Integrins are dimeric cell surface molecules composed of α and β subunits that promote cell–cell interactions and play important roles in the immune system. Within the integrin family, α4β7 displays relative specificity for the mucosal immunity and has multiple functions including lymphocyte homing, intestinal lymphoid tissue formation, and promoting inflammatory responses.1–6 Blockade of α4β7 is an appealing therapy for intestinal inflammatory diseases, as the effects of blocking this pathway will largely be limited to mucosal sites. While diverse functions for α4β7 in intestinal immunity are well described, the essential nature of α4β7 in these roles is less clear, and blockade of these essential functions could result in untoward effects of this therapy including inhibiting the homing of nonpathogenic lymphocytes and the formation of homeostatic lymphoid tissues.
Mucosal vascular addressin cell adhesion molecule (MAdCAM-1) is the binding partner for α4β7,1 and is expressed on high endothelial venules (HEVs) and postcapillary vessels in the intestine. In contrast, α4β7 is expressed by diverse populations of hematopoietic cells including lymphocytes, natural killer cells, monocytes, and lymphoid tissue inducer (LTi) cells.5–9 A role for α4β7/MadCAM-1 interactions in lymphocyte trafficking to the intestine is widely accepted; however, the lymphocyte subsets, cellular compartments, and conditions in which α4β7 plays an essential role, and therefore would be affected by α4β7 blockade, are less clear. Early studies using adoptive transfer of labeled lymphocytes identified a role for α4β7 in lymphocyte homing to Peyer's patches and mucosa in the uninflamed intestine.2 Later studies of using genetically modified mice confirmed that in the absence of β7 in the uninflamed intestine, Peyer's patches had diminished cellularity and mucosal lymphocyte populations were decreased.10 However, these studies were performed prior to the appreciation of the diversity of lymphocyte subsets, and consequently information regarding the essential nature of α4β7 for homing of specific lymphocyte populations in the uninflamed intestine is largely lacking. Moreover, blockade of the α4β7/MadCAM-1 pathway has had variable effectiveness in animal models of intestinal inflammation,3, 11–14 which could be explained by differences in the specificity of the blockade strategy, or the homing mechanisms used by lymphocyte subpopulations playing a pathogenic or protective role in each model.
In addition to the above effects, α4β7 blockade may affect the development of intestinal lymphoid tissues during inflammation. α4β7/MadCAM-1 interactions play a critical role in the development of isolated lymphoid follicles (ILFs),6 or intestinal lymphoid aggregates, in the uninflamed adult intestine. During pathogenic inflammatory diseases, including inflammatory bowel disease (IBD), increased numbers of lymphoid aggregates are also seen,15–18 and the above observations suggest that their development may also be dependent on α4β7/MadCAM-1 interactions. During physiologic inflammation, ILFs mediate homeostatic functions in intestinal immunity; conversely, the function of lymphoid aggregates during chronic intestinal inflammation is less clear and may be pathogenic, as these lymphoid aggregates have been proposed to be the sites containing the earliest manifestations of IBD.19, 20 Thus, α4β7 blockade could affect lymphoid aggregate development during uncontrolled intestinal inflammation, resulting in either detrimental or beneficial effects.
In order to better understand the processes lost in the absence of α4β7 function we investigated the effect of α4β7 blockade and β7 deficiency on lymphoid tissue development and lymphocyte subpopulations in the uninflamed and inflamed intestine. In contrast to current perceptions, we observed that T-lymphocyte populations in the diffuse lamina propria and lymphoid aggregates were unaffected in the absence of α4β7 function in normal mice and in the acute injury dextran sodium sulfate (DSS) colitis model. Concordant with prior observations we observed that lamina propria B-lymphocyte populations and the development of B-lymphocyte containing lymphoid aggregates were both affected by loss of α4β7 function in the normal intestine and in the acute injury DSS colitis model. Furthermore, we observed that Foxp3+ T-lymphocytes, like B-lymphocytes, were dependent on α4β7 function to localize to lymphoid aggregates. In contrast to the above observations, T-lymphocyte trafficking in T-lymphocyte-mediated colitis model was dependent on α4β7 function. Anti-α4β7 blockade did not alter intestinal T-lymphocyte populations or ameliorate acute injury DSS-induced colitis, but did improve chronic T-lymphocyte-mediated colitis as evidenced by decreased T-lymphocyte infiltration into the intestine and decreased production of inflammatory cytokines. These observations demonstrate differential dependence on α4β7 by intestinal lymphocyte subpopulations and suggest that α4β7 blockade may be a more targeted therapy than previously appreciated by selectively blocking T-lymphocyte trafficking to the intestine during chronic T-lymphocyte-mediated inflammation.
- Top of page
- MATERIALS AND METHODS
Biologic therapies targeting adhesion molecules have demonstrated efficacy in autoimmune and chronic inflammatory diseases. Currently, the most established of these therapies is an antibody directed against α4 integrins, natalizumab. α4 pairs with either β7 or β1 and therefore this strategy targets α4β7, α4β1, and their respective binding partners MadCAM-1 and VCAM-1. Natalizumab therapy effectively blocks lymphocyte recruitment into tissues and has shown promise for the treatment of multiple sclerosis and Crohn's disease.25–28 However, this therapy is complicated by a rare but significant incidence of progressive multifocal leukoencephalopathy in the treatment of both of these disorders.29, 30 MAdCAM-1 expression is relatively restricted to the mucosa, while VCAM-1 is more widely expressed, and this differential expression of the α4 integrin binding partners has raised promise that therapies selectively blocking α4β7/MadCAM-1 interactions may avoid some of the untoward effects of the more global α4 blocking strategies. Accordingly, α4β7 blocking strategies have demonstrated promise as therapies for IBD.23, 24 While it is widely accepted that α4β7/MadCAM-1 interactions play a role in lymphocyte trafficking to the intestine and that lymphocyte trafficking can affect diverse aspects of mucosal immunity, it is less well understood which aspects of mucosal immunity are entirely dependent on α4β7/MadCAM-1 interactions, and therefore would be affected by α4β7 blockade. Furthermore, recent investigations revealed a critical role for α4β7 in the development of intestinal lymphoid tissues in the uninflamed adult intestine,6 suggesting other potential effects of α4β7 blockade as a therapeutic approach. To better understand the potential effects of α4β7 blockade on the intestinal immune system we evaluated the loss of α4β7 function on intestinal lymphocyte populations and lymphoid tissues in the uninflamed and inflamed intestine.
Early studies demonstrated a role for α4β7 in the localization of lymphocytes to the intestine in the basal condition.2, 10 These studies were performed prior to the appreciation of the diversity of lymphocyte populations, and therefore did not address the dependence of specific lymphocyte subtypes on α4β7 for localization to the intestine. In addition, further observations suggested that the dependence of T-lymphocytes on α4β7 may not be universal, and that during inflammation T-lymphocytes can use alternative homing mechanisms.3 In contrast to these studies, we observed that intestinal T-lymphocyte populations in the uninflamed intestine were largely unaffected by the loss of α4β7 function. T-lymphocytes can play important roles in mucosal defense and repair during acute injury, and accordingly we evaluated intestinal T-lymphocyte populations during acute injury in the absence of α4β7 function. We also observed that intestinal T-lymphocyte populations were largely unaffected by the loss of α4β7 function during acute injury and, related to this, α4β7 blockade had little effect on the disease course in the acute injury DSS colitis model. One exception we observed related to the presence of Foxp3+ T-lymphocytes within intestinal lymphoid tissues. Foxp3+ T-regulatory cells are instrumental in controlling immune responses and maintaining tolerance in the intestine.31 Recent investigations demonstrate that human Foxp3+ T-regulatory cells express α4β7 early in life but lose α4β7 expression in adulthood,32 suggesting that this population may be less affected by α4β7 blockade. Related to this, we did not observe a decrease in Foxp3+ T-lymphocytes in the diffuse LP in the absence of α4β7 function; however, we did observe that the localization of Foxp3+ T-lymphocytes, but not Foxp3− T-lymphocytes, to intestinal lymphoid tissues was affected by the loss of α4β7 function. Therefore, in the uninflamed intestine and during acute injury α4β7 blockade had little effect on intestinal T-lymphocyte populations with the exception of inhibiting the localization of Foxp3+ T-lymphocytes to the intestinal lymphoid tissues.
The pathogenesis of IBD is incompletely understood, but largely regarded as being mediated by chronic inappropriate T-lymphocyte responses driven by luminal microbiota. To examine the effect of loss of α4β7 function in a chronic T-lymphocyte-dependent model, we evaluated intestinal T-lymphocytes in the splenocyte transfer model of chronic intestinal inflammation. We observed that loss of α4β7 function improved aspects of this disease model, including the infiltration of T-lymphocytes into the intestine and the production of inflammatory mediators. This correlates well with findings demonstrating efficacy of α4β7 blockade in the treatment of IBD,23, 24 and with prior studies demonstrating that adoptively transferred splenic T-lymphocytes use MadCAM-1-dependent interactions to adhere to the microvasculature of the chronically inflamed murine colon.11 A minority of Foxp3+ T-lymphocytes are CD25−, and therefore despite transferring T-regulatory cell depleted, CD4+ CD25−, splenocytes, a few Foxp3+ cells will be present in the recipients. Related to our above observations in the uninflamed and acutely injured intestine, we also observed that loss of α4β7 function reduced the population of Foxp3+ T-lymphocytes within the lymphoid aggregates in this chronic inflammation model.
The adult intestine has a unique ability to develop lymphoid tissues via a canonical lymphoid tissue inducer (LTi) cell-dependent pathway in response to physiologic inflammatory stimuli.33 These lymphoid tissues, termed isolated lymphoid follicles (ILFs) in the murine small intestine, resemble single-domed Peyer's patch and can act as inductive sites of the intestinal immune system by promoting homeostatic immune response in unchallenged animals.34 In the human intestine and murine colon these lymphoid tissues are generally referred to as lymphoid aggregates. How these structures develop and function during acute injury and chronic inflammation is largely unknown. Here we observed that, like the development of ILFs in the uninflamed murine small intestine, lymphoid aggregate formation in the murine colon at baseline and during acute injury is also dependent on α4β7, suggesting that the developmental requirements for ILFs or lymphoid aggregates are shared throughout the intestinal tract and during acute injury. This dependence closely correlates with the dependence of B-lymphocytes on α4β7 for their localization into the intestine. This may reflect the selectivity of intestinal B-lymphocytes for intestinal lymphoid tissues as opposed to the diffuse LP, or a requirement for B-lymphocyte priming within mucosal lymphoid tissues prior to their migration to the LP. Of note, Foxp3+ T-lymphocytes, but not Foxp3− T-lymphocytes, partially share this dependence, as in the absence of α4β7 function this population was reduced in the lymphoid aggregates, yet the overall Foxp3+ T-lymphocyte population in the diffuse LP was unaffected. This difference could be a result of differential homing properties of two Foxp3+ T-lymphocyte populations that localize to different compartments within the intestine, or alternatively could reflect a dependence on B-lymphocytes, or other cell types dependent on α4β7, within the lymphoid tissues for the localization or generation of Foxp3+ T-lymphocytes.
Lymphoid neogenesis is a feature of many chronic inflammatory conditions including infections and autoimmune diseases.35–37 The lymphoid tissues formed in these conditions could form via an LTi cell independent, noncanonical pathway, and are referred to as tertiary lymphoid tissues. These structures can be well organized, resembling lymph nodes, containing germinal centers, lymphatics, and high endothelial venules and can closely resemble the lymphoid aggregates present in the uninflamed intestine. There is also evidence suggesting that tertiary lymphoid tissues can support primary lymphocyte responses including oligoclonal B-cell expansion, somatic hypermutation, and terminal differentiation to plasma cells.38–40 Related to the current study, increased numbers of lymphoid aggregates are seen in inflammatory conditions in the human intestine including IBD,15–18 and several observations suggest that these are the sites containing the earliest manifestations of IBD.19, 20 Because of the nature of the splenocyte transfer model, we are unable to definitively address whether α4β7 blockade will block the development of all or some of the lymphoid aggregates during chronic inflammation. However, based on the observations with the acute injury model, it seems likely that α4β7 blockade will inhibit lymphoid tissue development during chronic inflammation, and this could contribute to the efficacy of this therapy.
α4β7 plays diverse roles in the intestinal immune systems, including lymphocyte homing and the development and maintenance of lymphoid tissues. Because of the selective expression of MadCAM-1, the binding partner of α4β7, blockade of this pathway has the promise to inhibit lymphocyte trafficking to the intestine while avoiding the untoward effects of more global antiadhesion molecule approaches. The findings presented here further expand this promise by demonstrating that intestinal T-lymphocyte populations are relatively unaffected by α4β7 blockade in the uninflamed intestine and during acute injury. Conversely, α4β7 blockade was effective at limiting the pathogenic T-lymphocyte population in the intestine during chronic T-lymphocyte dependent inflammation. Thus, α4β7 blockade may offer both a tissue specific and lymphocyte subset specific therapy during chronic T-lymphocyte-dependent intestinal inflammation.