|Home | About | Journals | Submit | Contact Us | Français|
Lymphocyte migration (homing) to specific tissues has an important role during protective and pathological immune responses, including inflammatory bowel diseases (IBDs). Lymphocytes use integrin α4β7 and the chemokine receptor CCR9 to localize to the gastrointestinal (GI) mucosa; their respective ligands, MAdCAM-1 and CCL25, are displayed on endothelial cells in intestinal post-capillary venules. Whereas GI-homing receptors are required for lymphocyte migration to the intestine in the non-inflamed steady state, their role during inflammation is a matter of debate. Reagents designed to block interactions between these receptors and their ligands have had variable degrees of success in animal models of IBD and patients. We discuss the mechanisms involved in lymphocyte localization to the intestinal mucosa and how they can be applied to therapy for IBD.
Lymphocytes localize to specific tissues during the protective immune response and in inflammatory disorders. Learning how these cells localize to different organs is important for understanding basic immunology as well as disease pathogenesis.
Circulating lymphocytes are exposed to extreme shear forces so they do not randomly adhere to endothelial cells; 1 instead, they express adhesion receptors for ligands expressed on endothelial cells. Adhesion usually takes place in post-capillary venules via a multistep process. First, lymphocytes are captured and loosely adhere to the endothelial cells (tethering and rolling, respectively), a step that usually requires selectins and their ligands, although the integrins α4β7 and α4β1 can also contribute to this step in some tissues. While lymphocytes are rolling they can be stimulated, generally via chemokine receptors (activation), which increases integrins' binding affinity and avidity. Integrin activation causes the lymphocytes to adhere to the endothelium (sticking) and then extravasation into non-inflamed or inflamed tissues.
Lymphocyte migration and adhesion to specific tissues are determined by the combination of receptors involved in each step, rather than a single receptor and adhesive molecule. The diversity of receptors use in each step of the adhesion process allows for versatile and tissue-specific localization of lymphocytes, making lymphocyte adhesion amenable to modulation for therapeutic purposes.
The mechanisms that regulate lymphocyte homing to different tissues have been reviewed; 2-4 we focus on lymphocyte migration to the gastrointestinal (GI) mucosa and discuss how this process might be modulated in patients, to reduce GI inflammation.
Naïve T and B cells constantly transit between the blood and secondary lymphoid organs (SLO), such as spleen, lymph nodes and Peyer's patches (PP). Upon activation in SLO, naïve lymphocytes become effector and/or memory T and B cells and express receptors that control their migration to extra-lymphoid tissues such as the skin, GI lamina propria, central nervous system (CNS), liver, and lungs 5.
Whereas migration to SLO occurs through the mechanism described above, lymphocyte migration to some extra-lymphoid tissues requires expression of specific receptors. T-cell localization the GI mucosa and the skin—the largest surfaces in the body that are exposed to the external environment—has been well characterized. T-cell migration to the skin requires ligands for P- and E-selectins, CCR4, and the integrin lymphocyte function antigen (LFA)-1 6.
In contrast to the skin, migration of T and B cells to the small intestine requires the integrin α4β7 and CCR9, whose induction depends on the vitamin A metabolite all-trans retinoic acid (RA) 3 (Figure 1). Localization to colon partially requires α4β7, but not CCR9; 7 the chemokine receptor(s) required for leukocyte migration to the colon have not been identified.
The ligand for CCR9, CCL25/TECK, is differentially distributed in a proximal-to-distal gradient in the small bowel; CD8+ T cells localize to the ileum partially via CCR9-independent mechanisms (Figure 1) 7. Alternative candidates for T-cell migration to the small bowel include CXCR3 and CXCR4, whose ligands (CXCL10 and CXCL12, respectively), are expressed in the GI mucosa 8. Consistent with an in vivo role for these alternative chemokine pathways, CXCR3-/- mice have lower numbers of CD8+ intestinal epithelial cells in the lamina propria; 9 blocking the interaction between CXCR4 and CXCL12 inhibits entry of T cells to the small intestine in steady-state and inflammatory conditions 10.
Localization of lymphocytes to the colon differs in some ways from migration to the small bowel—it requires either α4β7 or α4β1, but not CCR9 6, 11. The ligand for α4β7, MAdCAM-1, is expressed in small bowel and colon, whereas CCL25 is expressed in only the small bowel 12, 13. Moreover, whereas migration to the small intestine was impaired in CCR9-/- and β7 integrin chain-deficient (β7-/-) T-helper (Th)17, homing to the colon was reduced in only the β7-/- Th17 cells 14. Transfer of β7-/- Th17 cells into SCID mice induced less inflammation in the small and large bowel than transfer of wild-type Th17 cells (Table 1), whereas transfer of CCR9-/- Th17 induced less inflammation than wild-type cells in only the small bowel 14. Together, these data indicate that CCR9 is required for T-cell migration and pathogenicity primarily in the small intestine, whereas α4β7 is required for T-cell migration and pathogenicity in the small bowel and colon.
Migration of T cells to the intestinal mucosa also depends on their specific subset and phenotype. Recently activated CD8+ T cells require CCR9 for migration to the small bowel, whereas effector CD4+ T cells are less dependent on CCR9 for homing into this GI compartment15. Moreover, homing of CCR6-/- Th17 cells to PP and small bowel was significantly reduced, compared to wild-type Th17 cells, whereas homing of Th1 or Foxp3+ regulatory T cells (TREG) to these compartments did not require CCR6 16. RA induces α4β7 and CCR9 on TREG 17,18; however, mice given diets that did not contain vitamin A (and therefore lack RA synthesis) did not have decreased numbers of TREG in the small bowel, although Th17 cells were markedly reduced in this compartment 19. Moreover, TREG isolated from mice depleted of dietary vitamin A were equally efficient in suppressing ileitis as TREG from mice on a vitamin A sufficient diet (or that received extra vitamin A)20. However, these studies did not discriminate between thymus-derived or inducible TREG. Further studies are needed to determine the in vivo roles of RA in localization of TREG in the GI mucosa and their immunoregulatory functions there.
Another example of differential gut-homing requirements is cells that secrete immunoglobulin (Ig)A (IgA-secreting cells). CCR10 is expressed primarily by IgA-secreting cells, whereas cells that secrete antibodies against IgG or IgM do not express this receptor 21. Moreover, IgA-secreting cells require CCR10 to localize to the intestine, although CCR10 is not expressed on gut-associated T cells 4, 21. In addition to CCR10, IgA-secreting cells also express CCR9; mice deficient in this receptor have reduced numbers of these cells in the small intestine 22. Interestingly, blocking CCR9 prevented localization of IgA-secreting cells to the small bowel, whereas blocking either CCR10 or its mucosal ligand, CCL28, impaired their localization to small and large bowel 23. Similar to T cells, RA was sufficient to induce CCR9 and α4β7 (but not CCR10) on activated B cells and mice depleted of vitamin A had very low numbers of small bowel IgA-secreting cells. 19, 24, 25
IBDs, which include Crohn's disease (CD) and ulcerative colitis (UC), are associated with a massive influx of immune cells into the GI tract. During disease development, altered production of pro-inflammatory cytokines induces expression of alternative adhesion receptors and chemokines on intestinal endothelial cells, which might allow lymphocytes to migrate to the intestine without expression of the receptors that normally regulate their localization in that compartment 11. These alternative pathways of lymphocyte recruitment might have important implications for IBD therapy.
Studies of animal models and human tissues have indicated the role for gut-homing effector T cells in IBD pathogenesis. In experimental models of IBD, MAdCAM-1 is upregulated in the intestinal lamina propria.26-28,29 Similar MAdCAM-1 upregulation is observed in active inflamed tissues from patients with CD or UC, which is associated with increased numbers of α4β7+ T cells, compared with normal tissues 30, 31. Deficiency of β7 integrin subunit inhibits inflammation in a mouse model of CD 32 and antibodies against β7 or MAdCAM-1 reduced inflammation in mice with TNBS-induced or cell transfer-induced colitis 33,34 (Table 1). Blocking α4 or α4β7 reduced colitis in a non-human primate model of IBD 35, 36. Additional mechanistic insights have come from studies of SAMP1/YitFc mice, which spontaneously develop CD-like ileitis. Although SAMP1/YitFc mice deficient in β7 have reduced intestinal inflammation 37, antibodies that block α4β7 or MAdCAM-1 did not reduce the inflammation; only combined blockade of VCAM-1 and MAdCAM-1 significantly improved ileitis 29.
Blocking α1β1, a collagen-binding integrin that is upregulated in inflamed tissues, reduced TNBS-induced colitis in mice 38. Results from mice with dextran sodium sulphate (DSS)-induced colitis have varied results—some studies reported that blockers of MAdCAM-1 reduced inflammation,33, 39, 40 whereas others showed that development of colitis required α4β1–VCAM-1 and β2–ICAM-1 interactions, but not α4β7–MAdCAM-1 41, 42. Pathogenic effector T cells might not require only interaction between lymphocyte α4β7 and epithelial cell MAdCAM-1 to promote chronic inflammation, but other integrins that mediate immune cell localization during general inflammation might fulfill redundant roles in intestinal pathology.
SAMP1/YitFc mice have increased numbers of IgA-secreting cells in the MLN and LP 43. Adoptive transfer of B cells and T cells from SAMP1/YitFc into SCID mice increased ileitis, compared with transfer of only T cells 43. Moreover, B cells required α4β7 to exacerbate ileitis 37, indicating that B-cell localization to the GI tract might also be involved in IBD pathogenesis.
TREG cells are believed to prevent or even cure intestinal inflammation, based on studies from different models of IBD. However, the precise role of α4β7 and CCR9 in trafficking and function of TREG cells during GI inflammation is unclear. TREG cells seem to require β7-independent pathways—rather, those that involve CCR7 and CCR4 —for their immune suppressive functions and to prevent colitis 44-46. These alternative migratory pathways might allow TREG cells function in lymphoid compartments, such as prophylactic suppression before the onset of inflammation in MLN or PP. However, TREG cells might need GI homing receptors to suppress immune activity in the lamina propria during active inflammation 47, which is probably most relevant for development of therapeutics.
CCL5 and CCR5 are also up-regulated in experimental models of ileitis and mediate the specific recruitment of TREG cells and some subsets of effector CD4+ T cells48. These alternative chemokine pathways could account for the observation that blocking CCL25 or CCR9 is only effective at early stages of disease, even though expression of CCL25 increases in the small bowel of patients with CD. 49 Other, perhaps non GI-specific, chemokine signals might mediate lymphocyte homing at later stages during inflammation 50. Human lamina propria and intraepithelial lymphocytes also express CXCR3, CX3CR1, and CCR2, and levels of their ligands are increased in tissues of patients CD 51. Moreover, recruitment of T cells, monocytes, and DC to the inflamed mucosa might also involve CX3CL1 and its receptor CX3CR1, which contributes to pathogenesis of IBD 52-54.
Some of the extra-intestinal pathologies associated with IBD might arise from aberrant homing of immune cells. For instance, MAdCAM-1 and CCL25 are upregulated in the liver during primary sclerosing colangitis (PSC), a chronic disease characterized by progressive inflammation and scarring of the bile ducts. PSC has been associated with UC in epidemiologic studies 55.
Patients with IBD usually require life-long therapy with corticosteroids and other immunosuppressive drugs. Choice of therapy depends on the primary clinical goal (induction or maintenance of remission), the extent and severity of disease, the response to current or prior treatments, and the occurrence of side effects (summarized in Supplementary Table 1A). Many drugs for IBD can have serious adverse effects, and some patients become refractory to treatment during disease progression and require surgery. Therefore, new therapeutic approaches, that target specific inflammatory mediators, are needed.
Although the primary causes of IBD are not clear, many molecules that are involved in disease pathogenesis have been identified as targets for therapy. Therapeutics that have been developed include inhibitors of T-cell activation, co-stimulatory pathways, pro-inflammatory cytokine receptors, Th1 polarization, cytokines and their regulatory proteins, growth hormone, and growth factors (Supplementary Table 1B, C).
Although many of these biologic agents showed efficacy in preclinical studies, most of them have not shown efficacy in clinical trials, or have caused significant side effects 56. Antibodies to the cytokine TNF-α have been the most effective, and are currently used to treat patients with refractory moderate-to-severe active CD or UC. Infliximab binds the soluble bioactive and membrane-bound forms of human TNF-α and reduces its toxicity. Although Infliximab is effective in reducing the symptoms of IBDs, its immunosuppressive effects predispose patients to infections and increase risk of malignancies, such as lymphomas 57, 58.
Molecules that mediate lymphocyte localization to the GI mucosa, in the steady state or during development of inflammatory diseases such as IBDs, are attractive targets for drug development. Antibodies or compounds that selectively block homing receptors, 11 or reagents that sequester lymphocytes in secondary lymphoid organs (to prevent their migration to sites of inflammation), 59 have shown efficacy in animal models and in clinical trials for psoriasis 60, 61, asthma 62, graft-versus-host disease, 63 and multiple sclerosis 59.
Because α4β7 and CCR9 are the primary mediators of lymphocyte migration to the intestine, reagents that block their function should reduce inflammation in the intestinal mucosa, yet cause low levels of systemic immunosuppression. Agents developed for treatment of IBD disrupt interactions between LFA-1 and ICAM-1, α4β1 and VCAM-1, as well as α4β7 and MAdCAM-1 (Figure 2, Table 2).
The first successful clinical use of anti-ICAM-1 was in treatment of rheumatoid arthritis 64. Antibodies against ICAM-1 or antisense oligonucleotides that disrupt expression of ICAM-1 showed efficacy in mouse models of IBD, including DSS-induced colitis and SAMP-1/Yit mice ileitis65 66. Interestingly, in SAMP-1/Yit mice, anti-ICAM-1 was only effective when administered in combination with anti-VCAM-1 or anti-α4 integrins, indicating redundancy between LFA-1/ICAM-1 and α4β1/VCAM-1 pathways during inflammation in mice. Clinical trials that investigated the effects of reagents against ICAM-1 in patients with IBD included investigation of Alicaforsen (ISIS 2302), an antisense oligonucleotide that prevents expression of ICAM-1. Whereas an early-stage clinical trial suggested a therapeutic potential for Alicaforsen in patients with mild-to-moderate active CD 67, two subsequent, larger, multicenter trials failed to demonstrate significant efficacy 68 69. Despite this setback, patients treated with Alicaforsen, in an enema formulation, had significant improvements in distal UC in a randomized, placebo-controlled trial. 70 However, given the important role of LFA-1 interaction with ICAM-1 in leukocyte localization to many lymphoid and non-lymphoid tissues, as well as in T-cell activation, 11 it is likely that blockers of ICAM-1 will cause significant systemic immunosuppression.
Natalizumab is a recombinant, humanized, monoclonal IgG4 against α4; it inhibits MAdCAM-1 signaling through integrin α4β7 and VCAM-1 signaling through integrin α4β1 71. In placebo-controlled, randomized trials, 40% patients with moderate-to-severe CD responded to Natalizumab and went to remission, compared to 8% in the group that received placebo 72.
However, phase III clinical trials that included clinical response, remission, and maintenance as endpoints showed that the drug was more effective when given in combination with other immunosuppressants or before therapy with an anti-TNF-α reagent 73, 74. Since blockers of α4 probably do not affect T cells that have already localized to intestinal tissues, Natalizumab might not be sufficient to reduce ongoing inflammation—its combination with other immunosuppressant drugs might be required. Natalizumab has also shown potential for treatment of UC, 75 probably due to its ability to block α4β7 and α4β1 (integrins involved in localization of lymphocytes to the colon).
Natalizumab has also been used to treat patients with multiple sclerosis, 76 based on the role of α4β1 interaction with VCAM-1 in leukocyte homing to the CNS and experimental allergic encephalomyelitis 77. However, cases of progressive multifocal leukoencephalopathy (PML)—a rare and often fatal opportunistic infection of the CNS—have developed in some patients given Natalizumab (approximate incidence 1/1000), raising concerns about its safety 78, 79. Sporadic cases of melanoma have also been reported in patients treated with Natalizumab, which might be associated with impaired immunosurveillance of the skin following α4β1 blockade 80, 81. However, larger cohorts of patients that have received Natalizumab need to be studied to determine more precisely the incidence of these rare side effects. Similar safety concerns might also apply to an orally bioavailable inhibitor of α4 (AJM300), which has also shown to be effective in patients with active CD 82. Some cases of PML also occurred in patients with psoriasis who were treated with Efalizumab, a monoclonal antibody against LFA-1 83. Reagents that selectively block homing of lymphocytes to the GI without affecting immunosurveillance in other tissues (including the CNS), are urgently required for IBD.
Vedolizumab (MLN-02) is a humanized monoclonal IgG1 against integrin α4β7. In a phase II trial in 181 patients with UC, remission rates were significantly higher among subjects treated with Vedolizumab than those given placebo 84. Another placebo-controlled trial of 185 patients with mild-to-moderate active CD showed that Vedolizumab was significantly more effective than placebo at inducing remission in patients with CD 85.
A monoclonal antibody against MAdCAM-1 (Pf-0054,659, human IgG2) is being tested in a phase I/II clinical trial of patients with UC. Although the study includes a small number of patients, endoscopic examinations identified improvements among in patients treated with anti-MAdCAM-1, without major side effects 86.
Nevertheless, because α4β1 might also have a role in chronic intestinal inflammation 29, 41, 42, it is possible that selective targeting of α4β7 or MAdCAM-1 might not be as effective as reagents designed to block both α4 integrins.
The effects of blocking the interaction between CCL25 and CCR9 were demonstrated in SAMP-1/Yit mice with ileitis. Blocking either CCL25 or CCR9 during early stages of ileitis reduced inflammation, whereas no effect was observed when mice were given the reagents at later stages of disease progression 50. Moreover, administration of CCX282 (Traficet-EN), an orally bioavailable antagonist of CCR9, reduced the inflammatory response in mice when given before or after gut inflammation induced by TNF-α overexpression 87.
CCX282 is being tested in trials of patients with CD and refractory celiac disease (Figure 2, Table 2); preliminary efficacy and safety evaluations look promising 88. In a phase II trial, 74 patients with CD were given either CCX282 or placebo for 28 days. Fifty-eight percent of patients had a significant reduction in CD scores, compared with 31% in the placebo group; in the CCX282 group, the effect that was associated with reduced levels of pro-inflammatory cytokines and C-reactive protein 88. A subsequent phase II/III trial showed significant improvement in the CCX282 group; disease scores were reduced in 81% of patients and 41% experienced clinical remission—effects that were maintained even upon withdrawal of corticosteroids 88.
Results from larger phase III trials of CCX282 for CD and UC are pending. Although the drug was generally well tolerated and not associated with an increased risk of infection, a large cohort of patients must be followed for long time period to exclude risk for rare diseases such as PML. The fact that CCX282 is orally bioavailable offers a clear advantage to parenteral therapies, decreasing the cost of the treatment, eliminating morbidity associated with parenteral administration, and potentially increasing compliance.
Interestingly, tolerogenic plasmacytoid DC also express CCR9, 89 which seems to be required for localization of these cells to the small bowel 90. Inhibitors of CCR9 might affect migration of plasmacytoid DC and their tolerogenic functions in the intestine, with potential effects that should be considered and explored in models of IBD pathogenesis.
Although the exact cellular and molecular mechanisms of IBD pathogenesis are undefined, lymphocyte homing has an important role. Improved understanding of lymphocyte localization to the non-inflamed and inflamed intestinal mucosa has led to specific and effective therapies for IBD and improved the benefit–risk profile for patients.
Nevertheless, results from studies of animal models of IBD have identified alternative homing receptors that, in addition to α4β7 and/or CCR9, have roles in lymphocyte migration to the GI tract and might contribute to inflammation; these redundant homing pathways could account for the varying degrees of effectiveness of reagents that target GI-specific homing receptors in clinical trials. It is important to better define which receptors, adhesion molecules, and chemokine pathways contribute to chronic intestinal inflammation in humans. It is also important to determine the precise role of TREG cells in intestinal inflammation and whether GI-specific reagents that interfere with lymphocyte adhesion affect TREG functions. Combination therapies, which target more than one step in adhesion of lymphocytes to the intestinal epithelium, might be the most effective strategy for IBD. Combinations such as anti-α4β7 and antagonists of CCR9 could have additive effects to reduce inflammation in patients with IBD.
Funding: We are grateful to Allison McNulty for editorial assistance. JRM is indebted to Ingrid Ramos for constant support. EJV was supported by grants from Crohn's & Colitis Foundation of America (CCFA). UHvA was supported by grants from NIH. JRM was supported by grants from CCFA, Cancer Research Institute (CRI), Howard H. Goodman (MGH), Massachusetts Life Science Center (MLSC) and NIH DP2 2009A054301.
Author's contributions: EJV, BC, UHvA and JRM contributed equally in writing this review article.
Disclosures: The authors declare that no conflict of interest exists and have no other disclosures to make.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.