The tissue-resident regulatory T cell pool is shaped by transient multi-tissue migration and a conserved residency program
Burton O*, Bricard O* et al. (BioRxiv) DOI: 10.1101/2023.08.14.553196
Regulatory T cells (Treg)
Tissue-resident memory (Trm)
Immune cells found in non-lymphoid tissue have proven to be key players in tissue homeostasis. Regulatory T cells (Tregs), although found at a relatively low frequency compared to conventional T cells, are potent mediators of immune tolerance and have also been described in multiple non-lymphoid tissues. In addition to their role in establishing tolerance at barrier sites, some tissue Tregs also have roles in tissue repair and homeostasis. However, isolating and characterizing these rare cell populations for further study has been limited due to their low numbers, feasibility of processing these tissues and difficult purification.
Here, Burton and Bricard et al. led an enormous effort to isolate tissue-resident Tregs from 48 murine tissues in healthy mice. These cells were purified and phenotyped using high parameter flow cytometry. Surprisingly, the authors found that most Tregs shared phenotypic markers regardless of which tissue they originated from. To take an unbiased approach, RNA-sequencing was also performed on these purified Tregs and confirmed on the transcriptomic level that Tregs from different tissues do not have many distinguishable changes in gene expression. These findings contrasted with what has been reported in conventional T cells that seed and differentiate in tissues to become tissue-resident memory T cells (Trm) and have a distinct transcriptional program. Indeed, many T cell biologists would likely expect Tregs to fall into this “seed and specialization” model for tissue residency. The exception to this were gut associated Tregs, which were transcriptionally distinct and preferred gut homing upon transfer. Although somewhat surprising, these data provide an important direction for framing the way immunologists consider Treg residency in tissues, with a relatively short dwell time compared to what some may have previously assumed.
One major limitation is that all mouse models were on the same genetic background (B6). Of note, BALB/c mice exhibit higher numbers of Tregs and are more resistant to experimental autoimmune disease than B6 mice (PMID: 32330480). Although major effort was involved in adoptive transfer experiments, it would be important to know whether tissue Tregs remain mostly agnostic in other mouse backgrounds. Within the Treg clones that do show a tissue-preference (such as gut-associated Tregs), it would be interesting to see whether Tregs maintain their preference after multiple rounds of serial isolation and transfer to new recipients.
While testing the molecular determinants for Treg tissue residency, the authors employed an impressive array of knockout mice. However, the interpretation of “functionality” of the Tregs should be reconsidered instead as a test of the Tregs ability to migrate to those tissue sites. Experiments that induce an inflammatory challenge in a few of these tissue sites would be valuable to better distinguish between migration and anti-inflammatory roles of these knockout Tregs.
Aging experiments were notable and the microbiome aspect an important consideration to Treg biology. It was somewhat surprising to see such similarities among Treg phenotypes between the gnotobiotic, SPF, and cohoused mice. One suggestion for testing microbiome and age would be a fecal transplant of old vs. young mice and following Treg phenotypes before and after transplantation.
The TCR retrogenics experiments were novel and provide researchers with an improved method of tracking T cell clones by flow cytometry. Although the OT-II TCR clone did not differentiate into a Treg, it would be important to see this control TCR clone work successfully in an ovalbumin-specific model such as the RIPmOVA model of diabetes, or other antigen-specific model as a proof-of-concept.
A minor limitation is that in all flow cytometry sorts and sequencing experiments, the definition of a Foxp3+ Treg is limited to the detection of the Thy1.1 reporter. Although perhaps beyond the scope of this manuscript which does not induce inflammatory stimuli, it might be worth considering whether this Thy1.1 marker remains on exTreg cells that have lost their Foxp3 expression.
The amount of effort that was required to process all the murine tissues in parallel was remarkable. The consideration of 10 knockout mouse models for testing dependency of residency markers was also impressive. These findings are relevant to all specialities within autoimmunity and Treg biology. Additionally, the modification of the pro-code system for fixable flow cytometry was novel and will be valuable for the field of T cell biology and antigen-specificity. The probabilistic Markov chain model of cellular kinetics used to calculate Treg dwell times was innovative, although it does not allow for differentiation between cell death and dedifferentiation. Finally, the user-friendly website provided for gene expression data will be a valuable tool for other Treg biologists. Overall, this work alters our current understanding of Treg biology within tissues and convincingly challenges the seed and differentiation model for Tregs.
Reviewed by Kelsey Voss and Dörte Symmank as part of a cross-institutional journal club between the Vanderbilt University Medical Center (VUMC), the Max-Delbrück Center Berlin, the Charité Berlin, the Medical University of Vienna and other life science institutes in Vienna.