Hepatocyte growth factor suppresses interstitial fibrosis in a mouse model of obstructive nephropathy

Kidney International, Vol. 59 (2001), pp. 1304–1314

Hepatocyte growth factor suppresses interstitial fibrosis in amouse model of obstructive nephropathy

SHINYA MIZUNO, KUNIO MATSUMOTO, and TOSHIKAZU NAKAMURA

Division of Biochemistry, Department of Oncology, Biomedical Research Center, Osaka UniversityGraduate School of Medicine, Yamadaoka, Japan

Hepatocyte growth factor suppresses interstitial fibrosis in a animal models, focusing on glomerular injury [3, 4]. Inmouse model of obstructive nephropathy. addition to glomerulosclerosis, the molecular basis of

Background. As tubulointerstitial fibrosis (TIF) reflects the TIF has to be clarified as the severity of chronic tubuloin-prognosis of patients with various chronic renal diseases, theterstitial injury correlates best with impaired renal func-pathogenesis of TIF has to be clarified. Transforming growthtion and most accurately predicts renal dysfunction infactor-b (TGF-b) is a key mediator for renal fibrosis. We re-

ported that hepatocyte growth factor (HGF) prevents renal patients with CRF [5, 6]. It is important to address molec-fibrosis in nephrotic mice. However, the function of HGF in ular mechanisms whereby interstitial fibrosis occurs andchronic renal failure, except for nephrotic syndrome, remains progresses in chronically injured renal tissues in orderto be determined.

to minimize CRF-related pathological states [7, 8].Methods. Using mice subjected to unilateral ureter-ligatedHepatocyte growth factor (HGF) was originally iden-obstruction (UUO), we investigated the roles of HGF in TIF,

as induced by obstructive nephropathy. Pathophysiological tified and cloned as a potent mitogen for mature hepato-changes in the kidney after UUO treatment were analyzed cytes [9, 10]. It is now widely accepted that HGF is afocusing on expressions of renal HGF and TGF-b, TIF, tubular regenerative as well as a cytoprotective molecule in sev-proliferation, and apoptosis. Neutralizing antibody against ro-

eral organs with various injuries [11–13]. HGF is mainlydent HGF, or recombinant human HGF (rhHGF), was admin-produced in interstitial cells and targets epithelial andistrated to the UUO mice, and pathophysiological changes

after neutralization or supplements of HGF were analyzed. endothelial cells, thereby contributing to organogenesisResults. In this UUO model, TIF with tubular apoptosis and tissue repair through mitogenic, motogenic, and

became evident, and it was accompanied by a decrease in renal morphogenic activities [11–13]. In renal development,HGF expression and an increase in renal TGF-b expression.HGF is involved in morphogenesis and in the mesenchy-Neutralization of endogenous HGF accelerated the progres-mal-epithelial transition during nephrogenesis [13, 14].sion of TIF, accompanied by increases in TGF-b expression

and tubular apoptosis as well as by decreases in tubular prolifer- In adult kidney, HGF derives largely from nonparenchy-ation. In contrast, rhHGF attenuated TIF progression, and mal cells (that is, interstitial fibroblasts, macrophages,there were decreases in TGF-b expression and tubular apopto- and endothelial cells) in remnant intact areas around thesis, and an increase in tubular proliferation.

injured parenchyme [15, 16]. Of importance, exogenousConclusions. Endogenous as well as exogenous HGF atten-uated the progression of the fibrosis caused by obstructive HGF prevents tubular destruction and accelerates recov-nephropathy in these mice. Thus, local reduction in HGF levels ery from renal dysfunction in cases of acute renal failuremay account for TIF in chronic renal diseases. [16–19]. Thus, HGF may play a role as a cytoprotective

and renotropic factor [15, 16, 20].Recently, we found that endogenous as well as exoge-

Renal fibrosis, as evidenced by glomerulosclerosis and nous HGF prevents renal fibrosis in a mouse model oftubulointerstitial fibrosis (TIF), is a histologic hallmark nephrotic syndrome [21–23]. On the other hand, Taka-of chronic renal failure (CRF) [1, 2]. The molecular basis yama et al reported that transgenic mice, overexpressingof renal fibrosis has been studied using experimental HGF at an extremely high level (that is, nonphysiological

level) from an early embryonic stage manifested CRF(as characterized by glomerulosclerosis and tubular cystKey words: chronic renal disease, macrophages, renal fibrosis, trans-

forming growth factor-b, tubulointerstitial fibrosis. formation) [24]. Although HGF was clearly present inrenal biopsy specimens from CRF patients (includingReceived for publication May 3, 2000primary glomerulonephritis and drug-induced nephritis)and in revised form September 25, 2000

Accepted for publication October 30, 2000 [25, 26], little is known about the role of HGF in CRF,except in cases of spontaneous or transgenic CRF mice 2001 by the International Society of Nephrology

1304

Mizuno et al: Prevention of obstructive nephropathy by HGF 1305

[21–24]. This background led us to determine whether after obstructive treatment, the UUO mice were givenrhHGF (1 mg/kg/day, every 12 hours, subcutaneously)HGF may play a therapeutic or pathogenic role during

the progression of other types of CRF. for four consecutive days then killed five days after theinduction of obstructive nephropathy.Obstructive nephropathy occurs as a result of urinary

congestion, which is caused by intraluminal or extralumi-Immunohistochemical stainingsnal compressions (such as congenital hypoplasia, urinary

calculi, prostate tumor, and pregnancy) [27]. Rodents Renal tissues were fixed in cold 70% ethanol for 12hours and embedded in paraffin in a routine fashion.subjected to unilateral ureter-ligation obstruction (UUO)

have been widely used as a model to investigate the Tissue sections were cut at a thickness of 4 mm and weredewaxed and stained with hematoxylin and eosin. Topathogenesis of TIF [28–30]; however, potential involve-

ment of HGF in TIF has yet to be determined in this identify tubular growth, ethanol-fixed sections werestained with peroxidase-conjugated anti-human prolifer-model of fibrosis. We now provide evidence that HGF

may play a role in suppressing TIF, whereas a decrease ating cellular nuclear antigen (PCNA) monoclonal anti-body (EPOS System; Dako, Glostrup, Denmark) [21].in renal HGF level may facilitate the progression of TIF.To detect apoptotic cells, an in situ end-labeling methodwas applied on dewaxed sections using a kit (TACM

METHODSTdT system; Trevigen Inc., Gaithersburg, MD, USA)

Animal treatments [21]. To evaluate infiltration of interstitial macrophages,anti-mouse Mac-1 rat monoclonal antibody (1:400; Chemi-Six-week-old female ICR (22 to 24 g body wt; Slc

Japan, Hamamatsu, Japan) were anesthetized with keta- con, Temecula, CA, USA) was applied to the dewaxedsections as the primary reaction, followed by a secondmine chloride (80 mg/kg, subcutaneously) and xyladine

sulfate (8 mg/kg, subcutaneously). Under general anes- reaction with biotin-labeled anti-rat IgG goat IgG (Vector,Burlingame, CA, USA). Finally, an avidin-biotin couplingthesia, they underwent left proximal ureteral ligation, as

described elsewhere [28–30]. All experimental proce- (ABC) reaction was performed on the sections, using akit (Vectstain Elite; Vector). To identify myofibroblasts,dures were carried out with great care according to

guidelines for animal welfare set up by Osaka University monoclonal IgG against human a-smooth muscle actin(a-SMA; EPOS System; Dako) was used [21, 22]. ToGraduate School of Medicine.

To analyze the histologic and biochemical changes examine the extent of interstitial fibrosis, rabbit IgGagainst rat type I collagen (1:200; CosmoBio, Tokyo,during progression of TIF, 24 mice were killed at 0, 1.5,

3, and 7 days after the UUO treatment. At the time of Japan) was used for the primary reactions, followed bythe ABC technique. To visualize the expressions of growthnecropsy on six mice, renal tissues were fixed in cold

70% ethanol for histologic examinations, as described factors, rabbit IgG against porcine and human trans-forming growth factor-b1 (TGF-b1), -2, -5 (pan-TGF-b;later in this article. The remaining tissues were immedi-

ately frozen in liquid nitrogen and stocked at 2808C for 1:100; R&D Laboratory, Minneapolis, MN, USA), andrabbit IgG against rat HGF (1:1000; Institute of Immu-biochemical analyses, as discussed later in this article.

For neutralization of endogenous HGF, an anti-HGF nology, Tokyo, Japan) were used for the primary reac-tions, and the ABC technique was applied using theantibody was raised by immunizing rat HGF in healthy

rabbits, and the IgG fraction was purified using Protein Vectastain kit [21, 22].A-Sepharose (Pharmacia, Uppsala, Sweden). This anti-

Semiquantitation of renal phenotypes withrat HGF antibody shows cross-reactivity with mousehistologic scoresHGF (but not with human HGF) and modulates tissue

phenotypes in the CRF kidney [21], pancreatic islets Tubular lesions, characterized by tubular dilation andepithelial desquamation with interstitial expansions, were[31], and lung [32] in rodents subjected to parenchymal

injuries. The UUO mice were randomly divided into two graded according to the extent of cortical involvementon a scale from 0 to 4: 0 5 normal; 1 5 involvement ofgroups and injected with rabbit anti-rat HGF IgG (N 5

8) or normal rabbit IgG (N 5 8, 250 mg/day/mouse, less than 25% of the cortex; 2 5 involvement up to 25to 50% of the cortex; 3 5 involvement up to 50 to 75%intraperitoneally) for the initial three days. These mice

were killed three days after the UUO treatment. of the cortex; 4 5 extensive damage involving more than75% of the cortex. Then they were expressed as tubularTo evaluate the effect of exogenous HGF on the pro-

gression of TIF in UUO mice, 12 UUO mice were gener- injury scores [21, 22]. To evaluate the degree of tubulargrowth and apoptosis, tubular PCNA and TUNEL indi-ated and injected with recombinant human HGF (rhHGF).

rhHGF was purified from conditioned medium of Chi- ces were, respectively, determined by counting numbersof positive tubular cells in at least 20 randomly chosennese hamster ovary cells transfected with expression vec-

tor containing HGF cDNA of the five amino acids- nonoverlapping fields [3400 high-power field (hpf)]. Thedegree of macrophage infiltration was expressed as thedeleted type, as described [16, 17, 19, 21, 22]. One day

Mizuno et al: Prevention of obstructive nephropathy by HGF1306

Fig. 1. Natural course of tubular changes in unilateral ureter-ligated obstruction (UUO) mice after the onset of obstructive nephropathy. (A)Change in tubular injury score, and representative photomicrographs of tubular lesions after UUO manipulation (H.E. stain, 3130). (B) Changein tubular proliferating cellular nuclear antigen (PCNA) score, and representative photographs of epithelial cell proliferation in proximal tubules(PCNA stain, 3240). (C) Apoptotic changes in tubules during the progression of obstructive nephropathy (TdT stain, 3240). The overall mean 6SD histological scores were calculated based on individual values (N 5 6), which were determined from at least 20 high-power fields (hpf) per mouse.

Mac-1 score, determined by counting numbers of mono- man HGF as the antibody reacts with mouse or rat butnot human HGF [21]. Renal TGF-b1 levels were deter-nuclear cells immunoreactive for Mac-1, in at least 20

randomly chosen hpf. Immunoreactive signals for inter- mined in an ELISA system using a commercial kit(Quantikine TGF-b1; R&D) [21, 22]. Furthermore, typestitial a-SMA and for col (1) were observed in at least

20 randomly chosen hpf and scored on the scale of 0 5 I and type III collagen levels were measured by a dye-binding method using a kit (Sircol collagen assay; Bio-absent; 1 5 mild; 2 5 moderate; and 3 5 severe [21, 28].

To quantitate the renal expression of TGF-b or HGF, color Ltd., Belfast, Ireland) [21].sections of UUO kidney were examined under a light

Statistical analysesmicroscopy, and the degree of renal expression wasshown as a mean value of the following histochemical All data were expressed as mean 6 SD. An unpaired

two-tailed t-test was used to compare the means, and ascales: 0 5 no staining; 0.5 5 trace staining; 1 5 lightstaining; 2 5 moderate staining; and 3 5 intense staining value of P , 0.05 was considered to have statistical

significance.in at least 20 randomly chosen hpf [33]. These semiquan-titative analyses were all made in a blinded fashion.

RESULTSBiochemical analysis of renal tissuesChanges in the kidney after UUO treatmentRenal tissue extract was prepared as described pre-

viously [21]. HGF levels were determined in enzyme- Tubular dilation with epithelial atrophy and/or de-squamation occurred three days after UUO, and tubulo-linked immunosorbent assays (ELISAs) using a commer-

cial kit (HGF EIA; Institute of Immunology). In HGF interstitial lesions (including the interstitial expansionand mononuclear cell infiltration) were predominant inELISA, mouse or rat HGF is distinguishable from hu-


Fig. 2. Changes in interstitial events in UUO mice subjected to obstructive nephropathy. (A) Changes in macrophage infiltration. The distributionof Mac-1-positive macrophages in the interstitial regions and a semiquantitative Mac-1 score are shown (Mac-1 stain, 3240). (B) Changes ininterstitial myofibroblast formation [a-smooth muscle actin (a-SMA) stain, 3160]. (C) Fibrotic changes in tubulointerstitial regions, as estimatedby type I collagen [Col (1)] accumulation [Col (1) stain, 3220]. Histological scores are expressed as mean 6 SD (N 5 6).

the kidney on day 7 after UUO (Fig. 1A). Consistently, Expression of TGF-b and HGF in the UUO kidneythe tubular injury score increased for up to seven days Transforming growth factor-b may play an importantafter UUO (day 3, 1.17 6 0.39; day 7, 2.68 6 0.20). In role in the onset and progression of tubular atrophy andan early stage of the obstructive nephropathy, tubular TIF in UUO models [28, 29]. Interstitial cells positivegrowth became evident, and the PCNA score showed a for TGF-b were rare for up to three days after UUO,

whereas numerous TGF-b–positive cells were noted inpeak at three days after UUO (17.5 6 4.32 tubular cells/the interstitial areas of the kidney seven days after UUOhpf; Fig. 1B). In contrast to the tubular growth, tubular(Fig. 3A). To identify TGF-b–positive cells, immuno-cells positive for TdT (that is, an apoptotic marker) in-staining of serial sections with an anti–Mac-1 antibody (acreased, especially seven days after UUO (5.64 6 1.85marker for macrophages) was done. Most TGF-b–express-tubular cells/hpf), which suggested a rapid progressioning cells were immunoreactive for the marker antibodyto tubular cell death (Fig. 1C).(Fig. 3A, insert), thus indicating that macrophages are aIn parallel to the tubular apoptosis, macrophagesmajor source of TGF-b. Additionally, TGF-b–immuno-(identified as Mac1-positive cells) became evident, par-positive signals were also seen in spindle-shaped fibro-ticularly around the atrophied tubules (day 1.5, 2.38 6blastic cells as well as in tubular epithelial cells, as de-

0.56 cells/hpf; day 7, 8.63 6 1.68 cells/hpf; Fig. 2A). Inter- scribed elsewhere [28, 29]. The TGF-b staining scorestitial areas stained with anti–a-SMA antibody rapidly increased, especially seven days after UUO (day 1.5,expanded from three days after the onset of UUO, thus 0.59 6 0.24; day 7, 3.84 6 1.18; Fig. 3C, left). On theindicating myofibroblast hyperplasia (Fig. 2B). Conse- other hand, HGF was detected in peritubular cells (andquently, interstitial fibrosis, as evidenced by type I colla- partially on tubular cells) from three days followinggen overexpression, rapidly progressed from three days UUO (Fig. 3B). The HGF score reached a peak at threeafter UUO (type I collagen score on day 7, 2.25 6 0.15; days and decreased at seven days after UUO (Fig. 3C,

right).Fig. 2C).


Fig. 3. Differential expression of TGF-b and HGF in the obstructed kidney of UUO mice. (A) Immunohistochemical findings of renal TGF-bexpression after ureteral ligation (3240). The insert represents a positive signal for the Mac-1 antigen (a marker for macrophages), immunostainedon a serial section (3320). (B) Immunostaining for HGF expression in tubulointerstitium in the UUO kidney (3240). (C) Changes in renalexpression of TGF-b and HGF as estimated from the immunohistochemical scores. (D) Alterations in HGF, TGF-b1 and types I and III collagenlevels as determined by ELISA in the UUO kidneys. Immunohistochemical scores and biochemical values are expressed as mean 6 SD (N 5 6).

We quantitated HGF and TGF-b1 protein levels in 0.59 ng/g · tissue). As a result, the types I and III collagenlevels showed a parallel increase in TGF-b1 levels.the UUO kidneys using an ELISA method (Fig. 3D).

Renal HGF levels rapidly increased, reaching a peak onAdvanced fibrosis by neutralization of endogenous HGFday 3 after UUO (322.3 6 28.9 ng/g · tissue). However,

seven days following UUO, HGF levels declined to half To determine the role of intrinsic HGF in UUO mice,we examined renal phenotypes after anti-HGF IgG treat-of those seen three days after UUO. In contrast, renal

TGF-b1 levels gradually increased for up to seven days ment. PCNA-positive tubular cells were significantly re-duced in number in the anti-HGF IgG-treated mice asafter UUO (day 3, 1.34 6 0.45 ng/g · tissue; day 7, 2.38 6


Fig. 4. Aggravating effects of anti-HGF anti-body (a-HGF IgG) treatment on renal pheno-types in UUO mice. (A) Suppressive effect ofanti-HGF IgG treatment on tubular prolifera-tion (PCNA stain, 3260). (B) Aggravation oftubular apoptotic changes by anti-HGF IgGtreatment (TdT stain, 3260). (C) Accelera-tion of interstitial fibrosis by HGF-neutraliza-tion [collagen (1) stain, 3130]. Data are ex-pressed as mean 6 SD (N 5 8); *P , 0.05,**P , 0.01, and ***P , 0.001 vs. the normalIgG group (control).

Table 1. Effect of normal IgG or anti-HGF IgG on renal phenotypescompared with findings in normal IgG-injected micein UUO mice(8.79 6 2.27 vs. 17.2 6 4.07 cells/hpf, P , 0.01; Fig. 4A).

Normal Anti-HGFConversely, HGF-neutralized mice showed a higher TdTParameters IgG IgG P valuescore compared with findings in mice given a placeboTGF-b1 level ng/mL 1.260.4 2.060.5 ,0.05(2.81 6 1.03 vs. 1.13 6 0.42, P , 0.001; Fig. 4B). TheHGF level ng/mL 217.9624.3 161.3617.4 ,0.01renal TGF-b1 level was significantly higher in the HGF- TI score 1.8160.27 2.3860.31 ,0.05

neutralized mice than in mice given a placebo, while Mac-1 score 3.2160.58 4.4261.05 NSa-SMA score 1.0260.29 1.3460.39 NSanti-HGF IgG treatment decreased renal HGF levels toCollagen (1, 3) level mg/g 136.5648.3 193.9637.6 ,0.0574% over the control (Table 1). There was a definite trend

Data are mean 6 SD (N 5 8). Abbreviations are: HGF, hepatocyte growththat other fibrogenic parameters (including tubular injury,factor; UUO, unilateral ureter-ligated obstruction; TGF-b, transforming growth

interstitial Mac-1 and a-SMA scores, and renal collagen factor-b; TI, tubular injury; a-SMA, a-smooth muscle actin.

levels) were also increased in UUO mice given anti-HGF IgG (Table 1). Consequently, the HGF-neutralizedUUO mice manifested a rapid progression of TIF, as

ulate proliferation and apoptotic changes in tubular epi-evidenced by interstitial type I collagen score (Fig. 4C).thelial cells. Tubular growth in the rhHGF-treated UUO

Growth and protection of tubules by exogenous HGF mice was sustained at an 1.4-fold higher level over thatin saline-injected UUO mice (Fig. 5A). On the otherTo determine the preventive role of HGF in obstruc-

tive nephropathy, we asked whether rhHGF would mod- hand, tubular apoptosis was significantly suppressed in


Fig. 5. Beneficial effect of recombinant hu-man HGF (rhHGF) supplementation on tubu-lar events in UUO mice. (A) Enhancement oftubular proliferation by rhHGF administra-tion (PCNA stain, 3260). (B) Suppressiveeffect of rhHGF supplement on tubular apop-tosis (TdT stain, 3260). (C) Prevention of tu-bular epithelial desquamation by rhHGF ther-apy (H.E. stain, 3180). Histological scores areexpressed as mean 6 SD (N 5 6); *P , 0.05and **P , 0.01 vs. the saline group (control).

the rhHGF-injected mice as compared with findings in was reduced to 39% of the control level (Fig. 6B). Inaccordance with reduced TGF-b expression, the intersti-the control mice (TdT score, 2.49 6 0.78 vs. 4.01 6 1.15,

P , 0.05; Fig. 5B). Consistently, tubular destruction, tial a-SMA score was significantly lower in the rhHGF-treated mice than in control mice (Fig. 6C). Consequently,evaluated as the tubular injury score, was significantly

suppressed by rhHGF (1.23 6 0.42 vs. 2.56 6 0.13, P , rhHGF attenuated the progression of renal fibrosis, asevaluated by the type I collagen score (Fig. 6D).0.05; Fig. 5C). Overall, it was shown that not only endoge-

nous but also exogenous HGF have renotropic and anti- Attenuation of TIF by rhHGF led to the hypothesisthat a reciprocal balance between HGF and TGF-b1 mayapoptotic effects on tubular epithelial cells in this model

of obstructive nephropathy. occur in the progression of TIF in cases of obstructivenephropathy. In biochemical evaluations using ELISA,

Suppression of TIF by rhHGF there was a strong trend for exogenous HGF (that is,human HGF) to increase endogenous HGF (that is, mouseTo evaluate antifibrogenic functions of rhHGF, we

focused on interstitial changes in the UUO kidney. The HGF) levels (Fig. 7A). In contrast, the renal TGF-b1level was reduced in the rhHGF group with a significantnumber of Mac-1–positive macrophages was significantly

lower in the rhHGF group than in the saline group (3.61 6 difference (P , 0.05; Fig. 7B). Furthermore, the renaltypes I and III collagen levels in the UUO mice were0.73 vs. 6.85 6 1.05 cells/hpf, P , 0.01; Fig. 6A). Likewise,

rhHGF treatment led to a significant reduction in renal also reduced to a significant level when HGF was givenexogenously (Fig. 7C), and this was in agreement withTGF-b expression (including Mac-1–positive macro-

phages): The TGF-b score in the rhHGF-treated mice immunohistochemical evidence.


Fig. 6. Suppression of injury and fibrogenicevents in tubulointerstitium of UUO kidneyby HGF therapy. (A) Suppressive effect ofrhHGF injection on macrophage infiltration(Mac-1 stain, 3240). (B) Reduction of TGF-bexpression by rhHGF administration (TGF-bstain, 3240). The insert represents a positivesignal for the Mac-1 antigen expressed on aserial renal section (3320). (C) Suppressionof myofibroblast accumulation by rhHGFtherapy (a-SMA stain, 3130). (D) Attenuat-ing effect of rhHGF supplement on progres-sion of interstitial fibrosis [collagen (1) stain,3130). Immunohistochemical scores are ex-pressed as mean 6 SD (N 5 6); *P , 0.05and **P , 0.01 vs. the saline group (control).

DISCUSSION pression of TGF-b and the onset of glomerular and inter-stitial fibrosis [33–35]. Consistently, elevated expressionIn renal disease of diverse etiology, especially at anof TGF-b was noted in animal models of renal fibrosis,end stage of CRF [1, 2], TIF is a reliable histologic hall-including the UUO model [7, 8, 28, 29]. Importantly,mark. Thus, it is important to determine how TIF occurssuppression of TGF-b production in an UUO rat modeland progresses at a molecular level. Distinct lines of

evidence suggest a causal relationship between overex- by antisense oligonucleotide retards the progression of


syndrome [21], whereas a supplement of HGF decreasedthe renal TGF-b level and resulted in prevention/recov-ery from nephrosis-related CRF conditions [22]. Takentogether, we propose that the balance between TGF-band HGF may play a definite role in the regulation ofTIF in chronic renal diseases.

Given that a reciprocal balance between TGF-b andHGF regulates pathology of TIF, how HGF attenuatesTIF and how this reciprocal balance occurs had to begiven attention. Previous studies indicated that TGF-band HGF are counteracting in their biological activities.TGF-b strongly stimulates the synthesis of extracellularmatrix (ECM) proteins [37] and the production of inhibi-tors of proteinases involved in ECM breakdown, includ-ing plasminogen activator inhibitor-1 or tissue inhibitorof metalloproteinases [33, 38, 39]. In contrast, HGF stim-ulates or induces proteinases involved in breakdown ofECM proteins, including membrane type 1 matrix metal-loproteinase, urokinase-type plasminogen activator, and

Fig. 7. Changes in endogenous HGF, TGF-b1 and collagen levels in some matrix metalloproteinases in several types of cellsUUO mice by HGF supplement. (A) An increase in endogenous HGF [40–42]. TGF-b induces growth arrest [43] and apoptosislevel by rhHGF treatment. (B) Suppressive effects of rhHGF treatment

[44] in renal tubular cells and endothelial cells, whileon renal TGF-b1 levels. (C) Reduction of collagen (1, 3) level by rhHGFtreatment. Data are expressed as mean 6 SD (N 5 6); *P , 0.05 vs. HGF exhibits mitogenic [45, 46] and anti-apoptotic [47, 48]the saline group (control). activities in these cell types. HGF induces branching

tubulogenesis in renal epithelial cells, but inhibition isseen with TGF-b [49]. Thus, in addition to overexpres-sion of TGF-b, a reduction in renal HGF levels mightTIF [36], thereby suggesting a pivotal role for TGF-b

in fibrogenic events after obstructive nephropathy. Our allow for an increased susceptibility to TIF, while HGFneutralization could result in a rapid progression of TIF.present study also noted that expression of TGF-b in-

creased after UUO, and the increase in TGF-b expres- With regard to the reciprocal molecular balance, TGF-bstrongly inhibits gene expression of HGF in various typession correlated well with fibrotic changes of interstitial

regions in the kidney subjected to UUO. of cells [50–52], and this TGF-b1–induced suppressionof HGF expression seems to explain why the overexpres-During a rapid progression of TIF, expression of renal

HGF increased until three days after UUO, whereas it sion of TGF-b is associated with a decreased renal HGFexpression in a later stage after UUO. Since macro-decreased in an advanced stage of TIF. To determine

physiological functions of HGF, as related to UUO treat- phages are a major source of TGF-b in the UUO model[53] and the infiltration of macrophages was less in HGF-ment, we used two distinct approaches: biological neu-

tralization of HGF by antibody administration and aug- treated mice than in saline-injected mice, a reduction inTGF-b levels by HGF may be achieved, at least in part,mentation of HGF by supplement of exogenous HGF.

These two approaches clarified the potential role of HGF by suppression of macrophage infiltration. An alterna-tive possibility is direct inhibition of TGF-b productionin progression of TIF. Neutralization of HGF increased

renal expression of TGF-b, accompanied by increases in by HGF [41], although as yet we have not obtainedevidence in renal cells to support this possibility.interstitial type I collagen. In contrast, supplement of

HGF attenuated renal expression of TGF-b, accompa- In the UUO model, we found that changes in HGFexpression coincide with changes in tubular prolifera-nied with a decrease in the interstitial collagen accumula-

tion. Therefore, HGF is likely to play a role to retard tion: Tubular proliferation increased at an early stageof obstructive nephropathy, while it decreased in theor suppress the onset of TIF, and a decline of HGF,

which occurs in a reciprocal manner to a further increase advanced stage of TIF, in a reciprocal relationship tothe decrease in renal HGF level. These observationsin TGF-b expression in a later stage after UUO, may

allow for the rapid progression to TIF. We previously suggest an involvement of HGF as well as TGF-b in theregulation of tubular proliferation and apoptosis duringreported an association of the reciprocal balance be-

tween TGF-b and HGF regarding the onset and preven- the progression of TIF. More clearly, the modification ofHGF achieved by its neutralization or supplementationtion of CRF using nephrotic ICGN mice [21–23]: The

TGF-b–dominant balance achieved by neutralization of strengthens the involvement of HGF and TGF-b in regu-lating tubular proliferation and apoptosis. NeutralizationHGF accelerated the pathology of CRF after nephrotic


7. Eddy AA: Molecular insights into renal interstitial fibrosis. J Amof HGF decreased tubular growth, but enhanced renalSoc Nephrol 7:2495–2508, 1996

TGF-b expression and tubular apoptosis. In contrast, 8. Yamamoto T, Noble NA, Miller DE, et al: Sustained expressionsupplementation of HGF stimulated tubular prolifera- of TGF-b1 underlies development of progressive kidney fibrosis.

Kidney Int 45:916–927, 1994tion but decreased renal TGF-b expression and tubular9. Nakamura T, Nawa K, Ichihara A: Partial purification and charac-

apoptosis. Together with biological activities of these terization of hepatocyte growth factor from serum of hepatecto-growth factors on tubular cells [43–49], the balance be- mized rats. Biochem Biophys Res Commun 122:1450–1459, 1984

10. Nakamura T, Nishizawa T, Hagiya M, et al: Molecular cloningtween TGF-b and HGF may play a role in regulatingand expression of human hepatocyte growth factor. Nature 342:tubular proliferation and apoptosis during the progres- 440–443, 1989

sion of TIF in obstructive nephropathy. Although the 11. Matsumoto K, Nakamura T: Hepatocyte growth factor (HGF) asa tissue organizer for organogenesis and regeneration. Biocheminvolvement of a reciprocal balance of tubular epithelialBiophys Res Commun 239:639–644, 1997

proliferation and apoptosis in the progression of TIF has 12. Vargas GA, Hoeflich A, Jehle PM: Hepatocyte growth factoryet to be established [54–57], the reduction in parenchy- in renal failure: Promise and reality. Kidney Int 57:1436–1426, 2000

13. Balkovetz DF, Lipschutz JH: Hepatocyte growth factor and themal areas caused by impaired tubular proliferation andkidney: It is not just for the liver. Int Rev Cytol 186:225–260, 1999

increased apoptosis may allow for tubular atrophy or 14. Woolf AS, Kolatsi JM, Hardman P, et al: Roles of hepatocytegrowth factor/scatter factor and the met receptor in the early devel-disappearance and provoke reciprocal expansion of stro-opment of the metanephros. J Cell Biol 128:171–184, 1995mal components such as myofibroblasts (probably gener-

15. Nagaike M, Hirao S, Tajima H, et al: Renotropic functions ofated by macrophage-derived TGF-b) [53, 58] in the inter- hepatocyte growth factor in renal regeneration after unilateral

nephrectomy. J Biol Chem 266:22781–22784, 1991stitial regions. In this context, it is noteworthy that similar16. Igawa T, Matsumoto K, Kanda S, et al: Hepatocyte growth factorreciprocal changes in tubular proliferation and apoptosis

may function as a renotropic factor for regeneration in rats withwere seen in nephrotic mice with TIF [21, 22]. acute renal injury. Am J Physiol 265:F61–F69, 1993

17. Kawaida K, Matsumoto K, Shimazu H, et al: Hepatocyte growthObstructive nephropathy is assigned to a group offactor prevents acute renal failure and accelerates renal regenera-parenchymal and extraparenchymal renal disorders withtion in mice. Proc Natl Acad Sci USA 91:4357–4361, 1994

an impaired flow of urine or tubular fluid and one of 18. Miller SB, Martin DR, Kissane J, et al: Hepatocyte growth factorthe most frequent causes of end-stage renal disease in accelerates recovery from acute ischemic renal injury in rats. Am

J Physiol 266:F129–F134, 1994children [27, 55, 57]. The obstruction elicits tissue fibrosis19. Amaike H, Matsumoto K, Oka T, et al: Preventive effect of hepato-and impaired renal function. The potential therapeutic cyte growth factor on acute side effects of cyclosporin A in mice.

value of HGF for treatment of patients with obstructive Cytokine 8:387–394, 199620. Kopp JB: Hepatocyte growth factor: Mesenchymal signal for epi-nephropathy deserves ongoing attention.

thelial homeostasis. Kidney Int 54:1392–1393, 199821. Mizuno S, Matsumoto K, Kurosawa T, et al: Reciprocal balance

ACKNOWLEDGMENTS of hepatocyte growth factor and transforming growth factor-b1 inrenal fibrosis in mice. Kidney Int 57:937–948, 2000

This study was supported by grants from the Ministry of Education, 22. Mizuno S, Kurosawa T, Matsumoto K, et al: Hepatocyte growthScience, Sports, and Culture of Japan (No. 06283216 and No. 08408027 factor prevents renal fibrosis and dysfunction in a mouse modelto T.N.) and the Osaka Kidney Foundations (OKF 97-0003 to K.M. of chronic renal disease. J Clin Invest 101:1827–1834, 1998and OKF99-0006 to S.M.). We are grateful to M. Ohara for critical 23. Matsumoto K, Mizuno S, Nakamura T: Hepatocyte growth factorreadings of the manuscript and for pertinent comments. in renal regeneration, renal disease and potential therapeutics.

Curr Opin Nephrol Hypertens 9:395–402, 2000Reprint requests to Toshikazu Nakamura, Ph.D., Division of Bio- 24. Takayama H, Larochelle WJ, Sabnis SG, et al: Renal tubular

chemistry, Department of Oncology, Biomedical Research Center, hyperplasia, polycystic disease, and glomerulosclerosis in trans-Osaka University Graduate School of Medicine, Yamadaoka 2-2-B7, genic mice overexpressing hepatocyte growth factor/scatter factor.Suita 565-0871, Japan. Lab Invest 77:131–138, 1997E-mail: [email protected] 25. Taniguchi Y, Yorioka N, Yamashita K, et al: Localization of

hepatocyte growth factor and tubulointerstitial lesions in IgA ne-phropathy. Am J Nephrol 17:413–416, 1997REFERENCES

26. Yamaguchi K, Nalesnik MA, Michalopoulos GK: Expressionof HGF mRNA in human rejecting kidney as evidenced by in situ1. Ong AC, Fine LG: Loss of glomerular function and tubulointersti-

tial fibrosis: Cause or effect?. Kidney Int 45:345–351, 1994 hybridization. Urol Res 24:349–354, 199627. Klahr S: Urinary tract obstruction, in Diseases of the Kidney2. Nahas A, Muchaneta KE, Essawy M, et al: Renal fibrosis: Insights

into pathogenesis and treatment. Int J Biochem Cell Biol 29:55–62, (6th ed), edited by Schrier RW, Boston, Little, Brown, 1997, pp709–7381997

3. Bruijn JA, Roos A, Geus B, et al: Transforming growth factor-b 28. Kaneto H, Morrissey J, McCracken R, et al: Enalapril reducescollagen type IV synthesis and expansion of the interstitium in theand the glomerular extracellular matrix in renal pathology. J Lab

Clin Med 123:34–47, 1994 obstructed rat kidney. Kidney Int 45:1637–1647, 199429. Ishidoya S, Morrissey J, McCracken R, et al: Angiotensin II4. Wardle N: Glomerulosclerosis: The final pathway is clarified, but

can we deal with the triggers? Nephron 73:1–7, 1996 receptor antagonist ameliorates renal tubulointerstitial fibrosiscaused by unilateral ureteral obstruction. Kidney Int 47:1285–1294,5. Risdon RA, Sloper JC, Wardener HE: Relationship between

renal function and histological changes found in renal-biopsy speci- 199530. Moriyama T, Kawada N, Ando A, et al: Up-regulation of HSP47mens from patients with persistent glomerular nephritis. Lancet

2:363–366, 1968 in the mouse kidneys with unilateral ureteral obstruction. KidneyInt 54:110–119, 19986. Bohle A, Strutz F, Muller GA: On the pathogenesis of chronic

renal failure in primary glomerulopathies: A view from the intersti- 31. Nakano M, Yasunami Y, Maki T, et al: Hepatocyte growth factoris essential for amelioration of hyperglycemia in streptozotocin-tium. Exp Nephrol 2:205–210, 1994


induced diabetic mice receiving a marginal mass of intrahepatic b1-mediated effects on metabolism and viability in renal cells. AmJ Physiol 271:F689–F697, 1996islet grafts. Transplantation 69:214–221, 2000

32. Yamada T, Hisanaga M, Nakajima Y, et al: Enhanced expression 45. Igawa T, Kanda S, Kanetake H, et al: Hepatocyte growth factoris a potent mitogen for cultured rabbit renal tubular epithelialof hepatocyte growth factor after pulmonary ischemia reperfusion

injury in the rats. Am J Respir Crit Care Med 162:707–715, 2000 cells. Biochem Biophys Res Commun 174:831–838, 199146. Bowes R, Lightfoot RT, Warter BVD, et al: Hepatocyte growth33. Mozes MM, Bottinger EP, Jacot TA, et al: Renal expression

of fibrotic matrix proteins and of transforming growth factor-b factor induces tubulogenesis of primary renal proximal tubularepithelial cells. J Cell Physiol 180:81–90, 1999(TGF-b) isoforms in TGF-b transgenic mice. J Am Soc Nephrol

10:271–280, 1999 47. Yo Y, Morishita R, Nakamura S, et al: Potential role of hepatocytegrowth factor in the maintenance of renal structure: Anti-apoptotic34. Border WA, Okuda S, Languino LR, et al: Suppression of experi-

mental glomerulonephritis by antiserum against transforming action of HGF on epithelial cells. Kidney Int 54:1128–1138, 199848. Liu Y, Sun AM, Dworkin LD: Hepatocyte growth factor protectsgrowth factor-b1. Nature 346:371–374, 1990

35. Isaka Y, Fujiwara Y, Ueda N, et al: Glomerulosclerosis induced renal epithelial cells from apoptotic cell death. Biochem BiophysRes Commun 246:821–826, 1998by in vivo transfection of transforming growth factor-b or platelet-

derived growth factor gene into the rat kidney. J Clin Invest 49. Santos OF, Nigam SK: HGF-induced tubulogenesis and branchingof epithelial cells is modulated by extracellular matrix and TGF-b.92:2597–2601, 1993

36. Isaka Y, Tsujie M, Ando Y, et al: Transforming growth factor-b1 Dev Biol 160:293–302, 199350. Matsumoto K, Tajima H, Okazaki H, et al: Negative regulation ofantisense oligonucleotides block interstitial fibrosis in unilateral

ureteral obstruction. Kidney Int 58:1885–1892, 2000 hepatocyte growth factor gene expression in human lung fibroblastsand leukemic cells by transforming growth factor-b1 and glucocor-37. Border WA, Noble NA: Transforming growth factor-b in tissue

fibrosis. N Engl J Med 331:1286–1292, 1994 ticoids. J Biol Chem 267:24917–24920, 199251. Yo Y, Morishita R, Yamamoto K, et al: Actions of hepatocyte38. Baricos WH, Cortez SL, Deboisblanc M, et al: Transforming

growth factor-b is a potent inhibitor of extracellular matrix degra- growth factor as a local modulator in the kidney: Potential role inpathogenesis of renal disease. Kidney Int 53:50–58, 1998dation by cultured human mesangial cells. J Am Soc Nephrol

10:790–795, 1999 52. Ramadori G, Neubauer K, Odenthal M, et al: The gene of hepato-cyte growth factor is expressed in fat-storing cells of rat liver and39. Eickelberg O, Kohler E, Reichenberger F, et al: Extracellular

matrix deposition by primary human lung fibroblasts in response is downregulated during cell growth and by transforming growthfactor-b. Biochem Biophys Res Commun 183:739–742, 1992to TGF-b1 and TGF-b3. Am J Physiol 276:L814–L824, 1999

40. Pepper MS, Matsumoto K, Nakamura T, et al: Hepatocyte growth 53. Diamond JR, Kees FD, Ding G, et al: Macrophages, monocytechemoattractant peptide-1, and TGF-b1 in experimental hydro-factor increases urokinase-type plasminogen activator (u-PA) and

u-PA receptor expression in Madin-Darby canine kidney epithelial nephrosis. Am J Physiol 266:F926–F933, 199454. Truong LD, Petrusevska G, Yang G, et al: Cell apoptosis andcells. J Biol Chem 267:20493–20496, 1992

41. Taniyama Y, Morishita R, Nakagami H, et al: Potential contribu- proliferation in experimental chronic obstructive uropathy. KidneyInt 50:200–207, 1996tion of a novel anti-fibrotic factor, hepatocyte growth factor, to

prevention of myocardial fibrosis by angiotensin II blockade in 55. Chevalier RL, Goyal S, Wolstenholme JT, et al: Obstructivenephropathy in the neonatal rat is attenuated by epidermal growthcardiomyopathic hamsters. Circulation 102:246–252, 2000

42. Wang H, Keiser JA: Hepatocyte growth factor enhances MMP factor. Kidney Int 54:38–47, 199856. Hughes J, Johnson RJ: Role of Fas (CD95) in tubulointerstitialactivity in human endothelial cells. Biochem Biophys Res Commun

272:900–905, 2000 disease induced by unilateral ureteric obstruction. Am J Physiol277:F26–F32, 199943. Taipale J, Keski OJ: Hepatocyte growth factor releases epithelial

and endothelial cells from growth arrest induced by transforming 57. Klahr S: Obstructive nephropathy. Kidney Int 54:286–300, 199858. Diamond JR, Goor H, Ding G, et al: Myofibroblasts in experimen-growth factor-b1. J Biol Chem 271:4342–4348, 1996

44. Nowak G, Schnellmann RG: Autocrine production and TGF- tal hydronephrosis. Am J Pathol 146:121–129, 1995

Documents

Hepatocyte growth factor suppresses interstitial fibrosis in a mouse model of obstructive nephropathy