Natural CD4+CD25+Foxp3+ regulatory T cells (Treg) effectively prevent autoimmune disease development, but their role in maintaining physiological tolerance against self-Ag of internal organs is not yet defined. In this study, we quantified disease-specific Treg (DS-Treg) as Treg that preferentially suppress one autoimmune disease over another in day 3 thymectomized recipients. A striking difference was found among individual lymph nodes (LN) of normal mice; Treg from draining LN were 15–50 times more efficient than those of nondraining LN at suppressing autoimmune diseases of ovary, prostate, and lacrimal glands. The difference disappeared upon auto-Ag ablation and returned upon auto-Ag re-expression. In contrast, the CD4+CD25− effector T cells from different individual LN induced multiorgan inflammation with comparable organ distribution. We propose that peripheral tolerance for internal organs relies on the control of autoreactive effector T cells by strategic enrichment of Ag-specific Treg in the regional LN.
Self-tolerance is the fundamental process that guards against autoimmune disease. Maintained by self-Ags, tolerance is a physiological function that operates in the normal steady state. Although studies have implicated regulatory T cells (Treg)5 in peripheral tolerance, many details are not yet clarified. This is especially true for the internal organs, including endocrine organs, gonads, pancreatic islets, thyroid, CNS, and exocrine glands, all of which are major targets of human autoimmune disease. A current concept, lymph node (LN) surveyance by Ag-specific Treg, is supported by the finding that Ag-specific Treg are continuously capacitated by responding to auto-Ag in normal regional LN (1). Accordingly, Treg with transgenic TCR proliferate in regional LN of normal or lymphopenic hosts (2, 3), and regional LN are a location where Treg control effector autoimmune T cell responses (4). Recently, in mice with highly restricted TCR-VDJ expression, the Treg from individual LN showed distinctive TCR repertoires based on VDJ analysis (5). Collectively, these findings support LN-specific distribution of Ag-specific Treg; however, the functional capacity of the Treg in autoimmune disease suppression was not investigated in these studies; therefore, the relevance of the data to self-tolerance has remained undefined.
The ability to investigate tolerance mechanisms, based on detection and functional analysis of Ag-specific polyclonal T cells, is hindered by their rare occurrence. As an alternative approach, we investigated Ag-specific polyclonal Treg as a T cell population that suppresses one organ-specific autoimmune disease over another in the day 3 thymectomized (d3tx) mice (disease-specific Treg (DS-Treg)) (6, 7). In this study we describe a functional assay for the DS-Treg and document for the first time their regional and nonrandom distribution in the LN of normal mice.
Materials and Methods
Mice and breeders, obtained from the National Cancer Institute (Frederick MD) or The Jackson Laboratory, were kept in a pathogen-free facility. Thymectomies and orchiectomies were done under hypothermia (4, 8). 5α-Dihydrotestosterone pellets (20 mg per 60 day release; Innovative Research of America) were implanted s.c. Experiments were performed according to guidelines of the Animal Care and Use Committee of the University of Virginia (Charlottesville, VA).
Renal, cervical, and lumbar/sacral LN were identified and used as regional LN of the ovary, lacrimal gland, and prostate, respectively. CD4+CD25+ Treg or CD4+CD25− T cells were isolated to ∼95% purity via magnetic beads (autoMACS system; Miltenyi Biotec). Cells were transferred by i.p. injection to 5 day-old d3tx mice, which were studied 8 wk later, or by i.v. injection into syngeneic adult RAG knockout recipients, which were studied 3–4 wk later.
Histology and immunofluorescence
Pathology was graded in blinded samples as follows (4). For autoimmune ovarian disease (AOD), a grade of 1 referred to focal oophoritis, grades of 2 and 3 to increasing monocytic inflammation, and a grade of 4 to severe and diffuse inflammation. For experimental autoimmune prostatitis (EAP), a grade of 1 referred to focal monocytic inflammation, a grade of 4 to severe EAP with diffuse inflammation and loss of glands, and grades 2 and 3 to increasing severity between grades 1 and 4. Inflammation in the following organs was graded similarly as EAP: lacrimal gland, testis and epididymis, lung, liver, colon, skin, kidney, stomach, thyroid, eye, and small intestine. Indirect immunofluorescence was used to detect serum prostate and ovarian Abs (dilution 1/50). Pooled sera with prostate Abs from d3tx mice were used to detect prostate Ag expression on ethanol-fixed frozen prostate and ovarian sections.
Results and Discussion
Normal ovarian LN draining Treg have enhanced capacity to suppress AOD
D3tx B6AF1 mice develop dacryoadenitis (DA) concurrently with AOD or EAP (7) and auto-Ab against the respective tissue Ag (8, 9, 10). In each case, the disease and auto-Ab responses are fully suppressed by 0.5 × 106 or more Treg pooled from all the LN of normal donors (8).
To compare the Treg from regional vs nonregional LN in AOD suppression, CD4+CD25+ Treg from ovarian LN or pooled from nondraining LN of untreated B6AF1 female donors were transferred to groups of d3tx female mice at doses of 0.003–0.5 × 106, and the ovarian pathology was determined 8 wk later. AOD was suppressed only by 0.5 × 106 Treg from nondraining LN (Fig. 1⇓B and supplemental Fig. 1B).6 In stark contrast, AOD was completely suppressed by 0.03 × 106 Treg from the ovarian LN (Fig. 1⇓A and supplemental Fig. 1A). Thus, a 15-fold difference in AOD suppression was detectable between the Treg from draining and nondraining LN (Fig. 1⇓C). Serum oocyte auto-Ab was detected only in d3tx recipients of <0.01 × 106 Treg from ovarian LN and in d3tx recipients of <0.1.× 106 Treg from the nondraining LN (Figs. 1⇓, A and B, and supplemental Fig. 1C).
Prostate LN Treg and lacrimal gland LN Treg of normal mice preferentially suppress EAP and DA, respectively, in the same host
To determine whether the finding for AOD was a general phenomenon, we investigated EAP and DA. EAP was completely prevented by 0.01 × 106 Treg from prostate LN, whereas 0.5 × 106 Treg from the same LN were required to completely suppress DA (Fig. 2⇓, A and D, and supplemental Fig. 1, D and H). In contrast, Treg from LN draining the lacrimal gland completely suppressed DA at a dose of 0.03 × 106 Treg, whereas 0.5 × 106 Treg were needed to prevent EAP (Fig. 2⇓, B and E, and supplemental Fig. 1, E and G). Serum auto-Ab to the prostate gland was detected only in recipients of <0.003 × 106 Treg from the prostate LN and in recipients of <0.1 × 106 Treg from lacrimal gland LN (Fig. 2⇓, A and B, and supplemental Fig. 1F). Thus, the Treg from prostate LN were 50 times more potent in suppressing EAP over DA, whereas Treg from the lacrimal gland LN were 15 times more potent in suppressing DA over EAP (Fig. 2⇓, C and F).
We wish to emphasize that the 50-fold enrichment of EAP-specific Treg in the prostate-draining LN is 17 times greater than the 3-fold enrichment of the global Treg present in the pooled LN cells of prostate-bearing male mice over female mice determined earlier (8). Notwithstanding this marked difference in functional capacity, the Treg from LN draining the prostate, ovary, and lacrimal gland have similar expressions of Foxp3 and comparable abilities to suppress in vitro pan-T cell proliferation (supplemental Fig. 2).
The new bioassay described in this study is useful for semiquantitation of functional Ag-specific Treg. Notably, the assay is not dependent on T cell recognition of a specific T cell epitope that may or may not be relevant to or exclusive for a given autoimmune disease. Instead, it detects Treg that recognize all organ-derived autopeptides that are relevant to suppression of disease in that organ. Based on the assay, we have documented for the first time that Treg from normal regional LN are highly enriched in functional Treg with the capacity to suppress autoimmune disease of their draining organs. In our earlier reports, we studied DS-Treg from individual LN, but from d3tx mice with ongoing autoimmunity (4, 11) or DS-Treg from pooled LN of normal mice (8); therefore, they are not pertinent to self-tolerance.
Prostate-specific Ag expression influences LN enrichment of EAP-specific Treg
The DS-Treg enrichment in regional LN could result from an increase in Treg number and/or an enhancement of intrinsic Treg activity, and this may depend on the response of Treg to auto-Ag, their preferential homing to the regional LN, or both. To address the requirement of auto-Ag, we determined whether prostate Ag expression influences regional LN enrichment of EAP-specific Treg. When prostate Ag expression was eliminated in mice with neonatal orchiectomies (NOX) (8), the primacy of Treg from the prostate LN to suppress EAP was lost. When prostate Ag expression resumed in Treg donors that underwent NOX and 5α-dihydrotestosterone (DHT) treatment (8), the EAP-specific Treg activity in prostate LN was restored (Fig. 3⇓). Thus in the steady state, presentation of organ-derived tissue Ag to Treg is required for regional LN-specific enrichment of DS-Treg. A caveat of this study is that it has not ruled out a potential sex hormone effect on Treg function inherent in the experimental design.
Normal regional LN have enriched DS-Treg but not enriched DS-pathogenic T cells; correlation with reported TCR repertoire data
The Ag dependency or Ag specificity of a polyclonal autoimmune T cell response can also be extrapolated from TCR repertoire analysis of LN T cells. Lathrop et al. (5) recently documented the repertoire of Treg vs non-Treg based on thousands of individual T cells per LN in mice expressing a highly restricted TCR repertoire. Consistent with our finding of LN-specific DS-Treg, they also reported a distinctive repertoire among the Treg from individual LN. This was a finding unique to Treg, because comparable TCR repertoires were detected among the naive CD4+Foxp3− T cells of individual LN. To establish an additional correlate between the two studies and add functional support to the TCR repertoire data, we next determined the distribution of autoimmune inflammation in recipients of CD4+CD25− T cells from individual LN of normal donors (12, 13, 14). As shown in Fig. 4⇓ and supplemental Fig. 3, when CD4+CD25− T cells from renal LN, mesenteric LN, or cervical LN of normal mice were transferred to syngeneic RAG-knockout recipients, they elicited autoimmune inflammation with very similar organ distribution and severity. Monocytic inflammation and tissue destruction were noted maximally in the prostate, testis, and epididymis, lung, liver, skin, and colon, less so in the kidney, lacrimal gland, stomach, and thyroid, and the pancreatic islets, heart, eye, and small bowels were spared (supplemental Fig. 3). The finding of such a striking correlation between TCR repertoire and T cell function for both LN-specific Treg and effector T cells provides additional support for the conclusion that the normal LN enrichment of DS-Treg is Ag dependent.
The mechanism of LN-specific, Ag-specific Treg enrichment might depend on factors regulating T cell homing to LN, encounter with self-Ag, and their retention in the LN. Homing of naive T cells and Treg to normal LN are known to involve CD62L, CCR7, and the chemokines CCL19 and CCL21 (15). Autoimmune diseases occur in mice deficient in CD62L or CCR7 (16, 17), for which we can now add a potential explanation: the loss of DS-Treg enrichment in regional LN. Treg retention may result from up-regulation of CD69 on Ag-specific Treg that temporarily sequester sphingosine 1-phosphate receptor type 1, which is required for T cell egress from LN (18). Additional mechanisms may involve Treg response to antiapoptotic and/or cellular proliferation signals (19). Constrained by T cell homeostatic mechanisms (20), the number or activity of DS-Treg in the regional LN would be maintained at a threshold 15- to 50-fold greater than those in the nondraining LN.
Additional mechanisms participate in Ag-specific Treg homing to regional LN and the maintenance of abundant resident Ag-specific Treg in normal skin, lung, and mucosal sites (15, 21). It is conceivable that they also participate in Treg homing to internal organs and their regional LN. However, unlike mucosa such as intestinal lamina propria, very few Foxp3+ cells are detected in normal internal organs; for example, only 1% of normal prostate CD4+ T cells are Foxp3+ (our unpublished data). Moreover, TCR repertoire and functional capacity of T cells in a single LN are distinctive among the Treg but are shared among the T effectors; this finding is not consistent with the prevailing concept that the sharing of chemokine receptor expression allows for colocalization of Treg and T effectors with shared TCR specificity (15, 22, 23). Nonetheless, this is clearly an important area that requires further exploration.
In summary, we have documented major differences in the LN distribution of autoimmune disease-relevant effector T cells and Treg in normal mice. In the normal steady state, each regional LN is highly enriched for Ag-specific Treg with the capacity to inhibit autoimmune attack to the specific organ that drains to it, but contains very few Treg that control the autoimmune disease of other nondraining organs. These data provide strong and biologically relevant information for a critical and anticipatory role of Ag-specific Treg in peripheral tolerance. Positioned in this strategic location, maintained by self-Ag, we propose that DS-Treg can best control organ-specific tolerance by negating response to self-peptides in the event of infection, tissue necrosis, and other forms of local danger. Our finding also emphasizes that an investigation based on pooled LN Treg may not correctly inform the physiological status of Treg function in vivo.
We thank Yuefang Sun, Joshua Sparks, and Virginia Rubianes for expert technical assistance.
The authors have no financial conflict of interest.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
↵1 These studies were supported by National Institutes of Health Grants RO1 AI 41236 and RO1 AI51420.
↵2 K.M.W. and E.T.S. contributed equally to this work.
↵3 Current address: Hoffmann-La Roche Inc., Inflammation Discovery, Building 102/309A, 340 Kingsland Street, Nutley, NJ 07110-1199.
↵4 Address correspondence and reprint requests to Dr. Kenneth Tung, Department of Pathology, University of Virginia, Charlottesville, VA 22908. E-mail address:
↵5 Abbreviations used in this paper: Treg, regulatory T cell; AOD, autoimmune ovarian disease; DA, dacryoadenitis; DHT, 5α-dihydrotestosterone; DS-Treg, disease-specific Treg; d3tx, day 3 thymectomized (mice); EAP, experimental autoimmune prostatitis; LN, lymph node; NOX, neonatal orchiectomy.
↵6 The online version of this article contains supplemental material.
- Received December 22, 2008.
- Accepted October 13, 2009.
- Copyright © 2009 by The American Association of Immunologists, Inc.