Molecular Pathological Evaluation of Clusterin in a Rat Model of Unilateral Ureteral Obstruction as a Possible Biomarker of Nephrotoxicity

Abstract

In the present study, to determine the validity of considering clusterin as a possible biomarker of nephrotoxicity, the expression and distribution of clusterin in the rat UUO kidney were investigated. Real-time RT-PCR revealed an immediate increase in the clusterin mRNA level in the kidney, within 6 hours after UUO, and also maintenance of the mRNA expression level from day-1 to day-3 was 60-fold higher in the UUO kidney than in the sham kidney. ISH analysis revealed clusterin mRNA signals in the UUO renal tubular epithelium, whereas no signal was observed in the sham kidney. Detection of clusterin-α and -β was conducted using the subtype-specific antibodies, by both of western blotting and immunohistochemistry. Although clusterin-α was predominant in the UUO urine, only faint signals were noted at the brush border of the tubular epithelium or intraductal. On the other hand, strong signals of clusterin-β were detected in the UUO kidney homogenate, and the molecule was localized in the renal tubular epithelium. These results suggest that clusterin was translated in the renal tubular epithelium after de novo expression induced by renal injury. Thus, detection of clusterin mRNA and clusterin-β in the kidney or clusterin-α in the urine may be useful for predicting nephrotoxicity.

Introduction

It has been suggested that biomarkers are essential in the field of predictive toxicology. However, evaluation of nephrotoxicity using biomarkers has not yet been fully characterized. Recently, several studies have attempted to determine the gene expression frequencies associated with nephrotoxicity models using microarray technology (Huang et al., 2001; Luhe et al., 2003; Davis et al., 2004; Thukral et al., 2005). However, toxicogenomics using cDNA microarrays is not feasible for routine laboratory work because of the high cost of microarrays and also of the available gene expression databases related to toxicology. Furthermore, it is not easy to purify RNA from experimental animals or prepare labeled probes for routine toxicologic studies. On the other hand, if it were possible to predict toxicity using paraffin tissue sections or body fluids, it would be useful for routine toxicologic work.

It has been observed in toxicologic studies using experimental animals that the blood parameters of nephrotoxicity, namely, BUN and serum creatinine, are often not elevated in spite of the appearance of pathological changes in the kidney. In addition, no pathological findings can be found in most cases of early-stage renal failure. Therefore, it is necessary to establish the technology of predictive toxicologic pathology using biomarkers to diagnose renal toxicity in its initial phase (Chevalier, 2005; Thukral et al., 2005).

Clusterin is a disulfide-linked heterodimeric protein of 75 kDa, consisting of α- and β-subunits (Jones and Jomary, 2002; Kujiraoka et al., 2004). Clusterin has been shown to exhibit a protective gene function against gentamicin-induced renal tubular cell injury (Girton et al., 2002). Paradoxically, up-regulation of clusterin suggests the occurrence of renal injury. Therefore, clusterin may be a potential biomarker of nephrotoxicity.

The objective of this present study was to evaluate the potential for using clusterin as a biomarker for predicting nephrotoxicity using ISH, real-time RT-PCR, immunohistochemistry, and Western blotting.

Materials and Methods

Animals

Wistar rats were obtained from Charles River Japan (Yokohama, Japan) and acclimated until use. The animals were maintained in a 12-hour light/12-hour dark cycles, at a constant temperature of 23 ± 3°C and relative humidity of 50 ± 20%. Animal care and use were in conformity with the guidelines of the Institutional Animal Care and Use Committee of Taisho Pharmaceutical Co. Ltd. Thirty-six male rats, 8 weeks of age, were prepared. UUO was achieved by ligating the left ureter with a 5-0 silk suture. Sham-operated animals were used as controls. The rats were sacrificed by exsanguinations under anesthesia at 6 hours, and 1, 2, 3, or 7 days (n = 6 for each group) after the ureteric ligation and the obstructed kidneys were subjected to the following analyses.

For histological examination, the left kidney was removed and longitudinally divided, and one half was immediately fixed in neutralized 10% formalin buffer solution and then embedded in paraffin. For real-time RT-PCR and Western blot analyses, the other half was further cut in half and treated with RNALater (Ambion, Austin, TX, USA) to prevent RNA degradation, or frozen with liquid nitrogen.

Real-time RT-PCR

Total RNA purification was done using the TRIzol Reagent (Invitrogen, San Diego, CA, USA), in accordance with the manufacturer’s instructions. After the cDNA was prepared using the SuperScript preamplification system (Invitrogen) and oligo-dT primer, it was purified using a PCR purification kit (Qiagen, Valencia, CA, USA) and dissolved in 10 mM Tris-1 mM EDTA buffer at 10 μg/ml. Real-time RT-PCR was carried out using a QuantiTect SYBR Green RT-PCR Kit (Qiagen), according to the following protocol: denaturation at 95°C for 15 minutes, followed by 45 amplification cycles consisting of 10 seconds at 95°C, 20 seconds at annealing temperature, and 20 seconds at 72°C. The annealing temperatures were as follows: 54°C for clusterin, 57°C for fibronectin, and 61°C for β-actin. The primer sequences are shown in Table 1. β-actin was used as an internal control.

In situ Hybridization

The procedure for the ISH analysis was established at our laboratory (Ishii et al., 2004). Briefly, the following primer sequences were used to prepare RNA probes: forward, 5′-CAGGAGTTCTCTGACAATGAGC-3′; and reverse, 5′-AAGCTGTCCTGCATAGCATCTAGG-3′ (NIH accession: NM 053021). The predicted amplified PCR product size was 470 bp. The RNA probes were then labeled with digoxigenin (DIG), and the ISH was conducted using the Ventana HX automated ISH system (Ventana Japan, Yokohama, Japan).

Immunohistochemical Analysis

Three-μm-thick sections were boiled in 0.01 M citrate buffer (pH 6.0) in a microwave oven for 18 minutes to enhance the immunoreactivity of the anti-clusterin antibodies. For the immunohistochemical analysis, the sections were incubated with either anti-clusterin-α antibody (rabbit IgG, 10 μg/ml, Upstate biotechnology, Lake Placid, NY) which recognizes the rat clusterin precursor (70 kDa) and the α-subunit of clusterin (35 kDa), or anti-clusterin-β antibody (goat IgG, 0.25 μg/ml, Santa Cruz Biotechnology, Santa Cruz, CA) which recognizes the rat clusterin-β chain (43–64 kDa, due to glycosylation). Biotin-labeled secondary antibodies (Vector Laboratories, Burlingame, CA) were used at 10 μg/ml. Immunoreactivities were detected with di-aminobenzidine, and the nuclei were counterstained with hematoxylin. All the immunohistochemical analyses were performed using the Ventana NX Automated Immunohistochemistry System (Ventana Japan), in accordance with the manufacturer’s protocol.

Western Blotting

The kidney homogenate for the Western blot analysis was prepared with T-PER Tissue Protein Extraction Reagent (PIERCE, Rockford, IL, USA) containing Complete Proteinase inhibitor cocktail tablets (Roche Diagnostics, Indianapolis, IN, USA), in accordance with the manufacturer’s instructions. Then, after performing SDS-PAGE according to the standard method on 10% gels (Laemmli, 1970), the proteins from the kidney homogenate and urine were transferred to the Immuno-Blot PVDF Membrane for Protein Blotting (Bio-Rad Laboratories, Hercules, CA, USA). The membrane was blocked for 1 hour with Blocking Reagent (Roche Diagnostics) and rinsed three times for 5 minutes each in TTBS washing buffer (pH7.4) containing 20 mM Tris, 500 mM NaCl and 0.1% Tween 20. Detection of clusterin was performed using antibody against clusterin-α (rabbit IgG; 5 μg/ml) or clusterin-β (goat IgG; 0.2 μg/ml), the same as those used for the immunohistochemical analysis, and then incubated with alkaline phosphatase-conjugated anti-rabbit IgG (1:2000, Bio-Rad) or anti-goat IgG (1:20000, Santa Cruz) for 1 hour at room temperature. The blots were developed with Lumi-Imager F1 (Roche Diagnostics) using Immun-Star AP Substrate (Bio-Rad), in accordance with the manufacturer’s instructions.

Results

The time-course of changes in the mRNA expression levels of clusterin in the rat kidney after UUO was measured by quantitative RT-PCR. As shown in Figure 1A, the levels of clusterin mRNA increased immediately after the execution of UUO. The clusterin mRNA level was already 4.3-fold higher in the 6-hour-UUO kidney than in the sham kidney. At 24 hr after the UUO, the clusterin mRNA level was more than 60-fold that in the sham kidney, and continued to be elevated during and the expression was maintained at this level for the subsequent 3 days.

Expression of the fibronectin mRNA, the fibrosis marker, increased longitudinally. The highest expression in the UUO kidney was observed on day 3, at which time point, the fibronectin mRNA level was 10-fold higher than that in the sham kidney (Figure 1B). Histologically, interstitial fibrosis was observed on day 3 after the execution of UUO (data not shown). The expression of both clusterin mRNA and fibronectin mRNA decreased by day 7 in the UUO kidney, as the renal failure advanced.

ISH analysis was conducted to investigate the time-course of changes in the distribution of clusterin mRNA in the UUO kidneys. In the 6-hour-UUO kidney, clusterin signals were detected in the renal tubules in the outer stripe of the outer medulla (Figure 2B), whereas no signals were observed in these regions in the sham kidney (Figure 2A). The clusterin signals became progressively more marked in the renal tubular epithelial cells from day 1 to day 3 after the execution of UUO, to then decrease by day 7 after the execution of UUO (Figures 2C, 2D, 2E). No signals were detected using the sense probe in the day 3 UUO kidney (Figure 2F).

Western blotting was performed to measure the levels of clusterin protein in the kidney and urine. The following protein samples were examined; (1) homogenate of day-3 UUO kidney; (2) homogenate of sham kidney; (3) urine in the renal pelvis of day-3 UUO kidney (intra-kidney urine); (4) urine in the bladder of the sham-operated rat (intra-bladder urine). As shown in Figure 3A, the 70-kDa precursor protein of clusterin was detected in all the samples. Furthermore, the 35-kDa α subunit of mature clusterin was found in the intra-UUO kidney urine samples. There were no 35-kDa bands in the kidney homogenates or intra-bladder urine samples. On the other hand, two bands (lower band, 43 kDa; upper band, 63 kDa) were detected with anti-clusterin-β antibody in the homogenate of the day-3 UUO kidney. This difference appeared to be attributable to differences of glycosylation. Narrow smaller bands containing the 43-kDa and the lower band of the β subunit of clusterin were clearly observed in the sham-operated kidney, although no such bands were observed in either the intra-kidney or intra-bladder urine samples (Figure 3B).

The localization of the α and β subunits of clusterin was examined by immunohistochemistry using subtype-specific antibodies. As shown in Figure 4A, a faint signal of the α subunit of the protein was detected in the brush border of the tubular epithelium or intraduct in the UUO kidney using anti-clusterin-α antibody. No signals were observed in these regions in the sham kidney (data not shown). On the other hand, strong signals of the β subunit of clusterin were detected in the tubular epithelium (Figure 4B). The β subunit protein was immunohistochemically detected for extended periods after UUO, even on day 7 after the execution of UUO, by which time-point the clusterin mRNA expression level had decreased; on the other hand, no such signals were detected in the sham kidney (data not shown). The same localization of the clusterin-β subunit and clusterin mRNA in the tubular epithelium in the day-3 UUO kidney was confirmed in serial sections, by both immunohistochemistry and ISH (Figures 4C, 4D).

Discussion

The application of predictive pathology is essential to shorten the process of toxicology, because the process of drug discovery and development is cumbersome and expensive. To realize this, it is necessary to use biomarkers. Increased expression of clusterin has been established in several renal diseases and nephrotoxicity models (Darby et al., 1995; Silkensen et al., 1997; Huang et al., 2001; Davis et al., 2004), suggesting that clusterin may have a protective role against nephropathogenesis (Schwochau et al., 1998; Girton et al., 2002). Consistent with these reports, we confirmed that HEK-293 human embryonic kidney cells transfected with rat clusterin cDNA were protected from 2-bromoethylamine hydrobromide-induced cell toxicity (Ishii and Nakamura, 2005).

We conducted a detailed investigation of the distribution of clusterin mRNA by ISH, and demonstrated its increase in the UUO kidney by real-time RT-PCR. Clusterin mRNA expression was increased in the renal tubular epithelial cells in the UUO kidney, even from the very early stage of renal failure, that is, by 6 hr after the UUO procedure. This result suggests that detection of clusterin may be useful for early detection of renal tubular injury. Expression of clusterin mRNA was maintained at a high level up to three days after the execution of UUO, but by day 7 after the procedure, the expression level declined; this could be attributable to a decrease in the parenchymal volume because of progression of renal fibrosis, or advancing renal failure.

Western blotting with the α subtype-specific antibody of clusterin also revealed the presence of the 70-kDa precursor of the clusterin protein in the homogenate of both the sham and UUO kidneys. As the 70-kDa precursor of the clusterin protein was re-folded by boiling in the sample buffer with SDS, it could be detected with the α subtype-specific antibody of clusterin. However, no signals were detected in the sham kidney by immunohistochemistry. Because the 70-kDa precursor protein was folded in the fixed cells, the antigenicity recognized by the α subtype-specific antibody of clusterin might have been masked. On the other hand, clusterin-α after post-translational processing was noted at the brush border of the tubular epithelium or intraductally. These results suggest that while formalin-fixation remained very little clusterin-α secreted in the renal tubular lumen, so that the limiting detection of clusterin-α by immunohistochemistry was verified. In addition, the reason why the 35-kDa clusterin-α was not detected in the UUO kidney by Western blotting could be that there was little urine protein in the kidney homogenate samples.

On the other hand, Western blotting revealed accumulation of the clusterin-β in the UUO kidney homogenate, although no signal was found in the UUO urine with the β subtype-specific antibody of clusterin. These results were consistent with the intense signals in the renal tubular epithelium found by immunohistochemistry. In the sham kidney, the narrow smaller bands containing 43 kDa and lower band were recognized by chemiluminescence detection after Western blotting, but not by immunohistochemistry. These results suggest that the clusterin expressed constitutively in the sham kidney was less than the limit of detection by immunohistochemistry.

These results suggest that the clusterin protein was translated in the renal tubular epithelium after de novo expression of clusterin mRNA following renal injury, and then immediately converted into 2 subtypes by posttranslational processing. The α subtype was then excreted in the urine and the β subtype accumulated in the cytoplasm of the renal tubular epithelial cells. However, the mechanisms of posttranslational processing of clusterin remain to be clarified in detail.

Stimulation of the clusterin gene promoter region enhances clusterin expression. It has been demonstrated using ionizing radiation treated MCF-7 breast cancer cells that the expression of clusterin is up-regulated by activation of the insulin-like growth factor-1 (IGF-1)/IGF-1 receptor/Src/Mek/Erk signaling cascade (Criswell et al., 2005). IGF-1 is one of the various factors up-regulated in cases of obstructive nephropathy, and administration of IGF-1 was also shown to alleviate the tubular and interstitial pathology in the setting of UUO (Klahr, 2001). These observations suggest that the mechanism of clusterin expression might be associated with IGF-1.

As clusterin is immediately excreted into the extracellular space, interaction of clusterin with a putative clusterin receptor may be a possible mechanism underlying its protective role against nephropathogenesis. Megalin/Glycoprotein330 has been reported as a clusterin receptor (Zlokovic et al., 1996). However, Girton et al. (2002) concluded that the protective function of clusterin against gentamicin-induced renal tubular injury involved a megalin-independent mechanism. It has been suggested that clusterin exerts significant effects on physiological functions, including active cell death, immune regulation, cell adhesion, morphological transformation, and cytoprotection (Silkensen et al., 1995; Humphreys et al., 1999; Redondo et al., 2000; Caccamo et al., 2004). The reciprocal effects of clusterin functions might contribute to protection against renal tubular injury. Further studies are needed to clearly elucidate the functions of clusterin.

Renal interstitial fibrosis is an important process in chronic renal failure. In the UUO kidney, marked interstitial fibrosis after the expression of clusterin was confirmed by the up-regulation of fibronectin. It is known that TGF β 1 plays an important role in the process of fibrosis (Liu, 2006). It has been reported that the expression of clusterin is induced by TGF β 1 through the activation of AP-1 and protein kinase C (Jin and Howe, 1997). These observations suggest that the identification of clusterin mRNA expression might be predictive of a possible progression of renal failure.

Expression of clusterin mRNA in the renal tubular epithelium increased with progression of renal injury, while no clusterin signals were observed in the normal kidney. The detection of clusterin mRNA using ISH has the advantage that even a small number of injured cells can be confirmed. In this study, expression of clusterin mRNA was observed as early as within 6 hours after the execution of UUO. Furthermore, the detection of clusterin mRNA expression using ISH is feasible as a routine procedure in a toxicology laboratory, because paraffin sections fixed in neutralized 10% formalin-buffer solution are used in this histological examination. Furthermore, immunohistochemical detection of clusterin-β also appears to be useful. In addition, if a quantitative ELISA approach for the detection of clusterin-α in the urine is established, biochemical examination of urine samples may also be expected to be useful for predictive toxicology in the rat. Thus, clusterin may be a possible biomarker of nephrotoxicity.