Kidney International, Vol. 58 (2000), pp. 182–192

MCP-1 and EGF renal expression and urine excretion inhuman congenital obstructive nephropathy

GIUSEPPE GRANDALIANO,1 LORETO GESUALDO,1 FABIO BARTOLI, ELENA RANIERI,RAFFAELLA MONNO, ANTONIO LEGGIO, GUGLIELMO PARADIES, EUSTACHIO CALDARULO,BARBARA INFANTE, and F. PAOLO SCHENA

Division of Nephrology, Department of Emergency and Transplantation and Department of Developmental Age Biomedicine,University of Bari, Bari, Italy

in the urine collected from the obstructed pelvis and the MAG3MCP-1 and EGF renal expression and urine excretion in hu-clearance of the obstructed kidney (r 5 20.76). The presenceman congenital obstructive nephropathy.of recurrent UTI was associated with a significantly higherBackground. Obstructive nephropathy is characterized atMCP-1 excretion and a slight reduction in EGF urine concen-the histologic level by tubular atrophy and interstitial monocytetration. The surgical correction of UPJO was followed by aninfiltration. The molecular mechanisms underlying these histo-

logic changes are still poorly defined. Epidermal growth factor improvement of renal function together with a significant re-(EGF) produced by tubular cells seems to play a pivotal role duction in MCP-1 excretion and a marked increase in EGFin the modulation of tubular cell growth, while monocyte che- urine concentrations. Interestingly, EGF urine concentrationmotactic peptide-1 (MCP-1) is a powerful and specific chemo- measured before surgery was significantly correlated with thetactic and activating factor for monocytes. difference between the MAG3 clearance of the obstructed

Methods. Twenty-four patients with congenital ureteropel- kidney before and after surgery.vic junction obstruction [UPJO; 10 with recurrent urinary tract Conclusions. MCP-1 and EGF seem to be involved in theinfection (UTI) and 10 with no UTI] and 15 healthy children pathogenesis of tubulointerstitial damage in congenital ob-were studied. Diagnosis was made by renal ultrasound, intrave- structive nephropathy, and their urine excretion may representnous pielography, and MAG3 scan. Urinary samples were col- a powerful prognostic marker in this form of renal disease.lected before and after surgery. In 10 patients, urine was alsocollected directly from the affected pelvis at the time of surgery.Urinary EGF and MCP-1 levels were measured by enzyme-

Chronic obstructive nephropathy is characterized atlinked immunosorbent assay. MCP-1 and EGF gene expressionwere evaluated by in situ hybridization in 15 biopsies from the histopathological level by tubular atrophy and inter-congenital UPJO and in 10 normal kidneys. stitial fibrosis [1]. Although the acute hemodynamic and

Results. In normal kidneys, there was a high expression functional abnormalities associated with obstructive uro-of EGF mRNA, whereas MCP-1 mRNA was undetectable.pathy have been clarified, the cellular and molecularMCP-1 gene expression was strikingly increased at the tubulo-

interstitial level in UPJO biopsies compared with controls and mechanisms leading to the chronic histologic lesions arewas directly correlated with the extent of monocyte infiltration. still largely unknown [2–4].In addition, UPJO kidney sections showed a marked reduction In the last few years, the interest of different investiga-in EGF gene expression that was directly correlated with the

tors has been focused on the presence of infiltratingdegree of tubular damage. EGF urine concentration was sig-monocytes in the interstitial space of the obstructed kid-nificantly reduced in UPJO when compared with control and

directly correlated with its renal gene expression. On the other ney [5–7]. These bone marrow-derived cells may repre-hand, the MCP-1 urine concentration was strikingly increased sent a potential source of cytokines and growth factorsin UPJO patients. It is noteworthy that a significant and inverse promoting the activation of resident cells and thereforecorrelation was observed between the MCP-1 concentration

playing a pivotal role in the pathogenesis of tubulointer-stitial damage [8–10]. Indeed, Harris, Schreiner, and

1 Dr. Grandaliano and Dr. Gesualdo contributed equally to the present Klahr have clearly demonstrated that infiltrating mono-work. cytes contribute to the acute decline in renal function

after the onset of experimental hydronephrosis [11].Key words: chemokines, growth factors, congenital ureteropelvic junc-tion obstruction, MAG3, monocyte. Although the presence of monocytes within the inter-

stitial space in experimental hydronephrosis and theirReceived for publication August 12, 1999

role in the pathogenesis of tubulointerstitial injury areand in revised form January 25, 2000Accepted for publication February 10, 2000 well established, the mechanisms involved in the recruit-

ment of these immune cells are still poorly defined. The 2000 by the International Society of Nephrology

182

Grandaliano et al: Markers of obstructive nephropathy 183

Table 1. Main clinical features of the uretopelvic junctionlocal release of chemokines could play a key pathogenicobstruction (UPJO) patients enrolled in the study

role in this scenario [12]. Diamond et al showed an in-Age MAG3 clearancecreased monocyte chemotactic peptide-1 (MCP-1) pro-

Patients months Gender %tein expression after unilateral obstruction in rats, inC.D. 5 M 34temporal concordance with the increase in interstitialC.M. 72 M 55

macrophage number [6]. MCP-1 is a powerful and spe- C.G. 96 M 53C.F. 6 M NAcific chemotactic and activating factor for monocytesC.C. 156 M 64[13]. An increased expression of this chemokine has beenD.S. 11 M 51

demonstrated by our group in different acute and chronic D.A. 2 F NAD.G. 13 M 19human tubulointerstitial diseases [14]. Moreover, theD.M. 11 M 31MCP-1 renal expression and urine excretion were strictlyG.M. 6 M 49

correlated with the tubular damage and the extent of L.F. 12 M 71L.A. 2 M 58monocyte infiltration, supporting a role for this chemo-L.E. 13 F 49kine in the pathogenesis of tubulointerstitial injury [14].L.R. 32 M NA

On the other hand, epidermal growth factor (EGF), L.V. 18 M 44M.C. 1 F NAproduced by the ascending portion of the Henle’s loopM.F. 5 M 55and the distal convoluted tubule, is a peptide growthM.M. 48 F 74

factor that modulates epithelial cell growth and metabo- P.L. 24 M 66P.D. 15 M 74lism [15, 16]. It has been suggested as a mediator ofS.L. 2 M NAnormal tubulogenesis and tubular regeneration after in-S.A. 9 M 59

jury. A reduction in renal EGF expression and/or urinary V.M. 24 M 56Z.S. 2 M 12excretion has been reported during acute and chronic

tubular injury, including an experimental model of hy- NA is not available.

dronephrosis [17–19]. Moreover, the administration ofEGF in animal models of acute ischemic tubular necrosisand hydronephrosis has been shown to accelerate the

excreted) was 52.6% (range 19 to 75%). None of thetubular regeneration and to attenuate tubular damagepatients had clinical signs or laboratory markers of activeand apoptosis [20–22]. However, no information is avail-infection, while 10 of them had a positive history forable on the tubular expression and urine excretion ofrecurrent urinary tract infection (UTI), and 10 had nothis growth factor in human chronic obstructive ne-history of UTI. Urinary samples were collected beforephropathy.and between two and four months after surgery. In 10In the present study, we investigated EGF and MCP-1patients with UPJO, urine was collected directly fromrenal gene expression and EGF and MCP-1 urine ex-the obstructed pelvis with a 22-gauge needle cannula.cretion in human congenital obstructive nephropathyAll of the patients underwent pieloplasty, and in 16 pa-caused by ureteropelvic junction obstruction (UPJO),tients, a kidney biopsy was performed during surgery.and their correlation with the main histologic and clinicalTen apparently normal portions of kidneys removed forfindings.renal carcinoma were used as control. Ten patients werefollowed after surgery for a mean period of 16.7 months

METHODS (range 11 to 24), after which they repeated the MAG3Patients scan.

Twenty-four children affected by congenital UPJOIn situ hybridizationand 15 healthy children were included in the study after

Renal tissue was immediately included in OCT (Tis-their parents gave informed consent. The main clinicalsue-Tek, Elkhart, IN, USA, USA), snap-frozen in liquidfeatures of the patients enrolled in the study are summa-nitrogen, and stored in the same liquid until used. Frozenrized in Table 1. Diagnosis of UPJO was made by renalsections (6 mm thick) were collected onto polylysine-ultrasound, cistography, intravenous pielography, andcoated slides, dried briefly on a hot plate at 808C, andMAG3 scan. The last two tests were performed withfixed in 4% paraformaldehyde for 20 minutes. After twodiuretic stimulation (furosemide 0.5 mg/kg intravenouswashes in phosphate-buffered saline (PBS), dehydrationbolus) to discriminate obstructive from nonobstructivein graded ethanol, and short air drying, sections wereUPJO. Five patients had a prenatal diagnosis of hydrone-immediately used for in situ hybridization.phrosis. Vescico-ureteral reflux was ruled out in all of

Human MCP-1 [14] and EGF cDNA fragments [18]the cases. All cases had normal blood urea nitrogenwere subcloned into pGEM3Zf. After linearization ofand creatinine levels. The mean total MAG3 clearance

(expressed as the percentage of the radioactive tracer the plasmid with either BamH I or Xba I restriction

Grandaliano et al: Markers of obstructive nephropathy184

endonucleases, T7 or SP6 RNA-polymerases (Boeh- sin (immunohistochemistry), and silver grains (in situhybridization) were quantitated by a computer-basedringer Mannheim, Mannheim, Germany), respectively,

were employed to obtain run off transcripts of either the morphometric analysis system in 6 to 10 randomly se-lected interstitial areas. The video image was generatedantisense (complementary to mRNA) or sense (negative

control) strands. Transcription and labeling of RNA by a video camera (Hamamatsu, Milan, Italy) connectedthrough a frame grabber (Hamamatsu) to a Power PCprobes were performed as described [15]. Briefly, 80 mCi

of [35S]uridine-59-(a-thio)-triphosphate (specific activity, computer (Macintosh, Cupertino, CA, USA). Single im-ages were digitized for image analysis at 256 gray levels.1250 Ci/mmol; Amersham, Little Chalfont, UK) were

added to a 10 mL reaction mixture (0.5 mmol/L each of Imported data were analyzed quantitatively by an Opti-lab Pro 2.6.1 Software (Graftek, Villanterio, PV, Italy),adenosine-, cytosine-, and guanosine-59-triphosphate/

1 mmol/L dithiothreitol/10 U human placental RNase which operated a color-based pitch densitometry. Thestructures of interest were interactively discriminated byinhibitor/6 mmol/L MgCl2/10 mmol/L Tris-HCl, pH 7.5/

2 mmol/L spermidine/10 mmol/L NaCl), including 1 mg the operators using the cursor and then automaticallymeasured for total area. An optical threshold and filterof linearized plasmid and 16 U of either SP6 or T7 RNA

polymerase. The reaction was allowed to proceed for 60 combination were set to select only the nuclei of CD68-positive interstitial cells, the interstitial fibrosis (evalu-minutes at 388C. The plasmid DNA was removed by

digestion with 25 mg/mL RNase-free DNase I in a mix- ated by Masson’s Trichromic staining), the new Fuchsindeposits, or the silver grains.ture containing 2.5 mg/mL of yeast tRNA and 10 U of

RNase inhibitor for 10 minutes at 378C. Free ribonucleo- CD68-positive cells nuclei were then counted and ex-pressed in the number of cells per area, while areas oftides were removed by phenol-chloroform extraction fol-

lowed by ethanol precipitation. RNA probes were then green staining (interstitial fibrosis as identified by Mas-son’s trichromic) were extracted, quantitated, and ex-diluted in hybridization buffer, stored at 2808C, and

used within four weeks. The specific activity usually ob- pressed as a percentage of the total interstitial area.MCP-1 and EGF expression (silver grains) were quanti-tained was 1.2 to 1.4 3 109 cpm/mg of 35S-labeled RNA

probe. tated by computer-assisted densitometric procedure andwere expressed as arbitrary units (A.U.)/pixel. Finally,Prehybridization, hybridization, removal of non spe-

cifically bound probe by RNase A digestion, and further tubular damage was evaluated semiquantitatively in ablind fashion by three independent observers using awashing procedures were performed for both sense and

antisense probes as described previously [14, 15, 18]. scale ranging from 0 to 4, on the basis of the percentageof atrophic and/or dilated tubular sections. Using theAutoradiography was performed by dipping the dehy-

drated slides into Ilford G5 nuclear emulsion (Ilford, Statview II Software (Brain Power Inc., Calabasas, CA,USA), the individual data for each variable were tabu-Mobberley Cheshire, UK). The exposed slides were de-

veloped using Kodak D19 developer (Kodak, Hemel lated, and Pearson’s correlation coefficients were calcu-lated.Hampstead, UK), counterstained in hematoxylin and eo-

sin, and finally mounted.Enzyme-linked immunosorbent assay (ELISA)

Immunohistochemistry Urine samples were collected from the patients in-cluded in the study before surgery. In 14 congenitalThe immunohistochemical detection of infiltrating

monocytes was performed on frozen 4 mm thick kidney UPJO patients, a second urine sample was obtained be-tween the second and fourth months after surgery. Urinesections and was fixed in 4% paraformaldehyde, using

a chromatographically purified mouse anti-CD68 mono- samples were also obtained from 15 healthy controlsstrictly matched for age. Urine samples were centrifugedclonal antibody, specific for human monocytes (Clone

EBM11; Dako, Milan, Italy) as previously described [14]. at 600 3 g for five minutes and frozen at 2208C untiltested. Quantitative determination of MCP-1 and EGFImmobilized mouse antibodies were detected by the im-

munoalkaline phosphatase (APAAP) method with af- urine concentration were performed using a humanMCP-1 and EGF ELISA kit (Quantikine; R&D, Abing-finity-purified rabbit antimouse IgG antibodies (Dako)

and APAAP complex (1:50 dilution; Dako) following don, UK, and Biotrak, Amersham, UK, respectively), amultiple sandwich solid-phase enzyme immunoassay thata two-step technique as previously described. Alkaline

phosphatase was developed with New Fuchsin (Sigma, used monoclonal antibodies raised against humanMCP-1 and EGF. The sensitivity of the ELISA wasMilan, Italy). Negative controls were performed by omit-

ting the primary or secondary antibodies and employing 5 pg/mL for MCP-1 and 8 pg/mL for EGF. The enzymaticreaction was detected in an automatic microplate pho-nonimmune mouse antiserum as first layer.

The histologic lesions and the optical density of the tometer (Titertek; Flow Labs, Helsinki, Finland). TheMCP-1 and EGF concentration of the unknown samplessignals generated by the periodic acid-Schiff and Mas-

son’s Trichromic staining (light microscopy), New Fuch- was determined by interpolation into a standard curve

Grandaliano et al: Markers of obstructive nephropathy 185

Fig. 1. In situ hybridization for epidermal growth factor (EGF). Bright (A and C) and dark (B and D) field microphotographs of normal appearingkidney (A and B, 3400) and of biopsy specimens from patients with congenital ureteropelvic junction obstruction (UPJO; C and D, 3400) areshown. In normal kidney, a very high EGF mRNA expression can be observed, whereas in the UPJO sections, there is a striking decrease of thespecific signal.

developed with known amounts of recombinant human genesis of tubular damage [18, 24], the renal EGF geneexpression was evaluated by in situ hybridization in 16MCP-1 and EGF. Urine MCP-1 and EGF protein levelsbiopsy sections of congenital UPJO. Control kidneyswere normalized to urine creatinine excretion and wereshowed a very high basal expression of this growth factorexpressed as pg or ng/mg urine creatinine, respectively.(Fig. 1 A, B). The specific hybridization signal was mainly

Statistical analysis located within the loop of Henle and the distal tubules,as we have previously demonstrated [15]. In UPJO, theData were expressed as mean 6 SEM. Quantitativeintensity of the signal was strikingly reduced when com-data were compared among the groups by analysis ofpared with control kidneys (Fig. 1 C, D) and was inverselyvariance and paired t-test, as appropriate. The correla-correlated with the degree of tubular damage (r 5 0.65,tion coefficients are Pearson’s r values. Differences wereP , 0.01).considered to be significant when P , 0.05.

Recently, several studies reported that interstitial in-flammation is a constant feature in experimental models

RESULTS of chronic obstructive uropathy [6, 7, 11]. The inflamma-Tubular damage, mainly characterized by tubular dila- tory infiltrate has been characterized as constituted

tion and tubular cell atrophy, is one of the histologic mainly by monocytes and T lymphocytes [6, 7]. Thus,hallmarks of chronic obstructive nephropathy [1, 23]. the presence of interstitial monocytes in the biopsy coresSince in other acute and chronic tubulointerstitial dis- of congenital UPJO was investigated by immunohisto-eases the modulation of local EGF production and re- chemistry, using an anti-CD68 antibody, which preferen-

tially recognizes the monocytes coming from the circula-lease has been suggested to be directly involved in patho-

Grandaliano et al: Markers of obstructive nephropathy186

Fig. 2. Immunohistochemistry for CD68 in human renal biopsy specimens from normal (A, 3200) and obstructed kidneys (B–D, 3400). Sectionsdemonstrate a diffuse interstitial infiltrate primarily represented by CD68-positive cells.

tion rather than the tissue macrophages. In UPJO, an (Fig. 3 C–F), and was directly correlated with the extentof monocyte infiltration (r 5 0.90, P , 0.001). To confirminterstitial mononuclear cell infiltrate was present in each

biopsy and was mostly represented by CD68-positive the relevance of MCP-1 expression in the pathogenesisof interstitial monocyte influx, sequential sections werecells (Fig. 2). The glomeruli were free of specific staining

in all of the sections studied. The extent of monocyte stained for CD68 and were hybridized for MCP-1 (Figs.2D and 3 E–F). As shown, the tubules expressing MCP-1infiltration was strictly correlated with the degree of tu-

bulointerstitial damage, as expressed by tubular atrophy are surrounded by a CD68-positive cell infiltrate,strongly supporting a role for this chemokine in intersti-and interstitial fibrosis.

Diamond et al demonstrated an increase in MCP-1 tial monocyte recruitment.The reduced renal expression of EGF paralleled withprotein expression in an experimental model of hydrone-

phrosis and suggested this chemokine as the main media- its urine excretion evaluated before the surgical correc-tion of congenital UPJO. The EGF urine concentration,tor of monocyte influx into the interstitium [6]. Since no

similar observation has been reported in human obstruc- indeed, was significantly lower in UPJO then in healthysubjects (Fig. 4A), although no significant correlationtive nephropathy, the MCP-1 gene expression was evalu-

ated by in situ hybridization in the biopsy sections of could be observed between the renal functional parame-ters obtained by MAG3 scan and the EGF concentrationcongenital UPJO patients. The specific hybridization sig-

nal for this chemokine was barely detectable in the nor- measured either in urine spot or in the urine collecteddirectly from the obstructed pelvis.mal kidneys (Fig. 3 A, B). On the other hand, MCP-1

mRNA was markedly increased in the obstructed kid- Also, MCP-1 urine excretion mirrored the renal geneexpression. A striking increase of this chemokine in theneys (Fig. 3 C, D). The specific message was mainly

localized within the tubules, the infiltrating mononuclear urine was observed in the samples collected from thepatients prior to surgery (Fig. 4B). MCP-1 urine concen-interstitial cells and the endothelial cells of small vessels

Grandaliano et al: Markers of obstructive nephropathy 187

Fig. 3. In situ hybridization for monocyte chemotactic peptide-1 (MCP-1). A bright (A, C, and E) and dark (B, D, and F ) field microphotographof normal appearing kidney (A and B, 3200) and of biopsy specimens from patients with congenital UPJO (C–F, 3200) is shown. Normal kidneydoes not express MCP-1 mRNA, whereas in the UPJO sections, there is a striking increase of the specific signal, mainly localized at thetubulointerstitial level.

tration was significantly and directly correlated with the Ten of the 24 congenital UPJO patients enrolled inthe study presented with a history of recurrent UTI,mean time of transit of 99Tc-MAG3 through the cortex

of the obstructed kidney (Fig. 5A). Interestingly, the while 10 of them had no history of UTI. The recurrentUTI group was characterized by a MAG3 clearance sig-MCP-1 concentration in the urine collected directly from

the obstructed pelvis was inversely correlated with the nificantly lower than the group without history of UTI(UTI 40.0 6 8.7%; without UTI 59.3 6 3.6%, P 5 0.03).MAG3 clearance of the diseased kidney (Fig. 5B).

Grandaliano et al: Markers of obstructive nephropathy188

Fig. 4. EGF (A) and MCP-1 (B) urine excretion in healthy subjectsFig. 5. Correlation between MCP-1 excretion and functional parame-and in patients affected by congenital UPJO. MCP-1 and EGF urineters obtained by MAG3 scan. MCP-1 urine concentration was signifi-concentrations were measured as described in the Methods section andcantly and directly correlated with the mean time of transit (MTT) ofwere normalized to urine creatinine excretion. Data are represented asthe radioactive tracer through the cortex of the obstructed kidney (A;mean 6 SEM. *P , 0.01 vs. control.r 5 0.565, P , 0.03). On the other hand, the MCP-1 concentration inthe urine collected directly from the obstructed pelvis (pMCP-1) wasinversely correlated with the MAG3 clearance of the diseased kidney(B; r 5 0.769, P , 0.01).

EGF urine excretion, although slightly reduced, was notsignificantly lower in the recurrent UTI group when com-pared with the patients without a history of UTI (Fig. ance of the obstructed kidney (Fig. 7) and a statistically6A). MCP-1 urine concentration was significantly higher significant increase in EGF urine excretion that returnedin the group of patients with a history of recurrent UTI to normal levels within four months after surgery (Fig.than in the group with no history of UTI (Fig. 6B). 8A). MCP-1 excretion was markedly reduced after sur-However, in the latter group, MCP-1 urine concentration gery (Fig. 8B), reaching normal levels four months afterwas still higher than in the normal subjects (Fig. 6B). To the surgical correction of the obstruction. Interestingly,exclude the possibility of an increased production of EGF urine excretion measured before the surgical cor-this chemokine by epithelial cells along the urinary tract rection of UPJO was directly and significantly correlatedrather than by tubulointerstitial cells, renal MCP-1 ex- with the difference between MAG3 clearance of thepression was quantitated in the two groups of patients. obstructed kidney before and after surgery (r 5 0.90,MCP-1 tubulointerstitial expression in the patients with P , 0.001).recurrent UTI was significantly higher than in patientswithout infections (no UTI 60 6 9, UTI 160 6 19 A.U./

DISCUSSIONpixel, P , 0.001).Interestingly, the surgical correction of congenital Obstructive uropathy accounts for 2% of the end-stage

renal disease (ESRD) patients in renal replacement ther-UPJO was followed by an improvement in MAG3 clear-

Grandaliano et al: Markers of obstructive nephropathy 189

Fig. 7. MAG3 clearance in obstructed ( ) and nonobstructed (j)kidneys before and after surgical correction of congenital UPJO.MAG3 clearance is expressed as the percentage of radioactive tracerexcreted by each kidney. *P , 0.01 vs. obstructed kidney after surgeryand nonobstructed kidney before and after surgery.

injury in the animals is always synchronized, whereas inhuman studies the investigators have to deal with a cross-section picture in which the lesions are often representedat different stages in the different subjects studied. Nev-ertheless, confirming in humans the molecular informa-tion collected from the disease in experimental modelsis a necessary and unavoidable step. To our knowledge,the present study is the first attempt to elucidate the

Fig. 6. EGF (A) and MCP-1 (B) urine excretion in congenital UPJO molecular mechanisms involved in the development ofpatients with and without a history of recurrent urinary tract infection tubular damage and interstitial inflammation in human(UTI). MCP-1 and EGF urine concentrations were measured as de-

obstructive uropathy caused by congenital UPJO.scribed in the Methods section and were normalized to urine creatinineexcretion. Data are represented as mean 6 SEM. *P , 0.01 vs. control; The modulation of tubular epithelial cell trophism has**P , 0.01 vs. no UTI. been extensively studied in the experimental model of

obstructive uropathy. Gobe and Axelsen were the firstto observe the apoptosis of tubular cells in a rat model

apy in the U.S. Renal Data System, and this cause of of hydronephrosis [27]. Successively, Truong et al re-chronic renal failure is particularly relevant among the ported a rapid increase in tubular cell apoptosis associ-pediatric population [25, 26]. Although the etiology and ated with a decrease in kidney dry weight and meanthe acute hemodynamic and functional abnormalities are tubular diameter, suggesting apoptosis as a central eventwell characterized in most of the cases, the pathogenesis in the pathogenesis of tubular atrophy and renal weight

loss [28]. EGF may play a key role in this scenario. Thisof the chronic renal damage secondary to the urinarytract obstruction is still largely unknown [2–4]. growth factor has been suggested as a powerful trophic

factor for tubular epithelial cells [16, 29]. Indeed, it di-Several studies investigated the cellular and molecularmechanisms underlying the development of renal lesions rects renal cell proliferation during fetal life and inhibits

in vivo and in vitro tubular cell apoptosis [29–32]. More-induced by chronic urinary tract obstruction in an experi-mental model created by unilateral ureter ligation [3]. over, several studies have demonstrated the pathogenic

involvement of EGF in acute and chronic forms of renalHowever, the human disease is more complex than itsexperimental counterpart. Indeed, while the influence of damage [17–19]. Finally, the administration of recombi-

nant EGF has been shown to reduce tubular necrosissingle or recurrent UTI is excluded in the animal modelof hydronephrosis, this complication is often present in and/or apoptosis and to improve tubular cell regenera-

tion in different models of tubular injury, including ex-human obstructive uropathy. Moreover, as for other ex-perimental models of renal disease, development of the perimental hydronephrosis [20–22, 33].

Grandaliano et al: Markers of obstructive nephropathy190

clearance of the obstructed kidney after surgery. Inter-estingly, the local decrease in EGF production seems tobe primarily induced by the obstruction per se. Indeed,the surgical correction of the congenital UPJO resultedin a marked increase in EGF excretion that returned tonormal levels within four months from surgery. How-ever, EGF excretion may be also influenced, at least inpart, by the presence of recurrent infections, as suggestedby the observation of a lower, although not statisticallydifferent, EGF urine concentration in the patients witha history of recurrent UTI.

A histologic feature of obstructive uropathy that isreceiving more and more attention in the last few years isthe interstitial inflammatory infiltrate [7]. In our patientpopulation, this infiltrate is mainly constituted by mono-cytes, confirming the observation of Schreiner et al inexperimental hydronephrosis [7]. Monocytes, once with-in the interstitial space, may represent a source of cyto-kine and growth factors that can prime and maintain theactivation of resident cells [8]. Diamond et al demon-strated an increase in the amount of cortical transforminggrowth factor-b (TGF-b) mRNA and protein that paral-leled the elevations in interstitial macrophage numberafter unilateral obstruction [6]. The expression of thisprofibrotic cytokine was mainly localized within the in-terstitial infiltrate. This observation strongly supports therelevance of monocytes in the development of interstitialfibrosis. Moreover, Harris, Schreiner, and Klahr haveclearly shown that these infiltrating immune cells con-tribute to the acute decline in renal function after theonset of experimental hydronephrosis [11].

The chemotactic signals recruiting the monocytes intothe renal interstitium of the obstructed kidney are stillnot well characterized. We observed a striking increaseof MCP-1 gene expression and urine excretion in con-genital UPJO patients. The renal expression of this che-mokine was directly correlated to the extent of monocyteFig. 8. EGF and MCP-1 urine excretion in congenital UPJO patients

before and after surgery. MCP-1 and EGF urine concentrations were infiltration, suggesting its potential role in the recruit-measured before and between two and four months after surgery, as ment of these immune cells into the interstitial space.described in the Methods section and were normalized to urine creati-

This observation confirms the report by Diamond et alnine excretion. Data are represented as mean 6 SEM. *P , 0.01 vs.control; **P , 0.01 vs. presurgery. in a rat model of unilateral ureteral obstruction, who

described an increase in MCP-1 expression within thecortical tubules as early as 12 hours after the UUO thatwas in temporal correspondence with the surge of inter-

The present study demonstrates a decreased EGF re- stitial monocyte number [6]. However, in their study,nal expression and urine excretion in children affected MCP-1 expression was localized only by immunohisto-by congenital UPJO, confirming previous observations chemistry on the luminal surface of tubular cells. Thesein experimental models [19]. The potential pathogenic findings may raise some doubts about the tubular pro-role for EGF in renal damage secondary to UPJO is duction of the chemokine. The data collected in thesuggested by the observation of a significant correlation present study, using the in situ hybridization, excludebetween its gene expression and the degree of tubular the possibility of a nonspecific uptake of MCP-1 proteindamage. Moreover, our data suggest a role for EGF by tubular cells, confirming its local synthesis.in tubular cell regeneration after surgical correction of For both EGF and MCP-1, at least two variables couldUPJO. Indeed, basal EGF urine excretion strictly and influence their expression and urine excretion in UPJO

patients: obstruction and recurrent urinary tract infec-directly correlated with the improvement in MAG3

Grandaliano et al: Markers of obstructive nephropathy 191

2. Klahr S: Pathophysiology of obstructive nephropathy. Kidney Inttion. In contrast with EGF, MCP-1 excretion was signifi-23:414–426, 1983cantly higher in the group of patients with recurrent UTI 3. Klahr S: Obstructive nephropathy. Kidney Int 54:286–300, 1998

then in the UPJO patients with no history of infection. 4. Klahr S, Purkerson ML: The pathophysiology of obstructivenephropathy: The role of vasoactive compounds in the hemody-This observation strongly supports the hypothesis thatnamic and structural abnormalities of the obstructed kidney. AmUTI induces the expression of MCP-1. Indeed, bacterial J Kidney Dis 23:219–223, 1994

infection could increase MCP-1 expression by resident 5. Harris KPG, Klahr S, Schreiner G: Obstructive nephropathy:From mechanical disturbance to immune activation? Exp Nephrolcells either directly, through the effect of lipopolysaccha-1:198–204, 1993ride (LPS), or indirectly, through the action of tumor

6. Diamond JR, Kees-Folts D, Ding G, Frye JE, Restrepo NC:necrosis factor-a (TNF-a), interleukin-1 (IL-1), and in- Macrophages, monocyte chemoattractant peptide-1 and TGF-b1 in

experimental hydronephrosis. Am J Physiol 266:F926–F933, 1994terferon-g (IFN-g) released locally by infiltrating im-7. Schreiner GF, Harris KPG, Purkerson ML, Klahr S: Immuno-mune cells [12, 13]. It is conceivable that the increased

logical aspects of acute ureteral obstruction: Immune cell infiltrateurinary MCP-1 in the patients with recurrent infection in the kidney. Kidney Int 34:487–493, 1988could be due to an increased production of this chemo- 8. Nathan CF: Secretory products of macrophages. J Clin Invest

79:319–326, 1987kine by epithelial cells along the urinary tract rather9. Van Goor H, Ding G, Kees-Folts D, Grond J, Schreiner GF,than by tubular and interstitial cells. However, MCP-1

Diamond JR: Macrophages and renal diseases. Lab Invest 71:456–tubulointerstitial expression in these patients was sig- 464, 1994

10. Eddy AA: Interstitial macrophages as mediators of renal fibrosis.nificantly higher than in patients without infections. Al-Exp Nephrol 3:76–79, 1995though infection may influence MCP-1 production, the

11. Harris KPG, Schreiner GF, Klahr S: Effect of leukocyte deple-obstruction per se induces this chemokine expression tion on the function of the post-obstructed kidney in the rat. Kidneyand excretion, as suggested by the significant decrease Int 36:210–215, 1989

12. Stahl RA: Chemoattractants cytokines (chemokines) and immunein urine MCP-1 concentration following surgical correc-renal injury. Nephrol Dial Transplant 10:307–309, 1995tion of congenital UPJO. 13. Leonard EJ, Yoshimura T: Human monocyte chemoattractant

In addition, it is interesting to note that MCP-1 excre- protein-1. Immunol Today 11:97–101, 199314. Grandaliano G, Gesualdo L, Ranieri E, Monno R, Montinarotion significantly correlated with the MAG3 clearance

V, Marra F, Schena FP: Monocyte chemotactic peptide-1 expres-in UPJO patients, suggesting a link between the inflam-sion in acute and chronic human nephritides: A pathogenetic rolematory response and the tubular function and the possi- in interstitial monocytes recruitment. J Am Soc Nephrol 7:906–913,

bility of a direct involvement in the progression of renal 199615. Gesualdo L, Di Paolo S, Calabro A, Milani S, Maiorano E,damage in UPJO. This observation, once confirmed in

Ranieri E, Pannarale G, Schena FP: Expression of epidermala larger patient population, could suggest the use ofgrowth factor and its receptor in normal and diseased human

the urinary level of this chemokine as a diagnostic and kidney: An immunohistochemical and in situ hybridization study.Kidney Int 49:656–665, 1996prognostic marker.

16. Carpenter G, Cohen S: Epidermal growth factor. Annu Rev Bio-In conclusion, the present study suggests a role forchem 48:193–216, 1979MCP-1 and EGF in the development of tubulointerstitial 17. Safirstein R, Zelent AZ, Price PM: Reduced renal prepro-epi-

lesions in human congenital obstructive uropathy. Since dermal growth factor messenger RNA and decreased EGF excre-tion in ARF. Kidney Int 36:810–815, 1989MCP-1 expression could be modulated in vivo by cortico-

18. Ranieri E, Gesualdo L, Petrarulo F, Schena FP: Urinarysteroids [34] and local EGF depletion could be effec-IL-6/EGF ratio: A useful prognostic marker for the progression

tively corrected by systemic administration of the recom- of renal damage in IgA nephropathy. Kidney Int 50:1990–2001,binant protein [20–22, 33], these two peptides may be 1996

19. Storch S, Saggi S, Megyesi J, Price PM, Safirstein R: Ureteralalso considered to be potential therapeutic targets inobstruction decreases renal prepro-epidermal growth factor andhuman obstructive nephropathy. Tamm-Horsfall expression. Kidney Int 42:89–94, 1992

20. Humes HD, Cielinsky DA, Coimbra TM, Messana JM, GalvaoC: Epidermal growth factor enhances renal tubule cell regenerationACKNOWLEDGMENTSand repair and accelerates the recovery of renal function in post-

This study was partly supported by the following grants: Baxter ischaemic acute renal failure. J Clin Invest 84:1757–1761, 1989Extramural project (eight round, 1996–1998), Associazione per il Pro- 21. Kennedy WA, Buttyan R, Garcia-Montes E, D’Agati V, Ols-gresso Scientifico in Nefrologia e Trapianto (APSNT), Consiglio Nazio- son CA, Sawczuk IS: Epidermal growth factor suppresses renalnale delle Ricerche (C.N.R.; 98.513CT04), CNR-Target Project on

tubular apoptosis following ureteral obstruction. Urology 49:973–“Biotechnology,” Ministero dell’ Universita e della Ricerca Scientifica980, 1997e Tecnologica (60%). The authors acknowledge the skillful technical

22. Chevalier RL, Goyal S, Wolstenholme JT, Thornhill BA: Ob-help of Miss Tiziana Colucci, with the technical assistance of Mrs.structive nephropathy in the neonatal rat is attenuated by epider-Carmela Martino as well.mal growth factor. Kidney Int 54:38–47, 1998

23. Moller JC, Jorgenson TM, Mortensen J: Proximal tubular atro-Reprint requests to Loreto Gesualdo, M.D., Division of Nephrology,phy: Qualitative and quantitative structural changes in chronicDepartment of Emergency and Transplantation, University of Bari,obstructive nephropathy in the pig. Cell Tissue Res 244:479–491,Polyclinic, Piazza Giulio Cesare, 11, 70124 Bari, Italy.

E-mail: l.gesualdo@nephro.uniba.it 198624. Di Paolo S, Gesualdo L, Stallone G, Ranieri E, Schena FP:

Renal expression and urinary concentration of EGF and IL-6 inREFERENCESacutely dysfunctioning kidney transplanted patients. Nephrol DialTransplant 12:2687–2693, 19971. Nagle RB, Bulger RE: Unilateral obstructive nephropathy in the

rabbit. II. Late morphologic changes. Lab Invest 38:270–278, 1978 25. United States Renal Data System: Annual data report. II. Inci-

Grandaliano et al: Markers of obstructive nephropathy192

dence and prevalence of ESRD. Am J Kidney Dis 18(Suppl):S34– 30. Cybulsky AV, Goodyer PR, McTavish AJ: Epidermal growthfactor receptor activation in developing rat kidney. Am J PhysiolS47, 1996267:F428–F436, 199426. Kohaut EC, Tejani A: The 1994 annual report of the North Ameri-

31. Coles HSR, Burne JF, Raff MC: Large-scale normal cell deathcan Pediatric Renal Transplant Cooperative Study. Pediatr Neph-in the developing rat kidney and its reduction by epidermal growthrol 10:422–434, 1996factor. Development 118:777–784, 199327. Gobe GC, Axelsen RA: Genesis of renal tubular atrophy in exper-

32. Hammerman MR: Growth factors in renal development. Seminimental hydronephrosis in the rat: Role of apoptosis. Lab InvestNephrol 15:291–299, 199556:273–281, 1987 33. Coimbra TM, Cielinsky DA, Humes HD: Epidermal growth factor

28. Truong LD, Petrusevska G, Yang G, Gurpinar T, Shappel S, accelerates renal repair in mercuric chloride nephrotoxicity. AmLechago J, Rouse D, Suki WN: The kinetics of cell apoptosis and J Physiol 259:F438–F443, 1990proliferation in experimental chronic obstructive uropathy. Kidney 34. Noris M, Bernasconi S, Casiraghi F, Sozzani S, Gotti E, Re-Int 50:200–207, 1996 muzzi G, Mantovani A: Monocyte chemotactic protein-1 is ex-

29. Harris RC: Potential physiologic roles for epidermal growth factor creted in excessive amounts in the urine of patients with lupusnephritis. Lab Invest 73:804–809, 1995in the kidney. Am J Kidney Dis 17:627–630, 1991