Kidney International, Vol. 59 (2001), pp. 2146–2155

t-PA promotes glomerular plasmin generation and matrixdegradation in experimental glomerulonephritis

MASASHI HARAGUCHI, WAYNE A. BORDER, YUFENG HUANG, and NANCY A. NOBLE

Fibrosis Research Laboratory, Division of Nephrology, University of Utah, Salt Lake City, Utah, USA

t-PA promotes glomerular plasmin generation and matrix de- merular diseases [1, 2]. Matrix overproduction, increasedgradation in experimental glomerulonephritis. matrix deposition, and decreased degradation all appear

Background. In addition to its well-known role in degrading to contribute to matrix accumulation [3, 4]. Although ourfibrin, recent evidence suggests that plasmin degrades matrixunderstanding of the mechanisms underlying this processproteins and activates prometalloproteinases. Plasmin is gener-has dramatically increased in recent years, research hasated from plasminogen by tissue plasminogen activator (t-PA).

We hypothesized that t-PA treatment increases plasmin gener- emphasized what increases matrix production and howation in nephritic glomeruli and degrades pathological matrix we can stop it. As a result, a number of promising thera-leading to a therapeutic reduction in matrix accumulation. pies are on the horizon. However, even with a disease cure,Methods. Anti–Thy-1 nephritis was induced by injection of

normalization of matrix is likely to take many years.OX-7 antibody. Rats were given twice daily intravenous injec-Diabetic patients with proven glomerulosclerosis whotions of saline (disease control group) or human recombinant

t-PA (rt-PA; 1 mg/kg body weight) on days 3 through 5. Pro- received a successful pancreas transplant took 10 yearsteinuria, glomerular matrix protein staining, and glomerular to show significantly reduced matrix accumulation [5].mRNA levels for transforming growth factor-b1 (TGF-b1), When patients present with rapidly progressing fibrosis,fibronectin, and plasminogen activator inhibitor type 1 (PAI-1)

such as those with HIV-associated nephropathy, organwere evaluated at day 6. Localization of rt-PA, plasmin genera-failure may occur before the therapy takes effect. Intion by glomeruli in vitro, and glomerular production and con-

tent of active TGF-b1 were also investigated. these cases, it might be valuable to have therapies thatResults. Compared with disease control animals, proteinuria could quickly reduce accumulated matrix, just as many

and staining score for periodic acid-Schiff (2.75 6 0.17 vs. cancer therapies quickly reduce tumor load. Finally, the1.41 6 0.09), fibronectin-EDA1 (19 6 2 vs. 14 6 1), lamininprocess of fibrosis is closely related to normal wound(35 6 2 vs. 25 6 2), type I collagen (33 6 1 vs. 21 6 3), and

type IV collagen (27 6 2 vs. 23 6 1) were significantly reduced repair [3]. In terms of evolution, it may be that woundin treated rats (P , 0.01). Glomerular TGF-b1, fibronectin, repair was so important to survival that numerous back-and PAI-1 mRNA levels were unchanged. rt-PA colocalized up mechanisms are built in. If this is the case, the relatedwith fibrin along glomerular capillary walls and in the mesan-

process of fibrosis may be difficult to stop with one thera-gium. Nephritic glomeruli in vitro had decreased plasmin activ-peutic approach, and combinations of drugs may proveity, which was elevated by an in vivo presacrifice injection of

rt-PA. Glomerular production and content of active TGF-b1 extremely valuable. The present study takes an initialwere unchanged by the rt-PA injection. step to provide proof of the concept that enhancing ma-

Conclusions. These results show that injected rt-PA binds to trix degradation can be therapeutically efficacious.fibrin in nephritic glomeruli, thus increasing plasmin generationPlasmin and metalloproteinases (MMPs) are consid-and promoting pathological matrix degradation without acti-

ered to play key roles in matrix degradation [4, 6, 7].vating latent TGF-b. Agents that increase plasmin generation,such as t-PA, may have potential as antifibrotic therapies. Plasmin degrades matrix proteins fibronectin [8], laminin

[8], proteoglycan [9], and type IV collagen [10], as wellas fibrin. Plasmin also activates pro-MMP-1 (interstitial

Accumulation of extracellular matrix leading to glo- collagenase) [11] and pro-MMP-3 (stromelysin-1) [12].merulosclerosis is a central feature of progressive glo- MMP-3 then potently activates pro-MMP-9 (gelatinase B)

[13], which in turn degrades collagenous proteins. Plas-min is generated by the enzymatic activity of plasmino-Key words: tissue plasminogen activator, antifibrotic therapy, fibrin,

metalloproteinase, nephritic glomeruli, transforming growth factor-b. gen activators tissue-type (t-PA) and urokinase-type(u-PA). PA activity is inhibited by plasminogen activatorReceived for publication August 8, 2000inhibitors type 1 (PAI-1) and type 2 (PAI-2). Plasminand in revised form December 7, 2000

Accepted for publication January 11, 2001 generation is therefore the net result of the activity ofthe plasminogen activators and the PAIs. That matrix 2001 by the International Society of Nephrology

2146

Haraguchi et al: t-PA treatment of anti–Thy-1 nephritis 2147

degradation by cultured human mesangial cells depends logical matrix leading to a therapeutic reduction in ma-trix accumulation.on PAI-1 levels was clearly shown when a monoclonal

antibody to PAI-1 was added and matrix degradationincreased fourfold [14].

METHODSCultured human mesangial cells can produce t-PA [15]

Materialsand u-PA [16]. In normal human glomeruli, t-PA existsin the endothelium [17]. Tissue-PA binds to fibrin, and in Unless otherwise indicated, materials, chemicals, or

culture media were purchased from Sigma Chemical Co.doing so, its enzymatic activity is greatly increased [18].Tissue-PA and fibrin are found to be associated and (St. Louis, MO, USA).increased in many human glomerular diseases [19].

AnimalsOverexpression of transforming growth factor-b (TGF-b)is the most common molecular feature of fibrotic renal The studies were performed in male Sprague-Dawley

rats (200 to 250 g) obtained from Sasco (Omaha, NE,diseases [1, 20]. In addition to stimulating matrix synthe-sis and enhancing matrix-binding integrin expression, USA). Animal housing and care were in accordance with

the National Institutes of Health Guide for the Care andTGF-b inhibits matrix degradation [1, 3]. TGF-b1 mark-edly increases PAI-1 production and potently inhibits Use of Laboratory animals (NIH publication number

85-23, 1985).plasmin generation and matrix degradation by culturedhuman mesangial cells [21]. We have previously shown

Disease inductionthat PAI-1 synthesis dramatically increases and PA activ-ity decreases when TGF-b1 is added to normal glomeruli Glomerulonephritis was induced by tail vein injection

of the monoclonal anti–Thy-1.1 antibody OX-7 (2.5 mg/kgin culture [22]. TGF-b1 also increases production of tis-sue inhibitors of metalloproteinases (TIMPs) by cultured body weight) on day 0. The OX-7 monoclonal antibody

was produced in the ascitic fluid of adult female BALB/cmesangial cells [23, 24]. Isolated glomeruli from anti–Thy-1 nephritis rats show decreased PA [22] and MMP mice as described previously [34].activity [24, 25] in association with the overexpression

Time course of fibrinogen/fibrin staining inof TGF-b1. Finally, when neutralizing antibody to TGF-banti-Thy-1 nephritiswas administered to rats with Thy-1 nephritis, a dramatic

decrease in glomerular PAI-1 staining was observed [22]. Three rats were sacrificed as described later in thisarticle at each of eight time points from 0 to 28 daysAngiotensin II (Ang II) also participates in the control

of the PA-plasmin system by stimulating PAI-1 synthesis after OX-7 injection. Cortical tissue was stained for fi-brinogen/fibrin staining as described later in this article.directly and through the induction of TGF-b [26, 27].

The glomerular content of Ang II increases in anti–Thy-1rt-PA treatmentnephritis (abstract; Kanai et al, J Am Soc Nephrol 9:498A,

1998), and PAI-1 synthesis is decreased by Ang II block- A formulation of rt-PA suitable for injection into ratswas kindly provided by Genentech (South San Francisco,ade [27], a low-protein diet [28], and arginine supplemen-

tation [28] in this model. Recent studies have shown a CA, USA). rt-PA was administered intravenously by tailvein injection twice a day from day 3 to day 5 at a dosereduction in PAI-1 overexpression in response to Ang

II blockade in radiation-induced nephropathy [29] and of 1 mg/kg body weight (N 5 6). Control rats receivedan equal volume of saline (N 5 6).in aging [30].

Because increased Ang II, TGF-b, and PAI-1 are com-Sacrificemon features of fibrotic renal diseases, it is likely that

inhibition of normal matrix degradation contributes sig- Rats were sacrificed at day 6. Rats were anesthetizedwith ether. After blood was drawn from the lower ab-nificantly to the pathological process of matrix accumula-

tion [31–33]. It should be noted that currently used and dominal aorta, the kidney was perfused with 30 mL ofice-cold phosphate-buffered saline (PBS) and harvested.future therapies aimed at reducing Ang II and/or TGF-b

would be expected to increase plasmin generation and For histologic examination, cortical tissue was snap fro-zen and fixed in 10% neutral-buffered formalin. Glomer-matrix degradation through reductions in Ang II- and/or

TGF-b–induced increases in PAI-1 expression. The goal uli from individual rats were isolated by graded sieving(150, 125, 106, and 75 mm mesh metal sieves) as describedof the present study was to determine the therapeutic

efficacy of treatment with an agent that was not expected previously [35].to alter either Ang II or TGF-b levels, but only to en-

Light microscopyhance plasmin generation. We hypothesized that rt-PA,injected into nephritic rats, would bind to glomerular Three micrometer sections of paraffin-embedded tis-

sue were stained with periodic acid-Schiff (PAS) andfibrin leading to increased glomerular plasmin genera-tion. This plasmin would enhance degradation of patho- counterstained with hematoxylin. The total number of

Haraguchi et al: t-PA treatment of anti–Thy-1 nephritis2148

nuclei per glomerular cross section was calculated in 30 quent RNA extraction. For Northern analysis, RNA wasdenatured and fractionated by electrophoresis through arandomly selected glomeruli per sample.1.8% agarose gel (30 mg/lane) and transferred to a Hybond

Immunofluorescent staining for fibrinogen/fibrin nylon membrane (Amersham Corp., Little Chalfont, Buck-inghamshire, UK). Nucleic acids were immobilized byThree micrometer frozen tissue sections were fixed in

48C ethyl alcohol for three minutes. After washing in ultraviolet irradiation. Membranes were prehybridizedwith 50% formamide, 10% Denhardt’s solution, 0.1%PBS, FITC-conjugated rabbit anti-human fibrinogen/

fibrin antibody (Dako, Carpinteria, CA, USA) was incu- sodium dodecyl sulfate (SDS), 5 3 standard saline citrate(SSC), and 200 mg/mL denatured salmon sperm DNA,bated at a 1:100 dilution for two hours at room tempera-

ture. Intraglomerular deposition of fibrinogen/fibrin was and hybridized with DNA probes labeled with 32P-dCTPby random oligonucleotide priming (Boehringer Mann-semiquantitated by scoring 20 randomly selected glomer-

uli per section on a 0 to 4 scale as described previously heim Biochemicals, Indianapolis, IN, USA). The blotswere washed in 2 3 SSC, 0.1% SDS at room temperature[35, 36].for 15 minutes and in 0.1 3 SSC, 0.1 SDS at 558C for

Immunoperoxidase staining for matrix proteins 30 minutes. DNA probes used were as follows: (1) ratglyceraldehyde-3-phosphate dehydrogenase (GAPDH)For indirect immunoperoxidase staining, three micro-

meter sections of formalin-fixed, paraffin-embedded tis- cDNA (a gift from Drs. P. Kondaiah and M.B. Sporn,Dartmouth Medical School) [37], (2) TGF-b1 cDNAsue were incubated overnight at 48C with the following

antibodies: mouse anticellular fibronectin (fibronectin (kindly provided from Dr. D.L. Moses, Vanderbilt Can-cer Center) [38], (3) fibronectin-EDA cDNA (a generousEDA1, 1:100 dilution; Chemicon, Temecula, CA, USA),

rabbit anti-mouse laminin (1:100 dilution; Cappel, Au- gift from Dr. R.O. Hynes, Massachusetts Institute of Tech-nology) [39], and (4) PAI-1 cDNA (kindly provided byrola, OH, USA), rabbit anti-type I collagen (1:200 dilu-

tion; Rockland, Gilbertsville, PA, USA), and rabbit anti- Dr. T.D. Gelehrter, Michigan School of Medicine) [40].type IV collagen (1:200 dilution; Rockland). The Dako

Urinary protein excretionEnvisione System peroxidase (Dako) was used for asecondary antibody, and color was developed according Rats were housed in metabolic cages on day 5, and

24-hour urinary protein excretion was measured on urineto the manufacturer’s instructions.collected by the Bradford method (Bio-Rad Protein

Image analysis of matrix accumulation Assay; Bio-Rad Laboratories, Hercules, CA, USA).Quantitative measurement of glomerular area and

Double immunostaining for t-PA and fibrinogen/fibrinpositive staining area was performed on all experimentalrat tissues (N 5 6 per group). Three micrometer paraffin Two, day 3 nephritic rats were sacrificed 10 minutes after

rt-PA injection. The kidney was removed after perfusionsections of immunoperoxidase staining for fibronectinEDA1, laminin, type I collagen, and type IV collagen with PBS as described previously in this article. Three

micrometer frozen tissue sections were fixed in 48C ethylwere subjected to computer image analysis. For eachsection, 20 randomly selected glomeruli were evaluated, alcohol for three minutes. Staining was performed at

room temperature. First, rabbit anti-human recombinantand the mean index per section was calculated. The blackand white photographs of immunoperoxidase staining t-PA antibody (1:20 dilution; American Diagnostica Inc.,

Greenwich, CT, USA) was incubated with sections forwere scanned with a gray scale. The glomerular areaand positive staining area were processed with Adobe two hours, followed by TRITC-conjugated swine anti-

rabbit immunoglobulin (1:100 dilution; Dako) for onePhotoshop software. The glomerular area was selectedby tracing the border of Bowman’s capsule. The figure hour. Then after three washes in PBS, slides were incu-

bated with FITC-conjugated rabbit anti-human fibrino-obtained was changed to black by setting the outputlevel to 0. A binary figure of the positive staining area was gen/fibrin antibody (1:100 dilution; Dako) for two hours.gained by setting the threshold so that the background

Plasmin activity of isolated glomerulistaining disappeared. The areas obtained were measuredwith NIH image software, and then the percentage of Normal and day 3 nephritic rats were sacrificed 10 min-

utes after the injection of rt-PA or saline (N 5 4 forglomerular area with positive staining was calculated asa staining index (%). each group). Intact glomeruli from individual rats were

isolated as described previously in this article. The plas-RNA preparation and Northern hybridization min activity of isolated glomeruli was measured by chro-

mogenic substrate assay. Briefly, 1 3 104 freshly isolatedTotal RNA was extracted by a guanidinium isothiocy-anate method using Trizol (GIBCO BRL, Grand Island, glomeruli were placed in 100 mL PBS and 20 mL of

3 mmol/L plasmin specific chromogenic substrate solu-NY, USA) according to the manufacturer’s instructions.Isolated glomeruli from each rat were used for subse- tion (Tos-Gly-Pro-Lys-4-nitranilide-acetate; Chromozyme

Haraguchi et al: t-PA treatment of anti–Thy-1 nephritis 2149

PL, Roche Molecular Biochemicals, Indianapolis, IN,USA), and 100 mL of 100 mmol/L Tris-HCl buffer (pH 8.0)were added. After one hour of incubation at room tem-perature, glomerular supernatants were harvested. Theamount of p-nitroaniline released into supernatant byglomeruli was measured at 405 nm with a microplatereader. Standard curves were generated using porcineplasmin (Sigma, St. Louis, MO, USA). Plasmin activity ofisolated glomeruli was expressed as 1024 U/104 glomeruli.

Measurement of glomerular TGF-b1 productionand content

Day 6 nephritic rats were sacrificed 10 minutes afterthe injection of rt-PA (4 mg/kg body weight) or saline

Fig. 1. Time course of glomerular fibrinogen/fibrin staining in OX-7–(N 5 4 for each group). Glomeruli from individual rats induced nephritis. Data are from three rats at each time point. *P ,were isolated as described previously in this article. For 0.01 vs. normal; #P , 0.01 vs. day 3.measurement of glomerular production of TGF-b1, glo-meruli were resuspended at 2 3 103 glomeruli/mL/wellin RPMI 1640 supplemented with 0.1 U/mL insulin,

hours until at least 28 days, fibrinogen/fibrin staining was100 U/mL penicillin, 100 mg/mL streptomycin, and25 mmol/L HEPES buffer. After a 72-hour incubation significantly increased compared with normal animalsat 378C/5% CO2, the supernatant was harvested and (P , 0.01). This indicates that fibrin is present in thestored at 2708C until analysis. For measurement of glo- diseased glomerulus to which injected rt-PA may bind.merular content of TGF-b1, glomeruli were suspended Day 3 was chosen as the first treatment day because theat 2 3 104 glomeruli/mL in PBS. Glomeruli were homog- injury phase of disease has passed and fibrin depositionenized four times on ice by sonication. Each 15-second is dramatically increased.sonication was followed by a 15-second cool down. Aftercentrifugation at 400 3 g for 10 minutes at 48C, the Effects of rt-PA treatment on matrix accumulation insupernatant was stored at 2708C until analysis. Active anti–Thy-1 nephritisTGF-b1 was measured without acid treatment of samples Representative glomeruli stained with PAS are shownusing an inhibitory enzyme-linked immunosorbent assay

in Figure 2. The glomeruli from the disease control rats(ELISA) kit (Quantikine; R&D Systems, Minneapolis,showed marked accumulation of pathological matrix ex-MN, USA) according to the manufacturer’s instructionspressed as PAS-positive material at day 6 (Fig. 2A). Theas described previously [27]. Total (active plus latent)rt-PA–treated group showed reduced mesangial matrixTGF-b1 was measured after acid activation of cultureaccumulation in comparison to the disease control groupsupernatant.at day 6 (Fig. 2B). Figure 2C shows a graphical represen-

Statistics and calculation of percentage reduction in tation of PAS matrix scoring. PAS score increased fromdisease severity 0.57 6 0.10 in normal control rats to 2.75 6 0.17 in

disease control rats as a result of nephritis. rt-PA injec-Data are expressed as mean 6 SD. The significanceof differences in the measured values between groups tion decreased the matrix score significantly (P , 0.001)was analyzed by the Student t test or Welch t test. P , from 2.75 6 0.17 in the disease control group to 1.41 60.05 was considered statistically significant. The disease- 0.09 in the treated group. This represents a 62% reduc-induced increase in a variable was defined as the mean tion in the disease-induced increase in the PAS stainingvalue for the disease control group minus the mean value score. Total glomerular cell counts of the rt-PA treatedof the normal control group. The percentage of reduction group were not different from those of the disease con-in disease severity in a t-PA–treated group was calcu- trol group at day 6 (72 6 8 vs. 79 6 4).lated as follows:

To evaluate the effect of rt-PA on fibrinogen/fibrindeposition in nephritic glomeruli, direct immunofluo-1 2 1 Treated group mean 2 Normal control group mean

Disease control group mean 2 Normal control group mean2 3 100

rescent staining was performed at day 6 (Fig. 3). Thestaining score for fibrinogen/fibrin was reduced by 43%

RESULTS in the rt-PA treated group compared with the diseaseTime course of fibrinogen/fibrin staining in control group (P , 0.01).anti–Thy-1 nephritis Results of the image analysis of immunoperoxidase

staining for matrix proteins are shown in Figure 4. TheFigure 1 shows that glomerular fibrinogen/fibrin depo-sition increases rapidly after OX-7 injection. From 12 percentage of glomerular area with positive staining was

Haraguchi et al: t-PA treatment of anti–Thy-1 nephritis2150

Fig. 3. Immunofluorescent staining score for fibrinogen/fibrin at day 6.Glomerular staining for fibrinogen/fibrin was significantly lower in thert-PA treated group compared with the disease control group. *P ,0.01 compared with disease control (N 5 6 rats per group).

levels of TGF-b1, fibronectin, and PAI-1 were unchangedby rt-PA treatment.

Localization of administered rt-PA inFig. 2. Glomerular histology. Representative photomicrograph of glo- nephritic glomerulimeruli from anti–Thy-1 rats treated with saline (A) or recombinant

To check the localization of administered rt-PA intissue plasminogen activator (rt-PA) (B) at day 6; (C) periodic acid-Schiff (PAS) staining scores for each group is shown. rt-PA treatment nephritic glomeruli, double staining for t-PA and fibrin-resulted in a highly significant reduction in PAS staining score compared ogen/fibrin was performed. Glomeruli from a diseasewith disease control rats. *P , 0.001 vs. normal control; #P , 0.001 vs.

control rat showed weak staining for endogenous t-PAdisease control. Total glomerular cell counts did not differ between rt-at day 3 (data not shown). Glomeruli from two rt-PA–PA–treated and disease control glomeruli (N 5 6 rats per group).treated rats at day 3 showed intense staining for t-PA(Fig. 7A), and this staining pattern was similar to thatof fibrinogen/fibrin (Fig. 7B). These findings suggest that

calculated as staining index for quantitative analysis. injected rt-PA colocalized with fibrinogen/fibrin alongthe glomerular capillary loops and in the mesangium.Compared with the disease control group, the staining

index was significantly lower in the rt-PA–treated groupEffect of rt-PA on plasmin activity of glomeruliat day 6 for fibronectin-EDA1 (19 6 2 vs. 14 6 1, P , placed in culture

0.01), laminin (35 6 2 vs. 25 6 2, P , 0.01), type ITo evaluate the efficacy of rt-PA injection (1 mg/kgcollagen (33 6 1 vs. 21 6 3, P , 0.01), and type IV

body weight) on plasmin generation in glomeruli, wecollagen (27 6 2 vs. 23 6 1, P , 0.01). These representmeasured the plasmin activity of isolated glomeruli by

decreases in disease-induced matrix accumulation of chromogenic assay (Fig. 8). Day 3 nephritic glomeruli36% for fibronectin-EDA1, 63% for laminin, 36% for placed in culture showed decreased plasmin activity com-collagen I, and 50% for collagen IV. pared with normal rat glomeruli (8 6 1 vs. 11 6 1 3

1024 U/104 glomeruli, P , 0.01). Plasmin activity of theEffect of rt-PA on urinary protein excretion in nephritic glomeruli was elevated by in vivo presacrificeanti–Thy-1 nephritis injection of rt-PA (19 6 1 vs. 8 6 1, P , 0.01), whereas

the normal rat glomeruli did not show the significantTwenty-four–hour urinary protein excretion was mea-increase in plasmin activity by the same presacrifice rt-sured from day 5 to day 6 (Fig. 5). Urinary protein excre-PA treatment (12 6 1 vs. 11 6 1). rt-PA injection intotion was significantly lower in the rt-PA–treated groupnephritic rats resulted in plasmin activity that was 1.7-compared with the disease control group (46.5 6 15.6 vs.fold higher than normal glomerular plasmin activity and76.6 6 9.2 mg/day, P , 0.01).2.4-fold higher than that of nephritic glomeruli (Fig. 8).

Effect of rt-PA on glomerular mRNA Effect of rt-PA on TGF-b1 activation inlevels of TGF-b1, fibronectin, and PAI-1 in nephritic glomerulianti–Thy-1 nephritis Because plasmin has been shown to play a role in the

Glomerular RNA was isolated individually from each activation of latent TGF-b [41], we investigated glomeru-lar production and content of active and total TGF-b1rat on day 6. As shown in Figure 6, glomerular mRNA

Haraguchi et al: t-PA treatment of anti–Thy-1 nephritis 2151

Fig. 4. Image analysis of immunoperoxidasestaining for matrix proteins. Data present thepercentage of the glomerular area with posi-tive staining. Glomerular staining for fibro-nectin EDA1, laminin, type I collagen, andtype IV collagen were significantly lower inthe rt-PA–treated group at day 6. *P , 0.01compared with disease control (N 5 6 rats pergroup; 20 glomeruli per rat).

Fig. 5. Effect of rt-PA on 24-hour urinary protein excretion from day5 to day 6. Urinary protein excretion was significantly lower in thert-PA–treated group. *P , 0.01 compared with disease control (N 56 rats per group).

Fig. 6. Effects of rt-PA on glomerular mRNA expression in anti–Thy-1nephritis. (A) Representative Northern blot. The relative glomerularusing ELISA (Fig. 9). rt-PA was injected at fourfold the mRNA expression of TGF-b1, fibronectin, and PAI-1 is depicted in

dose used in other parts of this study to enhance any panels B, C, and D, respectively, for data on all blots (N 5 6 rats pergroup). Glomerular mRNA levels of TGF-b1, fibronectin, and PAI-1TGF-b1 activation that might be occurring. Day 6 waswere unchanged at day 6.also the peak of fibrin deposition (Fig. 1) and TGF-b1

expression in the model (data not shown). Rats weresacrificed 10 minutes after rt-PA injection (4 mg/kg bodyweight) on day 6, and glomeruli were isolated and either glomeruli. In one day 6 rat, the 4 mg/kg body weight

dose of rt-PA was injected. One day 6 disease controlsonicated to determine the glomerular content ofTGF-b1 or placed in culture for 72 hours to assess the rat served as the control. Glomeruli were isolated, and

glomerular plasmin generation was measured as de-glomerular production of TGF-b1. rt-PA treatment didnot affect the glomerular production of total TGF-b1 scribed previously in this article. A fivefold increase in

plasmin generation by nephritic glomeruli compared(455 6 118 vs. 521 6 102 pg/2 3 103 glomeruli/72 hours)or active TGF-b1 (5.6 6 1.0 vs. 5.9 6 0.3). rt-PA treat- with normal glomeruli was seen (37.0 vs. 7.5 3 1024 U/

104 glomeruli), suggesting that at this dose, plasmin gen-ment also did not affect total TGF-b1 (919 6 280 vs.922 6 272 pg/2 3 104 glomeruli) or active TGF-b1 (7.1 6 eration is greatly increased but does not lead to increased

activation of latent TGF-b1.2.3 vs. 6.7 6 0.9) in supernatant from sonicated, fresh

Haraguchi et al: t-PA treatment of anti–Thy-1 nephritis2152

Fig. 7. Representative immunofluorescent staining for fibrinogen/fibrin(A) and t-PA (B). Disease control rats showed weak staining for endoge-nous t-PA at day 3 (data not shown). (A and B) Double staining of aglomerulus from an rt-PA treated rat at day 3. Injected rt-PA colocalizedwith fibrinogen/fibrin along the capillary loops and in the mesangium.Sections from two disease control and two rt-PA–treated rats were

Fig. 8. Effect of rt-PA on the glomerular plasmin activity of normalexamined.and diseased (day 3) rats. Kidneys were removed 10 minutes after salineor rt-PA injection (1 mg/kg body weight). Nephritic glomeruli placedin culture had decreased plasmin activity, which was elevated by in vivopresacrifice injection of rt-PA (day 3, left and right bar, respectively;N 5 4 rats per group). *P , 0.01 compared with normal rats; #P ,DISCUSSION0.01 compared with disease control.

In anti–Thy-1 nephritis, fibrin deposition occurs alongthe glomerular capillary walls and in the mesangium.The time course of fibrinogen/fibrin staining shown inFigure 1 indicates that there is a great deal of fibrin to growth factors from the disease setting did not contribute

to the decrease in matrix accumulation is the finding thatwhich injected rt-PA can bind during the day 3 to day 5treatment period. In addition to its high affinity for fibrin, glomerular mRNA levels for key mediators of disease

were not altered at day 6 (Fig. 6). If the action of growtht-PA’s affinity for substrate plasminogen increases whenit binds to fibrin. This binding increases the availability factors released from the fibrin clot were eliminated by

fibrin degradation, one would expect disease to be dimin-of plasminogen locally, which results in more efficient(several hundred-fold) activation of plasminogen at the ished and to be reflected in lower mRNA levels for

TGF-b1, fibronectin, and PAI-1. This did not occur. Itfibrin surface [42]. Double immunostaining of t-PA andfibrinogen/fibrin showed colocalization in nephritic rat is possible that the administration of rt-PA from a time

earlier than day 3 might produce a reduction in diseaseglomeruli (Fig. 7). Furthermore, glomerular plasmin ac-tivity was elevated only when rt-PA was injected into severity, as reflected by a reduction in mRNA levels.

Antifibrotic effects of a mixture of orally administerednephritic rats and not when injected into normal ratswith trace fibrinogen/fibrin staining (Fig. 8). These data proteases have been reported in a number of experimen-

tal renal diseases [44, 45]. The mechanism(s) of thesestrongly suggest that when rt-PA is injected into ne-phritic rats from day 3 to day 5 of disease, the rt-PA is effects is unclear, but increased clearance of TGF-b by

protease/a2-macroglobulin complexes and immunomo-targeted to the glomerulus by fibrin that is present aspart of the response to injury, and when bound to fibrin, dulation by these proteases have been suggested [46, 47].

Zoja et al have demonstrated that rt-PA treatment de-rt-PA’s affinity for plasminogen increases resulting inplasmin generation. creased fibrin deposition and crescent formation in rabbit

antiglomerular basement membrane nephritis [48]. Oth-The therapeutic consequences of this increased plas-min were (1) highly significant decreases in PAS-positive ers found decreased expression of t-PA and increased

PAI-1 in anti-glomerular basement membrane (anti-material (Fig. 2); (2) decreased fibrinogen/fibrin accumu-lation (Fig. 3); (3) decreased accumulation of specific GBM) nephritis that was associated with reduced fi-

brinolytic activity [49]. In these studies, it was assumedmatrix proteins fibronectin EDA1, laminin, type I colla-gen, and type IV collagen (Fig. 4); and (4) decreased that the therapeutic effect arose directly from increased

fibrinolysis. However, there have been many disappoint-proteinuria (Fig. 5). It is uncertain exactly what the se-quence of events from increased plasmin generation to ing attempts to reduce disease severity by inhibiting co-

agulation, leading some to conclude that rather thandecreased matrix accumulation is in this setting. It ispossible that plasmin degraded primarily fibrin. Fibrin contributing to glomerular injury, coagulation is a sec-

ondary phenomenon [50]. In addition, none of theseclot is rich in platelets, and platelets are rich in platelet-derived growth factor and TGF-b [43]. It is possible studies asked whether the therapeutic effect observed

could be due to direct degradation of matrix by increas-that during fibrin degradation, these growth factors areinactivated and do not contribute to disease to the extent ing plasmin or other proteases, as our data suggest.

Although the mechanism of activation of TGF-b inthat they normally do. Arguing against this is the factthat treatment was started at day 3, relatively late in the vivo is poorly understood [41, 51], it has been shown that

plasmin activates TGF-b in vitro [41]. We found that adisease course. Stronger evidence that inactivation of

Haraguchi et al: t-PA treatment of anti–Thy-1 nephritis 2153

Fig. 9. Effect of rt-PA on glomerular produc-tion and glomerular content of total and activeTGF-b1. Kidneys were removed 10 minutesafter saline or rt-PA injection (4 mg/kg bodywt) at day 6 (N 5 4 for each group). The glo-merular production of total TGF-b1 (A) andactive TGF-b1 (B) and the glomerular contentof total TGF-b1 (C) and active TGF-b1 (D)were unchanged by rt-PA treatment.

large dose of rt-PA injected into a day 6 nephritic animal Fibrin deposition in the mesangium and the endothe-yielded a fivefold increase in glomerular plasmin genera- lium has been observed in a number of human glomeru-tion. When this dose was injected into other day 6 ne- lar diseases, including IgA nephropathy [19, 58], Hen-phritic animals, active TGF-b did not increase. These och-Shonlein purpura nephritis [58], and lupus nephritisresults suggest that increased plasmin alone is not suffi- [19, 59]. Although rt-PA treatment might be efficaciouscient to activate the latent TGF-b produced by nephritic in reducing matrix accumulation, rt-PA may cause systemicglomeruli. This conclusion cannot be a certainty, how- fibrinolysis and therefore may present a bleeding risk [60].ever, since the ELISA kit used to assay active TGF-b is We envision that if chronic administration of rt-PA werenot specifically designed for measuring active TGF-b in necessary to reduce matrix, smaller doses or the develop-nonacidified samples. Also, the values obtained for ac- ment of methods for local delivery would be necessary.tive TGF-b before acidification (to activate all TGF-b In conclusion, this study demonstrates that rt-PA treat-for measurement of total values) are at the lower limits ment attenuates the glomerular accumulation of fibrin,of sensitivity for the assay. Nonetheless, recent data in fibronectin, laminin, type I collagen, and type IV collagenthe bleomycin model of pulmonary fibrosis suggest that and decreases proteinuria without changing mRNA lev-a number of factors are required for activation of TGF-b els of TGF-b, PAI-1, and fibronectin in anti–Thy-1 ne-in vivo [52]. It is also likely that a substantial increase phritis. rt-PA and fibrin colocalize in nephritic glomeruli.in active TGF-b would have increased disease severity Plasmin activity of nephritic glomeruli is elevated by ato a detectable level. As demonstrated by all of the dis- presacrifice injection of rt-PA, but we find no evidenceease measures employed, this did not occur.

for increased activation of latent TGF-b.Recombinant tissue plasminogen activator decreasedThese results suggest that injected rt-PA binds fibrinproteinuria in this study. The mechanism of this effect

in nephritic glomeruli, where it increases plasmin gen-is uncertain but may be due to decreased glomerulareration and promotes pathological matrix degradation.matrix accumulation. A direct effect of plasmin on pro-While rt-PA may have limited applicability in human fi-teinuria would have been expected to have the oppositebrotic disease, these results show that therapeutic strate-effect since elastin and cathepsin G have been reportedgies aimed at increasing matrix degradation may haveto increase urinary protein excretion [53].great therapeutic potential.We measured glomerular cellularity in this study be-

cause it has been reported that MMPs have a mitogenicACKNOWLEDGMENTSactivity in vitro [54, 55] and in vivo [56]. Treatment of

anti–Thy-1 nephritis with the MMP inhibitor BB-1101 This work was supported by National Institutes of Health grantsDK 49374 (W.A.B.) and DK 43609 (W.A.B.).inhibited gelatinolytic MMP-2 activity, decreased cellu-

larity, and decreased matrix deposition [57]. In the pres-Reprint requests Nancy A. Noble, Ph.D., Fibrosis Research Labora-

ent study, rt-PA treatment did not alter glomerular cellu- tory, 391 Chipeta Way Suite E, Salt Lake City, Utah 84108, USA.E-mail: nancy.noble@hsc.utah.edularity at day 6.

Haraguchi et al: t-PA treatment of anti–Thy-1 nephritis2154

24. Poncelet AC, Schnaper HW: Regulation of human mesangial cellREFERENCEScollagen expression by transforming growth factor-b. Am J Physiol275:F458–F466, 19981. Border WA, Noble NA: Transforming growth factor-b in tissue

fibrosis. N Engl J Med 331:1286–1292, 1994 25. Sato Y, Fujimoto S, Hamai K, Eto T: Serial alterations of glomeru-lar matrix-degrading metalloproteinase activity in anti-thymocyte-2. Border WA, Okuda S, Nakamura T: Extracellular matrix and

glomerular disease. Semin Nephrol 9:307–317, 1989 induced glomerulonephritis in rats. Nephron 78:195–200, 199826. Kagami S, Kuhara T, Okada K, et al: Dual effects of angiotensin3. Border WA, Noble NA: Interactions of transforming growth fac-

tor-b and angiotensin II in renal fibrosis. Hypertension S31:181– II on the plasminogen/plasmin system in rat mesangial cells. KidneyInt 51:664–671, 1997188, 1998

4. Schnaper HW: Balanced between matrix synthesis and degrada- 27. Peters H, Border WA, Noble NA: Targeting TGF-b overexpres-sion in renal disease: Maximizing the antifibrotic action of angioten-tion: A determinant of glomerulosclerosis. Pediatr Nephrol 9:104–

111, 1995 sin II blockade. Kidney Int 54:1570–1580, 199828. Peters H, Border WA, Noble NA: Tandem antifibrotic actions5. Fioretto P, Steffes MW, Sutherland DER, et al: Reversal of

lesions of diabetic nephropathy after pancreas transplantation. of l-arginine supplementation and low protein diet during therepair phase of experimental glomerulonephritis. Kidney Int 57:N Engl J Med 339:69–75, 1998

6. Stetler-Stevenson WG: Dynamics of matrix turnover during 992–1001, 200029. Oikawa T, Freeman M, Lo W, et al: Modulation of plasminogenpathologic remodeling of the extracellular matrix. Am J Pathol 148:

1345–1350, 1996 activator inhibitor-1 in vivo: A new mechanism for the anti-fibroticeffect of renin-angiotensin inhibition. Kidney Int 51:164–172, 19977. Mignatti P: Extracellular matrix remodeling by metalloprotein-

ases and plasminogen activators. Kidney Int 47(Suppl 49):S12–S14, 30. Ma L-J, Nakamura S, Whitsitt JS, et al: Regression of vascular andrenal sclerosis in aging by angiotensin II type I receptor antagonist-1995

8. Liotta LA, Goldfarb RH, Brundage R, et al: Effect of plasmino- induced decrease in plasminogen activator inhibitor-1. Kidney Int58:2425–2436, 2000gen activator (urokinase), plasmin, and thrombin on glycoprotein

and collagenous components of basement membrane. Cancer Res 31. Yamamoto T, Noble NA, Cohen AH, et al: Expression of trans-forming growth factor-b isoforms in human glomerular diseases.41:4629–4636, 1981

9. Mochan E, Keler T: Plasmin degradation of cartilage proteogly- Kidney Int 49:461–469, 199632. Shihab FS, Yamamoto T, Nast CC, et al: Transforming growthcan. Biochim Biophys Acta 800:312–315, 1984

10. Mackay AR, Corbitt RH, Hartzler JL, Thorgeirsson UP: Base- factor-b and matrix protein expression in acute and chronic rejec-tion of human renal allografts. J Am Soc Nephrol 6:286–294, 1995ment membrane type IV collagen degradation: Evidence for the

involvement of a proteolytic cascade independent of metalloprotei- 33. Yamamoto T, Noble NA, Miller DE, et al: Increased levels oftransforming growth factor-b in HIV-associated nephropathy. Kid-nases. Cancer Res 50:5997–6001, 1990

11. He C, Wilhelm SM, Pentland AP, et al: Tissue cooperation in a ney Int 55:579–592, 199934. Ehrlich HJ, Klein Gebbink R, Keijer J, et al: Alteration of serpinproteolytic cascade activating human interstitial collagenase. Proc

Natl Acad Sci USA 86:2632–2636, 1989 specificity by a protein cofactor: Vitronectin endows plasminogenactivator inhibitor 1 with thrombin inhibitory properties. J Biol12. Nagase H, Enghild JJ, Suzuki K, Salvesen G: Stepwise activation

mechanism of the precursor of matrix metalloproteinase 3 (stromo- Chem 265:13029–13035, 199035. Okuda S, Languino LR, Ruoslahti E, Border WA: Elevatedlysin) by proteinases and (4-aminophenyl) mercuric acetate. Bio-

chemistry 29:5783–5789, 1990 expression of transforming growth factor-b and proteoglycan pro-duction in experimental glomerulonephritis. J Clin Invest 86:453–13. Ramos-Desimone N, Hahn-Dantona E, Shipley J, et al: Activation

of matrix metalloproteinase-9 (MMP-9) via a converging plasmin/ 462, 199036. Yamamoto T, Noble NA, Miller DE, Border WA: Sustainedstromelysin-1 cascade enhances tumor cell invasion. J Biol Chem

274:13066–13076, 1999 expression of TGF-b1 underlies development of progressive kid-ney fibrosis. Kidney Int 45:916–927, 199414. Baricos WH, Cortez SL, El-Dahr SS, Schnaper HW: ECM deg-

radation by cultured human mesangial cells is mediated by a PA/ 37. Fort P, Marty L, Piechaczyk M, et al: Various rat adult tissuesexpress only one major mRNA species from the glyceraldehyde-plasmin/MMP-2 cascade. Kidney Int 47:1039–1047, 1995

15. Lacave R, Rondeau E, Ochi S, et al: Characterization of a plasmin- 3-phosphate-dehydrogenase multigenic family. Nucleic Acids Res13:1431–1442, 1985ogen activator and its inhibitor in human mesangial cells. Kidney

Int 35:806–811, 1989 38. Derynck R, Jarrett A, Chen EY, Goeddel DV: The murinetransforming growth factor-b precursor. J Biol Chem 261:4377–16. Glass WF II, Kreisberg JI, Troyer DA: Two-chain urokinase,

receptor, and type 1 inhibitor in cultured human mesangial cells. 4379, 198639. Schwarzbauer JE, Tamkun JW, Lemischka IR, Hynes RO: ThreeAm J Physiol 264:F532–F539, 1993

17. Angles-Cano E, Rondeau E, Delarue F, et al: Identification different fibronectin mRNAs arise by alternative splicing withinthe coding region. Cell 35:421–431, 1983and cellular localization of plasminogen activators from human

glomeruli. Thromb Haemost 54:688–692, 1985 40. Zeheb R, Gelehrter TD: Cloning and sequencing of cDNA forthe rat plasminogen activator inhibitor-1. Gene 73:459–468, 198818. Ranby M: Studies on the kinetics of plasminogen activation by

tissue plasminogen activator. Biochim Biophys Acta 704:461–469, 41. Lyons RM, Keski-Oja J, Moses HL: Proteolytic activation of latenttransforming growth factor-b from fibroblast-conditioned medium.1982

19. Aya N, Yoshioka K, Murakami K, et al: Tissue-type plasminogen J Cell Biol 106:1659–1665, 198842. Lelievre Y, Bouboutou R, Boiziau J, et al: Low molecular weight,activator and its inhibitor in human glomerulonephritis. J Pathol

166:289–295, 1992 sequence based, collagenase inhibitors selectively block the inter-action between collagenase and TIMP (tissue inhibitor of metallo-20. Noble NA, Border WA: Angiotensin II in renal fibrosis: Should

TGF-b rather than blood pressure be the therapeutic target? Semin proteinases). Matrix 10:292–299, 199043. Abboud HE: Growth factors in glomerulonephritis. Kidney Int 43:Nephrol 17:455–466, 1997

21. Baricos WH, Cortez SL, Deboisblanc M, Xin S: Transforming 252–267, 199344. Sebecova K, Dammrich J, Fierlbeck W, et al: Systemic treatmentgrowth factor-b is a potent inhibitor of extracellular matrix degra-

dation by cultured human mesangial cells. J Am Soc Nephrol 10: with proteolytic enzymes in rat models of nonimmune mediatedrenal diseases. Int J Immunother 13:79–83, 1997790–795, 1999

22. Tomooka S, Border WA, Marshall BC, Noble NA: Glomerular 45. Emancipator SN, Chintalacharuvu SR, Urankar Nagy N,Stauder G: Oral enzymes in different animal models of glomerulo-matrix accumulation is linked to inhibition of the plasmin protease

system. Kidney Int 42:1462–1469, 1992 nephritis. Int J Immunother 13:97–103, 199746. Heidland AS, Paczek K, Teschner L, et al: Renal fibrosis: Role23. Marti HP, Kashgarian M, Lee L, Lovett DH: Transforming

growth factor-b1 stimulates glomerular mesangial cell synthesis of of impaired proteolysis and potential therapeutic strategies. KidneyInt 52(Suppl 62):S32–S35, 1997the 72-kd type IV collagenase. Am J Pathol 144:82–94, 1994

Haraguchi et al: t-PA treatment of anti–Thy-1 nephritis 2155

47. Lehmann PV: Immunomodulation by proteolytic enzymes. Nephrol 54. Kitamura M, Shirasawa T, Maruyama N: Gene transfer of metal-loproteinase transin induces aberrant behavior of cultured mesan-Dial Transplant 11:953–955, 1996gial cells. Kidney Int 45:1580–1586, 199448. Zoja C, Corna D, Macconi D, et al: Tissue plasminogen activator

55. Turck J, Pollock AS, Lee LK, et al: Matrix metalloproteinase 2therapy of rabbit nephrotoxic nephritis. Lab Invest 62:34–40, 1990(gelatinase A) regulates glomerular mesangial cell proliferation49. Malliaros J, Holdsworth SR, Wojta J, et al: Glomerular fibrino-and differentiation. J Biol Chem 271:15074–15083, 1996lytic activity in anti-GBM glomerulonephritis in rabbits. Kidney

56. Turck J, Pollock AS, Lovett DH: Gelatinase A is a glomerularInt 44:557–564, 1993mesangial cell growth and differentiation factor. Kidney Int 51:50. Border WA: Anticoagulants are of little value in the treatment1397–1400, 1997of renal disease. Am J Kidney Dis 3:308–312, 1984

57. Steinmann-Niggli K, Ziswiler R, Kung M, Marti HP: Inhibition51. Lawrence DA: Transforming growth factor-b: An overview. Kid-of matrix metalloproteinases attenuates anti-Thy 1.1 nephritis.ney Int 47(Suppl 49):S19–S23, 1995 J Am Soc Nephrol 9:397–407, 1998

52. Yehualaeshet T, O’Connor R, Green-Johnson J, et al: Activa- 58. Ono T, Muso E, Suyama K, et al: Intraglomerular depositiontion of rat alveolar macrophage-derived latent transforming growth of intact cross-linked fibrin in Ig A nephropathy and Henoch-factor b-1 by plasmin requires interaction with thrombospondin-1 Schonlein purpura nephritis. Nephron 74:522–528, 1996and its cell surface receptor, CD36. Am J Pathol 155:841–851, 1999 59. McCutcheon J, Evans B, D’Cruz DP, et al: Fibrin deposition inn

53. Johnson RJ, Couser WG, Alpers CE, et al: The human neutrophil SLE glomerulonephritis. Lupus 2:99–103, 1993serine proteinases, elastase and cathepsin G, can mediate glomeru- 60. Loscalzo J, Braunwald E: Tissue plasminogen activator. N Engl

J Med 319:925–931, 1988lar injury in vivo. J Exp Med 168:1169–1174, 1988