American Journal of Transplantation 2006; 6: 2292–2306 Blackwell Munksgaard
C 2006 The Authors C 2006 The American Society of Journal compilation Transplantation and the American Society of Transplant Surgeons
doi: 10.1111/j.1600-6143.2006.01493.x
Connective Tissue Growth Factor is a Biomarker and Mediator of Kidney Allograft Fibrosis O. Chenga , R. Thuilliera , E. Sampsonb , G. Schultzb , P. Ruizc , X. Zhanga , P.S.T. Yuend and R.B. Mannona, ∗ a Transplantation Branch and d Renal Diagnostics and Therapeutics Unit, National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health, Bethesda, Maryland, USA b Department of Cell Biology and c Pathology, University of Florida Medical Center, Miami, Florida, USA ∗ Corresponding author: Roslyn B. Mannon,
[email protected]
Chronic allograft nephropathy (CAN) is a leading cause of kidney graft failure following transplantation. Its causes are complex and include both immunological and nonimmunological factors. Here we have studied the development of CAN in a mouse model of kidney transplantation comparing isografts and allografts. Unlike the normal histology and normal serum creatinine of the uninephrectomized, nonrejecting isografted mice (0.219 ± 0.024 mg/dL), allografted mice demonstrated severe renal dysfunction (mean serum creatinine 0.519 ± 0.061 mg/dL; p < 0.005) with progressive inflammation and fibrosis of the kidney. These animals also showed an increased expression of connective tissue growth factor (CTGF), both systemically and within the graft. CTGF was highly expressed in tubuloepithelial cells of allografts, along with a smooth muscle actin, a marker of myofibroblasts, and transcriptionally associated with other markers of fibrosis. In vitro studies of tubular epithelium indicate that CTGF is capable of inducing EMT, independent of TGF-b . Finally, in human transplant recipients, serum and urine CTGF levels are significantly elevated compared to na¨ıve individuals. Urinary levels correlated with the histological presence of CAN. These studies suggest a critical role of CTGF in graft fibrogenesis, for both mouse and man. Thus, CTGF has potential as a biomarker of CAN, and also a therapeutic target in managing graft fibrosis. Key words: Chronic allograft nephropathy, CTGF, fibrosis, kidney, mouse, TGF-b , transplantation Received 30 November 2005, revised 31 May 2006 and accepted for publication 22 June 2006
2292
Introduction With the advent of stronger immunosuppressants employed in kidney transplantation over the last decade, acute rejection rates have been reduced by nearly 50%. However, long-term graft survival has not improved to a similar magnitude (1). Late graft loss in living patients is predominantly due to chronic allograft nephropathy (CAN), a disease characterized by hypertension, proteinuria and a progressive loss of kidney function occurring months to years after transplantation. Indeed, recent studies demonstrate that the histologic features of CAN may be seen within the first year of transplantation and are likely commonplace prior to the development of significant clinical renal dysfunction (2). The histologic manifestations of CAN are well described and include tubular atrophy and interstitial fibrosis (3), often accompanied by a chronic inflammatory cell infiltrate, as well as arteriosclerosis and glomerulosclerosis. The classification of CAN, moreover, encompasses a number of etiologies including both antigen-dependent and antigen-independent insults, and histologic evaluation may provide critical clues to cause and possible treatment (4). Regardless of the insult causing tissue injury following transplantation, the response to injury consists of both a proliferative response and an inflammatory response, culminating in extracellular matrix. Transforming growth factor beta (TGF-b) has long been recognized as a mediator of CAN in both humans (5–7) and rodents (8,9). Moreover, epithelial-mesenchymal transformation (EMT) is now appreciated as a critical event in fibrogenesis in kidney grafts (10,11), with TGF-b implicated in myofibroblast differentiation in kidney allografts (12). However, TGF-b has not only a fibrogenic effect, but important immunomodulatory properties in suppressing inflammation (13,14) and acute rejection (15,16). Taken together, these studies suggest that neutralizing TGF-b may be a less suitable strategy to prevent fibrosis in transplant settings and identifying an alternative mediator involved in fibrogenesis would provide a better target for ameliorating graft fibrosis (17). A recently recognized downstream effector of TGF-b is connective tissue growth factor (CTGF), a 38 kDa, heparinbinding cysteine-rich protein that can induce cell proliferation, collagen synthesis and chemotaxis in a variety of cells (18). Recent reports suggest that different domains in CTGF mediate collagen synthesis and myofibroblast differentiation (19). In vivo, CTGF appears to have
CTGF and Chronic Allograft Nephropathy
prosclerotic properties in rodent models of fibrotic kidney disease (20,21) and in human kidney disease (22). More recently, CTGF has been linked to chronic graft injury in rat kidney allografts during viral infection (23) and in rejecting mouse kidney allografts (24) and this response is not kidney specific, as heart allografts with chronic rejection and overexpression of TGF-b have elevated CTGF expression within the graft (25). However, the exact role of CTGF is not well understood nor is its potential as a target for amelioration of atrophy and fibrosis. To further explore the relationship between CTGF and chronic kidney allograft injury, we examined expression of CTGF using a mouse kidney transplant model. Here, we demonstrate that following transplantation of a kidney graft, gene and protein expressions for CTGF rise in allografted kidneys compared to isografts, and appear systemically in the recipient as well. Moreover, exposure of tubular epithelial cells to CTGF results in upregulation of markers of myofibroblast differentiation demonstrating that CTGF is capable of inducing EMT within renal tissue. Finally, human transplant recipients with tubular atrophy and fibrosis (i.e. ‘CAN’) have significantly elevated gene transcription for CTGF compared to biopsies from recipients with stable function and have elevated urinary CTGF levels compared to nonrejecting recipients. These studies strongly implicate that CTGF is a critical mediator of EMT and graft fibrosis in kidney transplant recipients.
Materials and Methods Animals C57BL/6 and Balb/c or (Balb/cX DBA/2)F 1 mice were purchased from Jackson Laboratory (Bar Harbor, ME, USA) or NCI (Frederick, MD, USA) at 6–8 weeks of age. For the allograft group, kidneys from C57BL/6 were transplanted into MHC-incompatible Balb/c or (Balb/cX DBA/2)F 1 mice. As an isograft control, kidneys from C57BL/6 were transplanted into their littermates. Mice were kept in the specific pathogen-free facility of the National Institute of Health and received standard laboratory chow and tap water ad libitum under approved institutional protocols in accordance with the NIH Guide for the Care and Use of Laboratory Animals.
Mouse kidney transplantation Vascularized kidney transplants were performed in mice as previously described (26). Briefly, mice were anesthetized with isoflurane, and the donor kidney, ureter and half of the bladder were recovered en bloc, including the renal artery with a small aortic cuff and the renal vein with a small caval cuff. These vascular cuffs were anastomosed to the recipient abdominal aorta and vena cava, respectively, below the level of the native renal vessels. Total cold ischemic time averaged 30–35 min. Donor and recipient bladders were attached dome to dome. The right native kidney was removed at the time of transplantation and a second nephrectomy to remove the left native kidney was performed 7 days later. Overall surgical mortality was approximately 15%, and there were no significant differences in perioperative mortality between the experimental groups. Allografted and isografted mice were sacrificed at 2, 4 and 6 weeks posttransplantation for further studies. Eight animals were studied in each group at each time point. In this model, no immunosuppression was given at any time point as the grafts have prolonged
American Journal of Transplantation 2006; 6: 2292–2306
survival with about 20% survival of all allografted animals to the latest time point of 6 weeks.
Measurement of kidney transplant function Serum creatinine levels were measured at each time point using the cation exchange HPLC method as previously described (27). Whole blood (100 lL) was removed from the tail vein via a hematocrit tube and 5 lL of the serum was assayed.
Histology evaluation The kidney transplant was removed at the time of sacrifice and the capsule was removed. The kidney was fixed in Accustain (10% formalin solution, Sigma, St. Louis, MO, USA), embedded in paraffin, sectioned and stained with hematoxylin-eosin and Masson’s trichrome by standard methods. Graft sections were evaluated in a masked fashion by an experienced transplant pathologist (P.R.) using Banff 1997 criteria (3). Individual Banff subscores ranged from ‘0’ for none to ‘3’ for severe changes in the Banff categories of tubulitis, inflammation, vasculitis, arteriolar hyalinosis and glomerulitis for acute injury and were totaled for a maximum score of 15. Chronic changes were similarly assessed for tubular atrophy, fibrosis, transplant glomerulopathy, mesangial matrix and vascular intimal fibrosis with a total maximum score of 15.
Immunohistology Detection and localization of CTGF as well as alpha-smooth muscle actin (aSMA) was performed to detect myofibroblasts and smooth muscle cells. Another eight isografts and seven allografts were transplanted to obtain additional tissue for study. Five-lm tissue sections were deparaffinized in xylene and rehydrated in water. For CTGF staining, antigen retrieval was performed by boiling in high pH antigen retrieval solution (DakoCytomation, Carpinteria, CA, USA) for 20 min. The sections were incubated with blocking buffer (3% H 2 O 2 ) for 5 min at room temperature. Blocking buffer was removed from the slides, and the slides were then incubated with 1:1000 diluted CTGF antibody (Abcam Inc., Cambridge, MA) or 1:100 diluted a-SMA antibody (Clone 1A4, DakoCytomation) for 60 min at room temperature. The slides were washed with tris-buffered saline (TBS), and incubated with HRPlabeled goat anti-mouse immunoglobulin (Envision+, DakoCytomation) for 30 min at room temperature. After washing with TBS, the slides were then incubated with DAB (DakoCytomation) for 10 min at room temperature. The slides were then washed with TBS again, and counterstained before mounting and evaluation.
Real-time PCR analysis A portion of the kidney was snap frozen in liquid nitrogen and stored at –80◦ C until the extraction of total RNA using RNAwiz (Ambion, Austin, TX, USA). Each cDNA template for quantitative real-time PCR was prepared by first-strand reverse transcription using random primers (First-strand cDNA synthesis kit; Roche, Indianapolis, IN). TaqMan probe and primer sequences for CTGF were designed using Primer Express software (version 1; PE Applied Biosystems, Foster City, CA, USA). Synthesized by PE Applied Biosystems, the probes were labeled with 6-carboxy-fluorescein (FAM) at the 5 end, and 6-carboxy-teteramethylrhodamine (TAMRA) at the 3 end. The reverse primer was designed to span exons 2 and 3 for CTGF to avoid genomic DNA amplification. TaqMan PCR assays were performed in duplicate with 10-ng total RNA templates in 96-well optical plates on an ABI Prism 7700 Sequence Detection System according to the manufacturer’s recommendations (PE Applied Biosystems). Murine TaqMan gene expression assays for 18S rRNA, fibronectin, vimentin, S100A4, BMP7, F4/80 and collagen type I were purchased from PE Applied Biosystems. For each sample, an amplification plot was generated, displaying an increase in the reporter
2293
Cheng et al. dye fluorescence (Rn) with each PCR cycle. From it, a threshold cycle (C t ) value was calculated, which was the PCR cycle number at which fluorescence was detected above threshold, based on the variability of baseline data in the first 15 cycles. Delta-C t for each gene target was obtained by subtracting the C t value of 18S rRNA from that of the gene target. Finally, the level of mRNA expression relative to expression in normal tissue was obtained by the following equation:
Image Station 1000 (Kodak, New Haven, CT, USA). Quantitation of the protein bands was performed using the Kodak 1D imaging analysis software (Kodak). G3PDH antibody (Chemicon International, Temecula, CA, USA) was used on the same blot, after membrane stripping using Tris-HCl/bmercaptoethanol buffer, to normalize the quantitation according to protein loading.
Relative expression =
Immunofluorescence staining for markers of EMT and quantitation
2−(dCt of target in treatment group−dCt of target in normal or isograft kidney) .
Mouse CTGF ELISA Serum samples were obtained in heparinized capillary tubes (VWR International, West Chester, PA, USA) and urine was collected by metabolic cage (Hatteras Instruments, Cary, NC, USA). Determination of CTGF concentration was performed by the use of a noncommercially available, quantitative ‘sandwich’ enzyme immunoassay technique as published earlier (28,29). In this assay, affinity-purified goat polyclonal antibody (10 lg/mL) specific for human CTGF was coated on 96-well microtiter MaxiSorp plates (NuncTM , Rochester, NY) at 4◦ C overnight. The plates were blocked with SuperBlock® Blocking Buffer in TBS (Pierce, Rockford, IL) per manufacturer’s directions. Standards and diluted samples (urine 1:10; serum 1:1) were brought up to a final volume of 50 lL in 50 mM Tris (Fisher Sci., Fair Lawn, NJ, USA), pH 8.0, 0.1 % BSA (Sigma), 0.1% Triton X (Sigma), added to the microtiter plates and incubated for 2 h at room temperature. After washing, a biotinlinked affinity-purified goat polyclonal antibody (2 lg/mL) specific for human CTGF was added to the wells to bind to the CTGF molecules bound by the first antibody. After an incubation time of 1 h at room temperature and subsequent wash, alkaline phosphate-conjugated streptavidin (Zymed, San Francisco, CA, USA) diluted 1:500 was added and incubated for 30 min. Alkaline phosphate substrate, p-nitrophenyl phosphate (Pierce), was added to the wells until the color change developed in the dark (about 20 min). Spectrophotometer readings were obtained at 405 nm. The detection limit of the assay was 0.1 ng/mL CTGF.
Cell culture of renal tubular epithelial cells Primary mouse renal proximal tubular epithelial cells (CS3.7—a gift from Anthony Jevnikar, U. of Western Ontario; (30)) were cultured in a 1:1 mixture of Dulbecco’s Modified Eagle Medium and Ham’s F12 medium (DMEM/F12, Invitrogen, Carlsbad, CA, USA), with 2 mM glutamine and 25 mM HEPES, supplemented with 5 lg/mL insulin, 1.25 ng/mL prostaglandin E 1 , 30 pg/mL thyroid hormone, 5 lg/mL transferrin, 1.73 ng/mL sodium selenite, 18 ng/mL hydrocortisone, 25 pg/mL epidermal growth factor, with or without 10% FBS. For gene and protein expression studies, subconfluent cell lines were serum starved for 24–48 h prior to being subjected to TGF-b (10 ng/mL; Sigma) and/or recombinant human CTGF (10 ng/mL; Abcam, Cambridge, MA, USA) diluted in regular culture media with 5% FBS. Cell were subsequently recovered at 48 h and processed for analysis. All studies were performed in triplicate and the results represent three independent experiments for each gene and protein studied.
Cells were cultured on four-well chamber slides (Nalge Nunc, Napierville, IL, USA). After reaching subconfluency, cells were washed in DMEM/F12 twice and cultured in medium without serum overnight followed by exposure to rCTGF, TGF-b or both for 48 h. Cells were fixed and permeabilized as previously described (31). Cells were blocked with 10% normal donkey serum (Jackson ImmunoResearch Laboratories, Inc., West Grove, PA, USA) in PBS for 1 h at room temperature. Mouse anti-human E-cadherin antibody (BD Bioscience, Clone 36; San Jose, CA, USA) or rabbit anti-human S100A4 (DakoCytomation) diluted 1:100 in PBS with 10% normal donkey serum was added at room temperature for 1 h. Slides were washed three times with fresh PBS. AlexaFluor 488-conjugated donkey anti-mouse or anti-rabbit IgG (Molecular Probes Inc., Eugene, OR, USA) diluted 1:200 in PBS with 2% normal donkey serum was then applied for 2 h at room temperature. Slides were washed again, and nuclei were stained with DAPI (Molecular Probes Inc.). Slides were mounted with Mount Coverslipping Medium (Polysciences Inc, Warrington, PA, USA). Images were captured through Axioskop 2 plus microscope with Zeiss AxioCamHRc digital camera (Carl Zeiss, Inc., Thornwood, NY, USA) using the manufacture’s image software, AxioVison. Quantification of immunofluorescent-positive stained cells was performed using a computerized method of point counting, as previously described (32). Briefly, stained cells were photographed as above. A 150-point grid was overlaid onto the photomicrograph using Adobe Photoshop (Adobe, San Jose, CA, USA). The number of points falling onto positive cells was counted and expressed as a percentage of the grid. Positive cells were counted in 30 photomicrographs from three independent samples for each antibody staining.
Patient evaluation Twenty-nine consecutive patients enrolled in institutional review board approved kidney transplant protocols at the National Institutes of Health Clinical Center were studied. Immunosuppression included standard triple therapy (n = 11), or depletional therapy using Alemtuzumab (n = 9) or rabbit ATG induction (n = 9) with tacrolimus and/or rapamycin. Samples were collected at the time of protocol or clinically indicated kidney biopsies. Thirty-three biopsies from these individuals were evaluated by a pathologist masked to the clinical course of the recipient and evaluated using Banff 1997 criteria (3). Seven milliliter of sera and 50 mL of urine were obtained at the time of acquisition. Twenty-eight serum samples from 25 patients and 24 urines from 21 patients were centrifuged at 4◦ C at 3200 rpm in a Sorvall RT7 plus centrifuge (Thermo Electron, Waltham, MA, USA). Supernatant was removed, aliquoted and immediately frozen at –70◦ C until assay.
Western blot analysis Media was removed and cells lysed in RIPA lysis buffer (Santa Cruz Biotechnology, Santa Cruz, CA, USA) per the manufacturer’s guidelines. Samples were then mixed 1:1 with 2× Laemmli Sample buffer (Bio-Rad, Hercules, CA, USA) and run on a NuPage 4–12% Bis-Tris gel (Invitrogen). Proteins were subsequently transferred onto PVDF membrane (GE Healthcare BioSciences, Uppsala, Sweden). Following blocking, the membrane was incubated with S100A4 antibody (DakoCytomation) overnight, washed and subjected to the appropriate horseradish peroxidase-conjugated secondary antibody. For detection, membranes were incubated with Supersignal West Femto reagents (Pierce, Rockford, IL, USA) and developed using a Kodak
2294
Human CTGF ELISA Human CTGF levels in serum and urine samples were measured by sandwich ELISA using a proprietary assay by FibroGen Incorporated (FibroGen, South San Francisco, CA, USA) as previously described (33). Urine values were expressed relative to creatinine concentration in the sample and recipient urine CTGF contained only N-terminal CTGF as previously determined in urine from nontransplanted recipients (34). Serum (n = 8) and urine (n = 20) samples for normal values were obtained from control subjects that had no history of renal disease or hypertension and were supplied by FibroGen Inc.
American Journal of Transplantation 2006; 6: 2292–2306
CTGF and Chronic Allograft Nephropathy
American Journal of Transplantation 2006; 6: 2292–2306
2295
Cheng et al. CTGF mRNA expression in human biopsy tissue Biopsies (16-gauge) were obtained based on protocol surveillance criteria (35) or as clinically indicated. Normal kidney tissue (NK; n = 15) was obtained from living donor kidneys under direct visualization prior to organ procurement by open nephrectomy and was also used for calibration and direct comparison. Patients were assigned to the CAN group (CAN; n = 16) when the biopsy demonstrated tubular atrophy and interstitial fibrosis with at least grade 2 by Banff criteria. Biopsies with concomitant acute rejection were excluded. Transcriptional findings were compared to those derived from a group of biopsies from recipients with stable allograft function (SF; n = 12), defined as having a protocol biopsy at least 1 month posttransplantation without a change in serum creatinine (
