The Role of CC Chemokine Receptor 5 (CCR5) in Islet Allograft Rejection

CCR5^+/+ and CCR5^−/− mice (B6129F2/J, H-2^b) (The Jackson Laboratory, Bar Harbor, ME) were housed under specific pathogen-free conditions in enclosed filter-top cages. The recipients had been rendered diabetic by intraperitoneal injection of 250 mg/kg streptozotocin (Upjohn, Kalamazoo, MI) and were considered diabetic when the tail vein blood glucose concentration was >250 mg/dl for 3 consecutive days.

Islets were isolated from major histocompatibility complex (MHC)-mismatched male BALB/c (H-2^d) donors. Pancreata were infused via the common bile duct with rodent Liberase RI (Sigma-Aldrich, St. Louis, MO) and digested for 30 min at 37°C. Islets were purified on a discontinuous Ficoll gradient (Sigma) and handpicked under a stereomicroscope. A minimum of 500 islets was transplanted beneath the capsule of the left kidney. Daily monitoring of tail vein blood glucose was used to assess islet graft function. Rejection was defined as the return of hyperglycemia (>250 mg/dl on two consecutive measurements). Occurrence of hyperglycemia after removal of the graft was used to confirm long-term allograft function (>90 days). The grafts were removed for pathological examination under a surgical microscope by excising the edge of the capsule containing the islets.

In mixed lymphocyte response (MLR) assays, 2 × 10⁵ of each responder and irradiated stimulator splenocytes were cultured in complete RPMI medium (90% RPMI 1640, 10% FCS, [Sigma] with l-glutamine and penicillin/streptomycin [BioWhittaker, Walkersville, MD], and 50 mmol/l 2-mercaptoethanol [Sigma]) in 96-well U-bottom plates (Corning-Costar, Cambridge, MA) (six wells per experimental group). Background proliferation was measured by incubating cells with culture medium alone. Cultures were pulsed with 1 μCi of tritiated thymidine after 72 h of incubation. The samples were harvested after 12 h (11).

The enzyme-linked immunosorbent spot (ELISPOT) assay has been recently described by us (11,12). Briefly, ELISAspot plates (Cellular Technology, Cleveland, OH) were coated with capture antibodies against interleukin (IL)-4, IL-5, IL-10, or interferon (IFN)-γ (PharMingen, San Diego, CA) in PBS and left overnight at 4°C. The plates were blocked for 1 h with PBS containing 1% BSA and were washed with PBS. A total of 1 × 10⁶ splenocytes were added to each well in 100 μl complete RPMI medium in the presence of the same number of irradiated syngeneic or allogeneic splenocytes. After 48 h, the plates were washed, biotinylated detection antibodies were added, and the plates were maintained at 4°C for an additional overnight incubation. After additional washing, horseradish peroxidase conjugate (Dako, Glostrup, Denmark) was added for 2 h at room temperature. To develop the spots, aminoethylcarbazole (10 mg/ml in N,N-dimethylformamide; Pierre Chemicals, Rockland, IL) freshly prepared in 0.1 mol/l sodium acetate buffer (pH 5.0) and mixed with 30% H₂O₂ was added to each well (200 μl per well). The resulting spots were counted on a computer-assisted ELISAspot image analyzer (Cellular Technology). The results were calculated as cytokine-producing cells per million splenocytes.

The ribonuclease protection assay (RPA) was performed as follows. Islet RNA was extracted in guanidine-thiocyanate with acid phenol/chloroform extraction and alcohol precipitation. RNA integrity was confirmed by agarose gel electrophoresis and quantitated by optical density measurement (260 nm). Intragraft expression of chemokines, chemokine receptors, and cytokines was assessed using the custom-designed Riboquant system (PharMingen). Samples of total RNA (5 μg) were hybridized with 3 × 10⁵ cpm/μl of probe labeled with [³²P]UTP (3,000 Ci/mmol; NEN Life Science Products) at 56°C overnight. RNase A and RNase Ti (PharMingen) digestion were carried out at 37°C for 45 min. RNA bands were quantitated by densitometric analysis with NIH Image software (National Institutes of Health, Bethesda, MD), PC environment by Scion Corp. (Windows 95 and NT). The results were normalized for L32 and GAPDH gene expressions.

One-half of each islet graft was embedded in OCT, frozen at − 20°C, and kept at −70°C until sectioning. Cryostat sections (2 μm) were fixed with acetone and dried. Indirect frozen section immunohistochemistry was performed using primary antibodies for T cells (CD4, GK1.5, and CD8, 53-6.7; PharMingen) and macrophages (F4/80). One-half of each graft was also fixed and paraffin embedded. Islets were stained for insulin and glucagon using guinea pig anti-insulin (1:10 dilution; Dako) and biotin-labeled goat anti-guinea pig Ig (1:200; Vector) and developed with Vector ABC (13). Each specimen was evaluated at a minimum of three different levels of sectioning. Images were taken with a RT Spot Camera (Diagnostic Instruments) and imported into Adobe Photoshop at 40× magnification.

Data were compared by nonparametric analysis (Instat software; GraphPad, San Diego, CA) using the log-rank test for graft survival data and Mann-Whitney test for data from proliferation assays and expression studies. Differences in graft survival were analyzed by a Kaplan-Meier test.

First, we investigated the effect of the targeted deletion of CCR5 on islet allograft survival in chemically induced diabetic recipients. Islet cells, isolated from complete MHC-mismatched BALB/c, were transplanted into CCR5^−/− (n = 15) or control CCR5^+/+ (n = 15) mice. Given that a small percentage of streptozotocin-induced diabetic mice recover spontaneously, long-term functioning grafts were verified by the reappearance of hyperglycemia following nephrectomy. As seen in Fig. 1, islet grafts survived for a mean of 38 ± 8 days in CCR5^−/− mice compared with the control group, which survived for 10 ± 2 days (P < 0.001). Twenty percent of the grafts in CCR5^−/− recipients survived for >90 days, as confirmed by the appearance of hyperglycemia after graft removal (Fig. 1). These data show that targeted deletion of CCR5 is associated with prolonged islet allograft survival in chemically diabetic mice, indicating that CCR5 plays an important role in islet allograft rejection.

On day 7, grafts from CCR5^−/− and CCR5^+/+ mice recipients showed a comparable degree of mononuclear cell infiltration, composed mostly of CD4⁺ T cells and macrophages (Fig. 2). Interestingly, long-term functioning grafts (>90 days) from CCR5^−/− showed persistent peri-islet mononuclear cellular infiltration of both CD4⁺ T cells and macrophages. Islets appeared intact, with strong insulin staining (Fig. 3).

The capacity of lymphocytes from CCR5^−/− mice to mount a proliferative response upon alloantigen stimulation in vitro was assessed in the one-way MLR. Splenocytes were recovered from transplanted CCR5^−/− and control CCR5^+/+ animals at the time of rejection. Lymphocytes from CCR5^−/− mice exhibited a degree of proliferation comparable to that of control CCR5^+/+ mice (n = 10/group; P = 0.8) (Fig. 4).

The incidence of cells producing IFN-γ, IL-4, IL-5, and IL-10 was measured by ELISPOT assay from cultured splenocytes of transplanted animals responding to donor cells in vitro (Fig. 5). All the panels showed significant response to concanavalin A (too numerous to count; not shown), indicating adequate viability of these cells. The number of spots in the wells with medium alone (background) or syngeneic cells (negative control) was <10/10 ⁶ spleen cells for IFN-γ, IL-5, and IL-10. IL-4 wells had an average background count of 30 spots. In all cases, the number of background spots was taken into consideration when analyzing the data. In transplanted animals, whereas the frequency of IFN-γ-producing cells was significantly higher in CCR5^+/+ recipients than in CCR5^−/− recipients, the frequency of IL-4- and IL-5-producing cells was significantly higher in CCR5^−/− than in CCR5^+/+ mice. Interestingly, the frequency of IL-10-producing alloreactive cells was not different between the two groups of mice (Fig. 5).

The frequency of IFN-γ-producing cells in naive CCR5^+/+ mice was relatively low (107 ± 16 spots/10 ⁶ spleen cells), yet significantly higher than in naive CCR5^−/− mice (43.3 ± 12 spots/10 ⁶ spleen cells; n = 3; P = 0.038). The frequency of IFN-γ-producing cells in response to third-party stimulator cells (C3H) was also relatively low in recipients of BALB/c islets and was not significantly different between CCR5^−/− and CCR5^+/+ mice (180.7 ± 59.76 and 131.7 ± 26.1, respectively; n = 6; P = 0.46).

We were intrigued by the pathological finding that islet allografts from CCR5^+/+ and CCR5^−/− mice exhibited a comparable amount of infiltration, yet the grafts in CCR5^−/− recipients had significant prolongation of graft survival. Intragraft expression of chemokines was assessed by RPA on islet grafts from CCR5^+/+ and CCR5^−/− recipients 8 days after transplantation. In our experience, at the time of rejection, the recovery of the islet cells is poor due to massive destruction of the islets in unmodified CCR5^+/+ recipients. Given that islets are rejected at 10 ± 2 days in controls, we harvested the grafts at day 8 to optimize islet recovery. Densitometric quantitative data for chemokines are shown in Fig. 6A. Whereas the expression of CCR5 and its ligands, RANTES and MIP-1β, was significantly higher in grafts from CCR5^+/+ recipients compared with CCR5^+/+ syngeneic controls, the level of these transcripts was significantly diminished in grafts from CCR5^−/− recipients and was comparable to the levels observed from CCR5^+/+ syngeneic controls. The expression of IP-10 and lymphotactin was upregulated in the allografts from CCR5^+/+ recipients. CCR5 and lymphotactin mRNA were not detected in syngeneic CCR5^+/+ grafts. Expression of IP-10 and lymphotactin was significantly reduced in CCR5^−/− recipients compared with CCR5^+/+ mice.

Intragraft cytokine expression data are shown in Fig. 6B. Whereas the expression of IFN-γ and IL-5 mRNA was not detected in syngeneic grafts, the intragraft expression of IFN-γ and IL-5 was significantly higher in allografts of CCR5^+/+ animals compared with CCR5^−/− recipients. Although intragraft IL-4 transcripts were detected at low levels with no significant difference between CCR5^+/+ allografts and syngeneics, grafts from CCR5^−/− recipients displayed a significantly higher amount of IL-4 mRNA. IL-10 mRNA was detected in all groups with no significant differences. The allografts from CCR5^−/− mice expressed insignificantly higher amounts of IL-10, consistent with our ELISPOT data in vitro. Of note, RPA of the islets just after isolation did not reveal expression of any of these chemokines (not shown).

CCR5 has been implicated in the regulation of Th1 lymphocyte function (14). CCR5 is the receptor for the proinflammatory chemokines: RANTES, MIP-1α, and MIP-1β (1). It has been shown that CCR5 and RANTES may play important roles in the pathogenesis of acute lung and renal allograft rejection in humans (9). Gao et al. (8) recently demonstrated that targeting CCR5 prolongs cardiac allograft survival in the mouse model. However, given a lack of intragraft IL-4, IL-5, or IL-10 mRNA in CCR5^−/− recipients, no evidence for immune deviation toward an intragraft Th2 regulatory cell population was apparent in their study (8). Studies on murine and human heart and skin models indicate that, in the process of acute rejection, the temporal expression of critical chemokines varies among different organs or tissues. For instance, CXCR3 and its ligands, IP-10 and Mig, are expressed by the T-cells infiltrating lung and heart allografts and mediate chemotaxis of T-cells at sites of rejection (15,16) (17). Targeting CXCR3 and its ligands has been shown to be associated with a significant prolongation of cardiac allograft survival (6, 18). Hence, each organ or tissue may require a unique set of chemokines to generate acute rejection (6,18). Despite the documented presence of chemokine in experimental and clinical allografts during acute rejection, the role of specific chemokines in the rejection process of islet allografts remains unclear. In our study, CCR5 mRNA was not detected in the grafts from syngeneic CCR5^+/+ recipients, and RANTES and MIP-1β expression in CCR5^+/+ recipients was significantly higher than that noted in syngeneic controls and CCR5^−/− recipients. The former finding indicates that production of CCR5 and its ligands is dependent on the infiltrating alloantigen-specific T-cells. Grafts from CCR5^−/− recipients also exhibited significantly lower expression of lymphotactin and IP-10. Islet graft rejection in wild-type mice was characterized by a Th1-type response with a strong induction of IFN-γ mRNA expression. In contrast, the prolonged islet allograft survival observed in CCR5^−/− mice was associated with an increase in intragraft and peripheral expression of IL-4 and IL-5 and a reduction in expression of IFN-γ. CCR5 ligands induce selective migration of Th1 cells in chemotaxis assays (5,19). Therefore, absence of CCR5 may explain the shift toward Th2 immune response in our study, and that the infiltrating cells in the CCR5^−/− grafts represent cells of a predominantly Th2 phenotype. It is possible that the small amount of RANTES expressed in the CCR5^−/− recipients may signal through CCR3 in the absence of CCR5, preferentially attracting Th2 cells (20). Decreased IFN-γ and increased IL-4 expression in turn may have led to suppressed macrophage activation (21,22). Because IFN-γ induces CCR5 ligands, lower expression of IFN-γ in CCR5^−/− may explain the lower expression of CCR5 ligands in this model (4). Furthermore, the ability to produce IFN-γ has been shown to be critical for efficient CD8⁺ T-cell-mediated rejection of islet allografts (23). Using a dextran sodium sulfate-induced colitis model, CCR5^+/+ mice were characterized by a Th1-type response with a strong induction of IFN-γ mRNA expression. In contrast, the reduced colonic damage observed in CCR5^−/− mice was associated with an increase in IL-4 and IL-5 mRNA expression, a reduction in IFN-γ mRNA expression, and an increase in the proportion of CD4⁺ T-cells in the lamina propria (24).

Interestingly, the intragraft mononuclear cell infiltrate in CCR5^−/− mice was comparable to that in wild-type recipients. Thus, the immunosuppressive effect of the targeted deletion of CCR5 is not necessarily related to the generalized recruitment and migration of CD4⁺ T-cells and macrophages into the graft. One could speculate that the Th2 switch observed in the graft and periphery of CCR5-deficient animals may be protective against islet allograft destruction. However, a regulatory function of Th2 cells in alloimmune responses remains controversial (25–27). Animals lacking the Th2 cytokine IL-4 can accept allografts (28), although this is not a universal finding in all models (29). In addition, Li et al. (30) showed that a Th2 switch is associated with graft acceptance in minor but not major mismatched graft recipients. The finding that CCR5^−/− mice ultimately reject their islet graft indicates that a Th2 switch alone is not sufficient for induction of long-term graft acceptance. Importantly, long-term functioning grafts that survive in CCR5^−/− mice in the presence of cellular infiltration indicate that not all types of inflammation are identical and that the amount of inflammation may not predict outcome. This infiltration appears to resemble other autoimmune settings, such as diabetes, in which T-cell recruitment does not universally lead to organ damage. Identifying the nature of this nonpathogenic infiltration may be extremely crucial in understanding similar presentations in various autoimmune models such as diabetes. In conclusion, our results demonstrate that the lack of CCR5 signaling results in prolongation of islet allograft survival and a switch to Th2 response. Targeting of CCR5 may have clinical application as a novel approach to clinical management of islet allograft recipients.

This work was supported in part by the Juvenile Diabetes Research Foundation (JDRF) Center for Islet Transplantation at Harvard Medical School.

We thank the Islet Care Laboratory of the JDRF Center for Islet Transplantation for providing mouse islets.