Role of AMPK in minimizing GvHD in allogeneic effector T cells

Graft-versus-host disease (GvHD) continues to limit the application of allogenic hematopoietic stem cell transplantation (allo-HSCT) as a treatment option for hematologic disorders, decreasing overall survival (OS) in patients. The primary treatment of acute GvHD (aGvHD) is steroids but includes a myriad of side effects. As allo-HSCT remains a promising option for hematologic disorders, novel approaches to impair GvHD-causing T-cells, and metabolic modulation of T cells to target GvHD responses, are required. AMP kinase (AMPK) is a cellular energy sensor, that regulates specific metabolic pathways within activated cells. Despite the growing knowledge of AMPK, there is limited understanding of its role in T cells early posttransplant. Monlish, et al., published findings from their latest study¹ on the role of AMPK in GvHD in JCI Insight and the key findings are summarized here.

Aims and Methods

• To assess the effect of AMPK activation in aGvHD, a major histocompatibility complex GvHD mismatch mouse model was used.
• To demonstrate the role of AMPK in T-cell development, an AMPK double knockout (AMPK dKO) mouse model was used.
• The role of AMPK in GvHD, either fl/fl or AMPK-dKO T cells were transplanted into recipient mice in a minor antigen mismatch model of GvHD.
• The role of AMPK in providing antitumor immunity was assessed by transplanting fl/fl or AMPK-dKO donor T cells with T cell–depleted bone marrow (BM) and a lethal dose of leukemia cells into recipient mice.
• To track AMPK effects in CD4⁺ and CD8⁺ cells, AMPK-dKO cells were transplanted into recipient mice and donor T-cells measured 7 days posttransplant.
• Target organ pathology was assessed using Day 22 liver and small intestine paraffin sections scored from mice who received AMPK-dKO T cells.
• The effect of AMPK loss on metabolic pathways was assessed by radiolabeling ³H-palmitate to study fat oxidation rates in AMPK-dKO T-cells and fl/fl T-cells.
• Any effect of AMPK deficiency on donor T cells after transplantation was assessed using competitive transplants by combining congenically marked fl/fl T cells with AMPK-dKO T cells followed by transplantation into recipient mice.
• Impact of AMPK deficiency on Tregs was measured at Day 7 using mice transplanted with WT and AMPK-dKO T cells.
• To determine the role of donor Tregs for AMPK-dKO benefit, comparison of donor grafts with or without pretransplant depletion of CD25⁺cells were used.

• To establish the occurrence of AMPK activation in human cells, T cells were enriched from healthy donors, labelled with CellTrace, and placed into an allogeneic MLR, and cell division status assessed at either 4 days (without calcineurin inhibitor or at 6 days (with calcineurin inhibitor; tacrolimus).

Results

Alloreactive T cells increased AMPK activation during aGvHD and induced liver kinase B1 expression:

• AMPKα increased >15-fold in allogenic CD4⁺ and CD8⁺ T cells alongside a 6.8-fold increase in levels of upstream liver kinase B1. In addition, a 3.5-fold increase in AMPK phosphorylation was found in response to antigen OT-1.

AMPK-dKO T cells maintain T cell development and in vitro proliferation:

• AMPKα1/β2 were efficiently deleted in all peripheral CD4⁺ and CD8⁺ T cells in the AMPK-dKO mice without affecting thymocytes, spleen cell numbers or CD4⁺CD8⁺ ratios, except for a minimal decrease in splenic TCRβ⁺ cells.
• AMPK-dKO T cells proliferated similarly to wild type (WT) fl/fl T cells.

Decreased severity of aGvHD following transplantation of AMPK-dKO T cells:

• AMPK-dKO T cell recipient mice showed improved survival of 85% vs 15% at 10 weeks, including decreased weight loss and improved clinical scores, compared with recipients of fl/fl T cells in the minor antigen mismatch model.
• A similar survival benefit was found with AMPK-dKO recipient mice in the major histocompatibility complex mismatch setting.

AMPK-dKO T cells retain robust antileukemic potential:

• Tumor burden increased significantly by Day 13 in control mice receiving leukemia and BM cells vs substantial leukemia clearance in mice receiving either fl/fl or AMPK-dKO T cells, respectively.
• Median survival in AMPK-dKO donors and control mice receiving only BM was 24 vs 13 days, respectively. Median survival was similar (27 vs 28 days, respectively; p = 0.79) when using lower doses of T cells in recipients of a 4-fold reduction in AMPK-dKO or fl/fl T cells receiving leukemia and BM cells while control mice survived a median of 13 days.  

AMPK-dKO T cells are recovered in lower numbers posttransplant:

• At 7 days posttransplant, 40% fewer donor cells, including a reduced number of CD4⁺ and CD8⁺, were recovered. Although some of the reduction in CD8⁺ population was attributable to enhanced apoptosis, AMPK necessity was more closely correlated with cell division.
• Most allogenic T cells were found to undergo ≥8 cell divisions by CellTrace analysis, and those cells that were undergoing higher numbers of cell divisions, decreased in number suggesting that AMPK deficiency results in impaired proliferation in CD4⁺, and to a lesser extent in CD8⁺, cells.

Decreased pathology and reduced immune cell infiltration in GvHD target organs:

• Recipients of AMPK-dKO T cells had lower liver pathology, with a trend towards decreased apoptosis in the small intestine.
• Pathologic changes in the liver were correlated with reduced hepatic infiltrates (2.5-fold fewer AMPK-dKO CD3⁺ cells) which was supported by 35% fewer T cells being recovered from the livers of recipients of AMPK-dKO T cells at Day 7.

Classic metabolic pathways are unaffected by AMPK loss:

• AMPK-dKO T cells, recovered on Day 7, oxidized ³H-palmitate, at rates equivalent to fl/fl cells and this equivalence remained despite variation in ³H-palmitate exposure, T-cell receptor  stimulation, and limitation of extracellular nutrients.
• APL formation and light chain 3 modification remained unchanged in AMPK-dKO T cells, even when autophagy in allogeneic T cells increased.
• High levels of interferon gamma and tumor necrosis factor alpha were expressed by AMPK-dKO T cells, with similar moderate to low levels of interleukin-4, IL-7, and IL-10.

AMPK deficiency decreases donor T cell numbers prior to Day 3:

• AMPK-dKO T cells decreased by Day 3 posttransplant and CD8⁺ remained consistently low. Although peak proliferation occurred on Day 5, there was no difference in CD8⁺ T cell proliferation.
• On the other hand, AMPK-dKO CD4⁺ proliferated on Day 3 and proliferation was equivalent to fl/fl cells on Days 5 and 7.
• AMPK-dKO cells decreased sharply by Day 5 and remained low with no difference in liver-associated AMPK-dKO T cells.

CD4⁺FoxP3⁺ cells increase in recipients of AMPK-dKO cells:

• High percentages of WT Tregs were observed if co-transplanted with AMPK-dKO conventional T cells.
• Most of original CD90.2 Tregs remained FoxP3⁺ on Day 7 in the presence of AMPK-dKO Tcons, suggesting Treg number and percentage following transplantation is independent of AMPK, and is instead dependent on the genotype of the accompanying effector cells.

Donor Tregs are dispensable for AMPK-dKO–derived benefits:

• While FoxP3⁺ percentage increased on Day 7 in AMPK-dKO Treg-replete grafts, Treg percentages decreased with pre-transplant removal of CD25⁺ compared to fl/fl levels.
• Survival at 10-weeks was much better in recipients with transplantation of Treg-replete, AMPK-dKO grafts versus Treg-replete fl/fl grafts (85% vs 30%; p  <0.0005) including decreased weight loss and lower clinical scores.
• GvHD was exacerbated with CD25⁺ depletion in fl/fl donors, leading to shorter median survival and increased weight loss. 

Human allogeneic T cells activate AMPK:

• Phosphorylation of AMPK revealed a >50% increase in the ratios of p-/total AMPK than in human T cells.
• Irrespective of TAC exposure, equivalent AMPK activation occurred in divided CD4⁺ and CD8⁺ T cells along with equal p-/total AMPK ratios.
• AMPK activation increased in human T cells in vivo by Day 7, followed by a further increase on Day 11 with a 3-fold increase in p-/total AMPK ratios.

Conclusion

The authors of this study demonstrated that the absence of AMPK in donor T cells reduced GvHD while preserving T cell-mediated cytotoxicity and maintaining immune reconstitution. Altogether, the findings suggest that improvements in GvHD with AMPK-dKO T cells were not driven by changes in the canonical AMPK-related pathways of fatty acid oxidation, autophagy, or mammalian target of rapamycin signaling, but rather a decrease in conventional Teff cells. These findings have implications for our understanding of in vivo T cell metabolism, the role of AMPK in driving fatty acid oxidation in T cells and the impact of AMPK signaling on Treg generation. The study suggests that inhibition of AMPK in T cells may prevent GvHD and therefore further investigation is needed.