Cholecystokinin Plays a Novel Protective Role in Diabetic Kidney Through Anti-inflammatory Actions on Macrophage: Anti-inflammatory Effect of Cholecystokinin

Male ICAM-1^−/− mice (C57BL/6J background) (15) were purchased from The Jackson Laboratory (Bar Harbor, ME). Male C57BL/6J (ICAM-1^+/+) mice were used as controls. WT and ICAM-1^−/− mice aged 8 weeks were divided into the following four groups (n = 5 each): 1) nondiabetic WT mice, 2) nondiabetic ICAM-1^−/− mice, 3) streptozotocin (STZ)-induced diabetic WT mice, and 4) STZ-induced diabetic ICAM-1^−/− mice. STZ was purchased from Sigma-Aldrich (St. Louis, MO). Mice in the diabetic groups received two intraperitoneal doses of STZ (each 100 mg/kg) given 7 days apart. Blood glucose levels were determined 7 days after STZ injection, and only mice with blood glucose concentrations >16 mmol/L were used in the study. Nondiabetic WT and ICAM-1^−/− mice received citrate buffer injections only. All animal procedures were performed according to the guidelines as described previously (16). Three months after the induction of diabetes, all mice were killed, and the kidneys were harvested.

Total RNA was extracted from each specimen of the renal cortex using the standard protocol included with the RNeasy Midi Kit (Qiagen, Valencia, CA) at 3 months. Preparation of biotin-labeled target cRNA and hybridization of probe arrays (CodeLink UniSet Mouse I Bioarray) were performed according to the manufacturer’s instructions (Amersham Biosciences, Uppsala, Sweden) (Gene Expression Omnibus accession numbers are available in the Supplementary Data).

The criteria for selecting genes that were induced or reduced by a diabetic state were as follows: 1) the gene flags were “true,” and 2) the ratio of the gene expression level in diabetic WT mice to that in diabetic ICAM-1^−/− mice was >2 or <0.5. We then selected 193 genes for further analysis. All normalized data values were replaced to log base 2 and subjected to hierarchical clustering as described previously (16).

CCK-1R^−/−, CCK-2R^−/−, and CCK-1R^−/−,-2R^−/− mice (C57BL/6J background) (17) were obtained from the Tokyo Metropolitan Institute of Gerontology. CCK-1R^−/− mice and CCK-2R^−/− mice were generated as described previously (18,19). C57BL/6J (CCK-1R^+/+,-2R^+/+) mice were used as controls. Male WT and CCK-1R^−/−,-2R^−/− mice aged 8 weeks were divided into four groups (n = 7 each): 1) nondiabetic WT mice, 2) nondiabetic CCK-1R^−/−,-2R^−/− mice, 3) STZ-induced diabetic WT mice, and 4) diabetic CCK-1R^−/−,-2R^−/− mice. Diabetes was induced as described above. Blood pressure, blood glucose, A1C, serum creatinine, urine creatinine, and urinary albumin were measured as described previously (16). Three months after the induction of diabetes, all mice were killed, and the kidneys were harvested as described previously (16). Male CCK-1R^−/− and CCK-2R^−/− mice aged 8 weeks (n = 7 each) were also used for comparison of albuminuria after induction of diabetes.

Bone marrow transplantation (BMT) was performed as described previously (20,21). Briefly, male WT and CCK-1R^−/− mice aged 7–9 weeks received 9 Gy of total body irradiation. Postirradiated male CCK-1R^−/− mice received a bone marrow transplant from WT mice (WT→1R^−/−; n = 6). Postirradiated WT mice received a BMT from CCK-1R^−/− mice (1R^−/−→WT; n = 6) or WT mice (WT→WT; n = 4). Four weeks after BMT, diabetes was induced in all mice by STZ as described above. Four weeks after the induction of diabetes, all mice were killed. DNA was isolated from bone marrow extracts of all recipient mice using a DNeasy Blood & Tissue Kit (Qiagen). The chimerism was confirmed by PCR (Supplementary Fig. 1) at the termination of the study as described previously (22). The specific oligonucleotide primer sequences are shown in Supplementary Table 1.

Male Sprague-Dawley (SD) rats were purchased from CLEA Japan (Tokyo, Japan). SD rats aged 4 weeks were divided into three groups (n = 7 each): 1) nondiabetic control group (NDM), 2) STZ-induced diabetic group (DM), and 3) CCK-8S–treated diabetic group (DM-CCK). At the age of 5 weeks, rats chosen for the DM and DM-CCK groups were injected intravenously with STZ (65 mg/kg body wt) in citrate buffer (pH 4.5). Rats in the NDM group received citrate buffer injections only. At the age of 6 weeks, Alzet osmotic minipumps (Durect Corporation, Cupertino, CA) were implanted subcutaneously in the backs of all the rats. Rats in the DM-CCK group were continuously infused with CCK-8S (Bachem, Bubendorf, Switzerland) dissolved in 0.9% saline and given at a rate of 5 μg CCK-8S/kg · h⁻¹. Animals in the NDM and DM groups were continuously infused with 0.9% saline only. Food intake was calculated as the average over 3 days. Serum CCK concentration in both diabetic groups was measured using the CCK Enzyme Immunoassay Kit (RayBiotech, Norcross, GA) according to the manufacturer’s instructions. Because the life expectancy of the osmotic pumps was 4 weeks, all pumps were replaced with new filled pumps when the rats reached the age of 10 weeks. Eight weeks after the induction of diabetes, all rats were killed, and the kidneys were harvested as described previously (16). Glomeruli were isolated from the left kidney by a previously developed sieving technique (23).

Periodic acid-methenamine silver (PAM)-stained sections were analyzed as described previously (24). To evaluate the glomerular size and mesangial matrix area, we examined 10 randomly selected glomeruli per mouse and 15 randomly selected glomeruli per rat under high magnification (×400). Quantitative analysis for all staining was performed in a blinded manner.

Immunoperoxidase staining was performed as described previously (4). Primary antibodies were monoclonal antibody against rat monocytes/macrophages (ED1, 1:50; Serotec, Oxford, U.K.), polyclonal antibody against WT-1 (1:50; Santa Cruz Biotechnology, Santa Cruz, CA), or polyclonal antibody against cholecystokinin octapeptide (1:500; Phoenix Pharmaceuticals, Belmont, CA), all of which were applied for 12 h at 4°C. Secondary antibodies were biotin-labeled goat anti-mouse IgG (Jackson ImmunoResearch Laboratories, West Grove, PA) or biotin-labeled goat anti-rabbit IgG (Vector Laboratories, Burlingame, CA) for 60 min at room temperature. Intraglomerular ED1-positive cells or WT-1–positive cells were counted in 20 glomeruli per animal (n = 4/group).

Immunofluorescence staining was performed using the methods described previously (5). Rabbit antitype IV collagen Ab (1:200; LSL, Tokyo, Japan) was used for the primary reactions for 60 min at room temperature, followed by a second reaction with fluorescein isothicyanate-conjugated goat anti-rabbit IgG (H+L; Zymed Laboratories, San Francisco, CA) for 30 min at room temperature. The immunofluorescence intensity of type IV collagen was quantified as described previously (24). We evaluated 15 glomeruli per animal (n = 4/group).

RNA extraction, real-time PCR, and visualization of gene expression were performed as described previously (24). The specific oligonucleotide primer sequences are shown in Supplementary Table 2.

Nuclear proteins were extracted from kidney tissues with a nuclear extract kit (Active Motif, Carlsbad, CA) according to the manufacturer’s instructions.

Nuclear factor-κB (NF-κB) p65-dependent DNA-binding activity was determined by TransAM NFκB p65 (Active Motif) according to the manufacturer’s instructions.

THP-1 cells were obtained from DS Pharma Biomedical (Osaka, Japan) and cultured according to the manufacturer’s instructions. PBS without calcium and magnesium [PBS (−)] was purchased from Invitrogen (Carlsbad, CA).

THP-1 cells were precultured in the RPMI 1640 (without glucose) medium supplemented with 10% FCS and 5.5 mmol/L d-glucose (Sigma-Aldrich) for 72 h. The cells were centrifuged, washed in PBS (−), centrifuged, and serum starved for 12 h in RPMI 1640 medium containing 5.5 mmol/L d-glucose. After starvation, the cells were adjusted to a cell density of 4 × 10⁵ cells/mL in RPMI 1640 medium containing 5.5 mmol/L d-glucose and 1% FCS and placed in six-well plates (Falcon, Franklin Lakes, NJ). A control scrambled peptide (H-Gly-Asp-Tyr-Asp-Met-Trp-Met-Phe-NH₂) and proglumide (a nonselective CCK receptor antagonist that interacts with both CCK receptors and can cross brain–blood barrier) were purchased from Sigma-Aldrich. The cells were exposed to the following stimuli (n = 5/group): 1) 5.5 mmol/L d-glucose (normal glucose [NG]); 2) 15 mmol/L d-glucose (high glucose [HG]); 3) 5.5 mmol/L d-glucose with 9.5 mmol/L mannitol (osmotic control [Mn]); 4) HG with scrambled peptide (10⁻⁶ M); 5) HG with CCK-8S (10⁻⁸ M); 6) HG with CCK-8S (10⁻⁷ M); 7) HG with CCK-8S (10⁻⁶ M); and 8) HG with CCK-8S (10⁻⁶ M) and proglumide (10⁻⁵ M).

CCK-8S and proglumide were added daily. After incubation for 72 h, total RNA was extracted, and quantitative real-time RT-PCR was performed as described above. Tumor necrosis factor-α (TNF-α) mRNA expression levels were normalized by β-actin in each sample. Values (means ± SEM) were expressed as the ratios of average values in HG.

THP-1 cell migration was analyzed with a modified Boyden chamber assay using a 24-well transwell with 5.0-μm pores (Corning Life Sciences, Corning, NY) as described previously (25,26). THP-1 cells were preincubated for 24 h in serum-free RPMI 1640 supplemented with 0.1% bovine serum albumin (Sigma-Aldrich). After starvation, CCK-8S, scrambled peptide, or proglumide was added to THP-1 cells at different concentrations, and the cells were added to the top chamber. CCK-8S or scrambled peptide was incubated from 15 min before addition to the top chamber, and proglumide was added 15 min before addition of CCK-8S. The medium in the lower well contained 100 ng/mL of recombinant human CC chemokine ligand 2 (CCL2; R&D Systems, Minneapolis, MN). Cells that migrated to the bottom side of the membrane were quantitated by CyQuant DNA-binding fluorescence (Invitrogen) according to the manufacturer’s instructions (n = 6 each).

All values are expressed as the means ± SEM. Differences between groups were examined for statistical significance using one-way ANOVA followed by Scheffe’s test. A P value <0.05 was considered statistically significant.

Hierarchical clustering identified 33 genes that were significantly upregulated only in diabetic WT mice but not remarkably changed in nondiabetic WT mice, nondiabetic ICAM-1^−/− mice, or diabetic ICAM-1^−/− mice (cluster 4; Fig. 1A). We focused on CCK because CCK is one of the most upregulated genes in cluster 4. Real-time RT-PCR revealed that the expression of CCK mRNA in the kidney cortices was significantly higher in diabetic WT mice than in diabetic ICAM-1^−/− mice (Fig. 1B), whereas there was no difference in CCK-1R or CCK-2R mRNA expression (Fig. 1C and D). We confirmed the distribution of CCK in the kidney tissues by immunoperoxidase staining. CCK was widely distributed in kidney tissues of nondiabetic WT mouse (Fig. 1E–G). In diabetic WT mice, the distal tubules and glomeruli were stained more intensely than in nondiabetic WT mice (Fig. 1H–J).

At 3 months after induction of diabetes, there was no significant difference in A1C, body weights, and creatinine clearance between the diabetic WT and diabetic CCK-1R^−/−,-2R^−/− groups (Table 1). Kidney weight per body weight was increased not only in the diabetic WT and diabetic CCK-1R^−/−,-2R^−/− groups but also in the nondiabetic CCK-1R^−/−,-2R^−/− group (Table 1). It was noteworthy that the urinary albumin/creatinine ratio was markedly increased in diabetic CCK-1R^−/−,-2R^−/− mice from 1 month to the end of the observation period compared with the diabetic WT mice (Fig. 2A). Furthermore, we compared levels of albuminuria among diabetic WT, CCK-1R^−/−, CCK-2R^−/−, and CCK-1R^−/−,-2R^−/− mice. Although there was no statistical significance, CCK-1R^−/−,-2R^−/− mice exhibited the most increased albuminuria at 3 months (Fig. 2A). Representative findings of the glomeruli in PAM-stained sections are shown in Fig. 2B. Glomerular hypertrophy was observed in both diabetic groups compared with the nondiabetic WT group at the end of the 3-month observation period (Fig. 2C). Mesangial matrix expansion was observed in both diabetic groups, but was more prominent in the diabetic CCK-1R^−/−,-2R^−/− group than in the diabetic WT group (Fig. 2D). Type IV collagen intensity was higher in both diabetic groups than nondiabetic groups, and the intensity in the diabetic CCK-1R^−/−,-2R^−/− group was further increased as compared with that in the diabetic WT group (Fig. 2E and F). Immunoperoxidase staining of WT-1, a normal podocyte marker, was performed to investigate the effect of CCK-8S in the progression of podocyte loss (Fig. 2G). The number of WT-1–positive cells per glomerulus was significantly decreased in the diabetic WT and both CCK-1R^−/−,-2R^−/− groups, whereas podocyte loss was more prominent in the diabetic CCK-1R^−/−,-2R^−/− group (Fig. 2H). CCK mRNA expression in the renal cortex was increased to the same extent in the diabetic WT group and diabetic CCK-1R^−/−,-2R^−/− group compared with the nondiabetic groups (Fig. 2I). In contrast, mRNA of CCL2, ICAM-1, cluster of differentiation (CD) 68, and kidney injury molecule-1 (KIM-1; a marker of tubular damage) were significantly upregulated in the diabetic CCK-1R^−/−,-2R^−/− group compared with the diabetic WT group (Fig. 2J–M). These findings suggest that diabetic renal injuries were exacerbated by deletion of both CCK-1R and CCK-2R via the inflammatory process. Furthermore, we performed BMT study to clarify whether deficiency of CCK-1R on infiltrating macrophages or resident renal cells is more important for the exacerbation of diabetic renal injury. BMT study showed that 1R^−/−→WT mice exhibited significantly increased relative kidney weight (Supplementary Table 3) and albuminuria (Fig. 2N) compared with WT→1R^−/− and WT→WT mice, suggesting the importance of CCK-1R on macrophages.

In the renal cortex of adult rats, the distribution of CCK was comparatively localized in distal tubules and glomeruli (Fig. 3A and B). In the renal medulla, the collecting ducts were stained intensely (Fig. 3C). We also identified CCK mRNA expression by real-time RT-PCR in the kidney cortex and isolated glomeruli obtained from normal adult rats (Fig. 3E).

At 8 weeks after induction of diabetes, systolic blood pressure, A1C, and kidney weight per body weight were elevated to the same level in both diabetic groups. However, there was no significant difference between DM and DM-CCK (Table 2). The body weight and serum creatinine of both diabetic groups were lower than that of the NDM animals. However, there was no significant difference between the DM and DM-CCK groups (Table 2). Fasting serum CCK levels in the DM-CCK group was increased ∼3.9-fold than that of the DM group (514 ± 89 vs. 132 ± 33 pg/mL).

It is noteworthy that CCK-8S treatment significantly reduced urinary albumin excretion compared with the DM group at 8 weeks (Fig. 4A). Food intake was increased to the same extent in both diabetic groups compared with the NDM group after induction of diabetes (Fig. 4B). Glomerular hypertrophy was observed in both diabetic groups as compared with the NDM group. There was no significant difference in glomerular size between the DM and DM-CCK groups (Fig. 4C and D). Mesangial matrix expansion was observed in the DM group; however, CCK-8S treatment significantly reduced mesangial matrix expansion compared with DM (Fig. 4E). Type IV collagen intensity was higher in the DM than the NDM group. CCK-8S treatment markedly reduced type IV collagen intensity compared with the DM animals (Fig. 4F and G). The average number of macrophages (ED1-positive cells) per glomerulus was markedly increased in the DM compared with the NDM group, whereas macrophage infiltration was significantly inhibited by CCK-8S treatment (Fig. 4H and I). The number of WT-1–positive cells per glomerulus was significantly decreased in the DM, whereas podocyte loss was significantly inhibited by CCK-8S treatment (Fig. 4J and K).

The mRNA expressions of CD68, ICAM-1, and TGF-β in the renal cortex were significantly upregulated in the DM group, and these increases were significantly suppressed by CCK-8S treatment (Fig. 5A–C). The increase of KIM-1 mRNA expression in the kidney of DM group was partially but significantly suppressed by CCK-8S treatment (Fig. 5E). In isolated glomeruli, CCK-8S treatment also decreased the mRNA expressions of CD68, ICAM-1, TGF-β, and TNF-α as compared with DM (Fig. 5F–I). Interestingly, CCK-8S treatment markedly increased the mRNA expression of nephrin in glomeruli as compared with DM (Fig. 5J). These findings suggest that CCK-8S inhibits the development of albuminuria via inhibition of proinflammatory genes in the diabetic kidney. NF-κB p65-dependent DNA-binding activity in the renal cortex was significantly increased in the DM compared with the NDM group. CCK-8S treatment significantly decreased the NF-κB p65-dependent DNA-binding activity (Fig. 5K).

TNF-α mRNA expression was significantly increased in the untreated HG group (Fig. 6A). Although TNF-α mRNA expression was not suppressed in the scrambled peptide-treated HG group, it was suppressed in the CCK-8S–treated (10⁻⁶ M) HG group compared with the untreated HG group (Fig. 6A). In addition, the anti-inflammatory effect of CCK-8S was largely abrogated by proglumide (an antagonist for both CCK receptors) (Fig. 6A). The number of THP-1 cells migrated into the lower chamber of the transwell was significantly reduced by CCK-8S treatment in a dose-dependent manner and was not reduced by scrambled peptide. These antimigratory effects of CCK-8S were completely abolished by proglumide (Fig. 6B).

In present study, we found that CCK is one of the significantly upregulated genes in the diabetic WT kidney compared with the diabetic ICAM-1^−/− kidney. We hypothesized that CCK might regulate inflammatory response in the diabetic kidney; however, little is known about the role of CCK and its receptors in the development of diabetic nephropathy.

Two types of CCK receptors have been identified (27,28). These receptors have been classified as CCK-1R and CCK-2R based on their highly distinctive ligand selectivities (18). The two types of CCK receptors are distributed in various cells or tissues, including the kidneys (29–33) and macrophages (34,35). CCK-1R^−/−,-2R^−/− mice are fertile and show no apparent developmental defects. It has been reported that the weights of the kidneys and liver are significantly increased in CCK-1R^−/−,-2R^−/− mice compared with WT mice, although no abnormality is visible in these organs (17). In this study, deletion of both CCK-1R and CCK-2R enhanced inflammatory reactions and exacerbated the development of albuminuria after induction of diabetes. Our results suggest that CCK is increased in the diabetic kidney of mice and may regulate macrophage-related proinflammatory genes via CCK receptors. Furthermore, BMT study revealed that CCK-1R on macrophages played a more important role in the early stage of diabetic nephropathy than CCK-1R on resident renal cells. In the BMT study, we used CCK-1R^−/− mice, because CCK-1R on macrophages plays a more dominant role in the anti-inflammatory effect of CCK-8S than CCK-2R (12). In contrast, CCK-2R is expressed on renal cells including murine mesangial cells (32). And the systemic absence of CCK-2R also exacerbated the development of albuminuria after induction of diabetes almost same extent as in CCK-1R^−/− mice. Therefore, although further BMT study is needed, endogenous CCK might act against not only infiltrating macrophages but also resident renal cells via CCK-2R.

Several transcription factors have been implicated in the glucose-mediated expression of genes involved in diabetic nephropathy (36). NF-κB is one of the key mediators in the inflammatory response and plays a pivotal role in the progression of diabetic nephropathy (37). Activation of NF-κB in both kidney tissues obtained by biopsies and human peripheral blood mononuclear cells has been shown to correlate with degree of diabetic nephropathy (38,39). And NF-κB is also involved in regulation of ICAM-1 expression in diabetic kidney (36). Li et al. (12) reported that CCK-8S inhibited lipopolysaccharide-induced cytokine production via suppression of NF-κB activity. We showed that CCK-8S significantly suppressed NF-κB activation in the diabetic kidney, suggesting that CCK-8S may inhibit ICAM-1 expression via inhibition of NF-κB activity and thus lead to suppression of macrophage infiltration in the diabetic kidney.

Guha et al. (40) reported that high glucose-induced TNF-α mRNA expression in THP-1 cells was mediated by NF-κB. Our results indicate that inhibition of both high glucose-induced TNF-α expression via NF-κB and CCL2-induced migration might be involved in the anti-inflammatory effects of CCK-8S in THP-1 cells. Because the nephrin gene expression in cultured podocyte is repressed by TNF-α at a transcriptional level (41), our results suggest that CCK-8S may prevent podocyte loss via inhibition of TNF-α mRNA expression in diabetic glomeruli.

Aunapuu et al. (42) recently reported that CCK overexpression was associated with renal morphological damage in transgenic mice without a significant difference in kidney weight or proteinuria, but with a thickened glomerular basement membrane. CCK is expressed from embryonic day (E) 8.5, whereas both CCK-1R and CCK-2R have been identified from E10.5; in addition, CCK is thought to affect tissue growth and differentiation (43). In another study, both CCK-1R and CCK-2R expression were confirmed from E14.5 in kidney tissues (33), and therefore, CCK might regulate renal microstructural growth. In nondiabetic CCK-1R^−/−,-2R^−/− mice, increased kidney and liver weight (17) and reduction in the number of glomerular podocytes also suggest that CCK may play a role in regulating the development or growth of these organs. In the current study, we examined the effect of CCK-8S using an STZ-induced diabetic model, but no nephrotoxicity was observed.

Because the half-time of CCK-8S in blood is short, it would be difficult to use a longer period for maintaining renoprotective effect in patients with diabetic nephropathy. Recently, León-Tamariz et al. (44) reported that PEGylated CCK-10, which did not cross the blood-brain barrier, maintained blood concentration longer than free CCK. Such drugs with a longer duration of action may be more suitable for clinical use.

In conclusion, we have shown that CCK is expressed in the kidney, and deficiency of both CCK-1R and CCK-2R accelerates development of albuminuria by enhancement of inflammation in the kidneys after induction of diabetes. Administration of CCK-8S confers protection against renal inflammation, leading to a reduction of albuminuria in diabetic rats. Furthermore, CCK-8S directly inhibits high glucose-induced TNF-α expression and migration in cultured THP-1 cells. Our findings may provide a novel strategy of therapy for the early stage of diabetic nephropathy and other inflammatory diseases.

This study was supported in part by Grants-in-Aid for Scientific Research from the Ministry of Education, Science, Culture, Sports and Technology of Japan (Grant 21591031 to K.S.) and the Ministry of Health, Labour and Welfare of Japan.

No potential conflicts of interest relevant to this article were reported.

S.M. researched data, contributed to discussion, and wrote the manuscript. K.S. contributed to discussion and reviewed and edited the manuscript. K.M. researched data and reviewed and edited the manuscript. S.O., M.S., R.K., D.H., N.K., T.T., S.N., C.S., A.F., H.N., and H.A.U. researched data. H.U.K., D.O., and H.M. contributed to discussion. K.S. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.