Regulatory T-cells in autoimmune hepatitis: An historical perspective with hints to the future

To the Editor:

We have read with interest the review by Fleming and colleagues, who discuss the central role of regulatory T-cells (Tregs) in the maintenance of immunotolerance in autoimmune liver diseases and liver transplantation.¹ We would like to raise a few points for discussion, focusing specifically on autoimmune hepatitis (AIH), as these may be relevant for future studies investigating the role of Tregs in AIH pathogenesis and for the development of immune therapies aimed at reconstituting this immunomodulatory cell subset.

When reviewing the literature on the involvement of Tregs in the pathogenesis of human AIH, the authors state that the role of Tregs in this context has been ‘difficult to discern’. We would like to emphasize that evidence about defective suppressor cell subsets in AIH dates to the late 1970s, with studies showing that concanavalin-A-induced suppression was absent in cells obtained from patients with AIH at remission and during relapse,² and that AIH is associated with a defect in suppressor lymphocytes controlling T cell immune responses to liver-specific protein.³ The consensus reached in these earlier studies in reporting defects in T-suppressor cells in AIH has not been achieved in subsequent investigations on CD4⁺CD25^highFOXP3⁺ Tregs, with some studies showing impairment and others reporting no phenotypic or functional defects of AIH Tregs.⁴ Although discussing these discrepancies might seem minor, they warrant attention because of the potential impact that these contrasting results could have on future studies and possibly on treatments. These discrepancies may denote gaps in the study of Tregs that have not yet undergone a thorough evaluation at the phenotypic, functional and epigenetic levels over time within the same patient. As Tregs in AIH display a high degree of functional plasticity, understanding whether these properties are stably maintained or change over time warrants additional investigations. This would not only help resolve the contrasting results but would likely provide insights into AIH pathogenesis and, importantly, inform the optimal timing for initiating Treg immunotherapy.

Along the same line, when discussing data obtained from experimental mouse models of AIH, Fleming and colleagues report conflicting results about Treg frequencies and function, ascribing these inconsistencies to different modalities of AIH induction and Treg detection. We agree with their consideration. However, we would like to highlight that in a humanized mouse model – generated by injecting a human cytochrome P4502D6 (CYP2D6)/formiminotransferase-cyclodeaminase (FTCD) fusion protein into HLA-DR3+ transgenic mice, which closely mimicked human AIH – reduced Treg frequencies were observed concomitantly with increased intrahepatic CD4+IL-17+ and CD8+IFNγ+ cells.⁵ In another model characterized by hepatocellular expression of an MHC class-II restricted immunodominant epitope of the lymphocyte choriomeningitis virus, accumulation of CD4-cells with epitope specificity was favored by selective depletion of Tregs sharing the same antigen specificity.⁶

The authors indicate that Treg/Th17 cell ratio has been used to predict the extent of liver inflammation and propose that the balance between the two cell types could represent a goal for immunotherapy in AIH. This is indeed an important point to consider when attempting to restore homeostasis in autoimmune diseases. As it is likely that, in AIH, impaired Tregs play a permissive role in liver damage, rather than being the initiating cause of it, immunotherapeutic strategies should interfere with T-effectors while at the same time boosting Treg immunity. Recent work has shown that alterations of aryl hydrocarbon receptor (AhR) signaling underlie both Treg dysfunction and persistence of Th17 cell effector properties in patients with AIH,⁷ as reflected by decreased levels and activity of CD39, an AhR-regulated ectoenzyme that is key to immunotolerance. Defective CD39 levels have been linked to Treg dysfunction and maintenance of Th17 cells in an inflammatory state. Blockade of factors that either inhibit AhR activation (e.g. hypoxia inducible factor 1-alpha) or serve as AhR non-canonical binding partners (e.g. Krüppel-like factor-6, estrogen receptor-alpha)⁷ might be critical to restore tolerance in AIH.

Finally, when discussing autoantigen-specific Tregs as an additional form of Treg-based immunotherapy, we would like to reiterate the suitability of both CYP2D6 and O-phosphoseryl-tRNA:selenocysteine-tRNA synthase (Sep-SecS) as autoantigenic targets in AIH-2 and AIH-1, as highlighted in previous work.^8,9 Since CD4 and CD8 immunodominant epitopes have been identified within these autoantigens, engineering Tregs that share the same autoantigen specificity should be considered for adoptive transfer purposes. As shown in murine and human studies in the context of AIH, antigen-specific Tregs display an advantageous functional profile when compared to polyclonal Tregs^6,10 and efforts should be allocated to develop and optimize strategies boosting this powerful Treg pool with the goal of restoring immunotolerance in a more effective and tailored manner.

Financial support

National Institutes of Health grant R0101 DK124408 (MSL).

Authors' contributions

Writing - review and editing: MSL, GMV, DV.

Conflict of interest

The authors declare no conflicts of interest pertaining to this manuscript.

Please refer to the accompanying ICMJE disclosure forms for further details.

Supplementary data

The following are the Supplementary data to this article: