Genetic Engineering of Glomerular Sclerosis in the Mouse via Control of Onset and Severity of Podocyte-Specific Injury

Transcription

1 Genetic Engineering of Glomerular Sclerosis in the Mouse via Control of Onset and Severity of Podocyte-Specific Injury Taiji Matsusaka,* Jing Xin, Suguri Niwa, Kazuto Kobayashi, Akira Akatsuka,** Hiroomi Hashizume, Qing-cheng Wang, Ira Pastan, Agnes B. Fogo,* and Iekuni Ichikawa* Departments of *Pediatrics, Medicine, and Pathology, Vanderbilt University School of Medicine, Nashville, Tennessee; Institute of Medical Science, Department of Pediatrics, Center of Excellence, and **Educational and Research Support Center, Tokai University Medical School, Isehara, Kanagawa, Japan; Department of Molecular Genetics, Institute of Biomedical Sciences, Fukushima Medical University School of Medicine, Fukushima, Japan; Department of Pediatrics, National Defense Medical College, Saitama, Japan; and Laboratory of Molecular Biology, Division of Basic Sciences, National Cancer Institute, National Institutes of Health, Bethesda, Maryland This study aimed to generate a mouse model of acquired glomerular sclerosis. A model system that allows induction of podocyte injury in a manner in which onset and severity can be controlled was designed. A transgenic mouse strain (NEP25) that expresses human CD25 selectively in podocytes was first generated. Injection of anti-tac (Fv)-PE38 (LMB2), an immunotoxin with specific binding to human CD25, induced progressive nonselective proteinuria, ascites, and edema in NEP25 mice. Podocytes showed foot process effacement, vacuolar degeneration, detachment and downregulation of synaptopodin, WT-1, nephrin, and podocalyxin. Mesangial cells showed matrix expansion, increased collagen, mesangiolysis, and, later, sclerosis. Parietal epithelial cells showed vacuolar degeneration and proliferation, whereas endothelial cells were swollen. The severity of the glomerular injury was LMB2 dose dependent. With 1.25 ng/g body wt or more, NEP25 mice developed progressive glomerular damage and died within 2 wk. With ng/g body wt of LMB2, NEP25 mice survived >4 wkand developed focal segmental glomerular sclerosis. Thus, the study has established a mouse model of acquired progressive glomerular sclerosis in which onset and severity can be preprogrammed by experimental maneuvers. J Am Soc Nephrol 16: , doi: /ASN Transgenic and gene-targeting technologies have made the mouse the most useful species for studying the function of a specific mammalian gene in vivo. However, for nephrologists, the mouse is clearly disadvantageous for studying glomerular diseases because it is resistant to experimental maneuvers that are widely used to induce glomerular sclerosis in rats. For example, puromycin aminonucleoside does not cause nephrosis in mice, probably because of low activity of adenosine deaminase activity in the mouse kidney (1). In the present study, we aimed to develop a mouse model in which glomerular sclerosis can be reproducibly induced by a simple procedure. Of note, glomerular sclerosis is characterized by expansion of extracellular matrix in the mesangial area; nevertheless, pure mesangial injury, typically observed in rat anti-thy1 nephropathy, causes only reversible changes but not glomerular sclerosis (2). Of further note, podocyte injury is closely associated Received August 31, Accepted January 6, Published online ahead of print. Publication date available at Address correspondence to: Dr. Taiji Matsusaka, Department of Pediatrics, Vanderbilt University School of Medicine, st Avenue South, MCN C4204, Nashville, TN Phoen: ; Fax: ; taiji.matsusaka@ vanderbilt.edu with glomerulosclerosis in many human and animal kidney diseases (3 6). An experimental attempt to induce injury in podocytes and parietal epithelial cells by microinjecting saponin solution into Bowman s space in rats resulted in glomerular sclerosis (7). Recent studies in human genetics and gene-targeted mice revealed that mutations in molecules that have important roles in podocytes, including nephrin, podocin, CD2-associated protein, and -actinin-4, result in proteinuria and glomerular sclerosis (8 15). These findings indicate that injury of podocytes is a key initial step that triggers the pathologic sequence of events that leads to glomerulosclerosis (12,16 18). On the basis of these findings, we designed a transgenic mouse line in which selective podocyte injury can be induced. For this purpose, we used a recombinant immunotoxin, anti- Tac (Fv)-PE38 (LMB2). LMB2 is a chimeric protein composed of the Fv portion of an anti-tac (human [h] CD25) antibody and PE38, a mutant form of Pseudomonas exotoxin that contains the translocation and ADP ribosylation domains. LMB2 selectively kills cells that express hcd25 in vitro and also causes complete regression in mice that bear hcd25-expressing tumor (19). In addition, LMB2 showed a major clinical response in patients with CD25-expressing malignancies in a phase I trial (20). Because of highly potent and selective cytotoxic activity, LMB2 Copyright 2005 by the American Society of Nephrology ISSN: /

2 1014 Journal of the American Society of Nephrology J Am Soc Nephrol 16: , 2005 has also been used to ablate specific cell types in transgenic mice that express hcd25 in the target cells (21 24). This technology is termed immunotoxin-mediated cell targeting (21). In the present study, we first established transgenic mouse lines that express hcd25 selectively in podocytes. Injection of LMB2 led to a selective and irreversible injury to podocytes in a time-specified manner. This system has allowed us to demonstrate a sequence of events initiated by podocyte injury. After the injection of LMB2, transgenic mice dose-dependently developed proteinuria and glomerular damages and later glomerular sclerosis. Materials and Methods The Animal Experimentation Committee at Tokai University Medical School and the Institutional Animal Care and Use Committee at Vanderbilt University Medical School approved the protocol, in accordance with the principles and procedures outlined in the National Institutes of Health Guidelines for the Care and Use of Laboratory Animals. Generation of the Transgenic Mice A 5527-bp DNA fragment of mouse Nphs1 gene (Genebank no. AF190638) was obtained by PCR and screening of a genomic library (BD Biosciences Clontech, Palo Alto, CA). Two ATG sequences in the first exon were disrupted by site-directed mutagenesis. A DNA fragment that contains hcd25 cdna, FLAG sequence, and a polyadenylation signal of the human growth hormone gene were combined with the Nphs1 DNA fragment. The resultant DNA (Figure 1) was injected into fertilized eggs that were obtained from BDF1 C57BL/6N mating. The integration of the transgene was identified by PCR and confirmed by Southern blot analysis. Two transgenic mouse lines (lines 9 and 18) were established. Both lines showed essentially similar phenotypes. Line 18 was maintained by backcrossing with the C57BL/6N strain. All data presented in this article were obtained from transgenic mice or their wild-type littermates in line 18 backcrossed more than four times. PCR Primers For identifying transgenic mice, primers 5 -GTTTATTATCAGT- GCGTCCAG-3 and 5 -CTTGTCATCGTCGTCCTTGTA-3 were used. For reverse transcription PCR detecting hcd25 mrna, primers 5 - AAGCGGGTCACTCTATATGC-3 and 5 -ATAAACCATCTGCCCCAC- CAC-3 were used. Figure 1. Structure of the transgene. The transgene consisted of a 5.5-kb promoter fragment of the mouse nephrin gene (Nphs1), cdna for human (h) CD25, FLAG sequence, and a polya signal from the human growth hormone (hgh) gene. The Nphs1 promoter fragment contains the first and a part of the second exons with disruptions of ATG by site-directed mutagenesis. Immunostaining The following primary antibodies and kit were used: Monoclonal anti-hcd25 (1:20; Lab Vision, Fremont, CA), monoclonal anti-rat synaptopodin antibody (Progen, Heidelberg, Germany), guinea pig antimouse nephrin antibody (1:100; Progen), rabbit anti-human WT-1 (1: 500; Santa Cruz Biotechnology, Santa Cruz, CA), rabbit anti-rat laminin (1:50, Lab Vision), rabbit anti-mouse collagen IV (1:500, Chemicon, Temecula, CA), and bromodeoxyuridine (BrdU) staining kit (Oncogene Research Products, Boston, MA). For synaptopodin and nephrin staining, tissues were microwaved, and for hcd25 staining, tissues were autoclaved. For laminin and collagen IV staining, tissues were digested with protease XXV. To determine the density of WT-1 positive cells, WT-1 immunostaining was visualized by Cy3-conjugated anti-rabbit antibody (Jackson Laboratories, Bar Harbor, ME), and the number of WT-1 positive cells was determined in all glomeruli on the sections by fluorescence microscopy. The degree of glomerular injury and the tuft area were determined in the corresponding glomeruli in the adjacent periodic acid-schiff (PAS)-stained sections. Tangential sections (glomerular area 1500 m 2 ) were excluded from the analysis. Transmission and Scanning Electron Microscopy Three transgenic and two wild-type mice each were analyzed 12 h and 1, 2, 4, and 6 d after the injection of LMB2 (25 ng/g body wt). More than four (for transmission electron microscopy [EM]) and 30 (for scanning EM) glomeruli were assessed blindly and qualitatively for each mouse. Morphometric Analysis For quantifying glomerular damage, injuries of epithelial cells and mesangial cells were graded separately for each glomerulus on PASstained 2- m sections, using a score of 0 to 4. For mesangial changes, score 0 represents no lesion, whereas 1, 2, 3, and 4 represent mesangial matrix expansion, hyalinosis, or sclerosis, involving 25, 25% to 50%, 50% to 75%, and 75% of the glomerular tuft area, respectively. Because podocytes and parietal epithelial cells in injured glomeruli often intermingle and cannot be distinguished reliably, all epithelial cells in Bowman s space were evaluated. The epithelial cell injury was graded as follows; 0, no lesion; 1, 2, 3, and 4, vacuolization, bleb, or proliferation of epithelial cells involving 25%, 25% to 50%, 50% to 75%, and 75% of the glomerulus, respectively. More than 50 sequential glomeruli from each mouse were evaluated. For both types of injury, the average index value and the percentage of affected glomeruli were calculated as follows: average index (n1 2 n2 3 n3 4 n4)/(n0 n1 n2 n3 n4), % affected (n1 n2 n3 n4)/(n0 n1 n2 n3 n4), where n0, n1, n2, n3, and n4 are the number of glomeruli with no lesion and grades 1, 2, 3, and 4 lesion, respectively. To quantify synaptopodin staining, we processed all kidney samples and sections at the same time. More than 30 glomeruli, excluding tangential sections, were photographed sequentially with a digital camera (AxioCam; Carl Zeiss Japan, Tokyo, Japan), using the same conditions. The ratio of synaptopodin-positive area to glomerular tuft area for each glomerulus was determined by image analysis software (Win Roof; Mitani Co., Maruoka, Fukui, Japan). The average ratio was determined for each mouse. Miscellaneous Methods Under anesthesia with diethyl ether, mice received an intravenous injection of LMB2 diluted by 0.1 ml of PBS that contained 0.1% BSA. Twenty-four-hour urine specimens were collected using metabolic cages. Urinary protein data shown in this study were obtained from

3 J Am Soc Nephrol 16: , 2005 Mouse Glomerulosclerosis 1015 adult (6- to 15-wk-old) female mice. Concentrations of total protein and creatinine in the urine and concentrations of creatinine, urea nitrogen, and glucose in the plasma were determined with an automatic analyzer (SRL, Tokyo, Japan). Plasma total protein concentration was determined using Protein Assay Kit (BioRad, Hercules, CA), with BSA as a standard. Plasma and urine samples (2 l) were analyzed by SDS- PAGE, using 2 to 15% gradient gel. BrdU labeling was done by continuous infusion of BrdU, and it was started just after the injection of LMB2 and continued for 14 d using mini-osmotic pumps (Alza, model 2002). Statistical Analyses Results are expressed as means SEM. Unpaired t test was used to analyze the difference between two groups. One-way ANOVA was used for multiple comparisons. Values were regarded as significant at P Results Establishment of the Transgenic Mouse Lines (NEP25) that Express hcd25 in the Podocyte Using mouse nephrin gene (Nphs1) promoter, we established two lines (lines 9 and 18) of transgenic mice that express hcd25 within the glomerulus. The staining pattern of hcd25 coincided with that of synaptopodin, examined in the adjacent sections (Figure 2), demonstrating that the transgene is expressed in the podocyte. Reverse transcription PCR revealed that the transgene was also expressed in the brain and pancreas but not in other organs, e.g., the heart, spleen, thymus, lung, or muscle (data not shown). Nephrotic Syndrome Induced by LMB2 in the Transgenic Mice Two to 3 d after injection of 1.25 ng/g body wt or higher doses of LMB2, transgenic mice (line 18) developed nonselective proteinuria (Figures 3 and 4). With higher doses of LMB2, transgenic mice developed proteinuria more rapidly, but the maximum amount of urinary protein was similar. The transgenic mice showed hypoproteinemia, ascites, edema, and renal Figure 2. Expression of hcd25 in podocytes in the transgenic mice (NEP25). Immunostaining of synaptopodin (A and C) and hcd25 (B and D) in adjacent sections of the glomeruli of the transgenic mice. The two staining patterns coincide with each other, demonstrating that hcd25 is expressed in the podocyte. Magnification, 400. Figure 3. Proteinuria induced by anti-tac (Fv)-PE38 (LMB2) in NEP25 transgenic mice. Total protein amount was determined in 24-h urine in transgenic (TG) mice or wild-type (WT) mice before or after the injection of the indicated dose of LMB2. Two to 3 d after the injection of 1.25 to 50 ng/g body wt LMB2, the TG mice developed massive proteinuria. TG mice died at approximately 5 to 6 d after the injection of 25 to 50 ng/g body wt LMB2, 7 to 8 d after 5 ng/g body wt, and 11 to 14 d after 1.25 ng/g body wt. With ng/g body wt LMB2, TG mice developed proteinuria 7 d after the injection and gradually recovered. The mice survived 28 d. The numbers of mice analyzed are indicated below the bars. *P 0.05 versus WT injected with 50 ng/g body wt LMB2. failure (Figure 4, Table 1). Because of severe edema and/or renal failure, most transgenic mice died at approximately 5 to 6 d after the injection of 25 to 50 ng/g body wt LMB2, 7 to 8 d after 5 ng/g body wt, 8 to 9 d after 2.5 ng/g body wt, and 11 to 14 d after 1.25 ng/g body wt. With ng/g body wt of the toxin, a majority of the transgenic mice showed mild and transient ascites and survived for 28 d. Proteinuria peaked at the seventh day and thereafter gradually decreased with time, returning nearly to the normal range by the 28th day. Wild-type littermates with LMB2 or transgenic mice without LMB2 (Figures 5A, 5B, and 7G) showed no abnormal functional or morphologic phenotype. Transgenic mice of line 9 showed essentially similar but slightly milder and more slowly progressive phenotypes when compared with line 18. All of the data described below are obtained from line 18. Light Microscopic Findings With 1.25 ng/g body wt or more of LMB2, glomerular injury first appeared 3 to 4 d after the injection, and thereafter glomerular morphology progressively deteriorated with time. Podocytes were severely injured with large blebs and vacuoles (Figure 5, D and F). Parietal epithelial cells and proximal tubular cells within Bowman s capsule but not those in the tubules

4 1016 Journal of the American Society of Nephrology J Am Soc Nephrol 16: , 2005 Figure 4. SDS-PAGE analysis of plasma and urinary proteins. Two microliters of urine before and 2 d after LMB2 in the indicated dose (ng/g body wt) and the same volume of plasma 4 d after LMB2 were analyzed. LMB2 induced dose-dependent nonselective proteinuria and reduction in plasma proteins in transgenic mice (T) but not in wild-type mice (W). also showed vacuolar degeneration. Podocytes and parietal epithelial cells adhered to each other and sometimes could not be distinguished from each other. These cells often contained protein reabsorption droplets. Parietal epithelial cells were often increased in number and occasionally showed mitosis (Figures 5D and 6A, arrows). When the proliferating cells were labeled with a continuous infusion of BrdU, a large number of parietal epithelial cells were found to have incorporated this marker (Figure 6E) compared to NEP25 without LMB2 (Figure 6B). The proliferating parietal epithelial cells formed up to two layers but never a crescent, and no necrotizing lesions were present. Glomeruli showed hyalinosis (Figure 5F), deposition of fibrin, occasional mesangial hypercellularity, matrix expansion (Figure 5, D and F), increases in laminin and collagens (Figure 7, A through F), and mesangiolysis (Figures 5F and 6D). The transgenic mice that survived for 3 wk had global or segmental glomerular sclerosis (Figure 7H). We separately semiquantified the injuries of epithelial (both parietal and visceral) and mesangial cells observed in PAS sections. On the fifth day, the index (an average injury score) and percentage affected (a ratio of injured glomeruli) for both epithelial and mesangial injuries were correlated with the dose of LMB2 (Figure 8). We next examined the time course of the glomerular injuries after the injection of 1.25 ng/g body wt LMB2. The epithelial injury and mesangial changes markedly deteriorated at approximately the seventh day and became very extensive and severe on the 14th day (Figures 5, E and F, and 9). In both studies, the majority of glomeruli showed similar scores for epithelial injury and mesangial changes at each time point. After the injection of LMB2, staining for synaptopodin, WT-1, nephrin, and podocalyxin decreased in injured glomeruli (Figure 10). In severely damaged glomeruli, all of these proteins disappeared completely. In some glomeruli, synaptopodin, nephrin, and WT-1 but not podocalyxin were downregulated before apparent injury was observed in the adjacent section stained with PAS. We measured the areas of synaptopodin staining in sections adjacent to those used for scoring epithelial injury and mesangial changes. The synaptopodin area decreased inversely with epithelial injury and mesangial change indices (Figures 8C and 9C). Collectively, 1.25 ng/g body wt or higher doses of LMB2 caused rapidly progressive glomerular injury, and the rate of progression was correlated with the dose of LMB2. In contrast, transgenic mice that received injections of ng/g body wt LMB2 showed mild and slowly progressive mesangial expansion, which was accompanied by less remarkable epithelial injury (Figure 11). Three weeks after the injection, the mice developed focal segmental sclerosis. Four weeks after the injection, a part of the glomeruli showed more advanced sclerosis, but the rest of the glomeruli showed less severe injury than those at the third week (Figures 5, G and H, and 11). In sclerotic glomeruli, synaptopodin staining was reduced, and the average density of WT-1 cells was less than that in those that were not exposed to LMB2 ( / m 2 [n 78] versus [n 184]; P 0.05). In glomeruli with no or very mild damage (mesangial injury score 1 and epithelial injury score 1), the average density of WT-1 cells was not decreased ( [n 140]), indicating that podocytes are not lost in these glomeruli. Sclerotic glomeruli sometimes clustered, but superficial and deep cortices contained a similar extent of sclerotic glomeruli. Within the glomerulus, no specific location was predisposed to the sclerosis. Thus, the location of the segmental sclerosis was observed without specific relationship to the vascular or urinary poles. Sclerotic glomeruli often had intense PAS positive deposition, which were stained for IgG, A, and M (data not shown). Table 1. Renal function 5 d after the injection of 25 ng/g body wt LMB2 a Genotype n Plasma Protein (g/dl) Ccr ( l/min) BUN (mg/dl) Transgenic Wild-type P a LMB2, anti-tac(fv)-pe38; Ccr, creatinine clearance; BUN, blood urea nitrogen (measured in plasma).

5 J Am Soc Nephrol 16: , 2005 Mouse Glomerulosclerosis 1017 Figure 5. Histology of kidney sections of NEP25 transgenic mice. Transgenic mice without LMB2 (A and B), 7 d after 5 ng/g body wt LMB2 (C and D), 14 d after 1.25 ng/g body wt LMB2 (E and F), and 28 d after ng/g body wt LMB2 (G and H). The arrow in D shows mitosis in a parietal epithelial cell. Periodic acid-schiff (PAS) staining. Magnifications, 200 in A, C, E, and G; 400 in B, D, F, and H.

6 1018 Journal of the American Society of Nephrology J Am Soc Nephrol 16: , 2005 Figure 6. Cell proliferation in NEP25 transgenic mice. (A) Mitosis in parietal epithelial cells (arrows) in transgenic mice 7 d after 5 ng/g body wt LMB2 (PAS staining). (B through E) In transgenic mice either without (B) or with (C through E) LMB2, proliferating cells were labeled by continuous infusion of bromodeoxyuridine (BrdU) for 14 d and revealed by immunostaining for BrdU (B, C, and E). (D) PAS staining of the section adjacent to E. Note the avid BrdU incorporation in tubular cells, interstitial cells, and parietal epithelial cells after LMB2 injection (C and E). A few cells within the glomerulus also incorporated BrdU (E). Magnification, 400 in A, D, and E; 200 in B and C. Protein reabsorption droplets in proximal tubular cells and proteinaceous casts first appeared at the onset of proteinuria in transgenic mice. Thereafter, tubulointerstitial injury progressively advanced in parallel with proteinuria (Figure 5, C and E). Some tubular cells were detached from the basement membrane. A large number of F4/80-positive macrophages were present in the interstitium (data not shown). In the late phase, there was interstitial fibrosis, and tubules were often dilated (Figure 5E). Labeling with BrdU demonstrated cell proliferation in a large number of tubular and interstitial cells (Figure 6C). Transferase-mediated dutp nick-end labeling staining revealed rare apoptosis in tubular and interstitial cells but not in glomeruli by 14 d (data not shown). No abnormality was discernible in other organs, including the brain and the pancreas. Figure 7. Mesangial changes in NEP25 transgenic mice. Kidneys from transgenic mice either without LMB2 (A, C, E, and G) or 14 d after 1.25 ng/g body wt LMB2 (B, D, F, and H) were analyzed by immunostaining for laminin (A and B) or collagen IV (C and D), Masson s staining (E and F), or silver staining (G and H). Magnification, 400. Three-Dimensional Structure of Podocytes Without LMB2, NEP25 mice showed normal glomerular structure (Figure 12A). Four days after the injection of LMB2 (25 ng/g body wt), numerous microvilli were found to arise from the cell body and the foot process of podocytes in the glomeruli of the transgenic mice (Figure 12B). Podocytes often had thinner and fewer primary processes, and the cell body was lifted from the capillary wall. Eventually, podocytes showed a tadpole-like appearance with only one extended primary process (Figure 12B), which was observed in 30% of

7 J Am Soc Nephrol 16: , 2005 Mouse Glomerulosclerosis 1019 Figure 8. Dose dependency of glomerular injury assessed 5 d after the injection of LMB2. (A and B) Index and percentage affected for epithelial cell injury (A) and mesangial changes (B). (C) Average ratio of synaptopodin-stained area/glomerular tuft area. The number of mice studied is in parentheses. Figure 9. Time course of glomerular injury after the injection of 1.25 ng/g body wt LMB2. Index and percentage affected for epithelial cell injury (A) and mesangial changes (B). (C) Average ratio of synaptopodin-stained area/glomerular tuft area. The number of mice studied is in parentheses. glomeruli. Foot processes were segmentally blunted or retracted. Occasionally, denuded glomerular basement membrane with detached foot processes was observed (Figure 12C). Ultrastructure of the Glomeruli Between 12 h and 2 d after the injection of LMB2 (25 ng/g body wt), podocyte cell bodies were segmentally degenerated with microvillous transformation and vacuoles (Figure 12E). The foot processes of podocytes were initially preserved (Figure 12D) and later blunted or effaced (Figure 12E). In glomeruli with podocyte degeneration, some endothelial cells were swollen (Figure 12E), but peritubular capillaries of the same kidney and capillaries in other organs, including the heart, the liver, the intestine, and the lung, remained intact (data not shown). Rare, segmental areas of mesangiolysis were found (Figure 12D). The changes in podocytes and glomerular endothelial cells became more severe and extensive 6 d after the injection (data not shown). Some parietal epithelial cells and proximal tubular cells lining Bowman s capsule were also injured. Most glomerular basement membrane and mesangial cells remained intact at this stage. No deposits were present.

8 1020 Journal of the American Society of Nephrology J Am Soc Nephrol 16: , 2005 Figure 11. Time course of glomerular injury after injection of ng/g body wt LMB2. Index and percentage affected for epithelial cell injury (A) and mesangial changes (B). The number of mice studied is in parentheses. Figure 10. Downregulation of podocyte-specific proteins. Kidneys from transgenic mice either without LMB2 (A, C, E, and G) or 14 d after 1.25 ng/g body wt LMB2 (B, D, F, and H) were analyzed by immunostaining for WT-1 (A and B), synaptopodin (C and D), nephrin (E and F), and podocalyxin (G and H). (A) Without LMB2, WT-1 protein is intensely stained in podocytes and faintly in parietal epithelial cells. (B) After LMB2, the intensity of WT-1 staining was decreased. (C) Without LMB2, synaptopodin is stained in podocytes, showing continuous linear pattern. (D) After LMB2, synaptopodin staining is markedly decreased in injured glomeruli. (E) Without LMB2, nephrin is stained in podocytes. (F) After LMB2, nephrin staining is markedly decreased in injured glomeruli. (G) Without LMB2, podocalyxin is stained intensely in podocytes and faintly in endothelial cells. (H) After LMB2, podocalyxin staining is segmentally decreased in severely injured glomeruli. Magnification, 400. Discussion As predicted by the incompatibility of the mouse IL2 ligand and hcd25, the transgenic mice, unless they received LMB2, showed no abnormal phenotype. Previous studies showed that LD50 of LMB2 in wild-type mice is 340 to 500 ng/g body wt (25,26). In wild-type mice, a lethal dose of LMB2 causes hepatic injury (27) but not renal injury, although it concentrates markedly more in the kidney than in the liver (28). In the present study, we used substantially lower doses of LMB2, ranging from to 50 ng/g body wt. We confirmed that wild-type mice showed no functional and morphologic abnormality after the injection of LMB2. Mouse Nphs1 promoter is active in extrarenal tissues, including the brain and the pancreas (29 31). In our transgenic mice, hcd25 was expressed in the brain and the pancreas. However, LMB2 injection caused no injury in these organs. This may be ascribed, among others, to the protection by the blood-brain barrier and/or to a low level of transgene expression in these tissues. Intravenous injection of LMB2 induced podocyte injury in the transgenic mice. Damage and progressive loss of podocytes were demonstrated by light microscopy, transmission and

9 J Am Soc Nephrol 16: , 2005 Mouse Glomerulosclerosis 1021 Figure 12. Scanning and transmission electron microscopy (EM), showing lesions induced in NEP25 transgenic mice. (A through C) Scanning EM of glomeruli before (A) and 6 d after (B and C) injection of LMB2 (25 ng/g body wt). Podocytes show tadpole-like appearance with single extended primary processes (C). Denuded glomerular basement membrane is observed (C, arrow). (D and E) Transmission EM of glomeruli after the injection of LMB2 (25 ng/g body wt). Twelve hours after injection (D), podocytes (P) have a few vacuoles (arrow), and endothelial cells (E) are swollen. The foot processes are preserved. There are rare segmental areas of mesangiolysis (arrow heads). Two days after the injection (E), podocytes (P) are degenerated with microvilli (short arrows), lysosomes, and foot process effacement (arrow heads). Endothelial cells (E) are swollen. Bars 1 m. scanning EM, and immunostaining for podocyte marker proteins. Podocyte damage, assessed by synaptopodin staining, was dependent on the dose of LMB2 and the time lapse after the injection. In some glomeruli, WT-1, synaptopodin, and nephrin were downregulated, whereas no obvious morphologic change was observed in the adjacent PAS-stained sections. Therefore, the decrease in podocyte marker proteins reflects both downregulation of these proteins in injured but still alive podocytes and a decrease in the number of surviving podocytes. In severely injured or sclerotic glomeruli, all podocyte markers, including podocalyxin, almost completely disappeared, suggesting that podocytes are lost in these glomeruli. Transferase-mediated dutp nick-end labeling staining detected only rare apoptosis in the glomerulus at various time points after various doses of LMB2, indicating that apoptosis is not the major mechanism of the podocyte loss. Within 2 d after the injection, abnormal morphology was largely confined to podocytes and glomerular endothelial cells. Podocytes were degenerated with vacuoles, microvillous transformation, and foot process effacement, and glomerular endothelial cells were swollen. Two sets of additional findings make it highly unlikely that the glomerular endothelial injury was a result of the direct effect of the toxin. First, hcd25 was found to be absent from the glomerular endothelial cells in the transgenic mice. Second, the swelling of endothelial cells was observed only in the glomerular capillaries that had injured podocytes but not in the peritubular capillaries or capillaries in any other organs examined, i.e., the endothelial swelling requires injured podocytes in proximity, an anatomic arrangement that is unique to the glomerulus. Glomerular endothelial cells are highly dependent on the continuous supply of vascular endothelial growth factor (VEGF) from podocytes as shown by Quaggin et al. (32). These investigators elegantly demonstrated that disruption of just one allele of the VEGF-A gene specifically in the podocyte resulted in swollen endothelial cells in a manner similar to that observed in the present study. In our NEP25 mice, VEGF in podocytes was found to be depressed after the injection of toxin (data not

10 1022 Journal of the American Society of Nephrology J Am Soc Nephrol 16: , 2005 shown). Taken together, it is reasonable to speculate that the endothelial swelling after the LMB2 injection in our NEP25 mice was a result of decreased VEGF production by injured podocytes. In addition to podocytes and glomerular endothelial cells, mesangial cells and parietal epithelial cells and, later, tubular cells became damaged. This is contrasting to the phenomenon in other transgenic mice with the LMB2-mediated cell ablation system. In those mice, LMB2 injured only the target neuronal cells that expressed hcd25 but never the adjacent cells that did not express hcd25 (21 24). Mesangial cells showed proliferation, matrix expansion, and mesangiolysis. Mesangial cell changes were well correlated with epithelial cell injury and with the decrease in synaptopodin staining. It seems that mesangial changes appeared after the podocyte injury but before podocytes were lost. The mechanism that links podocyte injury and mesangial changes across the glomerular basement membrane is intriguing. Both proximal tubule type and flat epithelial cells of Bowman s capsule had vacuolar degeneration, adhesion, necrosis, apoptosis, detachment, and proliferation. Damage and proliferation of parietal epithelial cells have also been demonstrated in a Thy-1.1 transgenic mouse model. In that model, podocytes ectopically express Thy-1.1, and injection of anti Thy-1.1 antibody induces proteinuria and damage and proliferation of parietal epithelial cells, followed by sclerosis (33). Although the glomerular phenotype is in many ways different from that of our model, both the Thy-1.1 and our studies echo the notion that damage of podocytes can induce parietal cell injury. Severe parietal cell injury was often found without involvement of tubular cells that are adjacent to the glomerulus, suggesting that parietal cell-specific mechanisms of injury are likely to exist. At approximately 2 wk after LMB2 injection, glomerular capillary lumens became obliterated, and glomeruli showed early sclerosis. Mice that survived for 3 wk developed wellestablished segmental or global glomerular sclerosis. Sclerotic areas lacked podocytes depicted by synaptopodin staining. It is interesting that 4 wk after the injection of ng/g body wt LMB2, the average scores of epithelial injury and mesangial changes were better than those measured at 3 wk after the injection. In addition, urinary protein was almost in normal range 4 wk after the injection. These suggest that glomeruli that are only mildly injured by LMB2 have the ability to recover. Our model is remarkably different from puromycin aminonucleoside nephrosis in rats (34 36). In that, injection of puromycin aminonucleoside induces mild proteinuria without a decrease in plasma protein and an extensive effacement of foot processes (37). Glomeruli soon recover from these changes, and then mild focal segmental glomerular sclerosis slowly develops. Our model, with injection of ng/g body wt LMB2, showed a somewhat similar pattern in terms of recovery from proteinuria, slow progression of focal segmental glomerular sclerosis, and unremarkable vacuolization of podocytes and parietal epithelial cells. In contrast, the transgenic mice that were given higher doses of LMB2 developed severe glomerular damages, which rapidly progressed to sclerosis. In conclusion, our model system carries several unique characteristics. First, uniform glomerular injury and sclerosis can be induced by a simple procedure, i.e., a single intravenous injection of LMB2. Thus, we know the precise time and site of disease onset. Second, transgenic mice are completely normal unless they receive an injection of LMB2, so they are easily maintained and reproduced. Third, the severity can be controlled in accordance with the dose of LMB2. Fourth, combination of this model with other genetically manipulated mice will enable us to explore the function of specific genes that participate in the progression of glomerular sclerosis. Acknowledgments This study was supported by National Institutes of Health Grants DK37868 and DK and the Research for the Future Program and Grant-in Aid for Scientific Research of the Japan Society for the Promotion of Science. A part of this study was presented in abstract form at the American Society of Nephrology meeting (San Diego, CA, November 14 to 16, 2003) and The International Society of Nephrology meeting (Berlin, Germany, June 8 to 12, 2003). We thank Dr. Hataba (Jikei University, Tokyo, Japan) for kind instruction for SEM and Dr. Kurihara (Juntendo University, Tokyo, Japan) for the generous gift of anti-podocalyxin antibody and discussions. References 1. Nosaka K, Takahashi T, Nishi T, Imaki H, Suzuki T, Suzuki K, Kurokawa K, Endou H: An adenosine deaminase inhibitor prevents puromycin aminonucleoside nephrotoxicity. Free Radic Biol Med 22: , Mauer SM, Fish AJ, Blau EB, Michael AF: The glomerular mesangium. I. Kinetic studies of macromolecular uptake in normal and nephrotic rats. J Clin Invest 51: , Rennke HG: How does glomerular epithelial cell injury contribute to progressive glomerular damage? Kidney Int Suppl 45: S58 S63, Kriz W, Gretz N, Lemley KV: Progression of glomerular diseases: Is the podocyte the culprit? Kidney Int 54: , Kim YH, Goyal M, Kurnit D, Wharram B, Wiggins J, Holzman L, Kershaw D, Wiggins R: Podocyte depletion and glomerulosclerosis have a direct relationship in the PANtreated rat. Kidney Int 60: , Ross MJ, Klotman PE: Recent progress in HIV-associated nephropathy. J Am Soc Nephrol 13: , Laurens WE, Vanrenterghem YF, Steels PS, Van Damme BJ: A new single nephron model of focal and segmental glomerulosclerosis in the Munich-Wistar rat. Kidney Int 45: , Kestila M, Lenkkeri U, Mannikko M, Lamerdin J, Mc- Cready P, Putaala H, Ruotsalainen V, Morita T, Nissinen M, Herva R, Kashtan CE, Peltonen L, Holmberg C, Olsen A, Tryggvason K: Positionally cloned gene for a novel glomerular protein nephrin is mutated in congenital nephrotic syndrome. Mol Cell 1: , Shih NY, Li J, Karpitskii V, Nguyen A, Dustin ML, Kanagawa O, Miner JH, Shaw AS: Congenital nephrotic syndrome in mice lacking CD2-associated protein. Science 286: , 1999

11 J Am Soc Nephrol 16: , 2005 Mouse Glomerulosclerosis Boute N, Gribouval O, Roselli S, Benessy F, Lee H, Fuchshuber A, Dahan K, Gubler MC, Niaudet P, Antignac C: NPHS2, encoding the glomerular protein podocin, is mutated in autosomal recessive steroid-resistant nephrotic syndrome. Nat Genet 24: , Kaplan JM, Kim SH, North KN, Rennke H, Correia LA, Tong HQ, Mathis BJ, Rodriguez-Perez JC, Allen PG, Beggs AH, Pollak MR: Mutations in ACTN4, encoding alphaactinin-4, cause familial focal segmental glomerulosclerosis. Nat Genet 24: , Pollak MR: Inherited podocytopathies: FSGS and nephrotic syndrome from a genetic viewpoint. J Am Soc Nephrol 13: , Kim JM, Wu H, Green G, Winkler CA, Kopp JB, Miner JH, Unanue ER, Shaw AS: CD2-associated protein haploinsufficiency is linked to glomerular disease susceptibility. Science 300: , Kos CH, Le TC, Sinha S, Henderson JM, Kim SH, Sugimoto H, Kalluri R, Gerszten RE, Pollak MR: Mice deficient in alpha-actinin-4 have severe glomerular disease. J Clin Invest 111: , Roselli S, Heidet L, Sich M, Henger A, Kretzler M, Gubler MC, Antignac C: Early glomerular filtration defect and severe renal disease in podocin-deficient mice. Mol Cell Biol 24: , Kerjaschki D: Caught flat-footed: Podocyte damage and the molecular bases of focal glomerulosclerosis. J Clin Invest 108: , Mundel P, Shankland SJ: Podocyte biology and response to injury. J Am Soc Nephrol 13: , Pavenstadt H, Kriz W, Kretzler M: Cell biology of the glomerular podocyte. Physiol Rev 83: , Kreitman RJ, Bailon P, Chaudhary VK, Fitzgerald DJ, Pastan I: Recombinant immunotoxins containing anti-tac(fv) and derivatives of Pseudomonas exotoxin produce complete regression in mice of an interleukin-2 receptor-expressing human carcinoma. Blood 83: , Kreitman RJ, Wilson WH, White JD, Stetler-Stevenson M, Jaffe ES, Giardina S, Waldmann TA, Pastan I: Phase I trial of recombinant immunotoxin anti-tac(fv)-pe38 (LMB-2) in patients with hematologic malignancies. J Clin Oncol 18: , Kobayashi K, Morita S, Sawada H, Mizuguchi T, Yamada K, Nagatsu I, Fujita K, Kreitman RJ, Pastan I, Nagatsu T: Immunotoxin-mediated conditional disruption of specific neurons in transgenic mice. Proc Natl Acad Sci U S A 92: , Sawada H, Nishii K, Suzuki T, Hasegawa K, Hata T, Nagatsu I, Kreitman RJ, Pastan I, Nagatsu T, Kobayashi K: Autonomic neuropathy in transgenic mice caused by immunotoxin targeting of the peripheral nervous system. J Neurosci Res 51: , Watanabe D, Inokawa H, Hashimoto K, Suzuki N, Kano M, Shigemoto R, Hirano T, Toyama K, Kaneko S, Yokoi M, Moriyoshi K, Suzuki M, Kobayashi K, Nagatsu T, Kreitman RJ, Pastan I, Nakanishi S: Ablation of cerebellar Golgi cells disrupts synaptic integration involving GABA inhibition and NMDA receptor activation in motor coordination. Cell 95: 17 27, Sano H, Yasoshima Y, Matsushita N, Kaneko T, Kohno K, Pastan I, Kobayashi K: Conditional ablation of striatal neuronal types containing dopamine D2 receptor disturbs coordination of basal ganglia function. J Neurosci 23: , Onda M, Kreitman RJ, Vasmatzis G, Lee B, Pastan I: Reduction of the nonspecific animal toxicity of anti-tac(fv)- PE38 by mutations in the framework regions of the Fv which lower the isoelectric point. J Immunol 163: , Tsutsumi Y, Onda M, Nagata S, Lee B, Kreitman RJ, Pastan I: Site-specific chemical modification with polyethylene glycol of recombinant immunotoxin anti-tac(fv)-pe38 (LMB-2) improves antitumor activity and reduces animal toxicity and immunogenicity. Proc Natl Acad Sci U S A 97: , Onda M, Willingham M, Wang QC, Kreitman RJ, Tsutsumi Y, Nagata S, Pastan I: Inhibition of TNF-alpha produced by Kupffer cells protects against the nonspecific liver toxicity of immunotoxin anti-tac(fv)-pe38, LMB-2. J Immunol 165: , Kreitman RJ, Pastan I: Accumulation of a recombinant immunotoxin in a tumor in vivo: Fewer than 1000 molecules per cell are sufficient for complete responses. Cancer Res 58: , Moeller MJ, Kovari IA, Holzman LB: Evaluation of a new tool for exploring podocyte biology: Mouse Nphs1 5 flanking region drives LacZ expression in podocytes. JAm Soc Nephrol 11: , Putaala H, Soininen R, Kilpelainen P, Wartiovaara J, Tryggvason K: The murine nephrin gene is specifically expressed in kidney, brain and pancreas: Inactivation of the gene leads to massive proteinuria and neonatal death. Hum Mol Genet 10: 1 8, Moeller MJ, Sanden SK, Soofi A, Wiggins RC, Holzman LB: Two gene fragments that direct podocyte-specific expression in transgenic mice. J Am Soc Nephrol 13: , Eremina V, Sood M, Haigh J, Nagy A, Lajoie G, Ferrara N, Gerber HP, Kikkawa Y, Miner JH, Quaggin SE: Glomerular-specific alterations of VEGF-A expression lead to distinct congenital and acquired renal diseases. J Clin Invest 111: , Smeets B, Te Loeke NA, Dijkman HB, Steenbergen ML, Lensen JF, Begieneman MP, Van Kuppevelt TH, Wetzels JF, Steenbergen EJ: The parietal epithelial cell: A key player in the pathogenesis of focal segmental glomerulosclerosis in Thy-1.1 transgenic mice. J Am Soc Nephrol 15: , Glasser RJ, Velosa JA, Michael AF: Experimental model of focal sclerosis. I. Relationship to protein excretion in aminonucleoside nephrosis. Lab Invest 36: , Velosa JA, Glasser RJ, Nevins TE, Michael AF: Experimental model of focal sclerosis. II. Correlation with immunopathologic changes, macromolecular kinetics, and polyanion loss. Lab Invest 36: , Diamond JR, Karnovsky MJ: Focal and segmental glomerulosclerosis following a single intravenous dose of puromycin aminonucleoside. Am J Pathol 122: , Andrews PM: A scanning and transmission electron microscopic comparison of puromycin aminonucleoside-induced nephrosis to hyperalbuminemia-induced proteinuria with emphasis on kidney podocyte pedicel loss. Lab Invest 36: , 1977