Intact extracellular matrix and the maintenance of immune tolerance: high molecular weight hyaluronan promotes persistence of induced CD4+CD25+ regulatory T cells

Abstract

The composition of the ECM provides contextual cues to leukocytes in inflamed and healing tissues. One example of this is HA, where LMW-HA, generated during active inflammation, is a TLR ligand and an endogenous “danger signal,” and HMW-HA, predominant in healing or intact tissues, functions in an inverse manner. Our data suggest that HMW-HA actively promotes immune tolerance by augmenting CD4+CD25+ T_Reg function, and LMW-HA does not. Using a human iT_Reg model, we demonstrate that HMW-HA but not LMW-HA provides a costimulatory signal through cross-linking CD44 which promotes Foxp3 expression, a critical signaling molecule associated with T_Reg. This effect, in part, may be mediated by a role for intact HMW-HA in IL-2 production, as T_Reg are highly IL-2-dependent for their survival and function. We propose that HMW-HA contributes to the maintenance of immune homeostasis in uninjured tissue and effectively communicates an “all-clear” signal to down-regulate the adaptive immune system through T_Reg after tissue matrix integrity has been restored.

Introduction

The inflammatory milieu is known to contribute strong activation stimuli to innate and adaptive immune responses through a variety of danger signals. However, regulatory contributions of ECM components, which potentially counterbalance the inflammatory cascade, are less understood. There is an immunologic dichotomy between intact and inflamed ECM [1, 2]. A number of ECM components exist in a nonproinflammatory form until they are fragmented or released from the matrix, at which time they become proinflammatory and chemoattractant [3,4,5,6,7]. In this context, ECM fragments have been suggested to function as endogenous “danger signals” analogous to exogenous pathogen-associated molecular pattern molecules derived from bacteria or viruses. Indeed, a number of ECM components, including fragments of versican [8], fibronectin [9], fibrinogen [10], biglycan [11], soluble heparin sulfate [12], and LMW-HA [13], are reported to signal through TLRs. Interestingly, TLR2 and/or TLR4 have been implicated in all of the aforementioned models, suggesting that common pathways for fragmentary matrix signals may exist. In addition, fragmentary ECM components promote immune activation and migration by a broad diversity of mechanisms other than TLR. These include interactions with a range of cytokines, chemokines, growth factors, MMPs, and other inflammatory mediators (reviewed in refs. [1, 14]).

Although several fragmentary ECM components are proinflammatory, the immunologic relevance of their whole parent molecules and of intact ECM in general is more ambiguous. Intact ECM has been viewed conceptually as an inert or immunologically passive foil to fragmentary ECM. However, experimental investigation is limited by the difficulty of recreating biologically plausible yet experimentally tractable model systems in vitro. Some of the salient considerations include the complex and dynamic interactions between ECM components, the extensive post-translational modification of the many ECM components, and the importance of matrix-bound soluble factors to the activity of many of these molecules in vivo.

Our recent work suggests that at least one intact ECM component, HMW-HA, may play an active role in the maintenance of immune tolerance. In this article, we discuss the role of intact HA in promoting the function of naturally occurring CD4+CD25+ T_Reg and present data from work in progress demonstrating that an analogous process occurs in T_Reg induced in vitro. In nT_Reg and iT_Reg systems, HMW-HA promotes persistence in vitro of Foxp3, a critical transcription factor for T_Reg. We propose a model for how intact HA may play an important role in the maintenance of immunological tolerance in uninjured or healing tissue.

HA IS AN IMMUNOLOGIC MEDIATOR

HA is an unadorned, repeating disaccharide of N-acetylglucosamine and D-glucuronic acid. It is not known to bind to cytokines or other small molecules, is not chemically modified postsynthesis, and exists in only one form: a chain of variable length ranging in molecular weight from 10⁴ to 10⁷ Da. Because of these features, HA has proved relatively amenable to immunologic investigation. The bioactive properties of HA vary with its size; the amounts and relative proportions of these have distinct effects on cellular metabolism and phenotype [15].

HA breakdown products predominate during injury and inflammation. LMW-HA (<15 saccharides; <3 kDa) are generated by endogenous catabolism, by bacterial hyaluronidases, and by mechanical forces and oxidative stress [16]. LMW-HA promotes angiogenesis [17], inflammation [18, 19], maturation of APCs [20, 21], and cell migration [22]. These effects of LMW-HA are mediated via MMP [23], plasminogen activator inhibitor [24], NO [25], and production of several proinflammatory cytokines, including MIP-1α, MIP-1β, MCP-1, IL-8, and IL-12 by macrophages [26,27,28,29] and IL-1β, TNF-α, and IL-12 by dendritic cells and endothelial cells [30].

HMW-HA (>2000 saccharides; >400 kDa) predominates in steady-state conditions and in healing tissues [31]. HMW-HA serves a variety of structural and regulatory functions in joints [32] and tissue repair [33]. HMW-HA has typically been reported to be inert or anti-inflammatory [19, 34]. HMW-HA inhibits phagocytosis by monocytes [35], limits antigen-antibody interactions [36], and has been suggested to impede lymphocyte activation [37]. HMW-HA is U.S. Food and Drug Administration-approved for injection into osteoarthritic joints and prevention of postsurgical abdominal adhesions and is under investigation in other settings including wound healing, burn treatment, and rheumatoid arthritis [38].

The best characterized HA receptor is CD44, a nearly ubiquitous cell surface receptor important in cell trafficking, activation, and apoptosis [39]. CD44 exists in multiple isoforms with distinct functional characteristics depending on alternative splicing and expression of different exons [40]. The CD44 expressed on resting T cells and monocytes is functionally inactive and binds HA only after TCR triggering or after activation by proinflammatory cytokines including TNF-α and IFN-γ [41]. Therefore, the ability of these cells to interact with HA is intrinsically related to their activation state. Signaling through CD44 improves with longer chains of HA up to 38 sugars in length [42].

CD4+CD25+ T_Reg CONSTRAIN ADAPTIVE IMMUNITY

CD4+CD25+ T_Reg are a relatively well-characterized, specialized subpopulation of T cells that expresses the transcription factor Foxp3 [43]. A variety of distinct mechanisms is involved potentially in T_Reg activity, and T_Reg cells display a range of regulatory functions [44, 45]. T_Reg arise during lymphocyte development from the thymus (nT_Reg), but it has now been shown by us and by other groups that they can also be generated in the periphery and induced ex vivo (iT_Reg) [46,47,48]. T_Reg are known to integrate diverse signals from the inflammatory milieu [49,50,51,52], including a key dependence on IL-2 to maintain viability and function. IL-2 is considered a proinflammatory cytokine, and the requirement for IL-2 fits with the model of T_Reg as a regulatory cell type that exists within inflammatory contexts to dampen overaggressive responses. The strength of TCR activation and the availability of IL-2 have been shown to be dominant factors in the triggering of T_Reg [53], consistent with the model that iT_Reg are a subset of T cells driven by overactivation into anergy and suppression of neighboring cells. An unaddressed question in T_Reg biology is how this highly IL-2-dependent cell type persists in healing or uninjured tissues, where IL-2 concentrations are presumably low.

HMW-HA PROMOTES THE FUNCTION OF NATURALLY OCCURRING CD4+CD25+ T_Reg

Most of the molecules that characterize T_Reg, including CD25 and Foxp3, are activation markers common to many highly activated T cells. The report that total Foxp3 expression and relative suppressor function are correlated with CD44 expression levels [54] fits within this paradigm. Furthermore, this study raised the possibility that CD44 interactions with HA may be integrally related to T_Reg function. Indeed, we reported subsequently [55] that HMW-HA promotes the suppressive function of human nT_Reg, and LMW-HA does not. In our follow-up work (manuscript submitted), we demonstrated that this effect is a result of a costimulatory function mediated by HMW-HA through CD44 expressed on activated T_Reg. Cross-linking of CD44 by HMW-HA and plate-bound anti-CD44 antibody promotes Foxp3 persistence, and neither LMW-HA nor soluble anti-CD44 antibody does so. Blocking antibodies directed against CD44 abrogate this effect. This is consistent with previous reports of cross-linking being integral to several functions attributed to CD44 [56, 57]. We found previously that T_Reg do not bind FITC-labeled HMW-HA unless they are preactivated via their TCR complex, at which point they up-regulate CD44v isoforms capable of binding HMW-HA [55]. The activation-induced expression of HA-binding isoforms of CD44 is correlated with Foxp3 expression and is consistent with similar activation-dependent HA binding seen in previous studies [58]. Perhaps consistent with a role for HMW-HA in promoting T_Reg function is the recent report that binding of HMW-HA but not LMW-HA causes the death of activated T cells [59]. Given the relationship between HA size and the stage of inflammation, the extent of CD44 cross-linking by HMW-HA could provide contextual cues to T_Reg and T cells regarding the inflammatory milieu.

HMW-HA PROMOTES FUNCTION OF CD4+CD25+ T_Reg INDUCED IN VITRO

As noted previously, T_Reg require IL-2 to maintain viability and function. In our previously published work, we found that HMW-HA promoted Foxp3 expression by activated T_Reg predominantly in settings of low IL-2 [55]. Given this observation, we then asked whether HMW-HA was impacting the availability of or requirement for this key cytokine. Consistent with reports of enhanced IL-2 production by other T cell subsets and NKT cells upon CD44 costimulation [60, 61] and with reports of requirements for HA and CD44 in certain IL-2-dependent processes [62, 63], we hypothesized that CD44 cross-linking promotes modest IL-2 production and thereby provides a tonic signal necessary for T_Reg persistence and function.

To examine the role of HMW-HA on T_Reg IL-2 interactions, we used a human iT_Reg model described previously [46] with minor variations. The induction of functional Foxp3+CD25+ T_Reg from CD4+CD25– precursors is known to be highly dependent on the availability of IL-2 [64, 65]. This model is pertinent to the role of the ECM in T_Reg physiology, as the induction of T_Reg from similar populations of cells is thought to occur in vivo in peripheral tissues, presumably where ECM integrity is most relevant. We found that HMW-HA but not LMW-HA promotes persistence of Foxp3 expression by such iT_Reg populations (Fig. 1). These effects were most pronounced when HMW-HA treatment occurs late but not early in induction cultures, when expression of activation markers and production of endogenous IL-2 could be expected to wane. Although soluble HMW-HA was used in the particular experiment shown in Figure 1, we find that the magnitude of the salutary effect on Foxp3 expression is typically most pronounced when plate-bound HMW-HA is used. This effect on Foxp3 expression was approximately equivalent to the addition of exogenous IL-2 added at 20 IU/ml (Fig. 2A). Upon sorting these cells on the basis of CD25 expression on Day 10 of culture, iT_Reg induced with HMW-HA or exogenous IL-2 treatment on Day 7 were found to suppress proliferative responses of autologous T cells better than control iT_Reg (Fig. 2B).

Figure 1.

HMW-HA treatment added late in induction culture promotes Foxp3 expression by iT_Reg. CFSE-labeled CD4+CD25– cells were activated and treated on Day 0 or Day 9 with HMW-HA or LMW-HA. Flow cytometry for Foxp3 expression on Day 10 is shown.

Figure 2.

HMW-HA treatment produces effects on iT_Reg number and function analogous to IL-2 addition. (A) Percentage of CD25+ cells positive for Foxp3 upon addition of HMW-HA or IL-2 at 20 IU/ml on Day 7 of the induction culture. Significance was determined via a Student’s t-test. ns, Not significant. (B) T_Reg suppression assay using the top 5% of CD25-expressing cells sorted from each induction condition and autologous CFSE-labeled CD4+CD25– T cells as the responder population. Data are expressed as percent suppression relative to proliferative population observed in the absence of iT_Reg. Data are representative of three independent experiments. (C) IL-2 was assessed in cell culture supernatants drawn on Day 8 following inception of the iT_Reg induction culture, 1 day after washing and transfer to plates coated with HMW-HA or controls. Data are for four experiments.

These effects on Foxp3 persistence and iT_Reg viability are consistent with HMW-HA, impacting the availability of or requirement for IL-2. Although our data do not rule out ancillary mechanisms, we find that HMW-HA but not LMW-HA treatment of cells in an iT_Reg induction culture promotes production of modest levels of IL-2 (Fig. 2C). We do not see a similar induction of IFN-γ (data not shown). Given the established role of IL-2 in T_Reg function and viability, we next asked whether HMW-HA promoted viability of iT_Reg populations. This does appear to be the case, as HMW-HA on Day 7 of induction culture indeed resulted in a larger percentage of viable cells on Day 10 of culture, as ascertained by staining for the apoptotic marker annexin V and the necrosis marker 7-amino-actinomycin (unpublished data). This is consistent with a previous report of a role for HA in cell survival [66].

A MODEL FOR HOW HMW-HA MIGHT PROMOTE THE MAINTENANCE OF IMMUNE TOLERANCE

We interpret the data reported here and in our related work as supporting the following model (Fig. 3). Upon activation in the setting of injury, a fraction of T cells is induced to become T_Reg. During the active immune response phase, the activity of these is thought to be inhibited by IL-6 (and other proinflammatory mediators) produced by subsets of activated T cells and macrophages. LMW-HA is likely one of these mediators, as it has been shown to promote IL-6 production, and HMW-HA does not [67]. Upon resolution of the injury response, the concentration of IL-2 produced by activated T cells diminishes. At the same time, tissue repair mechanisms begin to synthesize intact HMW-HA. As the concentration of ambient IL-2 wanes, the molar balance of HA molecules shifts in favor of HMW-HA forms capable of cross-linking CD44v isoforms. CD44 cross-linking by HMW-HA provides a second costimulatory signal in the setting of a primary antigen-specific signal through the TCR. Together, these two signals promote T_Reg activation, including low-level production of IL-2. Alternative sources of IL-2 in this context are also possible, as availability is likely to be limiting as the inflammatory response wanes. We propose that the HMW-HA molecular composition of the ECM thereby conveys a tissue integrity signal that functions as the opposite of a danger signal, in that adaptive T_Reg receive an important secondary signal in situ, promoting immune tolerance. This mechanism may be one way in which viable T_Reg populations persist in healing or uninjured tissues to bring inflammatory processes to a close and maintain homeostasis. The intermolecular interactions described here, between ECM and the adaptive T lymphocyte, may represent just one of many such pathways for providing contextual clues linking tissue status to the regulation of the immune response.

Figure 3.

A model for how HA size might contribute to immune regulation and the maintenance of immune tolerance. Upon injury, T cells proliferate in response to LMW-HA and other proinflammatory cues. A fraction of activated T cells (T) becomes T_Reg (T_R). As the concentration of ambient IL-2 wanes with the resolution of inflammation, the molar balance of HA molecules shifts in favor of HMW-HA forms capable of cross-linking CD44v isoforms. CD44 cross-linking by HMW-HA provides a second costimulatory signal in the setting of a primary antigen-specific signal through the TCR; together, these two signals promote production of IL-2 and thereby maintain T_Reg viability and function. The persistence of iT_Reg populations in peripheral tissues is likely to be important to the maintenance of immune tolerance.

ACKNOWLEDGMENTS

This work was supported by grants from NIH (DK46635, HL18645, and DK53004) and Juvenile Diabetes Research Foundation (The Center for Translational Research at BRI). P. L. B. is supported by NIH K-08 grant DK080178-01 and a NIH Loan Repayment Program grant. We thank Nathan Standifer and Alice Long for their helpful comments and suggestions.

DISCLOSURE

The authors have no conflicting financial interests.