STING regulates GvHD through CD8 and CD4 T cell subsets

Graft-versus-host disease (GvHD) remains a major cause of morbidity and mortality in patients that have received allogeneic hematopoietic stem cell transplant (allo-HSCT). Conditioning with pretransplant chemotherapy or irradiation causes widespread cell death, release of endogenous danger signals, and epithelial barrier dysfunction leading to bacterial translocation, all of which promote the induction of a proinflammatory cytokine storm. These signals drive the activation of antigen-presenting cells (APCs) and the differentiation of alloreactive donor T cells, which cause damage to various host tissues, a characteristic of GvHD.¹

Stimulator of interferon genes (STING) is an innate immune sensor that can contribute to the inflammatory response and was recently reported to protect against gastrointestinal GvHD after major histocompatibility complex (MHC)-mismatched allo-HSCT. However, its role in MHC-matched GvHD remains unexplored. Therefore, in in vivo/preclinical experiments, Cameron S. Bader and colleagues investigated whether STING could influence CD8- versus CD4-mediated GvHD and whether STING-dependent effects were related to the donor/recipient MHC-matched versus MHC-mismatched genetic disparity. The results were published in the journal Science Translation Medicine.¹ 

Methods

• Various MHC-matched (H2^b) donor/recipient strain combinations received allo-HSCT:
  • Recipient mice: C57BL/6J wild type (WT), C57BL/6J STING^−/− (STING-deficient), B6N-WT, or B6N-STING^HAQ/HAQ mice
  • Donor mice: Bone marrow (BM), and T cells from either LP/J (unseparated splenocytes) or C3H.SW (unseparated and pooled splenocytes and peripheral lymph node cells)
  • Hematopoietic chimeras between WT and STING-deficient mice
• Mismatched allo-HSCT was performed using transplanted BALB/c BM and unseparated lymph node cells into C57BL/6J -WT or -STING-deficient recipients, or transgenic mice.
• GvHD was assessed by
  • survival, weight loss, clinical scoring, flow cytometry, and histopathology;
  • frequencies of donor T cells expressing an effector memory or naïve phenotype; and
  • quantification of tissue RNA and serum cytokines.

Results

MHC-matched allo-HSCT

• STING-deficient recipients transplanted with MHC-matched BM and T cells had markedly decreased weight loss and a reduction in clinical GvHD score compared with WT recipients.
  • The difference in clinical score was maintained for 6–9 weeks.  
• At 5 weeks posttransplant, thymus evaluation showed that STING-deficient recipients had greater overall cellularity and a higher frequency of donor-derived CD4⁺CD8⁺ T cells compared with WT recipients.
• At 6 weeks posttransplant, STING-deficient recipients had
  • elevated inhibitory serum cytokine levels of IL-27 and IL-10 compared with WT recipients (p < 0.05);
  • a greater cellularity, a lower percentage of donor T cells expressing an activated effector memory phenotype (CD44^hiCD62L^lo), and a higher frequency of donor T cells with a naïve phenotype (CD44^loCD62L^hi) in peripheral lymphoid tissues compared with WT recipients (p < 0.001); and
  • reduced cellular skin infiltrate, thickening, and overall pathology scores of the ear skin compared with WT recipients (pathology score, p < 0.05).
• Pretransplant cohousing experiments revealed that any differences in microbiome were not likely to account for amelioration of GvHD in STING-deficient recipients.
• Examination of the colon tissue showed
  • reduced expression of the proinflammatory cytokines Ifnb1, Tnf, and Il6 in STING-deficient vs WT recipients 48 hours posttransplant (Table 1);
  • reduced colon pathology scores and number of donor T cells in the lamina propria 10 days posttransplant in STING-deficient vs WT recipients (pathology score, p < 0.05); and
  • an increased frequency of CD11b⁺ CD11c⁺K^b+I-A^b+ and CD11c⁺K^b+I-A^b+ APCs 10 days posttransplant in STING-deficient vs WT recipients, with reduced MHC class I expression.

Table 1. mRNA expression of proinflammatory cytokines in the colon of STING-deficent vs WT mice¹

Inflammatory cytokine	C3H.SW donor p value	LP/J donor p value
Ifnb1, interferon beta 1; Il6, interleukin 6; Tnf, tumor necrosis factor.
Ifnb1	< 0.01	< 0.01
Tnf	< 0.01	0.077
Il6	< 0.05	< 0.01

• Hematopoietic chimeric mice, in which the bone marrow was swapped between WT and STING deficient mice, that subsequently underwent MHC-matched allo-HSCT showed that mice with STING deficiency in nonhematopoietic cells had
  • decreased clinical scores of GvHD regardless of STING expression in the hematopoietic compartment (p < 0.05);
  • lower frequencies of activated donor CD4⁺ (p < 0.01) and CD8⁺ (p < 0.01) T cells expressed in lymphoid tissues 8 weeks after transplant;
  • reduced inflammation and focal ulceration of the skin (p < 0.05); and
  • reduced expression of Ifnb1, Tnf, and Il6.
• WT mice treated with STING agonist 5,6-dimethylxanthenone-4-acetic acid (DMXAA) developed worse GvHD compared with untreated controls (p < 0.01), while DMXAA treatment had no effect on STING-deficient mice.
• Transplantation of highly purified CD8⁺ donor T cells (C3H.SW donors) resulted in reduced GvHD clinical score (p < 0.01 from Day 14) and lower activated donor T cells (p < 0.05) in STING-deficient vs WT recipients.
• B6N-STING^HAQ/HAQ knock-in mice (with reduced STING activity) showed relatively lower GvHD clinical scores and a lower frequency of activated memory-type donor T cells compared with WT controls after allo-HSCT (p < 0.05). However, reduction in GvHD score was less compared with STING-deficient mice.

MHC-mismatched allo-HSCT

• In contrast to matched allo-HSCT, MHC-mismatched allo-HSCT using BALB/c BM and T cells showed increased GvHD clinical scores (p < 0.01 between Days 7–14) and decreased survival (p < 0.01) in STING-deficient recipients vs WTs.
• Similarly, transplantation of BALB/c BM and CD4⁺ donor T cells only resulted in no difference in GvHD between WT and STING-deficient recipients, as assessed by weight loss, clinical score, survival, and donor T cell phenotype.
• However, transplantation of highly purified donor CD8⁺ T cells (BALB/c donor) also reduced GvHD, as assessed by clinical score (p < 0.001), donor T cell phenotype (p < 0.05), and skin pathology (p < 0.01) in STING-deficient recipients vs WT.
• Transplantation of C57BL/6J-WT BM with or without unseparated lymph node cells (1.2 × 10⁶ T cells) into BALB/c-STING-deficient versus BALB/c-WT recipients resulted in
  • increased recipient splenic APCs (CD11b⁺CD11c⁺K^d+I-A^d+) with reduced MHC class I expression (p < 0.01);
  • higher recipient splenic CD11b⁺K^d+ cells (p < 0.05);
  • more CD11c⁺Kd⁺ cells in the colon lamina propria; and
  • lower numbers of splenic donor CD8⁺ T cells (p < 0.05).
• Allo-HSCT of BM from Nur77^GFP transgenic mice (GFP is expressed only in T cells activated via the TCR but not by inflammatory cytokines) showed a greater frequency and number of GFP positive donor CD4⁺ T cells (p < 0.001 and p < 0.01, respectively) in STING-deficient vs WT recipients 6 days posttransplant.

Conclusion

STING deficiency in the nonhematopoietic compartment of the recipient reduces GvHD after MHC-matched allo-HSCT by downregulation of MHC class I expression and reduced donor CD8⁺ activation. Conversely, in MHC-mismatched allo-HSCT, which is induced by donor CD4⁺ cells, deficiency of STING does not result in decreased GvHD. This indicates that STING deficiency reduces donor CD8⁺ T cell-mediated GvHD (as seen in matched allo-HSCT), while it has no impact on CD4⁺ T cell-mediated GvHD response (as seen in mismatched allo-HSCT). Therefore, the authors suggest that targeting the STING pathway in a clinical setting of matched allo-HSCT could potentially improve the outcomes of patients.