genes were significantly enriched in control relative to Kras-deficient CD4+ T cells, whereas, the IL-6–suppressed genes were enriched in Kras-deficient relative to control CD4+ T cells (Fig. 7G). Therefore, Kras deficiency clearly impairs IL-6–induced signaling and gene expression in CD4 T cells. In summary, targeting Kras in donor T cells subtly alters TCR-induced ERK pathway activation, specifically reduces the expression of various inflammatory cytokines and chemokines, and inhibits IL-6–mediated gene expression, consequently reducing aGVHD without affecting the antitumor capacity.
Donor T cells are the major pathogenic cells that are responsible for causing aGVHD. Once activated by host APCs, donor T cells undergo differentiation, proliferation, and acquisition of effector functions, and mediate target tissue damage. Targeting signaling pathways that are critically involved in donor T cell activation has been an area of active investigation. In this study, we examined the role of Kras in regulating T cell function in the context of aGVHD. We found that genetic ablation of Kras in donor T cells markedly reduced their ability to cause lethal aGVHD in an MHC-mismatched (C57BL/6 into BALB/c) allo-HSCT model. Importantly, the graft-versus-leukemia effect was largely preserved in the recipients of Kras-deficient donor T cells. Mechanistically, Kras deficiency reduced TCR-induced cell proliferation and ERK activation. RNA-Seq analysis revealed that Kras deficiency subtly altered TCR-induced gene expression profiles, especially decreasing the expression of a subset of inflammatory cytokines and chemokines. In addition, Kras deficiency impaired IL-6–induced ERK activation and the expression of the downstream gene RORγt, thus inhibiting the differentiation of Th17 cells, a specific subset of CD4 Th cells that can drive the initiation of cytotoxic T cell–mediated tissue damage and control the severity of aGVHD. Thus, our studies demonstrate that Kras signaling in donor T cells plays an important role in regulating GVHD and the graft-versus-leukemia response after allo-HSCT.
The Ras/ERK pathway is an important signaling pathway emanating from the pre-TCR and TCR and activates the downstream effectors to control T cell early development and late maturation, respectively (16, 17). Nras, Hras, and Kras have redundant functions in early T cell development (20, 22), whereas the individual Ras members appear to have distinct roles in TCR-mediated biological function (20, 26, 27). We found that Kras played an important and nonredundant role in T cell function, especially TCR-mediated response of donor CD4 and CD8 T cells to alloantigens. Clearly, the TCR uses different members of the Ras subfamily to regulate distinct functions in different subsets of mature T cells. Kras deficiency dramatically reduced aGVHD mortality and severity but largely preserved the antitumor capacity. Therefore, specifically targeting Kras or its downstream pathway could be a novel treatment that can effectively prevent aGVHD while maintaining antitumor function. Our findings indicate that using small interfering RNA or inhibitors to specifically block Kras as a therapy for preventing aGVHD following allo-HSCT warrants further investigation.
Although the three Ras members have distinct functional roles in mature T cells, these Ras members might also have an additional effect on TCR-mediated signaling and function. In support of this concept, Kras deletion in mice results in embryonic lethality, whereas compound deficiency of Kras plus Nras or Hras intensifies the embryonic lethality phenotype (25, 39). The expression of the Hras transgene or replacement of Kras with Hras is able to rescue embryonic lethality caused by Kras deficiency (39, 40). The Ras/MEK/ERK pathway is activated in alloreactive T cells during GVHD (41). Complete block or dramatic reduction of TCR-induced activation of the Ras/ERK pathway would decrease aGVHD severity but also could lose the antitumor effect. In fact, complete inactivation of the Ras/ERK pathway by ERK1/2 double deficiency impedes T cell development and blocks T cell proliferation induced by strong stimulation such as anti-CD3/PMA (42). Thus, fine manipulation of the Ras/ERK pathway might achieve the goal of preventing aGVHD while preserving antitumor activity. Indeed, pharmacologic inhibition of MEK preferentially suppresses alloreactive human T cells while sparing pathogen-specific T cells (43). In addition, MEK inhibition shows a beneficial effect on GVHD in murine models of allo-HSCT (43, 44). In line with this concept, we found that Kras deficiency subtly reduced the strength of TCR-induced ERK activation, which had no effect on T cell development but sufficiently inhibited the alloimmune response and reduced the ability of donor T cells to cause lethal aGVHD. This is consistent with our in vitro data, in which we found the proliferation of alloreactive Kras-deficient CD4 and CD8 T cells were significantly reduced in MLRs. However, strong T cell activation signals through anti-CD3/CD28 appeared to overcome the proliferative defect associated with Ras deficiency. Moreover, IL-6 is able to enhance T cell proliferation in MLRs (45), and impairment of IL-6 signaling by Kras deficiency could also contribute to the reduced T cell proliferation in MLRs. IL-6 is one critical cytokine for driving the differentiation of Th17 cells, a subset of CD4 Th cells that regulate the severity of aGVHD (46). Inhibiting IL-6 signaling has been shown to effectively mitigate GVHD in experimental hematopoietic stem cell transplantation (47, 48). More importantly, the results from recent clinical trials using tocilizumab, a humanized anti–IL-6R Ab, are very encouraging (49–51). Thus, suppressing IL-6–mediated pathogenic function of donor T cells by Kras deficiency might also contribute to the reduction of aGVHD severity and the preservation of antitumor effect.
The Ras pathway controls the activation of ERK, JNK, and p38, which ultimately leads to the upregulation of the transcription factor AP-1 (52, 53). AP-1 consists of dimers with different Fos and Jun protein family members and regulates expression of numerous genes and promotes a wide variety of cellular events (12–15, 54, 55). ERK phosphorylates and thus activates the transcription factor Elk that upregulates Fos expression (56, 57). JNK directly phosphorylates c-Jun, and the phosphorylation increases the transcriptional activity of c-Jun (58). We found Kras deficiency reduced ERK but not JNK or p38 activation. Consistently, our RNA-seq analysis demonstrated that Kras deficiency altered the expression of a relatively small number of AP-1 target genes. It has been shown that increasing TCR signal strength gradually activates AP-1 binding sites and positively correlates with AP-1 target gene expression output (59). At least, the epigenetic landscape can govern differential activation of enhancers with AP-1 binding sites and determine the sensitivities of these enhancers to the different levels of AP-1 activation (59). We found a significant reduction in the production of a small group of inflammatory cytokines and chemokines, including TNF-α, IL-17, and IFN-γ, in the spleen as well as GVHD target organs of the recipients, received Kras-deficient T cells. These inflammatory cytokine and chemokine genes might possess enhancers with a pre-established epigenetic landscape requiring high levels of AP-1 for their activation. A subtle reduction of AP-1 caused by Kras deficiency is able to reduce the expression of these inflammatory cytokines and chemokines.
Inflammatory cytokines are critically involved in mediating target organ damage during aGVHD. However, they may also participate in mediating the GVT response. Our results indicated that a significantly reduced cytokine storm characteristic of aGVHD might be responsible for aGVHD protection. Importantly, by employing a widely used A20 mouse lymphoma model, we were able to show that the GVT effect was preserved in the Kras-deficient T cell recipients. It is possible that Kras deficiency suppresses aGVHD–associated extensive donor T cell proliferation and cytokine production to a degree that is sufficient to control aGVHD but still allows for the induction of the GVT effect. In addition, the expression of the cytotoxic effector molecules that mediates tumor cell killing was intact in Kras-deficient T cells, which may also contribute to A-20 cell eradication.
Finally, our results indicated that inhibiting the Kras pathway may be an effective strategy to uncouple the GVT response from aGVHD. Given the availability of small molecule inhibitors for constitutively active Kras and/or its downstream signaling molecules and ongoing clinical trials using these inhibitors for cancer therapy, the translational potential of our research is very high. It is our expectation that inhibition of Kras will lead to a significantly decreased activation, proliferation, and cytokine production of donor T cells without affecting their cytotoxic properties. Pharmacologically targeting the Kras pathway represents a promising strategy to mitigate aGVHD without compromising the GVT effect following allo-HSCT.
This work was supported in part by National Institutes of Health (NIH), National Institute of Allergy and Infectious Diseases Grant AI079087 (to D.W.), NIH, National Heart, Lung, and Blood Institute Grants HL130724 (to D.W.), HL148120 (to R.W.), and HL126166 (to W.R.D.), and NIH, National Cancer Institute Grant CA152108 (to J.Z.).
The online version of this article contains supplemental material.
Abbreviations used in this article:
acute graft-versus-host disease
allogeneic hematopoietic stem cell transplantation
gene set enrichment analysis
T regulatory cell.
The authors have no financial conflicts of interest.