STING pathways and transplantation

Stimulator of interferon genes (STING) acts as a sensor for double stranded (ds)DNA from viral or bacterial sources in the cytoplasm. STING recently sparked interest as it was shown to play a role in the regulation of graft-versus-host disease (GvHD) in mice. How STING regulates CD8⁺ and CD4⁺ T-cell subsets was reported on by the GvHD Hub, and this article can be found here.

This article provides a summary of a review published in Blood by Cameron Bader and colleagues, which discusses how targeting STING can improve outcomes for transplant patients.¹

The STING signaling pathway

STING responds to the presence of cyclic dinucleotides (CDNs) which are produced by cyclic GMP-AMP synthase. Upon binding to the CDN, STING recruits TBK1 and moves to perinuclear areas. The transcription factors interferon regulatory factor 3 (IRF3) and NF-κB are activated as a result and induce the production of inflammatory molecules such as type 1 interferons (T1-IFNs).

Relationship between STING signaling and transplant outcome

Murine studies using STING knockout mice (B6-STING^−/−) reported more severe GvHD in this model after major histocompatibility complex (MHC)-mismatched allogeneic hematopoietic stem cell transplantation (allo-HSCT). However, in a matched unrelated donor model, STING deficiency resulted in an improved prognosis compared with wild-type mice.

Different T-cell subsets tend to predominate in the two transplant categories with CD4⁺ being increased in MHC-mismatched transplants compared with CD8⁺ in MHC-matched transplants. Notably, when only CD8⁺ cells were transplanted, STING deficiency was shown to be protective against GvHD development in either mismatch or matched SCT recipients.

Nonhematopoietic cells

Nonhematopoietic cells have a role in the initiation of gastrointestinal GvHD; however, which cell types are specifically important is unknown. Stimulation of STING in nonhematopoietic cells has been linked to autoimmune disease initiation involving activated T cells. Bader et al. hypothesized that STING activation in nonhematopoietic antigen presenting cells (APCs) led to inflammation that targeted distinct parenchymal cell groups compared with STING activation in hematopoietic cells. Production of cytokines/chemokines by nonhematopoietic APCs including downstream T1-IFNs may lead to the movement of hematopoietic APCs into lymphoid tissues, thereby increasing alloreactive T-cell responses. Following HSCT, nonhematopoietic APCs may activate resident memory cells by CD8⁺ effector cells.

Commensal bacteria impact

The intestinal microbiota may also be able to activate STING as a result of increased intestinal permeability following pre-transplant conditioning. This results in exposure of multiple cell populations to bacterial dsDNA. Gut bacteria have previously been noted to initiate lethal GvHD by upregulating MHC class II on intestinal epithelial cells. The amount of damage to the gut epithelium by preconditioning regimens may determine the degree of STING activation.

It is worth noting that the preclinical mouse studies performed, were carried out under sterile conditions so did not take into account the effect of opportunistic infections by viruses or bacteria that occur in allo-HSCT patients. Altering activation of STING could potentially change the progression of these infections.

STING in tumor cells

Patients who develop GvHD have a lower rate of leukemia relapse than patients who do not, as a result of the action of immune cells against the alloantigens expressed on both normal and malignant cells.

In different circumstances, STING activation appears to be either tolerogenic or immunogenic. STING antagonists may decrease the graft-versus-leukemia effect along with GvHD resulting in an increased risk of relapse, while STING agonists may increase the severity of GvHD. They may also promote the graft-versus-leukemia response by stimulating donor T cells that target tumor and alloantigens.

The degree of STING activation has been demonstrated to be linked to the immunogenicity and antigenicity of tumor cells. An increase in tumor-cell DNA, which can occur because of cell death following pre-HSCT chemotherapy, can cause increased STING activation and increased T1-IFN production.

The relationship between irradiation and STING is complicated by the presence of a 3’ repair exonuclease known as TREX1. Higher doses of irradiation attenuate STING as TREX1 is upregulated by radiation exposure. This enzyme reduces the CDN substrate by degrading dsDNA in the cytoplasm and provides another example of the conditioning regimen determining the level of STING activation. The impact of the type of tumor on STING activation remains unknown and may also be an influencing factor along with the method of pre-transplant conditioning used.

Proliferation of tumor cells has been reported to be decreased following STING activation through NF-κB and p53-driven activation of p21. On the other hand, STING activation can be immunosuppressive, by attraction of myeloid-derived suppressor cells via CCR2 to the tumor site.

STING in donor cells

In donor T-cells, STING activation can lead to natural killer cell antitumor action through increased levels of T1-IFNs. Therefore, STING looks to be an attractive target for posttransplant cellular therapy that may improve outcomes. A suitable method for stimulating antigen-specific immune responses may be pulsed dendritic cell vaccines. Specific dendritic cell populations have been identified to produce a strong response following treatment with STING-targeted therapy. T1-IFNs may also be suitable as an adjuvant with DNA vaccines.

STING agonists have been reported as successful adjuvants for tumor vaccines. STING delivery using nanoshells can increase antitumor effect by putting the tumor antigen and the STING-based adjuvant in the same location. Liposomal storage may also help movement through the cell membrane, and therefore, lead to increased activation of STING.

Outlook for STING-targeted agents

For allo-HSCT patients STING-targeted agents may be less favored in the future. As conditioning regimens that involve radiation or chemotherapy agents such as busulfan, fludarabine or cyclophosphamide are used less, less DNA damage will occur, and therefore, less STING activation will occur. Reduced intensity regimes may be favored more, leading to a reduction in the damage to the gastrointestinal tract, and a decrease in the amount of bacterial DNA reaching cells to activate STING.

In autologous hematopoietic stem cell transplantation (auto-HSCT), however, high-dose myeloablative conditioning regimens remain the treatment of choice which results in a high level of cell death and gastrointestinal damage. Therefore, STING regulation may be beneficial in this group of patients. However, the risk of decreasing inflammation and epithelial repair through STING-modulation should be considered carefully as a result of the potential impact on nonrelapse mortality.

Current agents in development are shown in Table 1. So far, these agents have only been tested in solid tumors.

Table 1. Therapeutic agents targeting the STING pathway tested*

Conclusion

While the STING pathway holds promise for the modulation of GvHD and transplant outcomes, further work is required to understand its significance in different tissues, whether hematopoietic in origin or not. The conditioning regimen pre-HSCT appears to be linked to the degree of activation of STING; however, the specifics require more study. It will be interesting to see the outcomes of testing the novel STING-targeted agents and see how this translates into clinical assessment and patient outcomes.