Mitogenic Signaling of Urokinase Receptor–Deficient Kidney Fibroblasts: Actions of an Alternative Urokinase Receptor and LDL Receptor–Related Protein

Home , Urokinase receptor

J Am Soc Nephrol 15: 2090–2102, 2004 Mitogenic Signaling of Urokinase Receptor–Deficient Kidney Fibroblasts: Actions of an Alternative Urokinase Receptor and LDL Receptor–Related Protein

GUOQIANG ZHANG, XIAOHE CAI, JESU´ SM.LO´ PEZ-GUISA, SARAH J. COLLINS, and ALLISON A. EDDY University of Washington, Children’s Hospital and Regional Medical Center, Division of Nephrology, Seattle, Washington

Abstract. The urokinase receptor (uPAR) attenuates myofibro- of functional uPA protein or LRP antisense RNA significantly blast recruitment and fibrosis in the kidney. This study exam- increased ERK signaling and cell mitosis in both genotypes. ined the role of uPAR and its co-receptor LDL receptor-related Enhanced uPARϪ/Ϫ fibroblast proliferation was reversed by a protein (LRP) in the regulation of kidney fibroblast prolifera- recombinant nonfunctional uPA peptide. The density of cell- tion and extracellular signal-regulated kinase (ERK) signaling. bound fluor-uPA was similar between uPARϪ/Ϫ and uPARϩ/ϩ Compared with uPARϩ/ϩ cells, uPARϪ/Ϫ kidney fibroblasts fibroblasts (78 Ϯ 6 versus 92 Ϯ 16 units). These data suggest that were hyperproliferative. UPARϪ/Ϫ fibroblast proliferation was uPAR-deficient kidney fibroblasts express lower levels of its 60% inhibited by an ERK kinase inhibitor. LRP protein was scavenger co-receptor LRP, resulting in greater extracellular ac- reduced and extracellular accumulation of urokinase-type plas- cumulation of uPA and PAI-1. Enhanced proliferation of minogen activator (uPA) and plasminogen activator inhibitor type uPARϪ/Ϫ fibroblasts seems to be mediated by uPA-dependent 1 (PAI-1) proteins were greater in uPARϪ/Ϫ cultures. Addition ERK signaling via an alternative urokinase receptor.

Progressive destruction of the kidney is a consequence of genesis. uPA acts primarily at extravascular sites, often local- excessive extracellular matrix accumulation (1). During renal izing to cell surfaces as a result of high-affinity binding to fibrogenesis, resident interstitial fibroblasts typically prolifer- uPAR, whereas tPA is typically active at intravascular sites of ate and undergo a phenotypic transformation associated with fibrin deposition and binds to annexin AII but not to uPAR. de novo expression of smooth muscle specific ␣-actin (␣- Along with uPA, the other known uPAR ligands are vitronectin SMA) (2). These myofibroblasts are the primary source of the and kininogen. interstitial matrix proteins that accumulate and ultimately de- In addition to enhancing pericellular uPA activity, although stroy nephron integrity and function (3). The molecular signals uPAR itself is a nonsignaling glycosylphosphatidylinositol- that regulate renal fibroblast activation and proliferation re- linked receptor, it serves multiple roles as a result of interac- main incompletely delineated. TGF-␤ and plasminogen acti- tions with cellular co-receptors that include ␣(v)␤3, ␣(v)␤5, ␤ vator inhibitor type 1 (PAI-1) are known to play important and 1 integrins (7). Another important binding partner is the roles (4,5). LDL receptor–related protein (LRP), a multifunctional scav- The urokinase receptor (uPAR) is a glycosylated 50- to enger and signaling receptor (8). PAI-1, the major uPA inhib- 60-kD protein that serves to concentrate serine protease activ- itor, interacts with uPAR-bound uPA, leading to internalization ity in pericellular regions (6). The serine proteases, urokinase- and degradation of PAI-1 and uPA via a process that requires type and tissue-type plasminogen activators (uPA, tPA), have LRP. In addition to PAI-1, several other LRP ligands have been implicated in diverse biologic processes, including cell been implicated in fibrogenesis, including protease inhibitors, proliferation and migration, tissue remodeling, wound healing, matrix proteins, and connective tissue growth factor. Both uPAR and LRP have been reported to be involved in morphogenesis, tumor development and metastasis, and angio- cellular mitogenic signaling pathways. Several studies have documented mitogenic effects of uPA, but the role of uPAR has often been inferred and not directly examined. Interactions Received November 18, 2003. Accepted May 11, 2004. between endogenous uPA, and uPAR and one of its co-recep- Correspondence to Dr. Allison A. Eddy, Children’s Hospital and Regional tors may stimulate basal levels of activated extracellular signal- Medical Center, Division of Nephrology, Mail Stop M1-5, 4800 Sand Point Way NE, Seattle, WA 98105. Phone: 206-987-2524; E-mail: allison.eddy@ regulated kinases (ERK), thereby regulating cellular behavior seattlechildrens.org (9,10). Through interactions with adapter and scaffold proteins, 1046-6673/1508-2090 LRP may modify mitogen-activated protein kinase (MAPK)/ Journal of the American Society of Nephrology ERK signaling initiated by other receptor(s), thereby regulating Copyright © 2004 by the American Society of Nephrology cell proliferation and differentiation (8). DOI: 10.1097/01.ASN.0000135057.41526.2C LRP and uPAR are often coordinately expressed (11). In our J Am Soc Nephrol 15: 2090–2102, 2004 Mitosis via an Alternative Fibroblast uPA Receptor 2091

recent studies, both receptors were expressed by renal intersti- Antibodies and cDNA Probes tial cells during fibrogenesis, and uPAR expression had impor- Primary antibodies that were used for immunochemical staining, tant fibrosis-attenuating effects (12). An important observation Western blotting, and selected antibody neutralization studies were was the presence of significantly fewer interstitial myofibro- FITC-conjugated anti-pan cytokeratin antibody (Sigma Chemical, St. blasts in the kidneys of wild-type mice compared with uPAR- Louis, MO); goat anti-mouse uPAR1 neutralizing antibody (R&D Systems, Minneapolis, MN); horseradish peroxidase (HRP)-conju- deficient mice. Given the presumed promitotic effects of uPAR gated mouse monoclonal antibodies for cow vimentin, human and evidence that resident interstitial fibroblasts proliferate and ␣-SMA, and proliferating cell nuclear antigen (PCNA) (Dako, differentiate into myofibroblasts during the early phase of Carpinteria, CA); goat anti–type I collagen antiserum (Southern Bio- chronic kidney damage (13), it seemed paradoxical that uPAR technology Ass., Birmingham, AL); rabbit anti-LRP antiserum from deficiency was associated with a greater number of myofibro- Dr. D.K. Strickland (American Red Cross, Rockville, MD) (16); blasts and worse fibrosis. Therefore, the present study was mouse anti-murine PAI-1 neutralizing monoclonal antibody designed to determine whether uPAR deficiency alters kidney (MA-33H1F7) from Dr. P. Declerck (Katholieke University, Leuven, Belgium) (17); and rabbit anti-mouse uPA polyclonal antibody fibroblast proliferation using a series of in vitro experiments (American Diagnostica, Greenwich, CT). Secondary antibodies were with primary cultures of kidney fibroblasts isolated from FITC-conjugated goat anti-rabbit IgG (Organon Teknika; West Ches- C57BL/6 wild-type and uPAR-deficient mice. The results ter, PA), rabbit anti-goat IgG (Southern Biotechnology Ass.), HRP- identify three important mechanisms: (1) increased expression conjugated goat anti-rabbit IgG (Chemicon International), and goat of LRP receptors by uPARϩ/ϩ fibroblasts leads to enhanced anti-mouse IgG (Sigma Chemical). cDNA probes used were rat PAI-1 clearance of uPA from the extracellular environment, (2) uPA from Dr. T.D. Gelehrter (University of Michigan, Ann Arbor, MI) is a fibroblast mitogen that seems to bind to an alternative (18), rat uPA from Dr. J. Degen (Children’s Hospital Research Foun- dation, University of Cincinnati, Cincinnati, OH) (19), human LRP receptor on uPARϪ/Ϫ fibroblasts, and (3) the ERK signaling from Dr. D.K. Strickland (American Red Cross) (20), and human pathway is involved in the mitogenic response of the alterna- glyceraldehyde-3-phosphate dehydrogenase (GAPDH; BD Bio- tive uPA receptor. sciences Clontech, Palo Alto, CA).

Materials and Methods Cellular Proliferation and ERK Activation Fibroblast mitotic activity was evaluated under several different Cell Cultures culture conditions using the CellTiter 96 Aqueous One Solution Cell Primary renal cortical fibroblast cultures were established as de- Proliferation Kit (MTS assay; Promega, Madison, WI) according to Ϫ Ϫ scribed previously (14) using uPAR-deficient (uPAR / ) and wild- the manufacturer’s instructions. The MTS Assay is a colorimetric type (uPARϩ/ϩ) mice on an identical C57Bl/6 background, provided method for determining the total number of viable cells. Our prelim- by Dr. Peter Carmeliet (Leuven, Belgium) (15). In brief, the renal inary study confirmed a significant correlation between counted cell cortex was separated from the medulla, minced, and differentially numbers and OD values at 490 nm (r ϭ 0.982, R2 ϭ 0.965). Each sieved to isolate tubular-rich interstitial fragments devoid of glomer- experiment was typically performed with n ϭ 12 separate wells of uli. The tubulointerstitial fragments were cultured at 37°C with 5% fibroblasts in 96-well plates under identical conditions. Nuclear mi-

CO2 in DMEM/F12 (1:1) supplemented with 5% FCS (vol/vol; pH totic activity was assessed as levels of PCNA protein measured by 7.35 to 7.45), hereafter referred to as full medium (FM), until a Western blotting. Culture medium was decanted, and the cells were fibroblast monolayer reached confluence. Early passage uPARϩ/ϩ treated with lysate buffer. Cell lysates and extracellular matrix were and uPARϪ/Ϫ fibroblasts were frozen in multiple aliquots, and all collected by scraping and sonication (21). Protein samples (20 ␮g) experiments were performed with cells at three to five passages. were separated by 7.5 to 12% SDS-PAGE and transferred to a Fibroblast identity was based on a typical fibroblastic spindle-like polyvinylidene difluoride membrane, and the immunoreactive protein shape; pan-cytokeratin negativity; and expression of ␣-SMA, vimen- was visualized using ECL enhanced chemiluminescence (reagents tin, and collagen. All initial descriptive studies were initiated at ~20% from Amersham Pharmacia Biotech, Piscataway, NJ). Amidoblack confluence and continued for 48 h, when cells were ~90% confluent. staining of the blots was used to determine loading equality. In our ␮ To avoid possible confounding effects of serum, we conducted inter- preliminary studies, protein loads between 5 and 30 g were within a linear range using amidoblack staining (R2 ϭ 0.9754). Band densities vention studies in serum-free media (SFM) or 1% FCS as established were quantified using the NIH Image Analysis Program, and data are by preliminary experiments. expressed in arbitrary densitometric units. The uPAR genotype was confirmed by reverse transcription–PCR ERK activation was measured by Western blotting as levels of (RT-PCR) using the SUPERSCRIPT First-Strand Synthesis System phosphorylated ERK (p-ERK) and phosphorylated retinoblastoma (Life Technologies BRL Life Technologies, Rockville, MD) and (Rb) protein using mouse anti–phospho-p44/42 MAPK (Thr202/ uPAR gene-specific primers (GSP) 5'-CCCTGGGTCCAGCCAG- Tyr204) E10 monoclonal antibody, rabbit anti-total p44/42 MAPK GGC-3' (GSP), 5'-GGGCCTGCAGTGCATGCAGTGTGAG-3' (1F), polyclonal antibody, and rabbit anti–phospho-retinoblastoma protein and 5'-GGTCCAGAGGAGGACGCCCC-3' (2R) to yield a predicted (p-Rb, Ser795; Cell Signaling Technology, Beverly, MA). Expression 912-bp product. Protein expression was confirmed by immunostaining of cytoskeletal protein ␣-SMA was also evaluated by Western blot- fibroblasts grown on coverslips. Slides that were stained with second- ting. Fibroblast mRNA levels for different cyclin-dependent kinases, ary antibodies only were run in parallel as negative controls. Slides including cdk1, cdk2, cdk4, cdk6, and cdk9, were determined using a were examined with a LEICA DMLB epifluorescence microscope and nonradioactive RiboQuant Multi-Probe RNase Protection Assay Sys- photographed using a computerized SPOT camera (Diagnostic Instru- tem (Pharmingen, San Diego, CA) according to the manufacturer’s ments, Sterling Heights, MI). instructions. 2092 Journal of the American Society of Nephrology J Am Soc Nephrol 15: 2090–2102, 2004

Phenotypic Comparison of uPARϩ/ϩ and uPARϪ/Ϫ Kidney Fibroblasts LRP mRNA levels were measured by RT-PCR using primer sequences 5'-GTATCTCAAAGGGCTGGCGGTG-3' (1F) and 5'-TG- CACCCAGCATACGGTCTC-3' (2R) with a predicted PCR product size of 619 bp (22). Primers for GAPDH were included in the PCR reaction as an internal control for RNA quantity and integrity (PCR product 339 bp). Protein levels were measured by Western blotting, and expression was confirmed in situ by immunostaining. Confocal microscopy was performed with fibroblasts in DMEM/ F12 supplemented with 1% FCS and 5.8 ϫ 10Ϫ9 M mouse recombinant PAI-1 (American Diagnostica). After a 5-min incubation at 37°C, the cells were fixed in 0.5% paraformaldehyde, permeabilized with 0.05% Triton X-100, and stained with anti-LRP antibody followed by goat anti-rabbit IgG-FITC. Cells were examined with an ACAS Ul- tima Laser Cytometer (Meridian Instruments, Okemos, MI). Fluores- cence excitation was at 488 nm, and FITC emission was detected with a 530/20 band pass filter. A ϫ60 oil immersion objective was used for scanning cells with a step (pixel) size of 0.2 ␮m in the x-y plane and a step size of 0.4 ␮m for serial optical sections in the z axis. Photographs were taken using a computerized SPOT camera. Levels of uPA and PAI-1 protein were examined by Western blotting, and steady-state mRNA levels were examined by Northern blotting as we previously described (12,23). For the Northern blots, total cellular RNA (15 ␮g/lane) was separated by a 1% agarose formaldehyde gel electrophoresis, and the results of GAPDH probing were obtained to evaluate RNA loading equality. The density of each hybridization band was quantified using the NIH Image program.

ERK kinase, LRP, and uPA Inhibition Studies The MEK1/2 (ERK1/2 kinases) inhibitor UO126 (1,4-diamino-2,3- dicyano-1,4-bis[2-aminophenylthio]butadiene; 20 ␮M; Cell Signaling Technology) was added to 48-h cultures to suppress ERK phosphorylation. Trypan blue stain (0.4%; GIBCO BRL Life Technologies, Grand Island, NY) indicated no cellular toxic effects when the cells were cultured in the presence of 20 ␮M UO126 for up to 72 h. An antisense RNA strategy was used to evaluate the role of LRP in fibroblast proliferation. Antisense oligonucleotide transfection experiments were performed using previously published sequences (24). Preliminary experiments determined that sense or antisense ODN (25 ␮M) in DMEM/F12 plus 1% FCS premixed with lipofectin (5 ␮g/ml) for 30 min before transfection were effective without toxicity up to 96 h. LRP expression was measured by RT-PCR to determine transfection efficiency. To determine the role of uPA as a mitogen for uPARϪ/Ϫ kidney fibroblasts, we performed studies using the full-length nonfunctional Figure 1. Urokinase receptor (uPAR) and LDL receptor-related pro- recombinant mouse uPA peptide (from American Diagnostica; cat. tein (LRP) expression. Fibroblasts were cultured at 37°C for 48 h in No. 119PC) (25). UPARϩ/ϩ and uPARϪ/Ϫ fibroblasts were cul- DMEM/F12 supplemented with 5% FCS (full medium [FM]). uPAR tured in SFM with or without uPA119 (5 ϫ 10Ϫ9 M) for 72 h, because was expressed exclusively by uPARϩ/ϩ renal fibroblasts as demon- this dosage demonstrated maximal inhibitory effect in the growth of strated by reverse transcription–PCR (RT-PCR; A) and immunofluo- these cells in our preliminary pilot studies. rescence staining (B). Expression of LRP was downregulated on uPARϪ/Ϫ fibroblasts as shown by immunostaining (C), RT-PCR (D), and Western blot analysis (E; n ϭ 4, P Ͻ 0.05, uPARϪ/Ϫ versus uPA Stimulation Studies uPAR ϩ/ϩ). Protein loading was evaluated by amidoblack staining. Parallel cell culture studies were performed to determine whether a The figures are representative data from one of two independent full-length functional uPA peptide (ProuPA from Dr. Jack Henkin; experiments. The ability of plasminogen activator inhibitor type 1 Abbott Laboratories, Abbott Park, IL) (26) in a concentration range (PAI-1) to induce LRP clustering on kidney fibroblasts was examined from 10Ϫ10 to 10Ϫ7M and the uPA amino terminal fragments (ATF; by confocal laser scanning microscopy. Within 5 min of addition of from Dr. Graham Parry; Attenuon Co., San Diego, CA) (25) in a PAI-1 (5.8 ϫ 10Ϫ9 M), LRP capping was observed on uPARϩ/ϩ concentration range from 10 to 1000 nM could modulate fibroblast cells but not on uPARϪ/Ϫ cells (F). Magnifications: ϫ400 in B and proliferation rates. Experiments were performed in SFM. C, ϫ600 in F. J Am Soc Nephrol 15: 2090–2102, 2004 Mitosis via an Alternative Fibroblast uPA Receptor 2093

Figure 2. uPAR and cell proliferation. UPAR-deficient kidney fibroblasts in FM (A through C) or wild-type fibroblasts incubated with uPAR neutralizing antibody in complete medium containing 1% FCS (D through F) demonstrated significantly higher proliferation rates than wild-type fibroblasts in FM or wild-type fibroblasts treated with control nonimmune IgG in complete medium containing 1% FCS as evaluated by the MTS assay (A and D). Re- sults are expressed as mean cell number Ϯ 1 SD OD units (n ϭ 12 per group). Western blot analysis for proliferating cell nuclear antigen (PCNA) protein at 48 h confirmed the differences in cell-cycle entry rates between uPARϩ/ϩ and Ϫ/Ϫ groups (n ϭ 5; B and C) or between the anti-uPAR antibody and control IgG-treated uPARϩ/ϩ cells (n ϭ 7; E and F). Results are mean Ϯ 1 SD densitometric units. *P Ͻ 0.05.

uPA Binding Study uPARϩ/ϩ and uPARϪ/Ϫ mice (n ϭ 5 each) with obstructive nephrop- The recombinant mouse uPA was labeled with Alexa Fluor 546 athy of 7 d duration (12,23). Cytoplasmic ␣-SMA and nuclear brdU were protein labeling kit (Molecular Probes) per the manufacturer’s instruc- labeled using the HRP substrate DABϩ nickel (dark blown/black color) tions. uPARϩ/ϩ and uPARϪ/Ϫ fibroblasts were incubated in bind- and the AKP substrate Fast red/Naphthol (red color), respectively. Pro- ing buffer (SFM containing 0.5% BSA) alone, binding buffer plus liferating fibroblasts were blindly counted in 10 consecutive cortical Alexa Fluor–labeled uPA119 (10Ϫ7 M), or binding buffer plus Alexa high-power fields (ϫ750) by a method previously reported (13). Fluor–labeled uPA and a 50-fold excess of unlabeled uPA119 at 37°C for 60 min (22). After washes, the cells were examined by comput- Statistical Analyses erized fluorescence microscopy and the luminance of the fluorescence All data are expressed as mean Ϯ 1 SD unless otherwise stated. was analyzed with the Image-Pro Plus image analysis software (Me- Results were analyzed by the t test or the Mann Whitney U test (binding dia Cybemetics, Silver Spring, MD). To exclude intracellular fluores- studies and histology data) using the SPSS or Excel software. P Ͻ 0.05 cence as a result of endocytosis of the Alexa Fluor–labeled uPA, we was considered statistically significant. All in vitro experiments were repeated binding studies at 4°C for 2 h and measured fluorescence dual, and data were not combined for statistics unless otherwise stated. intensity in situ by scanning the cells using a Typhoon 9410 Variable Mode Imager (Amersham Biosciences). Results uPAR-Deficient Kidney Fibroblasts Express Lower In Vivo Studies LRP Levels Staining for p-ERK and double staining for ␣-SMA and bromode- The uPAR genotype of the primary kidney fibroblast cul- oxyuridine (brdU) were performed on kidney paraffin sections of tures was verified by RT-PCR and immunostaining (Figure 1, 2094 Journal of the American Society of Nephrology J Am Soc Nephrol 15: 2090–2102, 2004

Figure 3. Effects of uPAR on extracellular signal-regulated kinase (ERK) mitotic signaling. Western blot analysis demonstrates significantly higher levels of phosphorylated ERK (p- ERK), total ERK, phospho-retinoblastoma (Rb) protein, and smooth muscle specific ␣-actin (␣-SMA) in 48-h FM cultures of kidney fibroblasts lacking uPAR compared with wild-type fibroblasts (A). Histograms are results of the densitometric analyses of Western blots for p-ERK and total ERK expressed as p-ERK-to- total ERK percent ratios (B), p-Rb protein, and ␣-SMA expressed as mean Ϯ 1 SD densitometric units (C and D; n ϭ 4 per group). Amidoblack staining was used to determine protein loading equality. *P Ͻ 0.05, uPARϪ/Ϫ versus ϩ/ϩ.

A and B). LRP, the important scavenger co-receptor of uPAR ern blot analysis, further documented the hyperproliferation of (8), was significantly downregulated in uPARϪ/Ϫ cells com- uPARϪ/Ϫ cells (3.9 Ϯ 0.6 versus 1 Ϯ 0.7 units, Ϫ/Ϫ versus pared with the uPARϩ/ϩ cells as detected by Western blotting ϩ/ϩ; P Ͻ 0.05 after 48-h culture; Figure 2, B and C). Addition (0.5 Ϯ 0.2 versus 1 Ϯ 0.2 units, Ϫ/Ϫ versus ϩ/ϩ; P Ͻ 0.05) of anti-uPAR neutralizing antibody to the culture medium (1% and RT-PCR (Figure 1, C through E). Confocal microscopy FCS) of the uPARϩ/ϩ fibroblasts increased proliferation rates Ϫ demonstrated that PAI-1 (5.8 ϫ 10 9 M) in 1% FCS medium compared with cells cultured with control nonimmune goat Ig stimulated LRP surface clustering on uPARϩ/ϩ cells but not (Figure 2, D through F). uPARϪ/Ϫ cells (Figure 1F), indicating the involvement of Given the importance of ERK signaling pathways in cellular uPAR in PAI-1/LRP interactions. proliferation (27,28), levels of ERK phosphorylation were evaluated. The basal levels of ERK phosphorylation, expressed UPAR-Deficient Fibroblasts Manifest a as p-ERK/total ERK after 48-h culture in FM, were threefold Hyperproliferative Phenotype and Increased ERK higher in the uPARϪ/Ϫ cells than in the ϩ/ϩ cells (Figure 3, A Activation and B). Phosphorylated retinoblastoma protein (pRb), the key During our initial studies, we observed that primary cultures cell-cycle regulator that promotes passage through the G1-S re- of uPARϪ/Ϫ fibroblasts in DMEM/F12 with 5% FCS (full striction site, was significantly increased (37-fold) in uPARϪ/Ϫ medium [FM]) grew quickly, reaching ~90% confluence by fibroblasts compared with the ϩ/ϩ cells after culture in FM for 48 h. Cell number increased to a greater extent in uPARϪ/Ϫ 48 h (Figure 3C). Evaluation of multiple cell-cycle regulatory than the ϩ/ϩ cells after 16, 40, and 48 h (all P Ͻ 0.05; Figure kinases using a multiprobe RNAase protection assay determined 2A). Cell cycle entry, measured as PCNA expression by West- that PITALRE (cdk9) was the predominant cyclin-dependent ki- J Am Soc Nephrol 15: 2090–2102, 2004 Mitosis via an Alternative Fibroblast uPA Receptor 2095

Figure 4. Effects of uPAR on kidney fibroblast uPA and PAI-1 expression. After 48 h in culture in FM, Northern blot analysis detected similar uPA mRNA levels between uPARϩ/ϩ and uPARϪ/Ϫ fibroblasts, but accumulated uPA protein levels measured by immunoblotting were greater in the uPARϪ/Ϫ cultures (A). Under the same conditions, PAI-1 protein levels were also significantly higher in the uPARϪ/Ϫ cultures despite significantly lower PAI-1 mRNA levels (B). Glycer- aldehyde-3-phosphate dehydrogenase mRNA bands and amidoblack-stained protein bands served to determine loading equality. Band densities are expressed as the mean Ϯ 1 SD arbitrary units. Gray bars are uPARϩ/ϩ, and white bars are uPARϪ/Ϫ fibroblasts. *P Ͻ 0.05. Immunohistochemical photomicrographs illustrates uPA staining (red reaction product) of uPARϪ/Ϫ fibroblasts (A) and PAI-1 staining (B) of uPARϩ/ϩ fibroblasts. Magnification, ϫ400.

nase mRNA expressed by the renal fibroblasts. Expression levels cells, despite similar mRNA levels (1.2 Ϯ 0.6 versus 1 Ϯ 0.4 were similar between uPARϩ/ϩ and Ϫ/Ϫ cells (137 Ϯ 5.1 units; P Ͼ 0.05; Figure 4A). Similarly, significantly more versus 134 Ϯ 10.4 units; P Ͼ 0.05). PAI-1 protein accumulated in the uPARϪ/Ϫ cells (5.3 Ϯ 1.5 The ␣-SMA cytoskeletal protein has been regarded as a versus 1 Ϯ 1.1 units; P Ͻ 0.05), despite lower PAI-1 mRNA marker of the “activated” or “transformed” state of renal fi- levels (0.41 Ϯ 0.07 versus 1 Ϯ 0.15 units; P Ͻ 0.05) when broblasts (myofibroblasts) and may be linked with ERK signal compared with the ϩ/ϩ cells (Figure 4B). transduction and cell proliferation (2). ␣-SMA protein levels were significantly higher in uPAR-deficient kidney fibroblasts ERK Inhibition Attenuates Proliferation of uPAR- (6.5 Ϯ 1.8 versus 1 Ϯ 0.2 units, Ϫ/Ϫ versus ϩ/ϩ after 48 h in Deficient Kidney Fibroblasts culture; Figure 3D). Because ERK phosphorylation levels were significantly el- evated in uPARϪ/Ϫ fibroblasts, the contribution of ERK ac- Excessive Extracellular Accumulation of uPA and PAI- tivation to the hyperproliferative phenotype of these cells was 1 Protein in uPAR-Deficient Fibroblast Cultures investigated using the ERK kinase inhibitor UO126. Prelimi- The primary uPAR ligand is uPA. PAI-1 binds to uPAR nary dose–response studies demonstrated complete inhibition indirectly via interactions with uPA. The entire uPA–uPAR– of fibroblast ERK activation with 20 ␮M UO126. After a 48-h PAI-1 complex undergoes endocytosis and degradation via a incubation, UO126 (20 ␮M in medium containing 1% FCS) process that requires the LRP receptor. When cultured in 5% completely suppressed the phosphorylation of ERK and was FCS medium for 48 h, the fibroblast uPA protein level (cellular associated with significantly lower PCNA protein levels (29 Ϯ ϩ extracellular protein) was twofold higher in uPARϪ/Ϫ cells 17 versus 70 Ϯ 7 units, UO126 versus control) and ␣-SMA (2.1 Ϯ 0.7 versus 1 Ϯ 0.6 units; P Ͻ 0.05) than in uPARϩ/ϩ protein levels (7 Ϯ 2 versus 20 Ϯ 11 units, UO126 versus 2096 Journal of the American Society of Nephrology J Am Soc Nephrol 15: 2090–2102, 2004

Figure 5. Effects of ERK inhibition on uPAR-null cells. uPAR- deficient renal fibroblasts were cultured in complete medium containing 1% FCS without (control) or with the ERK inhibitor UO126 (20 ␮M) for 48 h. Western blot analysis demonstrates that ERK inhibition significantly reduced PCNA and ␣-SMA protein levels (A). Band densities were quantified, and the data are expressed as mean Ϯ 1 SD, n ϭ 4/group (B). The inhibitory effect of UO126 on cell proliferation was further vali- dated in renal fibroblast cultures in the medium containing 1% FCS without (control) or with 20 ␮M UO126 for 16 and 48 h by the MTS assay (C). *P Ͻ 0.05, treated versus control.

control; Figure 5, A and B). Significantly fewer fibroblasts ulated fibroblast proliferation. Even when cultured without serum were observed in the UO126-treated groups of both uPARϪ/Ϫ (SFM) for 72 h, uPARϪ/Ϫ cells slowly proliferated, whereas and ϩ/ϩ cells after 16 and 48 h (Figure 5C). uPARϩ/ϩ cells remained quiescent (Figure 7, A and B). How- ever, when cells were co-cultured in SFM in the presence of LRP Deficiency Promotes Fibroblast Proliferation and uPA119 (5 ϫ 10Ϫ9 M), a full-length, nonfunctional recombinant ERK Phosphorylation mouse uPA (25), the low-level ERK phosphorylation and prolif- To evaluate whether the decreased expression of LRP on eration of uPARϪ/Ϫ cells were completely neutralized as mea- uPARϪ/Ϫ kidney fibroblasts played a functional role in fibro- sured by the MTS assay and Western blot analysis for PCNA and blast proliferation, we performed studies using an LRP anti- p-ERK (both P Ͻ 0.05, n ϭ 12 and 4, respectively; Figure 7, A sense oligonucleotide (ODN) (24). Antisense ODN transfec- through C). The uPARϩ/ϩ fibroblasts were quiescent in SFM; tion completely inhibited LRP expression in both uPARϩ/ϩ their PCNA expression level was not altered by co-culture in the and Ϫ/Ϫ cells as demonstrated by RT-PCR at 48 h (Figure presence of uPA119 (Figure 7, A and B). 6A). LRP inhibition significantly enhanced proliferation of Addition of functional recombinant ProuPA peptide (5 ϫ both uPARϩ/ϩ and Ϫ/Ϫ kidney fibroblasts compared with the 10Ϫ9 M) to the culture medium initiated early (initial 120 min) sense ODN-treated control groups of the same uPAR genotype ERK phosphorylation in uPARϩ/ϩ fibroblasts but not in the as measured by the MTS assay and by PCNA Western blot uPARϪ/Ϫ cells (Figure 7D). However, after 48 h, ProuPA analysis (Figure 6, B through D). Concurrently, p-ERK levels increased p-ERK levels as much as sevenfold in uPARϪ/Ϫ significantly increased by 1.4-and 1.6-fold in uPARϩ/ϩ and compared with nonstimulated fibroblasts (7.6 Ϯ 0.5 versus 1 Ϯ Ϫ/Ϫ cells, respectively, when compared with sense ODN- 0.3 in uPARϪ/Ϫ cells; 1.3 Ϯ 0.1 versus 1 Ϯ 0.1 in uPARϩ/ϩ treated groups (Figure 6, E and F). cells; both P Ͻ 0.05; Figure 7C). After a 48-h incubation, ProuPA stimulated PCNA expression 7.8- and 3.6-fold in uPA Stimulates Proliferation and ERK Activation in uPARϪ/Ϫ and ϩ/ϩ fibroblasts, respectively, compared with uPARϩ/ϩ and uPARϪ/Ϫ Cells unstimulated cells of the same genotype (Figure 7E). The Given that uPA protein levels were higher in cultures of mitogenic effect of ProuPA in renal fibroblasts was further uPARϪ/Ϫ kidney fibroblasts, we investigated whether uPA mod- confirmed by the changes in viable cell numbers by the MTS J Am Soc Nephrol 15: 2090–2102, 2004 Mitosis via an Alternative Fibroblast uPA Receptor 2097

Figure 6. LRP effects on cell proliferation. Kidney fibroblast LRP mRNA expression measured by RT-PCR is silenced after treatment with antisense oli- gonucleotides (AS; 25 ␮M) in media with 1% FCS for 48 h (A). LRP antisense augmented cell-cycle entry rates of both uPARϩ/ϩ and Ϫ/Ϫ cells when measured by PCNA protein expression by Western blot analysis at 48 h (B) and by the MTS proliferation assay performed at both 16 and 48 h (C and D; n ϭ 12 per group, *P Ͻ 0.05, antisense versus sense of the same uPAR genotype). The suppres- sion of fibroblast LRP was also associated with enhanced ERK signal activation. p-ERK1/2 West- ern blot analysis demonstrates significantly higher levels in antisense (AS) LRP-treated fibroblasts compared with sense (S) LRP- treated cells of the same uPAR genotype after culture for 48 h (E and F). Densitometric data of p-ERF1/2 bands are expressed as mean Ϯ 1 SD arbitrary units, n ϭ 3 each (F). *P Ͻ 0.05, AS versus S of same genotype.

cell proliferation assay (Figure 7F). The ATF of uPA is re- copy (37°C study) or by a Typhoon 9410 Imager (4°C study) ported to stimulate melanoma cell proliferation independent of using fluorochrome-labeled recombinant mouse uPA. The mean uPAR (25). The ATF fragment containing the growth factor– fluorescence luminance of kidney fibroblasts (after subtraction like domain and kringle domain induced a modest 1.3-fold of background fluorescence luminance of buffer-incubated increased proliferation of uPARϪ/Ϫ but not uPARϩ/ϩ fibro- control cells) was similar between uPARϩ/ϩ and Ϫ/Ϫ cells blasts at the dose of 5 ϫ 10Ϫ7 M (Figure 7G). in 37 or 4°C studies (30 Ϯ 13 versus 45 Ϯ 10 units and 92 Ϯ 16 versus 78 Ϯ 6 units, respectively; both P Ͼ 0.05). In the uPA Binds to Kidney Fibroblasts that Lack uPAR presence of a 50-fold excess of unlabeled uPA119, the Previous studies have suggested the existence of cellular uPA fluorescence luminance significantly decreased to 11 Ϯ 6 binding sites in addition to the high-affinity receptor uPAR, al- and 17 Ϯ 2 units in 37°C study and to 3 Ϯ 0.6 and 5 Ϯ 3 though this putative receptor(s) has not yet been identified units in 4°C study in the ϩ/ϩ and Ϫ/Ϫ cells, respectively (25,29,30). uPA binding was evaluated by fluorescence micros- (Figure 8). 2098 Journal of the American Society of Nephrology J Am Soc Nephrol 15: 2090–2102, 2004

Figure 7. The uPA-dependent effects on renal fibroblast proliferation. PCNA West- ern blotting at 72 h demonstrates active proliferation of the uPARϪ/Ϫ fibroblasts in serum-free medium (SFM) that is blocked by competitive inhibition with a nonfunctional mouse uPA peptide (uPA119; n ϭ 5 from two independent experiments; A). Results of MTS proliferation studies confirm that the enhanced proliferation of the uPARϪ/Ϫ fibroblasts (white bars) is reversed by uPA119 (n ϭ 12). *P Ͻ 0.05, uPA119 versus SFM of the same uPAR genotype; ϩP Ͻ 0.05, 72 h versus time 0 for fibroblasts of the same genotype (B). p-ERK Western blot analysis of protein isolated from 48-h fibroblast cultures demonstrates reduced levels in uPA119-treated cells and enhanced ERK phosphorylation when SFM was supplemented with prourokinase peptide (5 ϫ 10Ϫ9 M; C). Similar data were seen with uPARϩ/ϩ cells (not shown). p-ERK West- ern blotting demonstrates the very early activation of ERK within minutes of exposure of uPARϩ/ϩ fibroblasts to prouPA (initial 120 min; D). This early response was not detected with uPARϪ/Ϫ cells (not shown). Immunoblotting and MTS proliferation assays demonstrate significantly increased PCNA expression and higher viable cell numbers in both uPARϩ/ϩ and uPARϪ/Ϫ fibroblasts 48 h after prouPA stimulation. *P Ͻ 0.05, prouPA versus SFM of the same uPAR genotype; ϩP Ͻ 0.05, 48 h versus time 0 for fibroblasts of the same genotype (E and F). An MTS proliferation assay dose–response curve illustrates that the amino terminal fragment (ATF) of prouPA induced modest uPARϪ/Ϫ fibroblast proliferation at the dose of 500 nM, n ϭ 4 for each dose (G).

ERK Phosphorylation and Fibroblast Proliferation in Discussion Damaged Kidneys Is Enhanced in uPAR-Deficient Kidney tubules are a major source of uPA; its activity level Mice increases after injury caused by obstruction (5,12). Through its To determine whether these in vitro observations occur in proteolytic effects on extracellular matrix, uPA is predicted to vivo, we examined kidneys from uPARϩ/ϩ and uPARϪ/Ϫ attenuate fibrosis severity. More recently, it has become evi- mice after7dofureteral obstruction. Numerous ␣-SMA– dent that uPA also has important uPAR-dependent cellular positive activated fibroblasts were present in the interstitium of activities that dampen the fibrogenic response (23). The both kidneys, although they were more numerous in present study provides two important new insights. First, uPAR uPARϪ/Ϫ mice (Figure 9, A and B). Double immunostaining serves to maintain renal fibroblasts in an “inactive” phenotype demonstrated significantly more proliferating ␣-SMA ϩ fibro- by promoting the clearance of uPA–PAI-1 ligands via the blasts in uPARϪ/Ϫ mice (fivefold increase) compared with the uPAR-LRP degradation pathway. The uPAR-deficient kidney uPARϩ/ϩ, as assessed by in situ incorporated brdU (Figure 9, fibroblasts also expressed lower levels of the LRP scavenger A through C). Western blot analysis and immunostaining also co-receptor, resulting in greater extracellular accumulation of showed higher levels of phosphorylated ERK in the tubuloint- uPA and PAI-1 and higher levels of activated ERK. Second, in erstitium of the uPAR-deficient mice (Figure 9, D through G). the absence of uPAR, kidney fibroblast proliferation is en- J Am Soc Nephrol 15: 2090–2102, 2004 Mitosis via an Alternative Fibroblast uPA Receptor 2099 hanced and seems to be mediated at least in part by uPA- dependent ERK signaling via a currently uncharacterized alternative urokinase receptor. Downregulation of LRP on uPAR-deficient fibroblasts seems to promote their hyperproliferative phenotype by at least two mechanisms: extracellular accumulation of mitogens and altered basal ERK activity. Receptor-bound uPA–PAI-1 com- plexes undergo endocytotic degradation with recycling of the receptors (uPAR and LRP) to the cell membrane (31). The failure of PAI-1 to initiate LRP membrane clustering in uPARϪ/Ϫ fibroblasts illustrates the dysfunctional PAI-1 scav- enging pathway in uPARϪ/Ϫ cells. This pathway, together with a newly identified cell surface 200-kD protein (p200), is also responsible for the clearance of uPA (32). Levels of uPA protein were higher in cultures of uPARϪ/Ϫ fibroblasts, where it seemed to function as a fibroblast mitogen. LRP is linked through its intracellular domain to a group of recently identified adapter and scaffold proteins that are key organizers and regulators of the actin cytoskeleton and integrin-mediated MAPK/ERK1/2 signaling (8). MAPK/ERK1/2 signal transduction has been reported to initiate diverse cellular events, including proliferation, cytoskeletal reorganization, epithelial-mesenchymal transition, and LRP expression (27,28,33). The present study suggests that the hyperproliferative phenotype of the uPAR-deficient fibroblasts was mediated by ERK activation because the increased mitotic activity was reversed by the ERK1/2 kinase inhibitor, UO126. Decreased LRP expression by uPAR null fibroblasts seems to be associated with their ERK activity, given that fibroblast LRP inhibition alone enhanced cell proliferation and increased basal p- Figure 8. Alexa Fluor–uPA binding to kidney fibroblasts. Immuno- ERK levels, whether uPAR was expressed or not. However, fluorescence photomicrographs illustrate uPA binding to kidney fi- ϩ ϩ unlike uPAR / fibroblasts stimulated with exogenous uPA, broblasts using Alexa Fluor–labeled uPA as a probe in a study ERK phosphorylation was delayed until at least 120 min in the performed at 37°C (A). Control fibroblasts cultured on coverslips in uPARϪ/Ϫ cells. The absence of this very initial phase of the FM fail to fluoresce after incubation in binding buffer alone. After a uPAR-mediated MAPK/ERK1/2 signaling suggests one plau- 60-min incubation with Alexa Fluor–uPA (1 ϫ 10Ϫ7 M) in binding sible explanation for the downregulation of LRP in the buffer, both uPARϩ/ϩ and uPARϪ/Ϫ fibroblasts demonstrate uPA uPARϪ/Ϫ cells (34–36). This time lag before ERK activation binding. Competition experiments with an excess of nonlabeled uPA Ϫ6 in the uPARϪ/Ϫ fibroblasts suggests that cross-talk with other (5 ϫ 10 M) for 60 min blocks Alexa Fluor–uPA binding. A binding intracellular signaling pathways precedes ERK activation after study repeated at 4°C demonstrates similar binding results when ligand binding to the alternative uPA receptor. cellular fluorescence scanning was examined in situ using a Typhoon 9410 Variable Mode Imager (B). *P Ͻ 0.05, Alexa Fluor–labeled In addition to high-affinity binding of the growth factor–like uPA versus binding buffer alone or Alexa Fluor–uPA with excessive uPA domain to uPAR, previous studies reported specific uPA uPA119, n ϭ 4 per group. Magnification, ϫ400 in A. binding to an alternative, still unidentified low-affinity receptor (25,29,30,37). Whereas the mitogenic effects of uPA are widely known (25), the role of uPAR has not been as clearly ney fibroblasts is mediated via interaction with an alternative delineated. In fact, some previous uPAR blocking studies using uPA receptor(s). Surface binding of uPA to uPAR null fibro- a specific monoclonal antibody or uPAR antisense reported no blasts was confirmed by fluorescence microscopy using Alexa effect on uPA-induced proliferation, consistent with our cur- Fluor-labeled uPA. Furthermore, p-ERK levels and prolifera- rent findings that uPA-induced kidney fibroblast proliferation tive activity were induced in uPARϪ/Ϫ fibroblasts co-cultured is independent of uPAR expression (25,30,34). In fact, in the with recombinant functional uPA peptide, indicating intracel- present study, the opposite response was observed, with either lular signaling activity initiated by uPA bound to this alterna- genetic uPAR deficiency or pretreatment with a uPAR-block- tive receptor. ing antibody associated with higher rates of uPA-induced pro- The identity of this alternative uPA mitogenic receptor is not liferation than with control wild-type fibroblasts. The aug- yet known. The ATF of uPA expresses a domain that bound to mented mitotic activity of the uPARϪ/Ϫ cells was completely a unique receptor in studies by Koopman et al. (25). In the reversed by the nonfunctional uPA119 blocking peptide, sug- present study, the ATF peptide induced modest proliferation of gesting that the hyperproliferative state of uPAR-deficient kid- uPARϪ/Ϫ fibroblasts but nowhere near the magnitude of the 2100 Journal of the American Society of Nephrology J Am Soc Nephrol 15: 2090–2102, 2004

Figure 9. In vivo kidney fibroblast proliferation studies. Pro- liferating fibroblasts (A through C): Two hours before the ani- mals were killed, bromode- oxyuridine (brdU) was admin- istered intraperitoneally to mice with unilateral ureteral obstruction of 7 d duration. Dual immunohistochemical staining for ␣-SMA (dark brown cytoplasmic stain) and brdU (red nuclear stain) identified doubly stained fibroblasts undergoing mitosis in uPARϩ/ϩ kidneys (A) and uPARϪ/Ϫ kidneys (B). The white arrows highlight ex- amples of dual immunostaining of interstitial cells. (C) Histo- gram illustrates numbers of proliferating fibroblasts quantified from the double-staining studies. Gray bar represents uPARϩ/ϩ, white bar represents uPARϪ/Ϫ.*P Ͻ 0.05, Ϫ/Ϫ versus ϩ/ϩ, n ϭ 5 each. ERK signal activation (D through G): p-ERK immuno- blot analysis is stronger in uPARϪ/Ϫ kidneys compared with uPARϩ/ϩ kidneys 7 d after ureteral obstruction (D). Gray bar represents uPARϩ/ϩ kidneys, white bar represents uPARϪ/Ϫ kidneys; *P Ͻ 0.05, Ϫ/Ϫ versus ϩ/ϩ, n ϭ 5 each (E). Immunostaining demonstrates that ERK signaling is activated primarily in interstitial cells and is more extensive in uPARϪ/Ϫ kidneys (G) compared with uPARϩ/ϩ kidneys (F). Magnification, ϫ400.

response induced by the full single-chain uPA peptide (1.3- and uPAR-associated protein uPARAP (Endo 180), suggesting versus 7.8-fold increase). Perhaps preservation of the intramo- potential candidates for the alternative urokinase mitogenic lecular structure of uPA by internal disulfide bonds, known to receptor(s) (40). Ongoing microarray studies in our laboratory be essential for its proteolytic activity, is necessary to maxi- also support the presence of an alternative uPA-binding site on mize its mitogenic function as well (38,39). A recent phage kidney fibroblasts. After stimulation of uPAR-deficient fibro- display study identified a putative uPA kringle domain-binding blasts with pro-uPA for 48 h, 133 genes with an annotated consensus sequence that is expressed by at least 100 proteins, biologic function were significantly upregulated compared including 12 cell surface receptors, some of which are known with nonstimulated cells, including five genes involved in cell uPAR co-receptors: LRP, integrin ␣(v), membrane glycopro- proliferation and 24 involved in signal transduction (unpub- tein 130 (IL-6 receptor), insulin-like growth factor II receptor, lished data). What is also becoming apparent from these stud- J Am Soc Nephrol 15: 2090–2102, 2004 Mitosis via an Alternative Fibroblast uPA Receptor 2101 ies is that the mitotic response represents only one of several 5. Oda T, Jung YO, Kim H, Cai X, Lopez-Guisa J, Ikeda Y, Eddy biologic processes that are activated when uPAR fibroblasts AA: PAI-1 deficiency attenuates the fibrogenic response to ure- are stimulated with uPA, and many of these have been impli- teral obstruction. Kidney Int 30: 587–596, 2001 cated directly in fibrogenesis. 6. Mondino A, Resnati M, Blasi F: Structure and function of the The de novo expression of ␣-SMA by kidney fibroblasts not urokinase receptor. Thromb Haemost 82[Suppl 1]: 19–22, 1999 only is a marker of cellular activation and transition to a 7. Sitrin RG, Pan PM, Harper HA, Todd RF 3rd, Harsh DM, Blackwood RA: Clustering of urokinase receptors (uPAR; myofibroblastic phenotype, but also may serve as an alterna- CD87) induces proinflammatory signaling in human polymor- tive signal transduction pathway (2). This phenotypic change phonuclear neutrophils. J Immunol 165: 3341–3349, 2000 often occurs in association with fibroblast proliferation and 8. Herz J, Strickland DK: LRP: A multifunctional scavenger and synthesis of extracellular matrix proteins. It therefore is note- signaling receptor. J Clin Invest 108: 779–784, 2001 worthy that genetic uPAR deficiency was associated with 9. Webb DJ, Nguyen DH, Gonias SL: Extracellular signal-regu- enhanced ␣-SMA expression and filamentous actin disassem- lated kinase functions in the urokinase receptor-dependent path- bly (data not shown), illustrating a functional connection be- way by which neutralization of low density lipoprotein receptor- tween the uPAR/uPA/PAI-1/LRP membrane complex and the related protein promotes fibrosarcoma cell migration and actin cytoskeleton. It has previously been reported that inhibi- matrigel invasion. J Cell Sci 113: 123–134, 2000 tion of the formation of uPAR–uPA–PAI-1 tripartite com- 10. Aguirre-Ghiso JA, Liu D, Mignatti A, Kovalski K, Ossowski L: plexes significantly alters cytoskeletal signal transduction (41). Urokinase receptor and fibronectin regulate the ERK(MAPK) to The origin of kidney interstitial myofibroblasts during fibro- p38(MAPK) activity ratios that determine carcinoma cell prolif- genesis is a topic of great interest given their pivotal role in eration or dormancy in vivo. Mol Biol Cell 12: 863–879, 2001 kidney parenchymal destruction by fibrosis. Evidence cur- 11. Teesalu T, Blasi F, Talarico D: Embryo implantation in mouse: Fetomaternal coordination in the pattern of expression of uPA, rently suggests at least four possible sources of these cells: uPAR, PAI-1 and alpha 2MR/LRP genes. Mech Dev 56: 103– activation and proliferation of resident interstitial fibroblasts, 116, 1996 migrated perivascular cells, circulating mesenchymal cells, and 12. Zhang G, Kim H, Cai X, Lopez-Guisa J, Alpers C, Liu Y, transdifferentiated tubular epithelial cells. In vivo studies have Carmeliet P, Eddy A: Urokinase receptor deficiency accelerates documented that in situ proliferation of interstitial fibroblasts fibrosis in obstructive nephropathy. J Am Soc Nephrol 14: 1254– does occur after injury (13). Data from the present study 1271, 2003 suggest that enhanced mitotic activity accounts at least in part 13. Taneda S, Hudkins KL, Topouzis S, Gilbertson DG, Ophascha- for greater number of interstitial myofibroblasts and more roensuk V, Truong L, Johnson RJ, Alpers CE: Obstructive urop- severe fibrosis observed in mice that lack uPAR. The uPAR- athy in mice and humans: Potential role for PDGF-D in the deficient kidney fibroblasts express lower levels of their scav- progression of tubulointerstitial injury. J Am Soc Nephrol 14: enger co-receptor LRP, resulting in greater extracellular accu- 2544–2555, 2003 mulation of uPA and PAI-1. Enhanced proliferation of 14. Alvarez RJ, Sun MJ, Haverty TP, Iozzo RV, Myers JC, Neilson uPARϪ/Ϫ fibroblasts seems to be mediated by uPA-dependent EG: Biosynthetic and proliferative characteristics of tubulointer- ERK signaling via a currently uncharacterized novel urokinase stitial fibroblasts probed with paracrine cytokines. Kidney Int 41: receptor that is predicted to promote fibrosis. 14–23, 1992 15. Dewerchin M, Van Nuffelen A, Wallays G, Bouché A, Moons L, Carmeliet P, Mulligan RC, Collen D: Generation and character- Acknowledgments ization of urokinase receptor-deficient mice. J Clin Invest 97: 870–878, 1996 This work was supported by National Institutes of Health Grants 16. Ulery PG, Beers J, Mikhailenko I, Tanzi RE, Rebeck GW, DK54500 and DK44757 (A.A.E.). Part of this work was published as an abstract in J Am Soc Nephrol 14: 20A, 2003. Hyman BT, Strickland DK: Modulation of beta-amyloid precur- Our sincere thanks to Dr. Peter Carmeliet for providing breeding sor protein processing by the low density lipoprotein receptor- pairs of uPAR-deficient and wild-type mice, to Jack Henkin for related protein (LRP). Evidence that LRP contributes to the ProuPA, and to Dr. Graham Parry for the purified ATF peptide. pathogenesis of Alzheimer’s disease. J Biol Chem 275: 7410– 7415, 2000 17. Bijnens AP, Gils A, Knockaert I, Stassen JM, Declerck PJ: References Importance of the hinge region between alpha-helix F and the 1. Eddy AA: Molecular basis of renal fibrosis. Pediatr Nephrol 15: main part of serpins, based upon identification of the epitope of 290–301, 2000 plasminogen activator inhibitor type 1 neutralizing antibodies. 2. Gabbiani G: The biology of the myofibroblast. Kidney Int 41: J Biol Chem 275: 6375–6380, 2000 530–532, 1992 18. Zehab R, Gelehrter TD: Cloning and sequencing of cDNA for the 3. Chai Q, Krag S, Chai S, Ledet T, Wogensen L: Localisation and rat plasminogen activator inhibitor-1. Gene 73: 459–468, 1988 phenotypical characterisation of collagen-producing cells in 19. Rickles RJ, Darrow AL, Strickland S: Molecular cloning of TGF-␤1-induced renal interstitial fibrosis. Histochem Cell Biol complementary DNA to mouse tissue plasminogen activator 119: 267–280, 2003 mRNA and its expression during F9 teratocarcinoma cell differ- 4. Strutz F, Zeisberg M, Renziehausen A, Raschke B, Becker V, entiation. J Biol Chem 263: 1563–1569, 1988 van Kooten C, Muller G: TGF-beta 1 induces proliferation in 20. Strickland DK, Ashcom JD, Williams S, Burgess WH, Migliorini human renal fibroblasts via induction of basic fibroblast growth M, Argraves WS: Sequence identity between the alpha 2-mac- factor (FGF-2). Kidney Int 59: 579–592, 2001 roglobulin receptor and low density lipoprotein receptor-related 2102 Journal of the American Society of Nephrology J Am Soc Nephrol 15: 2090–2102, 2004

protein suggests that this molecule is a multifunctional receptor. and contribution to chemotaxis. J Biol Chem 275: 16450–16458, J Biol Chem 265: 17401–17404, 1990 2000 21. Seomun Y, Kim J, Lee EH, Joo CK: Overexpression of matrix 31. Czekay RP, Kuemmel TA, Orlando RA, Farquhar MG: Direct metalloproteinase-2 mediates phenotypic transformation of lens binding of occupied urokinase receptor (uPAR) to LDL receptor- epithelial cells. Biochem J 358: 41–48, 2001 related protein is required for endocytosis of uPAR and regula- 22. Xu W, Takahashi Y, Sakashita T, Iwasaki T, Hattori H, Yoshi- tion of cell surface urokinase activity. Mol Biol Cell 12: 1467– moto T: Low density lipoprotein receptor-related protein is re- 1479, 2001 quired for macrophage-mediated oxidation of low density li- 32. Poliakov A, Tkachuk V, Ovchinnikova T, Potapenko N, Bagry- poprotein by 12/15-lipoxygenase. J Biol Chem 276: 36454– antsev S, Stepanova V: Plasmin-dependent elimination of the 36459, 2001 growth-factor-like domain in urokinase causes its rapid cellular 23. Zhang G, Kim H, Cai X, Lopez-Guisa J, Carmeliet P, Eddy A: uptake and degradation. Biochem J 355: 639–645, 2001 Urokinase receptor modulates cellular and angiogenic responses 33. Montesano R, Soriano JV, Hosseini G, Pepper MS, Schramek H: in obstructive uropathy. J Am Soc Nephrol 14: 1254–1271, 2003 Constitutively active mitogen-activated protein kinase kinase 24. Hussaini IM, Brown MD, Weaver AM, Carpenter J, Karns LR, MEK1 disrupts morphogenesis and induces an invasive pheno- Vandenberg SR, Gonias SL: Stable antisense RNA expression type in Madin-Darby canine kidney epithelial cells. Cell Growth neutralizes the activity of low-density lipoprotein receptor-re- Differ 10: 317–332, 1999 lated protein and promotes urokinase accumulation in the me- 34. Chintala SK, Mohanam S, Go Y, Venkaiah B, Sawaya R, dium of an astrocytic tumor cell line. Antisense Nucleic Acid Gokaslan ZL, Rao JS: Altered in vitro spreading and cytoskeletal Drug Dev 9: 183–190, 1999 organization in human glioma cells by downregulation of uroki- 25. Koopman JL, Slomp J, de Bart AC, Quax PH, Verheijen JH: nase receptor. Mol Carcinog 20: 355–365, 1997 Mitogenic effects of urokinase on melanoma cells are indepen- 35. Degryse B, Sier CF, Resnati M, Conese M, Blasi F: PAI-1 dent of high affinity binding to the urokinase receptor. J Biol inhibits urokinase-induced chemotaxis by internalizing the uroki- Chem 273: 33267–33272, 1998 nase receptor. FEBS Lett 505: 249–254, 2001 26. Stefansson S, Kounnas MZ, Henkin J, Mallampalli RK, Chappell 36. Jo I, Im HM, Shin HJ, Won Cho K, Jung M, Kim SD, Kim Jeong DA, Strickland DK, Argraves WS: gp330 on type II pneumo- J, Ahn Jo S: Serum deprivation increases the expression of low cytes mediates endocytosis leading to degradation of pro-uroki- density lipoprotein receptor-related protein in primary cultured nase, plasminogen activator inhibitor-1 and urokinase-plasmino- rat astrocytes. Biochem Biophys Res Commun 299: 102–108, gen activator inhibitor-1 complex. J Cell Sci 108: 2361–2368, 2002 1995 37. Stepanova VV, Tkachuk VA: Urokinase as a multidomain pro- 27. Webb DJ, Thomas KS, Gonias SL: Plasminogen activator inhib- tein and polyfunctional cell regulator. Biochemistry (Mosc) 67: itor 1 functions as a urokinase response modifier at the level of 109–118, 2002 cell signaling and thereby promotes MCF-7 cell growth. J Cell 38. Kobayashi H, Gotoh J, Hirashima Y, Fujie M, Sugino D, Terao Biol 152: 741–752, 2001 T: Inhibitory effect of a conjugate between human urokinase and 28. Lukas J, Bartkova J, Bartek J: Convergence of mitogenic signal- urinary trypsin inhibitor on tumor cell invasion in vitro. J Biol ling cascades from diverse classes of receptors at the cyclin Chem 270: 8361–8366, 1995 D-cyclin-dependent kinase-pRb-controlled G1 checkpoint. Mol 39. Tanski WJ, Fegley AJ, Roztocil E, Davies MG: Domain-depen- Cell Biol 16: 6917–6925, 1996 dent action of urokinase on smooth muscle cell responses. J Vasc 29. Poliakov AA, Mukhina SA, Traktouev DO, Bibilashvily RS, Surg 39: 214–222, 2004 Gursky YG, Minashkin MM, Stepanova VV, Tkachuk VA: Che- 40. Liang OD, Chavakis T, Linder M, Bdeir K, Kuo A, Preissner KT: motactic effect of urokinase plasminogen activator: A major role Binding of urokinase plasminogen activator to gp130 via a for mechanisms independent of its proteolytic or growth factor putative urokinase-binding consensus sequence. Biol Chem 384: domains. J Recept Signal Transduct Res 19: 939–951, 1999 229–236, 2003 30. Mukhina S, Stepanova V, Traktouev D, Poliakov A, Beabealash- 41. Fincham VJ, James M, Frame MC, Winder SJ: Active ERK/ villy R, Gursky Y, Minashkin M, Shevelev A, Tkachuk V: The MAP kinase is targeted to newly forming cell-matrix adhesions chemotactic action of urokinase on smooth muscle cells is de- by integrin engagement and v-Src. EMBO J 19: 2911–2923, pendent on its kringle domain. Characterization of interactions 2000