Chemokine receptor 5 blockade modulates macrophage trafficking in renal ischaemic‐reperfusion injury

Abstract Chemokine receptor 5 (CCR5) is a pivotal regulator of macrophage trafficking in the kidneys in response to an inflammatory cascade. We investigated the role of CCR5 in experimental ischaemic‐reperfusion injury (IRI) pathogenesis. To establish IRI, we clamped the bilateral renal artery pedicle for 30 min and then reperfused the kidney. We performed adoptive transfer of lipopolysaccharide (LPS)‐treated RAW 264.7 macrophages following macrophage depletion in mice. B6.CCR5−/− mice showed less severe IRI based on tubular epithelial cell apoptosis than did wild‐type mice. CXCR3 expression in CD11b+ cells and inducible nitric oxide synthase levels were more attenuated in B6.CCR5−/− mice. B6.CCR5−/− mice showed increased arginase‐1 and CD206 expression. Macrophage‐depleted wild‐type mice showed more injury than B6.CCR5−/− mice after M1 macrophage transfer. Adoptive transfer of LPS‐treated RAW 264.7 macrophages reversed the protection against IRI in wild‐type, but not B6.CCR5−/− mice. Upon knocking out CCR5 in macrophages, migration of bone marrow‐derived macrophages from wild‐type mice towards primary tubular epithelial cells with recombinant CCR5 increased. Phospho‐CCR5 expression in renal tissues of patients with acute tubular necrosis was increased, showing a positive correlation with tubular inflammation. In conclusion, CCR5 deficiency favours M2 macrophage activation, and blocking CCR5 might aid in treating acute kidney injury.


| INTRODUC TI ON
Renal ischaemic-reperfusion injury (IRI) is a complicated orchestrated event that elicits diverse immunological responses. Monocytes/ macrophages, which exhibit great pliability, are important components of renal IRI. Their presence in the kidneys is closely correlated with a loss of renal function, [1][2][3] and plasticity of macrophages affects the incidence of acute renal damage owing to chronic fibrosis. [4][5][6] Chemokine receptor 5 (CCR5) is a G protein-coupled receptor that spans seven transmembrane domains and a co-receptor for macrophage-trophic human immunodeficiency virus (HIV) type 1 strains. 7 CCR5 is mainly associated with organ development, including angiogenesis, haematopoiesis, metastasis and chemotaxis. It is encoded on chromosome 3p21 and expressed by various immune cells such as resting T lymphocytes that have memory and effector T-cell phenotypes, monocytes, macrophages and immature dendritic cells. 7 Several ligands, including RANTES (regulated on activation, normal T cell expressed and secreted/CCL5), monocyte chemo-attractant protein 1 (MCP-1), macrophage inflammatory protein (MIP)-1α and MIP-1β, react with CCR5, are activated by CCR5 retroactive to CCR5 ligands.
CCR5 signalling plays various roles in inflammation and chemokine receptor expression because of macrophage heterogeneity. 8 A distinction between M1 and M2 macrophages suggests that the initiation and response to lipopolysaccharide (LPS)-or IFN-γ-induced stimulation are dependent on different signalling pathways of the Th1 or Th2 response. M2 macrophages are quite different from classically activated macrophages that produce trophic amines. 9,10 Phenotypic diversity of macrophages is important in acute ischaemic kidney injury development 2,3 and progression to chronic kidney disease. 4 Interestingly, several cytokines and chemokines are involved in the differentiation, recruitment and migration of monocytes and macrophages during this process. 11,12 CCR2 and MCP-1 play key roles in macrophage heterogenicity and plasticity, 13 but insufficient data are available on the association between CCR5 and the origin and subsets of macrophages. Moreover, data from post-transplantation biopsies show conflicting results because M2 macrophage deposition occurs during pro-inflammatory reactions rather than during tissue repair, 14,15 indicating that further research is necessary.
Here, we aimed to determine (a) the effect of macrophage phenotype on the expression of CCR5 and other chemokines, (b) the influence of macrophage phenotype on CCR5 signalling inhibition, and (c) the relevance of the CCR5 signalling pathway to IRI using in vivo and in vitro models. Finally, we analysed post-transplantation kidney biopsies to clarify the association between CCR5 and macrophages in acute kidney injury and clinical outcomes.

| MATERIAL S AND ME THODS
All experiments were performed with the approval of the Institutional Animal Care and Use Committee of the Clinical Research Institute of Seoul National University Hospital and in accordance with the Guidelines for the Care and Use of Laboratory Animals of the National Research Council. All experiments dealing with human specimens were also approved by the institutional review board of our institution (IRB number: H1910-011-1067). The experimental methods used in our study have been previously described. [16][17][18][19] All experiments were conducted in accordance with the guidelines of the 2013 Declaration of Helsinki.

| Experimental animals
Male, 8-week-old, C57BL/6 (B6) mice were purchased from Orient Company. B6.CCR5 −/− mice were originally produced by the Jackson Laboratory. All mice were raised in a pathogen-free animal facility.
For extended methods related to in vitro and in vivo manipulations, including induction of renal ischaemic-reperfusion injury (IRI), histologic analysis, confocal microscopy, quantitative real-time PCR, Western blot, cytokine assay, flow cytometry, mouse tubular epithelial cell and bone marrow-derived cell isolation, culture, in vitro cell migration assay, and statistical methods, as well as the functional study protocol and ischaemia-induced hypoxic condition procedures, please refer to online supplemental data.

| Depletion of CCR5 attenuates ischaemiareperfusion injury
To determine the specific role of CCR5 in IRI in the kidney, we induced renal ischaemia in mice. Renal function deteriorated significantly after IRI in the sham control and wild-type disease control mice (Cr 0.46 ± 0.03 mg/dL vs 2.05 ± 0.05 mg/dL; ***P < .001).

| Effects of CCR5 deficiency and changes in cytokine milieu
To examine how patterns of cytokine expression affect the extent of renal damage caused by IRI, we quantified mRNA levels using realtime PCR 24 hours after induction of IRI. Pro-inflammatory cytokines and chemokines such as interferon-γ (IFN-γ), IL-12 and CCR5 were increased by IRI in wild-type mice. However, in B6.CCR5 −/− mice, pro-inflammatory cytokines were significantly suppressed. Although IRI induced an increase in pro-inflammatory cytokines in wild-type mice, B6.CCR5 −/− mice showed enhanced expression of regulatory cytokines (IL-10) during IRI (Figure 2A). Enzyme-linked immunosorbent assay (ELISA), performed on whole kidney protein extracts, showed that IL-4, IL-10 and IL-13, but not IFN-γ, were increased in response to IRI in CCR5 −/− mice ( Figure 2B).

| Macrophage and CCR5 crosstalk
B6.CCR5 −/− mice showed less aggravated IRI in terms of apoptosis of tubular epithelial cells and creatinine (Cr) concentrations than did B6 wild-type mice. T-cell and macrophage infiltration decreased in B6.CCR5 −/− compared with that in wild-type mice. Next, we assessed whether monocytes and macrophages migrated to the inflammation site and were induced by IRI. After IRI, CD11b + CCR5 + cells were commonly expressed in wild-type mice and could be F I G U R E 1 The role of CCR5 regulation in kidney IRI. A, Comparisons between baseline kidney function in mice with CCR5 knockout and control mice after inducing IRI in the laboratory; blood urea nitrogen, creatinine, (B) PAS statin, (C) CD3, (D) F4/80, (E) MPO and (F) caspase 3 were detected using confocal microscopy. CCR5 null mice showed less extensive pathological changes and expressions of CD3, F4/80, MPO and caspase 3 than did B6 WT mice (n = 8/group; *P < .05, **P < .01). These results represent one of three independent experiments tracked by confocal microscopy ( Figure 4A). CXCR3 and CXCR4 are co-receptors of CCR5. Upon injury, CD11b + CXCR3 + cells were attenuated in B6.CCR5 −/− compared with that in wild-type mice ( Figure 4B). This pattern was similar to that in CD11b + CXCR4 + cells ( Figure 4C). In this IRI model, CXCR3 infiltration was more prominent than CXCR4 infiltration upon confocal microscopy examination.

| CCR5 modulation of macrophage phenotype
Classical activation of M1 macrophages caused enzymes such as inducible nitric oxide synthase (iNOS), which produces nitric oxide from arginine, to be attenuated in B6.CCR5 −/− compared with that in wildtype mice ( Figure 5A). Arginase-1 is a marker of M2 macrophages F I G U R E 2 The effect of CCR5 deficiency in the IRI model. A, The mRNA levels of inflammatory cytokines. The results of real-time PCR revealed that the expressions of various inflammatory cytokines, including CCR5 and IFN-γ, were significantly elevated in wild-type mice compared with those in control mice and were significantly attenuated in B6.CCR5 −/− mice (**P < .01). B, Whole kidney protein extract assay. On ELISA, the expressions of Th2 dominant cytokines, including IL-4, IL-10 and IL-13, were significantly elevated in B6.CCR5 −/− mice compared with those in control mice and were significantly attenuated in wild-type mice (n = 8 per each group, *P < .05 and **P < .01) F I G U R E 3 CCR5 inhibition significantly decreases damage in the kidney due to ischaemia-reperfusion injuries from T lymphocytes. A, Representative flow cytometry images. Flow cytometry analysis of CCR5 expression on CD3-positive T cells isolated from the kidneys of wild-type mice and CCR5-deficient mice 6 h after IRI. Values are represented as the means ± SEM of the cell counts from the kidney samples from 3 to 4 mice in each group. (*** P < .001). The proportions of intrarenal CCR5 expression gated on T cells in each animal group were as follows: 6.19 ± 0.47% in control mice, 34.44 ± 2.25% in wild-type mice and 0.90% in B6.CCR5 −/− mice. B, The proportions of intrarenal IFN-γ expression gated on T cells in each animal group were as follows: 6.79 ± 0.82% in wild-type mice and 3.65 ± 0.21% in B6.CCR5 −/− mice that metabolizes l-arginine through a non-fungicidal pathway and is involved in tissue repair during an inflammation cascade. Using immunofluorescence, arginase-1 stained more intensely in B6.CCR5 −/− than in wild-type mice ( Figure 5B). The intrarenal mRNA expression of proinflammatory cytokines such as TNF-α, IFN-γ and MCP-1 decreased, whereas arginase-1 expression increased after IRI in B6.CCR5 −/− mice ( Figure 5C). CCR5 protein levels increased in wild-type mice after induction of IRI ( Figure 5D). The M2 macrophage phenotype markers, arginase-1 and CD206, both increased in B6.CCR5 −/− compared with that in wild-type mice ( Figure 5E). Moreover, with CCR5 deficiency, there was a noteworthy increase in CD206 + macrophages ( Figure 5F). CD206 + CD11b + F4/80 + cells were induced in both wild-type and CCR5 knockout mice after IRI. The proportions of intrarenal CD206 + macrophages in each group were as follows: 1.09 ± 0.09% in control mice, 1.62 ± 0.40% in wild-type mice and 12.39 ± 1.43% in B6.CCR5 −/− mice ( Figure 5G).

| CCR5 inhibition and amelioration of hypoxiainduced injury
To demonstrate further the role of CCR5 inhibition in hypoxia, we conducted in vitro experiments using tubular epithelial cell lines.
We confirmed the expression of CCR5, and not HIF-1α or phospho (p)-CCR5, in normoxic conditions in human kidney-2 (HK-2) cells ( Figure 6A). Simultaneously, we cultured HK-2 cells in hypoxic conditions with TAK779. Six hours after hypoxic injury, p-CCR5 was expressed. A CCR5-CXCR3 inhibitor attenuated these changes in expression in a dose-dependent manner ( Figure 6A). Supernatant protein was quantified using ELISA to measure IL-6 and IL-8 levels ( Figure 6B). Hypoxia increased IL-6 and IL-8 protein levels, and TAK779 attenuated this increase in a dose-dependent manner.

F I G U R E 5
The expression of macrophage heterogeneity in injured B6.CCR5 −/− mice. A, The classical activation of M1 macrophages was representative of enzyme activity after inducing IRI; immunohistochemistry stain for iNOS was increased in wild-type mice. B, The immunohistochemistry stain for arginase-1 was intensified in B6.CCR5 −/− mice compared to that in B6 WT mice. C, The results of real-time PCR revealed that the mRNA levels of Th1 inflammatory cytokines such as MCP-1 and iNOS were significantly decreased in B6.CCR5 −/− mice compared to those in wild-type mice (***P < .001, n = 8 for each group). D, The protein levels of CCR5 were increased in wild-type mice after inducing IRI. E, In Western blot densitometry analyses, the expressions of M2 macrophage markers, such as CD206 and arginase-1, were substantially elevated in B6.CCR5 −/− mice and attenuated in wild-type mice. F, The comparison of expression for CD206 immunofluorescence staining on confocal microscopic examination. G, Representative flow cytometry images for CD206 + CCR5 + in CD11b + F4/80 + cells. The induction of CD206 + in CD11b + F4/80 + cells was increased in KO mice compared with that in wild-type mice. (**P < .01, n = 3 for each group)

| Increased CCR5 expression in the kidney biopsy samples
The kidney biopsy tissues from kidney transplant recipients with delayed graft function were assessed for biopsy-proven acute tubular necrosis (ATN) without evidence of cellular or humoral rejection.
Those with ATN were used to evaluate whether phosphorylation of CCR5 had a detrimental effect on IRI in patients undergoing transplantation. ATN in renal tissues from patients with glomerular

F I G U R E 7
In vitro-cultured RAW 264.7 macrophage cells and adoptive transfer to wild-type and B6.CCR5 −/− mice in the IRI model. A, Detailed schedule for in vitro-cultured RAW 264.7 macrophage cells and adoptive transfer has been presented. B, The M1 macrophage tendency was induced by stimulating RAW 264.7 cells with LPS, and 1 × 10 8 cells were adoptively transferred 2 h before the induction of IRI. The amplitude of the expression of INF-γ and iNOS was increased, which indicates that they reflect the characteristics of M1 macrophage subpopulation and the adoptive transfer of these macrophages. C, In both wild-type and B6.CCR5 −/− mice, injury was decreased after LC treatment (white bars). Undergoing the adoptive transfer of RAW cells with or without LPS showed an increase in the injury in wild-type mice, whereas there was no difference in the degree of injuries in B6.CCR5 −/− mice (grey bars). Ultimately, LC-pre-treated RAW cells with or without LPS were adoptively transferred, and injury increased in both wild-type and B6.CCR5 −/− mice, but predominantly in the latter (black bars). The detrimental effect of RAW cell adoptive transfer was weakened in B6.CCR5 −/− mice. D, The adoptive transfer of M1 macrophages decreased the mRNA expression of pro-inflammatory cytokines, such as MCP-1 and iNOS, in B6.CCR5 −/− mice (n = 5 per each group, *P < .05 and **P < .01) nephritis was also examined (  [26][27][28][29][30][31]. Numbers of p-CCR5 cells increased relative to those in the normal control sample ( Figure 9); in normal controls, p-CCR5 was scarcely detected. By immunohistochemical staining morphometry, kidney transplant recipients were more likely to have p-CCR5 cells than patients with glomerulonephritis (5.21 ± 1.32 vs 3.87 ± 1.71, P = .038). Renal tissue of patients with renal biopsy (n = 25) frequently contained p-CCR5 cells, and the number of these cells was positively correlated with tubule and vascular inflammation (tubulitis, P = .010, r = .508; intimal arteritis, P = .017, r = .474) ( Table 1).

| D ISCUSS I ON
In this study, B6.CCR5 −/− mice experienced less severe renal IRI than did wild-type mice; comparatively, they had lower Cr concentra-  Figure S1 that is proposed mechanism of the association between CCR5, Th1-type T cells and monocyte/macrophage. 20,25,26 Here, we focused on kidney damage and its up-regulation by CCR5, and explored the relevance and importance of macrophage infiltration.
Chemokines are 8-10 kD cytokines that stimulate leucocyte migration and chemotaxis to sites of inflammation. The chemokine receptor CCR5 is mediated by RANTES, monocyte inflammatory protein 1α (CCL3) and MIP-1β (CCL4), which are chemokine ligands produced in response to inflammation ( Figure S2). 7 Interestingly, in kidney transplant recipients with a CCR5-Δ32 base pair deletion, kidney graft survival is longer than that in recipients with wild-type CCR5, indicating the need for further studies to investigate the relationship between transplantation and CCR5 pathophysiology. 27,28 CCR5 inhibition protects against kidney damage in several animal models of kidney disease, and CCR5 antagonists reduce kidney damage by lessening the deposition of mononuclear cells in an experimental glomerular nephritis model. 29 CCR5 may attenuate renal damage by reducing T-cell deposition, as demonstrated in an acute kidney injury model. 26 Based on animal models, when CXCR3 and its chemokine co-receptor CCR5 are both inhibited, acute and chronic rejection of heart transplants is reduced through decreased T-cell activity. 30 Co-expression of CXCR3 and CCR5 is prominent in macrophages, as shown in previous HIV studies. 7,31 Taken together, these results suggest that macrophages also play a crucial role in the action of chemokines that is likely similar to that of T cells. 32 Therefore, in this study, we first observed crosstalk between CCR5 and other chemokines such as CXCR3 or CXCR4 ( Figure 4). After inducing IRI, CCR5 co-localized with CD11bpositive cells, and consequently, the CXCR3 chemokine, which binds to macrophages or T lymphocytes, 33 was expressed at a significantly lower level in B6.CCR5 −/− mice; however, CD11 + /CXCR4 + expression was not significantly modulated. The difference in expression was more pronounced for CXCR3, suggesting that co-signalling with CXCR3 may be important for CCR5 function. 30 Next, to investigate the functional role of macrophage phenotype and CCR5 in IRI ( Figure S3.), we compared heterogenic macrophage infiltration in wild-type and B6.CCR5 −/− mice ( Figure 5). As expected, macrophages in B6.CCR5 −/− mice, which had less injury, were highly expressed for the CD206 + M2 phenotype after IRI induction. In this study, we also showed that macrophage depletion with LC may reduce ischaemic acute kidney injury (AKI) in the con- Macrophage depletion and repletion have been studied previously in IRI mouse and rat models, 2,34-36 and macrophages also contribute to long-term fibrosis after IRI. 36 In hypoxic kidney conditions, macrophages secrete pro-angiogenic growth factors and pro-inflammatory cytokines. 37,38 Macrophages respond to hypoxia by increasing chemokine secretion, which accelerates the recruitment and chemotaxis of macrophages and exacerbates hypoxia of the microenvironment itself. [37][38][39] This is thought to be the main mechanism that attenuates damage in B6.CCR5 −/− mice.
When IRI occurs, immune cells are deposited and cytokine overproduction occurs because hypoxia is caused by a response to transcription factors that stimulate cytokines, including nuclear factor κB (NF-κB), heat shock factor protein 1 and HIF-1α. 38,40 Chemokines are also a major stimulant of chemotaxis, particularly when guiding neutrophils and M1 macrophages to the site of inflammation. 41,42 Chemokine receptor 2 (CCR2) promotes macrophage migration directly in IRI, 43  Previous studies have indicated a correlation between CCR5 expression on the surface of T cells and macrophages and cellular rejection in patients with biopsy-proven acute rejection. 44,45 In this study, to evaluate tissue expression of p-CCR5 in IRI, we performed IHC staining and morphometric analysis in pathologically confirmed ATN, but only in tissues from 25 patients, including kidney transplant recipients with no evidence of rejection and patients with glomerulonephritis. There was no difference in the renal survival rate in these patients based on a Kaplan-Meier curve for the occurrence of end-stage renal disease in accordance with morphometry positivity area (%)-based grouping (lower than the median value of p-CCR5 vs higher than the median value p-CCR5; data not shown). However, tissue p-CCR5-positive areas increased with the severity of tubulitis and intimal arteritis (Table 1). We confirmed that inflammatory cells, including those with the CCR5 receptor for chemokine ligands in tubulitis and endothelialitis, correspond to the distribution of appropriate ligands.
In summary, we identified the role of CCR5, which modulates inflammation and immunity, in macrophage induction of IRI. We showed that CCR5 blockade attenuated IRI via the macrophage heterogenic signalling pathway, irrespective of the anti-inflammatory effects of T cells. Modulation of the CCR5 pathway could thus prove useful as a new therapy for ischaemic AKI.

CO N FLI C T O F I NTE R E S T
The authors confirm that there are no conflicts of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author, SH Yang, upon reasonable request.