Highly specific blockade of CCR5 inhibits leukocyte trafficking and reduces mucosal inflammation in murine colitis

Targeted disruption of leukocyte trafficking to the gut represents a promising approach for the treatment of inflammatory bowel diseases (IBDs). CCR5, the shared receptor for MIP1α and β and RANTES, is expressed by multiple leukocytes. Here, we aimed to determine the role of CCR5 in mediating leukocyte trafficking in models of colitis, and evaluate the therapeutic potential of maraviroc, an orally active CCR5 antagonist used in the treatment of CCR5-tropic HIV. Acute and chronic colitis were induced by administration of DSS or TNBS to wild-type and CCR5−/− mice or adoptive transfer of splenic naïve CD4+ T-cells from wild type or CCR5−/− mice into RAG-1−/−. CCR5 gene ablation reduced the mucosal recruitment and activation of CCR5-bearing CD4+ and CD11b+ leukocytes, resulting in profound attenuation of signs and symptoms of inflammation in the TNBS and transfer models of colitis. In the DSS/TNBS colitis and in the transfer model, maraviroc attenuated development of intestinal inflammation by selectively reducing the recruitment of CCR5 bearing leukocytes. In summary, CCR5 regulates recruitment of blood leukocytes into the colon indicating that targeting CCR5 may offer therapeutic options in IBDs.

In addition to regulating leukocyte homing to the inflamed mucosa, CCR5 functions as a major co-receptor for HIV entry into target cells, and selective CCR5 antagonists have been developed that inhibit the replication of CCR5-tropic viral strains 7,8 . Maraviroc, is a small molecule that induces a non-competitive, slowly reversible, inhibition of CCR5 and displays therapeutic efficacy against CCR5-tropic HIV infection [9][10][11] . In addition, by blocking the signaling of all three CCR5 ligands, maraviroc effectively inhibits the migration and effector functions of CCR5-bearing leukocytes exerting anti-inflammatory and immune-modulatory effects [12][13][14][15] .
In the current report, we demonstrate that recruitment of CCR5-expressing leukocytes in the colon is essential for the onset and maintenance of inflammation in mouse models of colitis and that, maraviroc, a small molecule inhibitors of CCR5, protects against colitis development targeting CCR5-bearing leukocytes. These data suggest that CCR5 inhibitors could be used to modulate leukocyte trafficking in human IBDs.

Materials and Methods
Animals. CCR5 −/− (B6.129P2-Ccr5 tm1Kuz /J), CCR5 +/+ , CX3CR1 +GFP on C57BL6 background and Rag-1 −/− mice were from the SPF animal facility at A* STAR, Singapore. The study was carried out in strict accordance with the guidelines of the Institutional Animal Care and Use Committee (IACUC) of the Biological Resource Centre (BRC) of Biopolis in Singapore or University of Perugia, Italy. The BRC IACUC protocol was approved by the (F) Assessment of tissue biomarkers by Bio-Plex ® Multiplex immunoassay. Data were normalized to total tissue proteins. In each panel at least 5 animals per group are included. Markers for the Th1 phenotype are labeled in blue, for Th2 phenotype in violet, and for regulatory phenotype/chemokine in black. * P < 0.05 versus control; * * P < 0.05 versus DSS.  (MHC-II − /Gr-1 + /CD11b high ) in untreated and TNBS coliticmice that were treated or not by oral administration of maraviroc. Data shown mean ± SE n = 3 per group (* P < 0.05, * * P < 0.01, * * * P < 0.005; two-tailed, unpaired Student's t test).
Scientific RepoRts | 6:30802 | DOI: 10.1038/srep30802 DSS And TNBS Colitis Models. DSS colitis was induced by oral administration of 5% DSS in drinking water to weight matched, 4-6 weeks old male CD1 mice for 5 or 8 (not shown) days. Mice were randomized to receive DSS alone or DSS in combination with maraviroc at dose of 5, 25 or 50 mg/kg/day or prednisolone 5 mg/kg/day 16 , starting on day 1 after DSS. The dose of prednisolone used in this study has been demonstrated effective in reducing inflammation in rodent models of colitis and is roughly similar to the dose used in clinical setting (1 mg/kg) 16 . Animals were monitored daily for weight loss and fecal score (0, normal; 1, soft but still formed; 2, very soft; 3, diarrhea; and 4, liquid stool). The criteria/scoring system for assessing macroscopic damage was as follows; 0, no ulcer, no inflammation; 1, no ulcer, local hyperemia; 2, colon wall thickening/edema; 3, ulceration and inflammation at one site only; 4, two or more sites of ulceration and inflammation. Additional points were scored for the presence of indurations, intestinal adhesions, and each additional site of inflammation/ ulceration > 1 cm 2 in surface area.
For TNBS colitis, weight matched, 4-6 weeks old male wild-type and CCR5 −/− mice were pretreatedwith 100 μ l TNBS pre-sensitization solution (1% TNBS dissolved in 4:1acetone:olive oil solution) which was applied to shaved abdominal skin on day − 7 17 . Control mice were subjected to the same protocol but using pre-sensitization solution that lacked TNBS. After 7 d, the mice were anesthetized by i.p. injection with ketamine/xylazine solution (100 mg/ml ketamine and 20 mg/ml xylazine in saline administered at 100 μ l/10 g of body weight) and then treated with TNBS (0.5 mg per mouse) which was dissolved in 50% ethanol and administered intra-rectally using a 3.5 French catheter equipped with a 1 ml syringe. To ensure uniform distribution of TNBS throughout the colon and cecum, mice were held in a vertical position for 30 s after instillation. Mice were randomized into either the control group (TNBS only, n = 6) or treatment group (TNBS plus 50 mg/kg/d maraviroc per os, n = 5-6) from day 1 after TNBS administration, and then monitored daily as described above.
ELISA assay. Lamina propria mononuclear cells (LPMC) were obtained from mice with TNBS-induced colitis (6 days duration) and seeded into 96 well plates (200,000 cells/well) for stimulation with LPS (5 μ g/ml) in RPMI complete medium for 36 h at 37 °C 16,17 . Alternatively, CD4 + cells were sorted from the MLN after 3-4 d duration of TNBS colitis and then seeded into 96 well plates (100,000 cells/well) for stimulation with plate-bound anti-CD3 (5 μ g/ml) and soluble anti-CD28 (2 μ g/ml) in RPMI complete medium for 72 h at 37 °C. After incubation, the culture supernatant was analyzed for cytokine content by ELISA. Measurement of cytokine/chemokine release in the DSS model was achieved using a Bio-Plex ® Multiplex Immunoassay kit (BioRad, Italy).
Statistical analysis. Data are expressed as mean ± standard error. Two-tailed, unpaired Student's t tests were used to compare 2 groups of data, as indicated in the respective figures. P < 0.05 was considered significant. When more than two groups were considered, one way ANOVA followed by Bonferroni was used. GraphPad Prism software version 5.0 was used to prepare the graphics and perform all statistical analyses (GraphPad Software, San Diego, CA).

Maraviroc Induces A Dose-Dependent Decrease In Mucosal Inflammation In DSS Colitis.
In order to assess the efficacy of CCR5 blockade as a therapeutic approach in colonic inflammation, mice administered with DSS were treated with maraviroc (5, 25 or 50 mg/kg/day) or prednisolone (5 mg/kg/day) for the following 5 days. At a dose of 50 mg/kg/day, maraviroc attenuated the development of signs and symptoms of colitis with an efficacy that was comparable to 5 mg/kg/day prednisolone (Fig. 1). All three doses of maraviroc effectively blocked the recruitment of leukocytes into the colon, as assessed by measurement of MPO activity, although some of the therapeutic benefit of maraviroc treatment appeared to be lost at the lower doses ( Fig. 1A-D). Maravirocalso protected against development of macroscopic and microscopic damage and colonic shortening, and was as effective as prednisolone at preventing DSS-induced changes in colonic levels of IL-2, IL-4, IL-5, IL-6, IL-17 and GM-CSF thought the some of these effects were likely of minor mechanistic relevance (Fig. 1E,F). These protective effect persisted for up to 8 days (data not shown).

CCR5 Gene Ablation Or Maraviroc Therapy Reduces Inflammation In TNBS Colitis.
We next investigated the involvement of CCR5 in the pathology of a T-cell-driven colon inflammation using a TNBS-induced model of colitis in both wild-type mice and CCR5 −/− animals that received either placebo or maraviroc treatment (50 mg/kg for 5 days, starting one day after colitis induction) ( Fig. 2A-C). In wild-type animals, TNBS-induced colitis was associated with a robust influx of CD45 + CD11b + leukocytes into the LP (Fig. 2D,E). Analysis of the expression of MHC-II and myeloid differentiation antigen Gr-1 in these cells allowed the identification of three major subsets (Fig. 2F): population 1 comprised MHC-II + Gr-1 − CD11b int cells (P1; putative macrophages), whereas population 2 was made-up of MHC-II − Gr-1 − CD11b low cells (P2; putative monocyticcells), and population 3 cells were MHC-II − Gr-1 + CD11b high (P3; putative granulocytes). Further analysis of Ly6C, Ly6G, CD11c and CX3CR1 expression among the LP infiltrating leukocytes indicated that population P2 included a substantial fraction of Ly6C+ i.e. blood derived monocytes, and that exposure to TNBS increased substantially the proportion of these cells (Table 1 and Supplementary Figure 1) 19,20 . In wild-type animals, leukocyte infiltration of the colon was associated with development of phenotypic features of acute colitis ( Fig. 2A-C). In contrast, CCR5 −/− mice exhibited milder weight-loss, reduced severity of diarrhea, and only limited thickening of the colon wall compared with CCR5 +/+ animals ( Fig. 2A-C 1, 2 and 3. Flow-cytometric analysis of the myeloid cell markers Ly6-G, Ly6-C, CD11c and CX3CR within population P1 cells (MHC-II + GR-1 + CD11b int ), P2 cells (MHC-II − GR-1 − CD11b low ) and P3 cells (MHC-II − GR-1 + CD11b high ). CX3CR1 expression was determined by analysis of the colonic leukocyte infiltrate in CX3CR1 +GFP mice. Values indicate mean ± standard error of n = 3 per group (* P < 0.05, two-tailed, unpaired Student's t test). % of positive cells; ND, not detected.
rescued wild-type animals from the development of wasting disease, diarrhea and macroscopic inflammation ( Fig. 2A-C; P < 0.05 versus DSS wild type). Consistent with the concept that CCR5 blockade achieves these beneficial effects by impairing the mucosal recruitment of multiple leukocyte lineages, we also observed that disruption of CCR5 signaling either by gene ablation or small molecule antagonism significantly attenuated the influx of myeloid cells into the LP (Fig. 2D-F).

CCR5 Expression In Both The Innate And Adaptive Immune Compartments Contributes To
Murine Intestinal Inflammation. Further analysis of the CD11b + myeloid cells detected in wild-type colitic mice revealed an increase in CCR5 expression among these cells after exposure to TNBS (Fig. 3A). This increase in CCR5 expression in the myeloid cell compartment was accompanied by a shift away from the predominant macrophage population detected in the steady-state (CCR5 + CD11b int MHC-II + Gr-1 − cells), towards an inflammatory infiltrate made up primarily of monocyte-like cells (CCR5 + CD11b low MHC-II − Gr-1 − ) ( Fig. 3B; * p < 0.05, * * * p < 0.005). CCR5 antagonism using maraviroc therapy abrogated the accumulation of CCR5 + CD11b + leukocytes in the inflamed colonic mucosa (Fig. 3A and Supplementary Figure 2, * p < 0.05), and was associated with a robust reduction of pro-inflammatory mediators including TNFα , IL-6 and IL-1β (Fig. 3C). Importantly, CCR5 also contributed to the mucosal recruitment of pro-inflammatory T cells. After 6 days of TNBS administration, we detected a substantial number of CD4 + cells in the LP, but only in mice that were sufficient in CCR5 (Fig. 4A). CD44 expression has previously been described as a marker of memory T-cells with tissue homing potential 21 .
CCR5 is considered to be a phenotypic marker for effector/memory T-cells 22,23 . Consistent with this view, CD4+ T-cells with a CD44 high 'memory' phenotype exhibited stable high expression of CCR5 in the mouse LP (Fig. 4B). Furthermore, TNBS colitis associated with increased levels of CCR5 expression within the T-cell pool, but this was largely restricted to the CD44 low immature subset, suggesting maturation of these cells towards an effector/memory phenotype during inflammation (Fig. 4B). CCR5 upregulation by colitis in CD4+ CD44 low cells was abrogated by co treatment with maraviroc (Fig. 4B,C).

Mucosal Inflammatory Responses Of CCR5 + Leukocytes Exhibit An Effector/Memory Th17
Profile. Within the CD44 high subset of LP CD4 + T-cells, we further identified that CCR5 + cells were substantially larger in size than CCR5 − cells in both the steady-state and during inflammation, suggesting selective cell cycle entry and clonal expansion of the CCR5 + population in this model 24 . Similarly, while the CD44 low subset of CD4 + LP exhibited fairly uniform size in the steady-state, there was a robust increase in the size of the CCR5 + cells following exposure to TNBS (Fig. 5A). These data are consistent with the known ability of chemokines to influence T-cell differentiation 25 .
Previous reports suggest that CCR5 is selectively expressed on IFN-γ producing Th1 cells and allowing targeted recruitment of these cells to inflammatory sites 23,24 . Therefore we next sought to determine whether CCR5-expressing leukocytes in the TNBS model produce cytokines consistent with a Th1 phenotype. For this purpose, we FACS-sorted CCR5-and CCR5+ populations of CD4+ T-cells from the LP and MLN of colitic mice and then stimulated these cells with anti-CD3/CD28 mAb. Of relevance, we found that almost all colonic CCR5 + CD4 + T-cells died during the 3-day stimulation period (data not shown), likely due to activation-induced cell death, whereas there was no difference in the rate of cell death and proliferation detected among MLN-derived cells (data not shown and Fig. 5D, respectively). Assessment of cytokine levels in culture supernatants of MLN-derived cells demonstrated that CCR5 + CD4 + cells displayed a profile consistent with Th17 polarization during TNBS colitis (Fig. 5E).

Leukocyte Expression Of CCR5 Is Required For Both Acute And Chronic Murine Colitis.
The role of CCR5 in the context of chronic inflammation was investigated using the adoptive transfer model of T-cell-mediated colitis in animals grafted with either CCR5 +/+ or CCR5 −/− T-cells. Adoptive transfer of naïve CD4 + CD45RB high T-cells from wild-type miceinto RAG1 −/− recipients conferred gut pathology characterized bya severe weight loss, diarrhea, macroscopic inflammation and colonic thickening (Fig. 6A-D). In contrast, mice that were transferred with naïve CD4 + CD45RB high T-cells that lacked CCR5 exhibited milder disease and fewer total colonic leukocytes as well as reduced infiltration of CD4 + effector/memory T-cells into the gut mucosa (Fig. 6E). Indeed, unlike CD4 + T-cells from wild-type donors, CCR5 −/− CD4 + T-cells appeared unable to efficiently mature into effector cells and exhibited significantly reduced production of IFNγ and IL-17 when analyzed ex vivo (Fig. 6F). A similar immune profile was also observed in the MLN of mice transferred with CCR5 −/− CD4 + T-cells ( Supplementary Figures 3 and 4). Together, these data indicated that CCR5 ablation in CD4 + T-cells abolishes a Th1/Th17-type inflammatory pathology of the gut which resembles that of human IBD 26 .

CCR5 blockade attenuates sign and symptoms of chronic colitis.
To investigate whether the therapeutic effects observed in acute models of colitis was maintained over time, Rag1 −/− rendered colitic by transfer of CD4 + T cells from wild type donors, were administered with maraviroc 50 mg/kg/d for 3 weeks starting on day 34 after colitis induction. As shown in Fig. 7 and Supplementary Figure 5, we found that maraviroc attenuates colon inflammation in this setting and reduced signs and symptoms of colitis, as well as the inflow of CD45+ cells into the LP. Analysis expression of markers of activation demonstrated that long term administration of maraviroc reduced the number of IFNγ + cells, but had no effect on IL-10 and TGFβ stained cells. Altogether these data confirm that the maraviroc attenuates inflammation in chronic models of colitis.

Discussion
In the present study we provide evidence that CCR5, a chemokine receptor expressed by different leukocytes exerts a critical role in the development of acute and chronic inflammation in murine models of colitis. Analysis of CCR5 expression on LP leukocytes in mice administered TNBS revealed that colitis development in this modelis mediated by the recruitment of three major leukocyte populations: MHC-II + Gr-1 − CD11b int cells (putative macrophages), MHC-II − Gr-1 − CD11b low cells (likely monocytes/macrophages), and MHC-II − Gr-1 + CD11b high cells (putative granulocytes). A further characterization of these cells, revealed that the monocytes/macrophages compartment in the TNBS model is largely composed by Ly6C+ cells. Previous studies have shown that in the steady state the intestinal monocytes/macrophages population in mice is maintained by the continuous recruitment of Ly6C+ monocytes (CD14+ cells in humans) from the blood stream 25 . Under resting conditions, these monocytes undergo local differentiation into reparative, anti-inflammatory, macrophages (MHCII+ CD11b Int and CX3CR1 + ) that are highly phagocytic, hypo-responsive to pro-inflammatory stimuli producing large amounts of IL-10 in inflammatory milieus. The maturation of these Ly6C high blood monocytes into tissue-resident macrophages (MHCII pos Gr1 neg CD11b + ) is governed by signals received from the local micro environment, and involves cell transition through several intermediate phenotypes. However, the normal pattern of monocyte differentiation is disrupted during inflammation, leading to the recruitment and accumulation of Ly6C + MHCII pos CX3CR1 int effector monocytes in the intestine, where theycan generate TNFα and other soluble mediators. In the present study, we observed that in the steady state themonocytic cell population of the colonic LP is largely made up byLy6C + cells (up to 60%), and the predominance of Ly6C+ cells in this compartment increases further in the TNBS colitis (> 90% of the monocytic cell infiltrate), strongly indicating the selective recruitment of Ly6C+ effector monocytes from the blood stream 23,24 . Consistent with this view, while only 10% of gut monocytic cells expressed CCR5 in the steady state, the proportion of CCR5+ cells increased to ~70% in response to TNBS, supporting the notion that colitis development in this model is supported by recruitment of circulating Ly6C + monocytes via a mechanism that involves CCR5. Importantly, these Ly6C + CCR5 + cells displayed a strong polarization towards an effector phenotype and produced pro-inflammatory cytokines including IL-1β , IL-6 and TNFα .
In addition to regulating the trafficking of innate leukocytes, CCR5 is also critically required for the recruitment of circulating T-cells to the intestinal mucosa. Thus CCR5 −/− mice were essentially unable to recruit CD4 + cells into the LP in response to TNBS 27 . Phenotypic and functional characterization of LP lymphocytes demonstrated that ~15% of CD4 + cells in the gut express CCR5 in the steady state, and that exposure to TNBS promoted the recruitment of additional CCR5+ CD4+ T-cells with robust polarization towards a Th17 profile. These data underlined a broad role of CCR5 in regulating the mucosal recruitment and functional maturation of both innate immune cells and T-cells arriving from the blood. Consistent with the data obtained from chemical colitis models, the adoptive transfer of naive CD4 + CD45RB high T-cells into RAG-1 −/− mice demonstrated that CCR5-deficent T-cells were unable to replicate the chronic colitis that was induced by CCR5-sufficient T-cells. Transplantation of CCR5 −/− T-cells associated with a substantial reduction in the number of CD4 + and CD45 + cells recruited in the LP of grafted animals, as well as reduced weight loss, lower diarrhea score, milder macroscopic inflammation and reduced colonic shortening. In contrast to the acute chemical colitis models, attenuation of gut inflammation by CCR5 gene ablation, however, was due largely to the impaired infiltration of CD4 + T-cells and did not depend on changes in the CD11b + myeloid compartment. In addition to disrupting the mucosal recruitment of CD4 + T-cells, genetic ablation of CCR5 also reducedthe proportion of cells with an activated phenotype (CD44 high CD62L − ), and impaired the ability of these cells toproduce IFNγ and IL-17 without any discernable loss of immuno regulatory function (IL-10+ TGF-β + CD4+ T-cell frequency).
The critical role of CCR5 in regulating leukocyte trafficking in intestinal inflammation was confirmed by the protective effects exerted by maraviroc in three model of colitis, i.e. the acute colitis induced by DSS and TNBS and in the chronic colitis induced by transfer of CD4+ T cells into Rag1 −/− mice. Maraviroc is an orally active, allosteric modulator of CCR5 that binds to the extracellular portion of the receptor, thereby preventing the conformational changes required for signal transduction after engagement of physiological ligands (CCL3, 4 and 5) or viral proteins (HIV GP120) 8,9 . Here, we have shown that administration of maraviroc attenuates gutinflammation in rodent models of colitis by selectively depleting CD11b + CCR5 + cells from the colonic mucosa (in the TNBS model) and modulating the production of signature cytokines such as IL -1β , IL-6 and TNFα in both DSS-induced and TNBS models of colitis. Of relevance, the therapeutic effect of maraviroc was maintained in the chronic model of colitis resulting in a robust attenuation of recruitment of CD45+ /CD11b+ cells in the LP.
While the data we report strongly support a role for CCR5 in rodent colitis, CCR5 gene mutations do not overtly modulate human IBD phenotypes, as demonstrated by the fact thata common 32-bp 'loss-of-function' deletion (Δ 32) of CCR5 fails to confer protection against IBD [28][29][30][31][32][33] . In contrast, the Δ 32 mutation confers protection against development of primary sclerosing cholangitis in IBD 34 . These data suggest that alternative mechanisms of leukocyte trafficking to the gut may compensate for the absence of CCR5 signaling in IBD patients with this mutation, and conversely that the benefits of maraviroc therapy may be restricted to IBD patients with a functional CCR5 gene.
In conclusion, we have provided compelling evidence that CCR5 mediates the trafficking of both innate and adaptive immune cells in rodent models of colitis. Pharmacological inhibition or genetic ablation of CCR5 rescued mice from colitis in both acute and chronic models, hence the clinically-approved small molecule antagonist of CCR5, maraviroc may represent a novel approach to reducing the mucosal trafficking of blood leukocytes in human IBDs.