Increased urinary levels of CXCL5, CXCL8 and CXCL9 in patients with Type 2 diabetic nephropathy

Journal of Diabetes and Its Complications 23 (2009) 178 – 184 WWW.JDCJOURNAL.COM

Increased urinary levels of CXCL5, CXCL8 and CXCL9 in patients with Type 2 diabetic nephropathy☆ Mayumi Higurashi a,b,c,⁎, Yoshiyuki Ohya d , Kensuke Joh d,e , Masahiro Muraguchi f , Motonobu Nishimura b,e , Hiroyuki Terawaki b , Kazuo Yagui a , Naotake Hashimoto a , Yasushi Saito a , Kenichi Yamada b,e a

Department of Clinical Cell Biology, Graduate School of Medicine, Chiba University, Chiba, Japan b Division of Clinical Research, Sakura National Hospital, Chiba, Japan c The Japanese Association Clinic, Singapore, Singapore d Division of Pathology, Sakura National Hospital, Chiba, Japan e Clinical Research Center, National Hospital Organization Chiba-East National Hospital, Chiba, Japan f Tokushima Research Institute, Otsuka Pharmaceutical Co. Ltd., Tokushima, Japan Received 17 August 2007; received in revised form 12 November 2007; accepted 7 December 2007

Abstract CXC chemokines are particularly significant for leukocyte infiltration in inflammatory diseases. Recent reports have shown that inflammation is one of potential pathogenic mechanisms for diabetic nephropathy. However, information on inflammation related with CXC chemokines in human Type 2 diabetic nephropathy still remains scarce. We measured urinary and serum levels of three CXC chemokines, CXCL5, CXCL8 and CXCL9, in 45 Type 2 diabetic patients (DM), 42 primary renal disease (PRD) patients and 22 healthy controls by enzyme-linked immunosorbent assay. Urinary levels of CXCL5, CXCL8 and CXCL9 in DM were significantly elevated compared to those in controls (Pb.0001, Pb.01, Pb.001; respectively). They increased consistent with urinary albumin excretion rate (UAER) and correlated with UAER in partial correlation analyses (r =0.41, Pb.01; r =0.40, Pb.01; r =0.45, Pb.01; respectively). Urinary levels of CXCL5 in DM were significantly interrelated to HbA1c (r =0.42, Pb.01). On the other hand, PRD showed significant increased levels of urinary CXCL8 and CXCL9 compared to controls (Pb.001, Pb.01; respectively), and so did PRD as UAER increased. However, there were no significant elevations of urinary levels of CXCL5 in PRD in spite of the increased UAER. We found significant associations of UAER in DM with diabetes duration, 1/serum creatinine, urinary CXCL5 (adjusted R2=0.67, Pb.0001) or CXCL9 (adjusted R2=0.69, Pb.0001) in a stepwise multiple regression analysis. These results suggest that these three CXC chemokines may be involved in the progression of human Type 2 diabetic nephropathy and that CXCL5 may be of use for telling diabetic nephropathy from primary renal diseases. © 2009 Elsevier Inc. All rights reserved. Keywords: Diabetic nephropathy; Inflammation; CXC chemokine

☆ This work was supported by a Health and Labour Science Research Grant for Clinical Research in Evidenced Based Medicine (0209001) from the Ministry of Health, Labour and Welfare of Japan. A portion of this study was presented by Mayumi Higurashi in the poster presentation entitled “Involvement of Activated Macrophages in the Development of Diabetic Nephropathy” at the Renal Week 2003 Meeting, San Diego, CA, USA, November 14–17, 2003. ⁎ Corresponding author. The Japanese Association Clinic, Singapore, 120 Adam Road, Singapore 289899, Singapore. Tel.: +65 64696488; fax: +65 64671298. E-mail address: [email protected] (M. Higurashi).

1056-8727/07/$ – see front matter © 2009 Elsevier Inc. All rights reserved. doi:10.1016/j.jdiacomp.2007.12.001

1. Introduction Chemokines are a family of proinflammatory peptides that play an important role in the migration of leukocytes to sites of tissue injury. There are two major families of chemokines, and these are referred to as CXC and CC chemokines, respectively, according to the position of the first two of four conserved cysteine residues (Baggiolini, Dewald, & Moser, 1997). In humans, CXC chemokines and

M. Higurashi et al. / Journal of Diabetes and Its Complications 23 (2009) 178–184

CC chemokines differ in the chromosomal location of their genes and their target cell selectivity. CXC chemokines mainly attract neutrophils and lymphocytes, whereas CC chemokines act on monocytes, lymphocytes, natural killer cells and dendritic cells. As most of CXC chemokines have high affinity for single receptors [CXCL8/interleukin (IL)-8, which binds to CXC chemokine receptor (CXCR) 1 and 2, is one of the exceptions] while CC chemokines recognize two or more receptors that differ in ligand specificity and cellular distribution, it is thought that CXC chemokines elicit more selective leukocyte responses than CC chemokines (Baggiolini et al., 1997). Chemokines are of great importance in inflammatory renal diseases, and they play a key pathogenic role in the events leading to kidney injuries. That is, a number of CC and CXC chemokines induce leukocyte infiltration into the glomeruli and the interstitium in human glomerulonephritis, tubulointerstitial diseases and renal transplant rejection (Schlöndorff, Nelson, Luckow, & Banas, 1997; Segerer, Nelson, & Schlöndorff, 2000), and the accumulation of leukocytes results in the synthesis of various pro-inflammatory or inflammatory substances that aggravate renal injuries. Such chemokines are synthesized by leukocytes and cells resident in the kidney (Schlöndorff et al., 1997; Segerer et al., 2000). It has been demonstrated that inflammation is one of potential pathogenic mechanisms for diabetic nephropathy since monocyte/macrophage infiltration was observed in human diabetic kidneys (Furuta et al., 1993; Tuttle, 2005; Navarro & Mora, 2005; Galkina & Ley, 2006). Previous studies have reported that serum or urinary levels of inflammatory parameters, including sialic acid (Chen et al., 1996), fibrinogen (Dalla Vestra et al., 2005), C-reactive protein (Dalla Vestra et al., 2005), IL-6 (Dalla Vestra et al., 2005) and tumor necrosis factor-alpha (TNF-α) (Navarro, Mora, Macía, & Garcia, 2003) are elevated in patients with Type 2 diabetic nephropathy. Drugs that have renoprotective effects for experimental or clinical diabetic nephropathy, such as pentoxifylline, mycophenolate mofetil and ruboxistaurin, are thought to work through their antiinflammatory activities (Navarro et al., 1999; Utimura et al., 2003; Kelly, Chanty, Gow, Zhang, & Gilbert, 2005). Schmid et al. (2006) recently described the activation of nuclear factor-κB transcriptional programs which control the expression of the genes activated during inflammation in human diabetic nephropathy. Moreover, many reports have focused attention on the relationship between the development of diabetic nephropathy and CCL2/monocyte chemoattractant protein-1 (MCP1, a member of CC chemokine family), which is considered to contribute to the recruitment of macrophages to the kidneys (Galkina & Ley, 2006). As regards the relationship between CXC chemokines and diabetic nephropathy, increased urinary excretion of CXCL8 was described in diabetic nephropathy in 2002 (Tashiro et al., 2002), and one recent report has demonstrated the elevated plasma levels of

179

CXCL8 and CXCL9/monokine induced by interferon γ in diabetic nephropathy (Wong et al., 2007). However, information on inflammation related with CXC chemokines in diabetic nephropathy still remains scarce compared to that in other inflammatory renal diseases. Therefore, we measured the serum and urine concentrations of CXC chemokines in patients with Type 2 diabetes, which are already known to be implicated in other inflammatory renal diseases, and investigated the associations of these levels with renal injuries. In this study, we examined three CXC chemokines whose ways of measurement had been commercially available in our laboratory as of 2001: CXCL5/epithelialderived neutrophil activating factor-78, CXCL8 and CXCL9. CXCL5 acts primarily on neutrophils and CXCR2 is the receptor for CXCL5 (Baggiolini et al., 1997), but it is considered that CXCL5 differs from CXCL8 in their mechanisms of induction (Walz et al., 1991). CXCL5 is present in human renal tubule epithelial cells and implicated in human renal allograft rejection (Schmouder, Strieter, Walz, & Kunkel, 1995) and urinary tract infection (Olszyna et al., 2000). CXCL8 is known to be a central chemotactic factor for neutrophils and is also able to activate neutrophils and induce granule release (Baggiolini et al., 1997). In human glomerulonephritis, CXCL8 production was found in endothelial cells, mesangial cells and tubular epithelial cells (Schlöndorff et al., 1997; Segerer et al., 2000). CXCL9 selectively attracts T cells via its receptor, CXCR3 (Baggiolini et al., 1997; Segerer et al., 2000). It is produced by resident glomerular cells and thought not only to act as a chemoattractant but also to induce the proliferation of mesangial cells via CXCR3 in human proliferative glomerulonephritis (Romagnani et al., 1999, 2002).

2. Material and methods 2.1. Subjects We selected 45 patients with Type 2 diabetes mellitus (DM) and 42 patients with primary renal diseases (PRD) as disease controls [16 with minimal-change nephrotic syndrome, 10 with membranous nephropathy, 8 with IgA nephropathy, 4 with focal glomerulosclerosis and 4 with membranoproliferative glomerulonephritis] from October 2001 to July 2004 at Sakura National Hospital. Twenty-two volunteers who were confirmed healthy in regular medical checkups including oral glucose tolerance test were used as healthy controls. Table 1 shows descriptive data for DM, PRD and the control group. Patients with allergic disease, liver disease, cancer, infectious disease, cardiovascular disease or cigarette smoking history were excluded from the study. In particular, patients with urinary tract infections were carefully eliminated by means of urine cultures and/or based on

180

M. Higurashi et al. / Journal of Diabetes and Its Complications 23 (2009) 178–184

Table 1 Profile of patients, divided according to levels of s-Cr, and control subjects Type 2 diabetes (DM) n Male/female Age (years) BMI (kg/m2) Diabetes duration (years) Mean BP (mmHg) ACEI therapy (%) ARB therapy (%) Statin therapy (%) s-Cr (μmol/L) LDL (mmol/L) HbA1c (%) UAER (mg/24 h) Urinary CXCL5 (pg/μmol Cr) Urinary CXCL8 (pg/μmol Cr) Urinary CXCL9 (pg/μmol Cr)

45 25/20 59±9 a 22.4±2.4 10.7±3.0 94±9 4 (9) 12 (27) 12 (27) 174±150 a 2.5±0.6 c 7.3±1.2 d 602 [189, 2189] c 3 [0, 16] d 4 [0.5, 14] b 4 [1, 11] a

Primary renal diseases 42 25/17 58±10 b 22.6±1.7 93±11 3 (7) 10 (24) 13 (31) 231±162 c 2.8±0.6 c 5.0±0.4 1050 [202, 1758] c 0.1 [0, 1] 3.5 [0.5, 23] a 2.2 [0, 7] b

Control 22 13/9 51±10 22.9±1.7 92±8

39±16 1.7±0.4 5.1±0.4 12 [3, 21] 0.1 [0, 0.5] 0 [0, 0] 0.1 [0, 0.1]

Values are shown as mean±S.D., number with a percentage in parentheses, or median with first quartile and third quartile in brackets. a b c d

Pb.001 vs. control subjects. Pb.01 vs. control subjects. Pb.0001 vs. control subjects. Pb.0001 vs. PRD and control subjects.

microscopic findings. Determination of antinuclear antibodies and rheumatoid factor in DM, C-reactive protein and tumoral markers including carcinoembryonic antigen, cancer antigen 125 and α-fetoprotein in DM and PRD was negative or within the normal range. None of the patients received any corticosteroids, immunosuppressive agents or anti-platelet drugs at the moment of entering the study. Diabetic nephropathy was diagnosed clinically if the following criteria were fulfilled: presence of diabetic retinopathy or diabetes duration of ≥8 years when persistent urinary albumin excretion rate (UAER) was ≥30 to b300 mg/24 h; presence of diabetic retinopathy when persistent UAER was ≥300 mg/24 h; no clinical or laboratory evidence of other kidney or renal tract disease; or diabetic nephropathy diagnosed by renal biopsy. Twenty-four patients with DM were treated with insulin, 4 with insulin plus oral hypoglycaemic agents, 6 with one or more oral hypoglycaemic agents, and 11 on control only by diet. No patients were on treatment with glitazones. Primary renal diseases without DM were diagnosed by renal biopsies, clinical characteristics and/or oral glucose tolerance test by the experienced nephrologists. The patients were grouped according to the degree of UAER: UAER b30 mg/24 h [DM Subgroup 0 (D0), n=7; PRD Subgroup 0 (P0), n=6]; UAER ≥30 to b300 mg/24 h [DM Subgroup 1 (D1), n=9; PRD Subgroup 1 (P1), n=9]; UAER ≥300 mg/24 h to proteinuria b3.5 g/24 h [DM Subgroup 2 (D2), n=16; PRD Subgroup 2 (P2), n=18]; or proteinuria ≥3.5 g/24 h [DM Subgroup 3 (D3), n=13; PRD Subgroup 3 (P3), n=9] (Table 2). All subjects provided written informed consent, and this study was approved by the Human Research Ethics Committee of Sakura National Hospital.

2.2. CXC chemokine measurements Spontaneously voided midstream urine and venous blood were collected after overnight fasting. Urine or serum samples obtained after centrifugation were stored at −20°C for no more 1 month until measurement. CXCL5, CXCL8 and CXCL9 levels were measured by an enzymelinked immunosorbent assay, using a human CXCL5specific polyclonal goat IgG (R&D systems, Inc., Minneapolis, MN, USA), a mouse anti-human CXCL8 monoclonal antibody (Otsuka Pharmaceutical Co. Ltd., Tokushima, Japan) and a mouse anti-human CXCL9 monoclonal antibody (BD Biosciences PharMingen, San Diego, CA, USA), with intra-assay coefficients of variation of 3.9% or less and a sensitivity of 20 pg/ml, respectively. We had verified beforehand that our way of measurements can detect ex vivo production of stimulant (lipopolysaccaride, phytohemagglutinin, OK432 or polysaccharide krestin)induced CXCL5, CXCL8 and CXCL9 in normal whole blood (data not shown). All concentrations of urinary CXC chemokines were expressed as values corrected by urinary creatinine levels. 2.3. Statistical analyses Data are presented as the mean±S.D., except for angiotensin-converting enzyme inhibitor (ACEI) therapy, angiotensin receptor blocker (ARB) therapy, statin therapy, which are presented as the number with a percentage in parentheses; UAER, urinary CXCL5, CXCL8 and CXCL9, which are expressed as the median with first quartile and third quartile in brackets. Comparisons among DM, PRD and controls (Table 1), four subgroups of DM (D0, D1, D2

M. Higurashi et al. / Journal of Diabetes and Its Complications 23 (2009) 178–184

181

Table 2 Profile of patients stratified by UAER Type 2 diabetes (DM)

n Male/female Age (years) BMI (kg/m2) Diabetes duration (years) Mean BP (mmHg) ACEI therapy (%) ARB therapy (%) Statin therapy (%) s-Cr (μmol/L) LDL (mmol/L) HbA1c (%) UAER (mg/24 h) Urinary CXCL5 (pg/μmol Cr) Urinary CXCL8 (pg/μmol Cr) Urinary CXCL9 (pg/μmol Cr)

Primary renal diseases

D0

D1

D2

7 4/3 59±12 23.0±1.2 8.8±2.6

9 5/4 60±10 23.4±0.9 9.7±2.7

16 9/7 60±9 22.2±2.4 11.0±2.8

D3 13 7/6 58±8 21.8±3.3 12.0±3.3

92±8.6 93±9.8 94±10 95±8.3 1 (14) 2 (22) 1 (6) 0 (0) 1 (14) 1 (11) 5 (31) 5 (38) 2 (29) 3 (33) 4 (25) 3 (23) 60±42 69±39 170±110 314±168a 2.2±0.2 2.4±0.3 2.4±0.5 2.9±0.8 7.4±0.8 7.5±0.5 7.2±1.1 7.0±1.8 12 [6, 24] 158 [107, 209] 604 [440, 1044]b 2698 [2035, 2839]c 0.1 [0, 0.1] 0.2 [0.1, 1] 7 [2, 13]d 12 [3, 24]b 0.1 [0, 0.1] 1 [0, 2]

9 [2, 23]e

14 [5, 23]f

0.1 [0, 0.2] 0.5 [0.1, 2]

6 [2, 13]g

9 [5, 24]a

P0

P1

P2

P3

6 4/2 52±10 21.8±2.8

9 6/3 56±11 22.5±2.0

18 10/8 60±10 22.9±1.7

9 5/4 58±12 23.1±1.9

92±8 93±10 94±14 94±10 1 (17) 2 (22) 0 (0) 0 (0) 1 (17) 1 (11) 5 (28) 3 (33) 2 (33) 3 (33) 5 (28) 3 (33) 70±47 98±62 268±119h 355±208i 2.3±0.4 2.3±0.4 2.9±0.4j 3.4±0.8h 5.0±0.7 5.1±0.4 5.0±0.3 4.8±0.5 10 [5, 14] 211 [175, 288] 845 [434, 1432]i 2976 [2501, 3789]k 0.1 [0, 0.1] 0.1 [0.1, 1.5] 0.1 [0, 1] 0.1 [0, 0.1] 0.1 [0, 0.5] 0.1 [0, 1]

9 [3, 29]l

0.2 [0, 3]

2 [1, 8]

0.2 [0, 6]

12 [2, 23]m 4 [1, 23]m

D0, P0: UAER b30 mg/24 h; D1, P1: UAER ≥30 to b300 mg/24 h; D2, P2: UAER ≥300 mg/24 h to proteinuria b3.5 g/24 h; D3, P3: proteinuria ≥3.5 g/24 h. Values are shown as mean±S.D., number with a percentage in parentheses or median with first quartile and third quartile in brackets. a Pb.0001 vs. D0 and D1, Pb.01 vs. D2; bPb.0001 vs. D0 and D1; cPb.0001 vs. D0, D1 and D2; dPb.01 vs. D1; ePb.01 vs. D0 and D1; fPb.001 vs. D0 and D1; g Pb.01 vs. D0; hPb.001 vs. P0 and P1; iPb.0001 vs. P0 and P1; jPb.01 vs. P1; kPb.0001 vs. P0, P1 and P2; lPb.01 vs. P0 and P1; mPb.01 vs. P0.

and D3; Table 2), or four groups of PRD (P0, P1, P2 and P3; Table 2) were performed by post hoc test (Bonferroni's method) after using one-way analysis of variance or Kruskal–Wallis test when appropriate. Statistical significance of the use of ACEI therapy, ARB therapy and statin therapy between DM and PRD was analyzed by means of Mann–Whitney U test (Table 1). Correlations between all variables were calculated using Spearman's correlation tests, and partial correlation analysis was performed adjusting for gender, age, body mass index (BMI), duration of diabetes, mean blood pressure (BP), ACEI therapy, ARB therapy, statin therapy, low-density lipoprotein (LDL) and HbA1c, to determine the extent to which a relationship was altered after adjusting for the other variables. Finally, to determine the relative contribution of variables to UAER, a forward stepwise multiple regression analysis was performed with UAER as the dependent variables. Independent variables were gender, age, BMI, duration of diabetes, mean BP, ACEI therapy, ARB therapy, statin therapy, 1/serum creatinine (s-Cr), LDL, HbA1c, urinary CXCL5, CXCL8 and CXCL9. P values of b.05 were considered statistically significant.

3. Results As shown in Table 1, male/female, age, BMI, mean BP, ACEI therapy, ARB therapy or statin therapy, s-Cr, LDL and UAER did not show any significant differences between DM

and PRD. DM had higher HbA1c than both PRD and controls (Pb.0001, respectively). DM and PRD exhibited higher age, s-Cr, LDL and UAER than controls. Urinary levels of CXCL5, CXCL8 and CXCL9 in DM were significantly elevated, respectively, compared with controls (Pb.0001, Pb.01, Pb.001; respectively). Urinary levels of CXCL8 and CXCL9 in PRD were also significantly greater than those in controls (Pb.001, Pb.01; respectively). PRD, however, showed no significant elevation of urinary CXCL5 compared with controls, and DM exhibited significantly higher levels of urinary CXCL5 compared with PRD (Pb.0001). No statistically significant differences were observed among all the subgroups of DM or PRD shown based on UAER status as regards gender, age, BMI, mean BP and percentage of ACEI therapy, ARB therapy or statin therapy (Table 2). s-Cr was significantly higher in D3 than in D0, D1 (Pb.0001, respectively) and D2 (Pb.01). UAER in D2 was greater than that in D0 and D1 (Pb.0001, respectively), and so was UAER in D3 than that in D0, D1 and D2 (Pb.0001, respectively). s-Cr was significantly higher in P2 than in P0 and P1 (Pb.001, respectively), and in P3 than in P0 and P1 (Pb.0001, respectively). UAER in P2 was greater than that in P0 and P1 (Pb.0001, respectively), and so was UAER in P3 than that in P0, P1 and P2 (Pb.0001, respectively). Urinary levels of CXCL5, CXCL8 and CXCL9 were elevated in DM, consistent with increased UAER. D3 showed significantly higher urinary levels of CXCL5 than D0, D1 (Pb.0001, respectively), and so did D2 than D1

182

M. Higurashi et al. / Journal of Diabetes and Its Complications 23 (2009) 178–184

(Pb.01). Urinary levels of CXCL8 were significantly increased in D2 and D3 than in D0 (Pb.01 and Pb.001, respectively) and D1 (Pb.01 and Pb.001, respectively). Urinary levels of CXCL9 were elevated in D3 than in D0, D1 (Pb.0001, respectively) and D2 (Pb.01). Urinary levels of CXCL8 and CXCL9 in PRD increased with the development of urinary albumin excretion, not in the stepwise manner seen in DM. P2 showed significantly higher urinary levels of CXCL8 compared with P0 and P1 (Pb.01, respectively). P3 showed significantly higher urinary levels of CXCL8 and CXCL9 compared with P0 (Pb.01, respectively). However, there were not any significant elevations in urinary levels of CXCL5 in PRD despite increased UAER. In contrast to the above data, CXCL8 was not detected in serum in either patients or controls. CXCL5 and CXCL9 in serum were also undetectable in all controls (0.0 [0–0] pg/ μmol) and most patients (0.0 [0–0.01] pg/μmol), and serum levels of each CXC chemokine showed no significant difference between any patient group and controls. There was no correlation between urine and serum levels of CXC chemokines in any group. Table 3 shows significant unadjusted and partial Spearman correlation coefficients between the variables in DM. UAER in DM showed significant correlations with duration of diabetes, 1/s-Cr, urinary CXCL5, CXCL8 and CXCL9 in unadjusted correlation analysis, and these relationships also remained significant after adjusting for the effect of gender, age, BMI, duration of diabetes, mean BP, ACEI therapy, ARB therapy, statin therapy, s-Cr, LDL and HbA1c by partial correlation analysis. Urinary CXCL5, CXCL8 and CXCL9 were not significantly correlated with 1/s-Cr after adjusting the variables, although they were correlated with it before adjusting them. HbA1c was significantly associated with urinary CXCL5 both before and after adjusting the variables. In multiple regression analysis, diabetes duration (Pb.01), 1/s-Cr (Pb.0001) and urinary CXCL5 (Pb.01) or CXCL9 (Pb.01) from the many independent variables described in the Material and Methods section were significant determinants of UAER in DM as follows: UAER=763 +(99×duration of diabetes)+(−1024×1/s-Cr)+(24×urinary CXCL5) (adjusted R2=0.67, Pb.0001), and UAER=674 +(97×duration of diabetes)+(−957×1/s-Cr)+(31×urinary CXCL9) (adjusted R2=0.69, Pb.0001).

4. Discussion The present study revealed a progressive rise in urinary CXCL5, CXCL8 and CXCL9 levels with increasing UAER, with no CXCL8 and little CXCL5 and CXCL9 elevation in the serum, in Type 2 diabetic patients. Moreover, we found significant correlations between UAER in DM and urinary CXCL5, CXCL8 or CXCL9 in partial correlation analyses. It is reported that expression of intercellular adhesion molecule (ICAM)-1 is increased in the kidneys of patients with diabetic nephropathy (Hirata et al., 1998), and ICAM-1deficient mice are resistant against diabetic renal injury (Chow, Nikolic-Paterson, Ozols, Atkins, & Tesch, 2005). Seeing that CXC chemokines as well as CC chemokines play a central role in the migration of leukocytes to sites of tissue injury, our results may also support the idea that there is a highly possible involvement of leukocyte infiltration into the kidneys in the progression of diabetic nephropathy. Few previous studies, however, have reported the relation between CXC chemokines and diabetic nephropathy, and the mechanism underlying the increased concentration of urinary CXC chemokines in diabetic patients is unknown at present. This study is a cross-sectional one, so that we can only propose the relation between the progression of diabetic nephropathy and CXC chemokines. Referring to the hypothetical model for the role of CXC chemokines in acute renal injury (Schlöndorff et al., 1997), a possible explanation is that some characteristic stimuli in the diabetic milieu, such as hyperglycemia (Ihm et al., 1998), the renin– angiotensin system (Mezzano et al., 2003) or the advanced glycation end products (Schwedler, Schinzel, Vaith, & Wanner, 2001), which have been reported to contribute to renal inflammation in diabetic nephropathy, activate production of CXC chemokines by renal resident cells, and, moreover, leukocytes attracted into the kidney parenchyma by CXC chemokines produce additional CXC chemokines. Infiltrating leukocytes and renal resident cells can themselves cause various types of renal damages by secretion of inflammatory mediators such as TNF-α (Navarro et al., 2003), growth factors (Cooper, 1998), proteolytic enzymes (Schreiner, 1991), reactive oxygen species (Fridovich, 1978) and various chemokines (Segerer et al., 2000). We believe that the report on the relationship between urinary CXCL5 and diabetic nephropathy is a new finding. In this study, urinary levels of CXCL5 in PRD did not

Table 3 Significant unadjusted and partial Spearman correlation coefficients between the variables in patients with type 2 diabetes

UAER 1/s-Cr HbA1c

Unadjusted Adjusted Unadjusted Adjusted Unadjusted Adjusted

ns, not significant.

Duration of diabetes

1/s-Cr

Urinary CXCL5

Urinary CXCL8

Urinary CXCL9

0.41 (Pb.01) 0.45 (Pb.01) ns ns ns ns

−0.76 (Pb.0001) −0.75 (Pb.0001)

0.61 (Pb.0001) 0.41 (Pb.01) 0.53 (Pb.001) ns 0.37 (Pb.05) 0.42 (Pb.01)

0.65 (Pb.0001) 0.40 (Pb.01) 0.61 (Pb.0001) ns ns ns

0.68 (Pb.0001) 0.45 (Pb.01) 0.62 (Pb.0001) ns ns ns

ns ns

M. Higurashi et al. / Journal of Diabetes and Its Complications 23 (2009) 178–184

increase significantly in spite of the elevation of UAER. On the contrary, significant correlations were noted between urinary levels of CXCL5 and UAER, or HbA1c in DM. The expression of CXCL5 has so far been demonstrated in human renal tubular cells in case of urinary tract infection and renal allograft rejection, most of which reveal primarily peritubular injury (Schmouder et al., 1995; Olszyna et al., 2000). That may be the reason why we did not find a significant rise in urinary CXCL5 in PRD. Interestingly, CXCL5 is also reported to be present in human platelets and plays an important role in platelet activation as well as in neutrophil migration (Power, Clemetson, Clemetson, & Wells, 1995). Since patients with diabetic nephropathy show activation of platelet function and up-regulation of the coagulation system (Young et al., 1995; Omoto et al., 1999; Carr, 2001), it is also interesting to consider that CXCL5 might be associated with the pathogenesis of diabetic nephropathy through platelet activation. In our study of diabetic patients, we observed a significant increase in urinary CXCL8 and a significant correlation between UAER and urinary levels of CXCL8. On the other hand, Tashiro et al. (2002) have shown that the levels of urinary CXCL8 in patients with diabetic nephropathy increased in microalbuminuric stage, but that the levels in macroalbuminuric stage were slightly lower than those in microalbuminuric stage. They concluded rather that CXCL8 may play a role at a comparatively early stage in diabetic nephropathy. The cause of discrepancy between our results and theirs is not clear although there could be some differences in the characteristics of the patients studied. One report has recently demonstrated the elevated plasma levels of CXCL9 in diabetic nephropathy (Wong et al., 2007), but the mechanism for the increased CXCL9 in diabetic nephropathy has remained unclear. In human glomerulonephritis, the number of CXCR3-positive cells, mainly interstitial T cells, was correlated with renal function, proteinuria and percentage of globally sclerosed glomeruli (Segerer et al., 2004). Seeing that previous reports have shown interstitial T-cell infiltration in diabetic kidneys (Mezzano et al., 2003) and that CXCL9 selectively attracts T cell via CXCR3, we may think of CXCL9 as one of the candidates that play a role in T-cell infiltration into the diabetic kidneys as well as in other inflammatory renal diseases. Incidentally, our divergent results in blood samples from those of Wong et al. (2007) may be caused by the differences in the subjects, blood specimens or way of measurements (they analyzed not serum but plasma samples of diabetic patients with normo- and micro-albuminuria using flow cytometry), but there have been few other reports regarding CXC chemokines in diabetic patients yet and we need more studies to elucidate it. Mezzano et al. (2003) have shown that there is a significant association between interstitial cell infiltration, including T cells and macrophages, and interstitial fibrosis in Type 2 diabetic nephropathy. Furthermore, patients with diabetic nephropathy have been reported to show the

183

activated status of neutrophil, such as enhanced spontaneous adherence (Takahashi et al., 2000) and elevated TNFα-stimulated superoxide production (Ohmori et al., 2000). Therefore, increased infiltration of both neutrophils and T cells in the kidneys may be an important contributor to the progression of diabetic nephropathy. Our finding of a significant correlation between urinary levels of CXC chemokines and UAER in DM suggests that neutrophils and T cells attracted by CXC chemokines also have the possibility to cause or enhance the injuries in diabetic kidney in addition to macrophages. However, further studies, such as experiments using kidney biopsy specimens to inquire into CXC chemokines production, longitudinal clinical investigations and so forth, are required to lend support to the pathogenic significance of CXC chemokines in diabetic nephropathy. References Baggiolini, M., Dewald, B., & Moser, B. (1997). Human chemokines. An update. Annual Review of Immunology, 15, 675−705. Carr, M. E. (2001). Diabetes mellitus: a hypercoagulable state. Journal of Diabetes and Its Complications, 15, 44−54. Chen, J., Gall, M. A., Yokoyama, H., Jensen, J. S., Deckert, M., & Parving, H. H. (1996). Raised serum sialic acid concentrations in NIDDM patients with and without diabetic nephropathy. Diabetes Care, 19, 130−134. Chow, F. Y., Nikolic-Paterson, D. J., Ozols, E., Atkins, R. C., & Tesch, G. H. (2005). Intercellular adhesion molecule-1 deficiency is protective against nephropathy in type 2 diabetic db/db mice. Journal of the American Society of Nephrology, 16, 1711−1722. Cooper, M. E. (1998). Pathogenesis, prevention, and treatment of diabetic nephropathy. Lancet, 352, 213−219. Dalla Vestra, M., Mussap, M., Gallina, P., Bruseghin, M., Cernigoi, A. M., Saller, A., Plebani, M., & Fioretto, P. (2005). Acute-phase markers of inflammation and glomerular structure in patients with type 2 diabetes. Journal of the American Society of Nephrology, 16, S78−S82. Fridovich, I. (1978). The biology of oxygen radicals. Science, 201, 875−880. Furuta, T., Saito, T., Ootaka, T., Soma, J., Obara, K., Abe, K., & Yoshinaga, K. (1993). The role of macrophages in diabetic glomerulosclerosis. American Journal of Kidney Diseases, 21, 480−485. Galkina, E., & Ley, K. (2006). Leukocyte recruitment and vascular injury in diabetic nephropathy. Journal of the American Society of Nephrology, 17, 368−377. Hirata, K., Shikata, K., Matsuda, M., Akiyama, K., Sugimoto, H., Kushiro, M., & Makino, H. (1998). Increased expression of selectins in kidneys of patients with diabetic nephropathy. Diabetologia, 41, 185−192. Ihm, C. G., Park, J. K., Hong, S. P., Lee, T. W., Cho, B. S., Kim, M. J., Cha, D. R., & Ha, H. (1998). A high glucose concentration stimulates the expression of monocyte chemotactic peptide 1 in human mesangial cells. Nephron, 79, 33−37. Kelly, D. J., Chanty, A., Gow, R. M., Zhang, Y., & Gilbert, R. E. (2005). Protein kinase C beta inhibition attenuates osteopontin expression, macrophage recruitment, and tubulointerstitial injury in advanced experimental nephropathy. Journal of the American Society of Nephrology, 16, 1654−1660. Mezzano, S., Droguett, A., Burgos, M. E., Ardiles, L. G., Flores, C. A., Aros, C. A., Caorsi, I., Vio, C. P., Ruiz-Ortega, M., & Egido, J. (2003). Renin– angiotensin system activation and interstitial inflammation in human diabetic nephropathy. Kidney International, 64(Suppl 86), S64−S70. Navarro, J. F., & Mora, C. (2005). Role of inflammation in diabetic complications. Nephrol Dial Transplant, 20, 2601−2604.

184

M. Higurashi et al. / Journal of Diabetes and Its Complications 23 (2009) 178–184

Navarro, J. F., Mora, C., Macía, M., & Garcia, J. (2003). Inflammatory parameters are independently associated with urinary albumin in type 2 diabetes mellitus. American Journal of Kidney Diseases, 42, 53−61. Navarro, J. F., Mora, C., Rivero, A., Gallego, E., Chahin, J., Macia, M., Méndez, M. L., & Garcia, J. (1999). Urinary protein excretion and serum tumor necrosis factor in diabetic patients with advanced renal failure: effects of pentoxifylline administration. American Journal of Kidney Diseases, 33, 458−463. Ohmori, M., Harada, K., Kitoh, Y., Nagasaka, S., Saito, T., & Fujimura, A. (2000). The functions of circulatory polymorphonuclear leukocytes in diabetic patients with and without diabetic triopathy. Life sciences, 66, 1861−1870. Olszyna, D. P., Opal, S. M., Prins, J. M., Horn, D. L., Speelman, P., van Devennter, S. J., & van del Poll, T. (2000). Chemotactic activity of CXC chemokines interleukin-8, growth-related oncogene-α, and epithelial cell-derived neutrophil-activating protein-78 in urine of patients with urosepsis. The Journal of Infectious Diseases, 182, 1731−1737. Omoto, S., Nomura, S., Shouzu, A., Hayakawa, T., Shimizu, H., Miyake, Y., Yonemoto, T., Nishikawa, M., Fukuhara, S., & Inada, M. (1999). Significance of platelet-derived microparticles and activated platelets in diabetic nephropathy. Nephron, 81, 271−277. Power, C. A., Clemetson, J. M., Clemetson, K. J., & Wells, T. N. (1995). Chemokine and chemokine receptor mRNA expression in human platelets. Cytokine, 7, 479−482. Romagnani, P., Beltrame, C., Annunziato, F., Lasagni, L., Luconi, M., Galli, G., Cosmi, L., Maggi, E., Salvadori, M., Pupilli, C., & Serio, M. (1999). Role for interactions between IP-10/Mig and CXCR3 in proliferative glomerulonephritis. Journal of the American Society of Nephrology, 10, 2518−2526. Romagnani, P., Lazzeri, E., Lasagni, L., Mavilla, C., Beltrame, C., Francalanci, M., Rotondi, M., Annunziato, F., Maurenzig, L., Cosmi, L., Galli, G., Salvadori, M., Maggi, E., & Serio, M. (2002). IP-10 and Mig production by glomerular cells in human proliferative glomerulonephritis and regulation by nitric oxide. Journal of the American Society of Nephrology, 13, 53−64. Schlöndorff, D., Nelson, P. J., Luckow, B., & Banas, B. (1997). Chemokines and renal disease. Kidney International, 51, 610−621. Schmid, H., Boucherot, A., Yasuda, Y., Henger, A., Brunner, B., Eichinger, F., Nitsche, A., Kiss, E., Bleich, M., Gröne, H-J., Nelson, P. J., Schlöndorff, P. J., Cohen, C. D., & Kretzler, M. (2006). Modular activation of nuclear factor-κB transcriptional programs in human diabetic nephropathy. Diabetes, 55, 2993−3003. Schmouder, R. L., Strieter, R. M., Walz, A., & Kunkel, S. L. (1995). Epithelial-derived neutrophil-activating factor-78 production in human

renal tubule epithelial cells and in renal allograft rejection. Transplantation, 59, 118−124. Schreiner, G. F. (1991). The role of the macrophage in glomerular injury. Seminars in Nephrology, 11, 268−275. Schwedler, S., Schinzel, R., Vaith, P., & Wanner, C. (2001). Inflammation and advanced glycation end products in uremia: Simple coexistence, potentiation or causal relationship? Kidney International, 59(Suppl 78), S32−S36. Segerer, S., Nelson, P. J., & Schlöndorff, D. (2000). Chemokines, chemokine receptors, and renal disease: From basic science to pathophysiologic and therapeutic studies. Journal of the American Society of Nephrology, 11, 152−176. Segerer, S., Banas, B., Wöornle, M., Schmid, H., Cohen, C. D., Kretzler, M., Mack, M., Kiss, E., Nelson, P. J., Schlöndorff, D., & Gröne, H. J. (2004). CXCR3 is involved in tubulointerstitial injury in human glomerulonephritis. The American Journal of Pathology, 164, 635−649. Takahashi, T., Hato, F., Yamane, T., Inaba, M., Okuno, Y., Nishizawa, Y., & Kitagawa, S. (2000). Increased spontaneous adherence of neutrophils from type 2 diabetic patients with overt proteinuria. Diabetes Care, 23, 417−418. Tuttle, K. R. (2005). Linking metabolism and immunology: diabetic nephropathy is an inflammatory disease. Journal of the American Society of Nephrology, 16, 1537−1538. Tashiro, K., Koyanagi, I., Saitoh, A., Shimizu, A., Shike, T., Ishiguro, C., Koizumi, M., Funabiki, K., Horikoshi, S., Shirato, I., & Tomino, Y. (2002). Urinary levels of monocyte chemoattractant protein-1 (MCP-1) and Interleukin-8 (IL-8), and renal injuries in patients with type 2 diabetic nephropathy. Journal of Clinical Laboratory Analysis, 16, 1−4. Utimura, R., Fujihara, C. K., Mattar, A. L., Malheiros, D. M., Noronha, I. L., & Zatz, R. (2003). Mycophenolate mofetil prevents the development of glomerular injury in experimental diabetes. Kidney International, 63, 209−216. Walz, A., Burgener, R., Car, B., Baggiolini, M., Kunkel, S. L., & Strieter, R. M. (1991). Structure and neutrophil-activating properties of a novel inflammatory peptide (ENA-78) with homology to interleukin 8. The Journal of Experimental Medicine, 174, 1355−1362. Wong, C. K., Ho, A. W., Tong, P. C., Yeung, C. Y., Kong, A. P., Lun, S. W., Chan, J. C., & Lam, C. W. (2007). Aberrant activation profile of cytokines and mitogen-activated protein kinases in type 2 diabetic patients with nephropathy. Clinical and Experimental Immunology, 149, 123−131. Young, B. A., Johnson, R. J., Alpers, C. E., Eng, E., Gordon, K., Floege, J., Couser, W. G., & Seidel, K. (1995). Cellular events in the evolution of experimental diabetic nephropathy. Kidney International, 47, 935−944.