Establishment of a Mouse IgA Nephropathy Model With the MBP-20-Peptide Fusion Protein.код для вставкиСкачать
THE ANATOMICAL RECORD 293:1729–1737 (2010) Establishment of a Mouse IgA Nephropathy Model With the MBP-20-Peptide Fusion Protein LEI ZHANG,1 FEI YE,1 YAN HE,1 DAN KONG,2 CHANGSONG HAN,1 ZHIJIE ZHAO,1 JIANG ZHU,3 HONGXUE MENG,1 XINGHAN LIU,4 AND XIAOMING JIN1* 1 Department of Pathology, Harbin Medical University, Hei Longjiang Province, People’s Republic of China 2 Molecular and Cellular Pathology, Graduate School of Medical Science, Kanazawa University, Kanazawa, Japan 3 Department of Orthopedics, First Hospital of Harbin Medical University, Hei Longjiang Province, People’s Republic of China 4 Department of Biochemistry, Harbin Medical University, Hei Longjiang Province, People’s Republic of China ABSTRACT Here, we aimed to determine whether immunoglobulin-A nephropathy (IgAN) could be induced in Balb/c mice by immunizing them with a fusion protein (MBP-20 peptide) comprising the maltose-binding protein (MBP) and a 20amino-acid peptide derived from Staphylococcus aureus. A recombinant plasmid encoding the fusion protein was constructed and expressed in bacterial cells. The synthetic 20-peptide was used to prepare the monoclonal antibody. Balb/c mice were immunized with the MBP-20-peptide fusion protein over a 21-week course before renal histology was examined at the light and electron microscopic levels. Direct immunoﬂuorescence staining with the anti-20-peptide monoclonal antibody was also performed using renal biopsy tissue from human IgAN patients as a comparison. IgA and IgG speciﬁc for the 20-peptide in human and mice serum were detected. The IgAN experimental mice developed a clinical and pathological proﬁle that closely resembled that of human IgAN patients, including the induction of hematuria and numerous histopathological features. Levels of IgA and IgG speciﬁc for the 20-peptide were signiﬁcantly increased in serum from the IgAN experimental mice and IgAN patients compared with control mice and non-IgAN patients. In IgAN model mice, the anti-20-peptide antibody labeled glomeruli, while the antibody strongly labeled glomeruli and weakly labeled tubular epithelial cells in renal tissue from human IgAN patients. In conclusion, immunization with an MBP-20-peptide fusion protein is able to induce clinical and pathological features closely resembling IgAN in Balb/c mice, indicating a potentially useful role for the model in the study of IgAN and related C 2010 Wiley-Liss, Inc. diseases. Anat Rec, 293:1729–1737, 2010. V Key words: IgA nephropathy; Staphylococcus aureus; MBP-20peptide fusion protein; 20-peptide; IgA nephropathy model Grant sponsor: Heilongjiang Postdoctoral Grant; Grant number: LRB-07-329; Grant sponsor: Harbin Special Fund for Technological Innovation; Grant numbers: 2006RFXXS035, 2007RFLXS017; Grant sponsor: The Innovation Foundation in Harbin Medical University. Lei Zhang and Fei Ye are contributed equally to this work. *Correspondence to: Xiaoming Jin, Department of Pathology, Harbin Medical University, Baojian Road 157, Nangang DisC 2010 WILEY-LISS, INC. V trict, Harbin, China. Fax: 86-451-86669472. E-mail: email@example.com Received 10 November 2008; Accepted 15 May 2010 DOI 10.1002/ar.21225 Published online 20 August 2010 in Wiley Online Library (wileyonlinelibrary.com). 1730 ZHANG ET AL. Immunoglobulin-A nephropathy (IgAN) is the most common primary and chronic glomerulonephritis worldwide (Sharmin et al., 2004). IgAN is typically diagnosed in young adults and is more common in males. Approximately 40% of patients experience recurrent episodes of macroscopic hematuria, frequently preceded by infection 1 or 2 days earlier (Van der Boog, 2005). Characteristic histopathological features of the disease include mesangial deposition of IgA and other immunoglobulin isotypes, such as IgG and IgM, in addition to complement components; leading to mesangial proliferation and ultimately glomerular ﬁbrosis (Widstam-Attorps et al., 1992). As many as 30% of patients progress to end-stage renal failure (Montinaro et al., 1999). Some research groups have reported a relationship between infection with methicillin-resistant Staphylococcus aureus (MSRA) and IgAN. For example, Koyama et al. (2004) observed a correlation between the development of IgAN and a history of MSRA infection. These and other MSRA-related glomerulonephritis have been found to be associated with the glomerular deposition of immune complexes containing IgA, C3, IgG, and sometimes IgM (Fridkin et al., 2005; Koyama et al., 1995; Long and Cook, 2006; Satoskar et al., 2006; Shimizu et al., 2005, 2007). Sharmin et al. (2004) demonstrated that the IgAN pathological immune response was centered on Staphylococcus superantigens and found that S. aureus cell membrane antigen, a 30–35 kDa protein, was able to induce IgAN in Balb/c mice when administered subcutaneously. Otherwise, a 20-amino-acid peptide (NVGGDNVDIHSIVPVGQDPH) on a 30–35 kDa protein was found to be the antigenic determinant responsible for the immune manifestations of the disease (Koyama et al., 2004). Therefore, we speculated that this 20-amino-acid-peptide has a potential role in the pathogenesis of IgAN. Several research groups have attempted to develop a mouse model for IgAN based on an induced immune response to BSA and Staphylococcus enterotoxin B (Liu et al., 1989), outer membrane protein of Escherichia coli (E. coli; Endo et al., 1993; Han et al., 1998) and dextran G200 (Gesualdo et al., 1990; Isaacs et al., 1981). Our group has attempted many IgAN models using methods from the above references. Unfortunately, no study to date has produced the stable and accurate model that will be required for studying IgAN experimentally. In this study, we established an IgAN animal model with the 20-peptide as an antigen on the basis of S. aureus antigens inducing IgA-type glomerulonephritis in Balb/c mice (Sharmin et al., 2004). We investigated whether using the 20-peptide as an antigen determinant of S. aureus can induce IgAN-like changes in Balb/c mice. MATERIALS AND METHODS Materials A synthetic DNA fragment encoding the 20-aminoacid-peptide cell membrane antigen of S. aureus, the pMAL-c2G/irrelevant 20 peptide plasmid and a synthetic 20-peptide was supplied by Shanghai Sangon Biological Engineering Technology & Services. The pMAL-c2G plasmid, which was used to generate the maltose-binding protein (MBP) fusion protein, was obtained from New England Biolabs, Beverly, MA, as were the amylose afﬁnity columns, restriction endonucleases, T4 DNA ligase, and T4 polynucleotide kinase. Taq DNA polymerase and DNA markers were purchased from TaKaRa Bio. HiTrap Protein A HP was from Amersham Biosciences, Piscataway, NJ. Balb/c mice were obtained from the Second Afﬁliated Hospital of Harbin Medical University. Frozen renal biopsies from 250 patients who had received a renal biopsy at the Pathology Department of Harbin Medical University (Harbin, China) between January 2009 and March 2010 were randomly selected. The diagnosis of nephropathy was conﬁrmed by histopathology. Sections from all biopsy specimens were also stained routinely for IgA, IgG, IgM, and complement component C3. Three investigators judged the ﬂuorescence intensity of the staining independently; intensity was graded semiquantitatively on a scale of 0 (no staining) to 4. Frozen renal IgAN biopsies were obtained from 50 patients (22 men and 28 women; average age 29.08 7.33 years). Two hundred biopsies from proven nonIgAN glomerulonephritis patients were included as controls, which included 50 cases of mesangial proliferative glomerulonephritis (MsPGN), 50 cases of membranoproliferative glomerulonephritis (MPGN), 50 cases of membranous nephropathy (MGN), and 50 cases of focal segmental glomerulosclerosis (FSGN). The protocol of the study was approved by the ethics committee in Harbin Medical University, and informed consent was obtained for sampling renal biopsy tissues. Cloning of the pMAL-c2G/20 Peptide Plasmid A synthetic DNA fragment encoding the 20 peptide amino acid sequence (NVG GDNVDIHSIVPVGQDPH) was synthesized (Koyama et al., 2004). It was ligated between the Hind III and SnaBI sites of the pMAL-c2G plasmid at a ratio of 3:1 using 1 lL T4 DNA ligase at 16 C overnight. The resulting plasmid was named pMAL-c2G/20 peptide. The entire length of the recombinant plasmid was sequenced by the Shanghai Sangon Biological Engineering Technology & Services, which conﬁrmed that the insert lay at the correct site and in the correct orientation. The plasmid was extracted with R Plus SV Minipreps DNA Puriﬁcation Systhe WizardV tem (Promega, Madison, WI) according to the manufacturer’s instructions. In addition, there is a 12aa fragment between the Hind III and SnaBI sites of the pMAL-c2G plasmid and we inserted another 8aa fragment on the Hind III site to form the pMAL-c2G/irrelevant 20 peptide plasmid. The pMAL-c2G/20 peptide and pMAL-c2G/irrelevant 20 peptide plasmids were used to transform E. coli BL21 cells, which were cultured in LB medium with antiaminobenzyl penicillin overnight at 37 C. The pMAL-c2G/20 peptide and pMAL-c2G/irrelevant 20 peptide plasmids respectively express MBP-20peptide fusion protein and MBP-irrelevant-20 peptide fusion protein. The sequence of irrelevant-20-peptide is YVEFGSSRVDLQSFPRKASE. Expression and Puriﬁcation of Fusion Proteins Induction of the Ptac promoter was accomplished by incubating bacteria carrying the pMAL-c2G/20 peptide plasmid or pMAL-c2G/irrelevant 20 peptide plasmid in 0.5 mM isopropylthiogalactoside (IPTG, Sigma-Aldrich, St. Louis, MO) for 6 hr at 37 C. Following expression of the plasmid in bacteria, the fusion protein was puriﬁed on ESTABLISHMENT OF A MOUSE IgAN MODEL 1731 Fig. 1. Expression of the MBP-20-peptide fusion protein (A) SDSpolyacrylamide gel electrophoresis of amylose column fractions from the puriﬁcation of the MBP-20-peptide fusion protein using maltose. Lane 1: puriﬁed protein; lane 2: hybrid protein; lane 3: precipitate after ultrasound and centrifugation; lane 4: supernatant after ultrasound and centrifugation; lane 5: induced bacterium liquid; lane 6: uninduced bacterial liquid. (B) Western blot analysis of MBP-20-peptide fusion protein and mouse monoclonal anti-20-peptide. Top two lanes: MBP20-peptide fusion protein; bottom two lanes: isolated 20-peptide. In each case, both denatured (D) and nondenatured (ND) protein samples were loaded. The antibody detected a fragment of the correct size in each case. an afﬁnity column containing maltose. In this expression system, the protein or peptide of interest is normally intended to be cleaved from the MBP moiety using a protease site engineered into the sequence of the vector. However, we predicted that the 20-peptide alone would not be of sufﬁcient size for an optimal immune response, so we therefore recovered the MBP-20-peptide-fusion protein intact by elution with free maltose (Fig. 1A). We conﬁrmed the afﬁnity of the monoclonal antibody for the 20-peptide by ELISA (data not shown). The sensitivity and speciﬁcity of the antibody for the 20-peptide was conﬁrmed by Western blotting (Fig. 1B). Preparation of Mouse Monoclonal Anti-20-Peptide Keyhole limpet hemocyanin (KLH, Sigma-Aldrich) is a commonly used carrier for peptide coupling in antibody production. The monoclonal antibody was produced by the following method: ﬁrst, an injection was made of KLH coupled synthetic 20-peptide (0.5 mg) and 250 lL deionized water, emulsiﬁed in 250 lL of complete Freund’s adjuvant. Balb/c mice were immunized by injection at two sites and the surplus was administrated intraperitoneally. Second, Mice were immunized after 3 weeks with the same dose of KLH coupled 20-peptide which was emulsiﬁed in incomplete Freund’s adjuvant. Balb/c mice were administered by intraperitoneal immunization, with increasing doses every 2 weeks. Third, from the third increased dose immunization, tail blood was collected to measure the antibody titer (valence) every 7 days. When valence was accorded with standards, nonemulsiﬁed KLH coupled 20 peptide was used for the increasing dose immunizations by intraperitoneal administration. Spleen cells were fused after 3 days with traditional methods to prepare monoclonal antibodies against KLH coupled 20peptide. Monoclonal antibodies were identiﬁed using indirect enzyme-linked immunosorbent assay (ELISA) methods, 1 lg mL1 BSA coupled 20-peptide coated ELISA ﬂask to select monoclonal hybridomas. Balb/c mice were administered by intraperitoneal injection of the hybridoma and monoclonal antibody was puriﬁed with HiTrapTM Protein A HP Columns (Amersham Biosciences) for use in the following experiments. Western Blot Analysis The MBP-20-peptide fusion protein and 20-peptide concentration was quantitated using the Bio-Rad protein assay (Bio-Rad Laboratories, Hercules, CA). Denatured MBP fusion protein was resolved on 12% gradient SDSpolyacrylamide gels and nondenatured MBP fusion protein on 12% gradient polyacrylamide gels and blotted onto nitrocellulose membrane for Western blot analysis. Denatured 20 peptide resolved on 16% gradient SDSpolyacrylamide gels and nondenatured 20 peptide on 16% polyacrylamide gels for Western Blot analysis. The blot was probed with suitable monoclonal anti-20-peptide and goat antimouse antibodies and was developed by the ECL chemiluminescent method (Amersham Biosciences) according to the manufacturer’s instructions. Immunization Mice One hundred and seven four-week old male Balb/c mice (weighing 20-–22 g) were acclimatized to standard laboratory conditions for 7 days prior to experimentation. A total of 30 mice received no treatment. Mice in the remaining groups were injected subcutaneously every 14 days. Of these, 30 mice received column elution buffer, 30 received 3 mg/kg of the MBP-irrelevant-20peptide fusion protein (derived from the expression of pMAL-c2G/irrelevant 20 peptide plasmid), and 80 mice received 3 mg/kg of the MBP-20-peptide fusion protein (IgAN experimental group). Injections were in incomplete Freund’s adjuvant (Sigma-Aldrich) in all groups. Injections continued for 21 weeks, during which time urine was collected weekly and analyzed for the presence of protein and red blood cells. Mice in each group were sacriﬁced at the end of 21 weeks and sections of 1732 ZHANG ET AL. the major organs were processed for light and electron microscopy. All morphological ﬁndings from experimental animals were observed independently by three pathology researchers (LZ, FY, and XM, J). Ten microscopic ﬁelds of two different areas (total 20 ﬁelds per animal at 400 magniﬁcation) were randomly chosen. The microscopic scores were set based on the sizes of the lesion involved. In our study, the IgAN animal model only showed proliferation in the mesangial region, thus we used the extent of mesangial proliferation as the evaluation standard of tissue injury. Normal histology with no mesangioproliferation was 0; mild mesangioproliferation, 1; moderate mesangioproliferation, 2; strong mesangioproliferation, 3. The severity of each variable was also graded as from 0 to 3. The overall histological injury scores were calculated by a summation of the scores relating to size and other variables. Urinalysis Blood and urine specimens were collected weekly and at the time of autopsy. Urinalysis was undertaken before freezing at 80 C. Urine erythrocytes were detected using a dipstick system (MULTISTIX (SBA)-610 multifunctional half-automatic biochemical analysis apparatus; Ji lin, China). Urine erythrocytes were also observed under light microscopy to distinguish between erythrocytes and hemoglobin. Therefore, our results demonstrated the level of erythrocytes in urine of the IgAN animal model, but not hemoglobin. Urine protein and creatinine were measured respectively using the Bradford protein assay and the Jaffe Creatinine Assay. Enzyme-Linked Immunosorbent Assay (ELISA) The concentrations of IgG and IgA in mice and human sera were assayed by ELISA. Each well of a polystyrene microtiter plate (Corning, NY) was coated with 20-peptide in carbonate buffer (0.05 M, pH 9.6) overnight at 4 C. After washing with PBS containing 0.05% Tween 20 (PBS-T), the plates were incubated with 1% fetal bovine serum for 60 min to block nonspeciﬁc reactivity of the sera. Plates were then incubated with mouse serum samples from control and IgAN experimental mice and serum from IgAN and non-IgAN patients at room temperature for 2 hr, and then washed with PBS-T. Peroxidase-conjugated goat antimouse and antihuman IgG and IgA were added to the plates and incubated at RT for 1 hr. After washing with PBS-T, TMB peroxidase substrate was added, and the reaction stopped with 2 N H2SO4. Absorbance was measured with BIO-RAD 550 microplate reader at a wavelength of 450 nm. Immunoﬂuorescence Examination Frozen slices from renal tissue of experimental mice were ﬁxed in acetone for 1 min. After ﬁxation, nonspeciﬁc protein binding sites were blocked with 5% normal goat serum in PBS (pH 7.4). IgA, IgG, IgM, and C3 in mouse renal tissues were detected with Fluorescein-labeled goat antimouse IgA, IgG, IgM (all Invitrogen, CA), and FITC antimouse complement component C3 (Cedarlane, Canada). A total of 20 glomeruli per mouse at 400 magniﬁcation were randomly chosen and images were acquired with a ﬂuorescence microscope (Nikon E800) using a digital camera (1200F; Nikon) and software (ACT-1; Nikon). Frozen slices from human renal biopsy specimens (50 cases of IgAN and 200 cases non-IgAN) and renal tissue of experimental mice after ﬁxation were blocked, and the slices were then incubated with the mouse monoclonal anti-20-peptide (1:100 dilution for 12 hr at 4 C), washed three times in PBS and incubated with FITC-labeled antimouse IgG antibody (Vector Laboratories, CA). All results were observed using a ﬂuorescence microscope (Nikon E800). Statistical Analysis Data are expressed as the mean SD. Serological statistical analysis of differences between the IgAN experimental group and controls were calculated using a oneway ANOVA. Immunoﬂuorescence statistical analysis was calculated using Kruskal-Wallis test for IgM and C3 and t-tests for IgA and IgG. Different statistical analyses were used to evaluate the immunoﬂuorescence differences between the IgAN experimental group and controls because the mean square of the IgA and IgG immunoﬂuorescence value was irregular. Histological injury statistical analysis was performed using a t-test. Statistical analysis of the differences in 20-peptide between IgAN patients and non-IgAN patients were performed using v2 analysis. P < 0.05 was considered statistically signiﬁcant. RESULTS Urinalysis Figure 2 shows that mice in the IgAN experimental group began to develop hematuria during the 11th week. It rose sharply on the 15th week before reaching a plateau on the 20th week (the IgAN experimental group; n ¼ 80, control groups; n ¼ 90) (P < 0.05). The urine P/C ratio began to remarkably increase after the 15th week, and then climbed again at 20 weeks to reach a maximum of approximately 4.2 0.46 (the IgAN experimental group; n ¼ 80, control groups; n ¼ 90) (P < 0.05). Neither hematuria nor proteinuria was seen in either of the control groups. Light and Electron Microscopy Light microscopy revealed a mild and moderate increase in the amount of mesangial matrix as well as proliferation of mesangial cells in the IgAN experimental mice while showing no other pathological changes (Fig. 3B). In contrast, in the other three control groups, mesangial matrix expansion and mesangial cell proliferation were not observed (Fig. 3A) (P < 0.05). Electron microscopy showed numerous large, electron-dense deposits in the mesangium and subendothelium in the immunized group (Fig. 3D,E), whereas there were no electron-dense deposits in the other three control groups (Fig. 3C). Immunoﬂuorescence Findings Immunoﬂuorescence microscopy, using anti-IgA, -IgG, -IgM, and -C3 antibodies showed deposits in the glomeruli in the IgAN experimental group, with particularly intense deposition of IgA and IgG only in the mesangium, there is no speciﬁc immunoﬂuorescence in the ESTABLISHMENT OF A MOUSE IgAN MODEL 1733 Fig. 2. Measurement of hematuria and proteinuria in mice chronically administered with MBP-20-peptide fusion protein (A) Hematuria. A small degree of hematuria was seen in the IgAN experimental group until the 15th week, when it increased sharply to reach a maximum at 17 weeks (P < 0.05, IgAN experimental group, n ¼ 80; the control groups, n ¼ 90). (B) Proteinuria. The urine P/C ratio began to remarkably increase after the 15th week and then climbed again at the 20th weeks to reach a maximum of approximately 4.2 0.46 (P < 0.05, IgAN experimental group, n ¼ 80; the control groups, n ¼ 90). Fig. 3. Histology and electron microscopy of glomeruli in the experimental mice. (A, B) Light micrographs of sections from a control mouse treated with the MBP-irrelevant-20-peptide fusion protein and the immunized mouse treated with MBP-20-peptide fusion protein stained with H&E (glomeruli indicated by arrows) (magniﬁcation: 400). (A) MBP-irrelevant-20-peptide fusion protein did not induce any pathological change. (B) The MBP-20-peptide fusion protein induced mild mesangial cell proliferation and mesangial matrix expansion in glomeruli. (C, D, E) Electron micrograph of glomeruli from a control mouse treated with the MBP-irrelevant-20-peptide fusion protein and the immunized mouse treated with MBP-20-peptide fusion protein (electron dense deposits indicated by arrows; 8000 magniﬁcation:). (C) In the MBP-irrelevant-20-peptide fusion protein control group, there was no pathological change. (D) An IgAN model mouse showing numerous large, electron-dense deposits in the mesangium. (E) An IgAN model mouse showing numerous large, electron-dense deposits in the subendothelium. capillary walls, blood vessels or interstitium (P < 0.001) (the IgAN experimental group; n ¼ 80, control groups; n ¼ 90). In contrast, the glomeruli in the control group where mice were immunized with the MBP-irrelevant20-peptide fusion protein showed slight IgG and IgM deposits (Fig. 4). The other two control groups of mice showed an absence of immunoﬂuorescence. We next examined renal tissue from both experimental mice and renal biopsies of IgAN and non-IgAN patients using the mouse monoclonal anti-20-peptide antibody in this study. The 20-peptide antigen was primarily associated with glomeruli in the IgAN experimental group but was not expressed in the other three control groups (Fig. 5A,B). In patient tissue biopsies, 1734 ZHANG ET AL. Fig. 4. Deposition of immunoglobulins and complement in glomeruli. A. Immunoﬂuorescence microscopy, using anti-IgA, -IgG, -IgM and -C3 antibodies showed glomerular deposits in the IgAN experimental group. Of these, glomerular deposition of IgA and IgG in the mesangium was particularly intense. In contrast, the glomeruli in the MBP-irrelevant-20peptide fusion protein control group showed slight IgG and IgM deposits. The other two control groups of mice were negative (immunized; n ¼ 80, control; n ¼ 30 in each group) (magniﬁcation: 200). B. The immunoﬂuorescence signal of IgA, IgG, IgM and C3 in the IgAN experimental group is stronger than in the control groups (IgAN experimental group, n ¼ 80;the control groups, n ¼ 90) (P < 0.001). Images were captured using a Nikon E800, and photomicrographs were quantiﬁed using ACT-1. mild and moderate 20-peptide antigen deposition was detected in 82% of IgAN, 16% of MsPGN, 14% of MPGN, 18% of MGN and 20% of FSGN patients. The expression of the 20-peptide in IgAN is signiﬁcantly higher than that in non-IgAN (P < 0.001) (n ¼ 50 in each group). The 20-peptide antigen was mainly deposited in glomeruli, with very little 20-peptide detected in renal tubular epithelial cells in biopsy tissue from IgAN patients (Fig. 5C,D). ferative glomerulonephritis with a severe glomerular IgA deposition in later life was described (Launay et al., 2000). Researchers then selected a strain from the ddY mice with a high incidence and an early onset of glomerular IgA deposition to develop a model of IgAN (Miyawaki et al., 1997). Now, some researchers focus on the relationship between respiratory tract infections and IgAN, so the outer membrane antigens of Haemophilus parainﬂuenzae (OMHP; Yamamoto et al., 2002) and S. aureus (Sharmin et al., 2004) antigens were used to establish an experimental model of IgAN in C3H/HeN mice and Balb/c mice (Sharmin et al., 2004) respectively. On the basis of the IgAN model induced by S. aureus, the 20-peptide of S. aureus was used to establish IgAN model in Balb/c mice for the ﬁrst time in our study. Our study used the MBP-irrelevant 20-peptide fusion protein as a control for the MBP-20-peptide fusion protein to immunize Balb/c mice. MBP-20-peptide fusion protein induced glomerular deposition of IgA, IgG, IgM, and C3, whereas MBP-irrelevant-20-peptide fusion protein only induced glomerular deposition of IgG and IgM. Immunoﬂuorescence showed IgA antibody and 20-peptide antigen co-deposition in the glomeruli of the IgAN experimental mice, demonstrating that using the 20-peptide as an antigenic determinant of S. aureus can induce experimental IgAN in Balb/c mice. The antibody to the 20-peptide antigen also mainly labeled glomeruli of human IgAN patients (82% in IgAN). Therefore, we believe this study illustrates that a close relationship Serological Findings Levels of serum anti-20-peptide IgA and IgG antibodies in the IgAN experimental group were signiﬁcantly higher than in the control groups (P < 0.001) (the IgAN experimental group; n ¼ 30, control; n ¼ 30 in each group), and the IgA and IgG concentration in serum from IgAN patients were higher than those from non-IgAN patients (P < 0.001) (n ¼ 30 in each group) (Fig. 6). DISCUSSION In the previous study, we tried to establish the mouse models using several previously described methods (Endo et al.,1993; Gesualdo et al., 1990; Han et al., 1998; Isaacs et al., 1981; Liu et al., 1989). However, we found that the result was not satisfactory. With the exception of the above-mentioned models, existing animal models of IgA are described below. Initially, a ddY mouse that can spontaneously develop mesangioproli- ESTABLISHMENT OF A MOUSE IgAN MODEL 1735 Fig. 5. Immunoﬂuorescence detection in renal tissue from IgAN model mouse and patients using the mouse monoclonal anti-20-peptide antibody (A) Section of kidney from a control mouse treated with the MBPirrelevant-20-peptide fusion protein. No expression of 20-peptide was detected (magniﬁcation: 400). (B) Section of kidney from an IgAN mouse. The glomeruli were intensely stained (glomeruli indicated by arrows; 400 magniﬁcation). (C) Renal biopsy section from a non-IgAN patient. No expression was detected in the glomeruli and tubular epithelial cells (200 magniﬁcation) of most non-IgAN cases. (D) Renal biopsy section from an IgAN patient. There is strong labeling of glomeruli and weak labeling of tubular epithelial cells in 82% cases from IgAN (glomeruli indicated by arrows; 200 magniﬁcation) (n ¼ 50 in each group) (P < 0.001). exists between the 20-peptide antigen and the pathogenetic development of IgAN. The IgAN experimental mice showed mild and moderate mesangial cell proliferation and mesangial matrix expansion, exhibiting electron-dense mesangial deposits after immunization with the MBP-20-peptide fusion antigens which have similarities to the pathological changes of human IgAN. IgA immune complexes deposited in glomeruli can induce leukocyte inﬁltration and inﬂammatory reactions, leading to damage of the nephric tubule and interstitium but not the glomeruli. Tubulointerstitial damage can inﬂuence glomerular hemodynamic changes resulting in hematuria and proteinuria (Sánchez-Lozada et al., 2003). In our study, hematuria was detected by the 11th week, which rose sharply on the 15th week before reaching a plateau by the 20th week. The urine P/C ratio began to remarkably increase after the 15th week, and then peaked at the 20th week. This result demonstrated that glomerular hemodynamic damage is aggravated when immunization times were increased. These ﬁndings were signiﬁcantly different from those of mice treated with the MBP-irrele- vant-20-peptide fusion protein and the other two control groups. In addition, our study demonstrated that anti-20 peptide IgA and IgG levels in mouse serum increased in the IgAN experimental Balb/c mice, resembling our observations in human IgAN. IgA in serum may conjugate the 20-peptide to form immune complexes which deposit in the glomeruli resulting in IgAN-like changes. However, the IgA system of mice signiﬁcantly differs from that of the human. In humans, several research groups have found that glomerular IgA deposition might occur not only due to IgA immune complexes but also due to the nonimmunological formation of macromolecular IgA1 induced by abnormal O-glycosylation in the IgA1 hinge (Sano et al., 2002). Altered O-glycosylation might favor self-aggregation of IgA1 (Kokubo et al., 1997) or act as an autoantigen in immune complexes with IgG (Tomana et al., 1999). Besides, the abnormal physiochemical properties of circulating IgA1, such as size, charge and glycosylation, might be one of the key pathogenesis factors of IgAN. (Hashim et al., 2001; Iwase et al., 2002; Leung et al., 2001, 2002; Sano et al., 1736 ZHANG ET AL. Fig. 6. Anti-20-peptide immunoglobulins are increased in IgAN patient and the immunized mice sera The concentration of IgA (A) and IgG (B) antibodies against the 20-peptide in the serum of IgAN patients was higher in the non-IgAN patients (P < 0.001) (n ¼ 30 in each group). The levels of IgA (C) and IgG (D) antibodies against 20-peptide in the serum of IgAN experimental mice was signiﬁcantly higher than in control mice (n ¼ 30 in each group) (P < 0.001). 2002). The interaction between IgA1 and human mesangial cells (HMC) via some special receptors (Tamouza et al., 2007; Wang et al., 2004) is one of the most important aspects in the pathogenesis of IgAN, resulting in the enhancement of HMC proliferation, inﬂammation, sclerotic cytokine release and extracellular matrix production to induce the renal injury of IgAN (Wang et al., 2004). Mice have only one form of IgA and lack the hinge region. Human IgA is mostly monomeric, whereas murine IgA is mostly polymeric. Moreover, human IgA has O- and N-glycans, whereas murine IgA has only Nglycans (Suzuki et al., 2005). Therefore, we propose that the 20-peptide antigen itself and/or this immune complex may be more important for the glomerular deposition of IgA than the nature of IgA in the immunized mice. Some studies demonstrated that the genetic background of IgAN patients could contribute to disease susceptibility (Galla, 2001; Hsu et al., 2000; Schena, 1995; Scolari, 2003). Resembling the human situation, the ddY mouse is a spontaneous animal model of human IgAN with a highly variable incidence and extent of mesangial proliferation and extracellular matrix expansion with paramesangial IgA depositions as a result of the heterogeneous background (Imai et al., 1985; Suzuki et al., 2005). Whether the induction of IgAN induced by the 20-peptide antigen is somehow regulated by speciﬁc genetic factors, or if the 20-peptide has changed the disease susceptibility in our study is unknown, but is worth investigating in the future. We chose to express the 20-peptide antigen as a fusion protein with MBP. One potential problem with this approach was that the antigen might have become hid- den within the foreign protein sequences during folding, thus becoming inaccessible to the immune system. However, our immunoblot data demonstrated that the monoclonal antibody against the 20-peptide was able to label the fusion protein under both denaturing and nondenaturing conditions. In addition, the fusion protein could be used to produce a faithful mouse model of the human disease, and the monoclonal antibody against 20-peptide distinguished IgAN from non-IgAN biopsy tissues. Nonetheless, the potential for the antigen to be obscured remains a consideration for other studies and should not be ruled out. A major difference between our approach and those used historically is that previous studies used whole membrane antigens of S. aureus in establishing an immunological IgAN model (Sharmin et al., 2004). The advantage of our model is that using the 20-peptide removes the inﬂuence of other S. aureus membrane antigens that do not normally participate in IgAN pathogenesis. We selected 21 weeks as the observation time to establish the IgAN animal model in our experiments, although this is more prolonged than in the previously established model (normally 15–16 weeks; Sharmin et al., 2004). The IgAN experimental mice showed mild and moderate mesangial cell proliferation and mesangial matrix expansion while showing no other morphological pathological changes. In summary, the 20-peptide of S. aureus induced mesangial deposition of IgA and C3, mesangial cell proliferation and mesangial matrix production in Balb/c mice. Our study has demonstrated that the 20-peptide-IgA complex induced glomerular and tubulointerstitial ESTABLISHMENT OF A MOUSE IgAN MODEL damage resulting in hematuria and proteinuria. Our study is the ﬁrst to establish an experimental model of IgAN with the 20-peptide of S. aureus. LITERATURE CITED Endo Y, Kanbayashi H, Hara M. 1993. Experimental immunoglobulin A nephropathy induced by gram-negative bacteria. Nephron 65:196–205. Fridkin SK, Hageman JC, Morrison M, Sanza LT, Como-Sabetti K, Jernigan JA, Harriman K, Harrison LH, Lynﬁeld R, Farley MM. 2005. Methicillin-resistant Staphylococcus aureus disease in three communities. N Engl J Med 352:1436–1444. Galla JH. 2001. Molecular genetics in IgA nephropathy. Nephron 88:107–112. Gesualdo L, Ricanati S, Hassan MO, Emancipator SN, Lamm ME. 1990. Enzymolysis of glomerular immune deposits in vivo with dextranase/protease ameliorates proteinuria, hematuria, and mesangial proliferation in murine experimental IgA nephropathy. J Clin Invest 86:715–722. Han QF, Fan MH, Zou WZ. 1998. Bacterium coli induced IgA nephropathy model. J Beijing Med Univ 30:85. Hashim OH, Shuib AS, Chua CT. 2001. Neuraminidase treatment abrogates the binding abnormality of IgA1 from IgA nephropathy patients and the differential charge distribution of its alphaheavy chains. Nephron 89:422–425. Hsu SI, Ramirez SB, Winn MP, Bonventre JV, Owen WF. 2000. Evidence for genetic factors in the development and progression of IgA nephropathy. Kidney Int 57:1818–1835. Imai H, Nakamoto Y, Asakura K, Miki K, Yasuda T, Miura AB. 1985. Spontaneous glomerular IgA deposition in ddY mice: an animal model of IgA nephritis. Kidney Int 27:756–761. Isaacs K, Miller F, Lane B. 1981. Experimental model for IgA nephropathy. Clin Immunol Immunopathol 20:419–426. Iwase H, Katsumata T, Itoh A, Hiki Y, Ikuko N, Sano T, Takatani T, Kobayashi Y. 2002. Detection of enriched Thomsen-Friedenrich antigen on IgA1 from IgA nephropathy patients. J Nephrol 15:703–708. Kokubo T, Hiki Y, Iwase H, Horii A, Tanaka A, Nishikido J, Hotta K, Kobayashi Y. 1997. Evidence for involvement of IgA1 hinge glycopeptide in the IgA1-IgA1 interaction in IgA nephropathy. J Am Soc Nephrol 8:915–919. Koyama A, Kobayashi M, Yamaguchi N, Yamagata K, Takano K, Nakajima M, Irie F, Goto M, Igarashi M, Iitsuka T. 1995. Glomerulonephritis associated with MRSA infection: a possible role of bacterial superantigen. Kidney Int 47:207–216. Koyama A, Sharmin S, Sakurai H, Shimizu Y, Hirayama K, Usui J, Nagata M, Yoh K, Yamagata K, Muro K, Kobayashi M, Ohtani K, Shimizu T, Shimizu T. 2004. Staphylococcus aureus cell envelope antigen is a new candidate for the induction of IgA nephropathy. Kidney Int 66:121–132. Launay P, Grossetête B, Arcos-Fajardo M, Gaudin E, Torres SP, Beaudoin L, Patey-Mariaud de Serre N, Lehuen A, Monteiro RC. 2000. Fcalpha receptor (CD89) mediates the development of immunoglobulin A (IgA) nephropathy (Berger’s disease). Evidence for pathogenic soluble receptor-Iga complexes in patients and CD89 transgenic mice. J Exp Med 191:1999–2009. Leung JC, Tang SC, Lam MF, Chan TM, Lai KN. 2001. Charge-dependent binding of polymeric IgA1 to human mesangial cells in IgA nephropathy. Kidney Int 59:277–285. Leung JC, Tsang AW, Chan LY, Tang SC, Lam MF, Lai KN. 2002. Size-dependent binding of IgA to HepG2, U937, and human mesangial cells. J Lab Clin Med 140:398–406. Liu ZH, Li LS, Li L. 1989. Staphyloentero-toxin induced IgA nephropathy model. Chin J Nephrol 5:6. Long JA, Cook WJ. 2006. IgA deposits and acute glomerulonephritis in a patient with staphylococcal infection. Am J Kidney Dis 48:851–855. 1737 Miyawaki S, Muso E, Takeuchi E, Matsushima H, Shibata Y, Sasayama S, Yoshida H. 1997. Selective breeding for high serum IgA levels from noninbred ddY mice: isolation of a strain with an early onset of glomerular IgA deposition. Nephron 76:201–207. Montinaro V, Gesualdo L, Schena FP. 1999. The relevance of experimental models in the pathogenetic investigation of primary IgA nephropathy. Ann Med Interne (Paris) 150:99–107. Rifai A, Small PA Jr, Teague PO, Ayoub EM. 1979. Experimental IgA nephropathy. J Exp Med 150:1161–1173. Sánchez-Lozada LG, Tapia E, Johnson RJ, Rodrı́guez-Iturbe B, Herrera-Acosta J. 2003. Glomerular hemodynamic changes associated with arteriolar lesions and tubulointerstitial inﬂammation. Kidney IntSuppl(86):S9–S14. Sano T, Hiki Y, Kokubo T, Iwase H, Shigematsu H, Kobayashi Y. 2002. Enzymatically deglycosylated human IgA1 molecules accumulate and induce inﬂammatory cell reaction in rat glomeruli. Nephrol Dial Transplant 17:50–56. Satoskar AA, Nadasdy G, Plaza JA, Sedmak D, Shidham G, Hebert L, Nadasdy T. 2006. Staphylococcus infection-associated glomerulonephritis mimicking IgA nephropathy. Clin J Am Soc Nephrol 1:1179–1186. Schena FP. 1995. Immunogenetic aspects of primary IgA nephropathy. Kidney Int 48:1998–2013. Scolari F. 2003. Inherited forms of IgA nephropathy. J Nephrol 16:317–320. Sharmin S, Shimizu Y, Hagiwara M, Hirayama K, Koyama A. 2004. Staphylococcus aureus antigens induce IgA-type glomerulonephritis in Balb/c mice. J Nephrol 17:504–511. Shimizu Y, Sakurai H, Hirayama K, Seki M, Yoh K, Yamagata K, Koyama A. 2005. Staphylococcal cell membrane antigen, a possible antigen in postmethicillin resistant Staphylococcus aureus (MRSA) infection nephritis and IgA nephropathy, exhibits high immunogenic activity that is enhanced by superantigen. J Nephrol 18:249–256. Shimizu Y, Seki M, Kaneko S, Hagiwara M, Yoh K, Yamagata K, Koyama A. 2007. Patients with IgA nephropathy respond strongly through production of IgA with low avidity against Staphylococcus aureus. Contrib Nephrol 157:139–143. Suzuki H, Suzuki Y, Yamanaka T, Hirose S, Nishimura H, Toei J, Horikoshi S, Tomino Y. 2005. Genome-wide scan in a novel IgA nephropathy model identiﬁes a susceptibility locus on murine chromosome 10, in a region syntenic to human IGAN1 on chromosome 6q22-23. J Am Soc Nephrol 16:1289–1299. Tamouza H, Vende F, Tiwari M, Arcos-Fajardo M, Vrtovsnik F, Benhamou M, Monteiro RC, Moura IC. 2007. Transferrin receptor engagement by polymeric IgA1 induces receptor expression and mesangial cell proliferation: role in IgA nephropathy. Contrib Nephrol 157:144–147. Tomana M, Novak J, Julian BA, Matousovic K, Konecny K, Mestecky J. 1999. Circulating immune complexes in IgA nephropathy consist of IgA1 with galactose-deﬁcient hinge region and antiglycan antibodies. J Clin Invest 104:73–81. Van der Boog PJ, van Kooten C, de Fijter JW, Daha MR. 2005. Role of macromolecular IgA in IgA nephropathy. Kidney Int 67:813– 821. Wang Y, Zhao MH, Zhang YK, Li XM, Wang HY. 2004. Binding capacity and pathophysiological effects of IgA1 from patients with IgA nephropathy on human glomerular mesangial cells. Clin Exp Immunol 136:168–175. Widstam-Attorps U, Berg U, Bohman SO, Lefvert AK. 1992. Proteinuria and renal function in relation to renal morphology. A clinicopathological study of IgA nephropathy at the time of kidney biopsy. Clin Nephrol 38:245–253. Yamamoto C, Suzuki S, Kimura H, Yoshida H, Gejyo F. 2002. Experimental nephropathy induced by Haemophilus parainﬂuenzae antigens. Nephron 90:320–327.