OOefH

T

Fmr

FoMS

Sr

RSSn

1

©

d

2

RIGINAL ARTICLESsteopontin as an injury marker expressing inpithelial hyperplasia lesions helpful in prognosis of

ocal segmental glomerulosclerosisAO-AI SHUI, SHUK-MAN KA, SHUN-MIN YANG, YUH-FENG LIN, YI-FEN LO, and ANN CHEN

AIPEI, TAIWAN, ROC

Focal segmental glomerulosclerosis (FSGS) is characterized by typical sclerosis butalso shows other non-sclerotic lesions that provide prognostic informations. The glo-merular epithelial hyperplasia lesion (EPHL) that develops earlier than the scleroticlesions is a key determinant of progression of FSGS. However, the relationship amongEPHL, glomeular sclerosis, and macrophage infiltration in FSGS is unclear, and theEPHL-associated markers helpful for prognosis of FSGS have still not been completelyidentified. Here, we performed clinicopathologic, immunochemical, and molecularanalyses to examine whether osteopontin (OPN), a macrophage chemokine, is aninjury marker of EPHLs correlating with glomerular sclerosis and macrophage mobili-zation. First, the FSGS model was induced in Balb/c mice by a single injection ofadriamycin, and consecutive sclerosis changes were evaluated. In parallel, we usedreverse transcription-polymerase chain reaction and Western blot analyses to deter-mine levels of OPN in isolated glomeruli and urine, respectively. Immunohistochemistrywas applied to assess the OPN expression in EPHLs and macrophage infiltration aroundthe glomeruli. Our results showed that, within glomeruli, OPN expressed restrictedlywithin EPHL; the OPN mRNA and protein of glomeruli increased on day 11, correlatingwell with the early EPHL, and following sclerosis and macrophage infiltration. In addition,immunohistochemistry (IHC) staining of OPN greatly highlighted early glomerular EPHLs,helping microscopic identification of EPHLs. We propose that the OPN expression inEPHLs could contribute to the progression of FSGS by recruiting macrophage toward thecompromised glomeruli. Detection of OPN in glomeruli and urine could be helpful inprognosis of FSGS. (Translational Research 2007;150:216–222)

Abbreviations: ANOVA � analysis of variance; BSA � bovine serum albumin; EPHL � epithelialhyperplasia lesion; FSGS � Focal segmental glomerulosclerosis; HE � hematoxylin and eosin;IHC � immunohistochemistry; OPN � osteopontin; PBS � phosphate-buffered saline; RT-PCR �

reverse transcription-polymerase chain reaction; TBST � Tris-buffered saline containing Tween

rggacs

hoalgborH

ocal segmental glomerulosclerosis (FSGS) showssclerosis of some glomeruli (focal disease) or ofparts of a glomerular capillary tuft in some glo-

eruli (segmental disease).1 Although glomerular scle-otic lesion is a key feature of FSGS, other non-scle-

rom the Graduate Institute of Medical Sciences; the Departmentf Pathology, Tri-Service General Hospital; the Department of Internaledicine, Tri-Service General Hospital; and Graduate Institute of Life

ciences, National Defense Medical Center, Taipei, Taiwan, ROC.

ubmitted for publication October 20, 2006; revision submitted Feb-uary 28, 2007; accepted for publication April 4, 2007.

eprint requests: Dr. Ann Chen, Department of Pathology, Tri-ervice General Hospital, National Defense Medical Center, No. 325,ec. 2, Cheng-Gung Road, Taipei, Taiwan, ROC; e-mail: doc31717@dmctsgh.edu.tw.

931-5244/$ – see front matter

2007 Mosby, Inc. All rights reserved.

soi:10.1016/j.trsl.2007.04.003

16

otic histopathologic lesions can also appear inlomeruli and affect progression of FSGS.2 Withinlomeruli, molecular markers that express prominentlynd restrictedly in the glomerular non-sclerotic lesionsould help histologically identify the lesions and giveignificant prognostic information.

Epithelial hyperplasia lesion (EPHL), a non-scleroticistologic feature, has been frequently observed in bi-psies of FSGS patients and tissue sections of FSGSnimal models.2–7 Some EPHLs in FSGS share histo-ogic features with the crescent of typical crescenticlomerulonephritis2–7; the EPHL observed in FSGS haseen considered to be a key determinant of progressionf FSGS,3,4,6 implicated by the well-known pathogenicole of crescent in glomerular sclerosis and fibrosis.8–10

owever, EPHLs in FSGS usually show focal and

egmental distribution, which make the use of the

Eimmwiig

moaisimdshritOmc(m

iiecmcp

M

mTa(fwrtdwatOisPp

tcL

sc((suuLaA

bttd(k

lepgw1

www2rwrsfcB(td

mbp(li1zplg

f

Translational ResearchVolume 150, Number 4 Shui et al 217

PHLs as a histologic feature for prognosis of FSGSmpractical. Furthermore, unlike typical crescentic glo-erulonephritis, FSGS is not classified as an immune-ediated acute nephropathy1,11; it is still unclearhether the pathogenic mechanism of EPHLs in FSGS

s similar to that in typical crescent glomerulonephritisn terms of mechanisms for macrophage infiltration andlomerular sclerosis and fibrosis.10

Osteopontin (OPN), a well-known monocyte che-oattractant, has been regarded as a crescent biomarker

f immune-mediated acute nephritis in patients andnimal models,8,9,12 in which it plays a pathogenic rolen the formation of crescentic lesions and the progres-ion of the nephritis through its chemotactic property tonduce macrophage infiltration.8–10 However, OPN isore appropriate to act as an injury marker than as a

isease-specific biomarker, as OPN upregulation is ob-erved in several kidney lesions such as glomerularypertension-induced and immune-induced renal inju-ies,13–15 and a variety of physiological and patholog-cal conditions such as myocardiac infarction and ven-ricular remodeling.16,17 So far, the equivalent role ofPN in serving as a pathogenic factor and/or an injuryarker of FSGS is still unclear, despite that signifi-

antly increased levels of secreted phosphoprotein 1ie, OPN) has been noticed in the kidney of an adria-ycin-induced FSGS mouse model.18

In the current study, we performed clinicopathologic,mmunohistochemical, and molecular analyses to clar-fy the relationships among EPHL formation, OPNxpression, and macrophage infiltration during the timeourse of FSGS in an adriamycin-induced FSGS mouseodel. We conclude that OPN is an injury marker that

ould help histologically identify EPHLs and estimaterognosis of FSGS.

ETHODS

Treatments of animals and analyses of samples. Experi-ents were performed on 8-week-old female BALB/c mice.he mice were injected intravenously with a single dose ofdriamycin (0.1 mg/10 g body weight) or normal salinecontrol group) as described previously.19 Based on the dif-erent severity of proteinurea, days 0, 4, 7, 11, 15, and 20ere chosen for collecting blood and urine samples through

etroorbital venous plexus and by metabolic cages, respec-ively.19 Total urine protein and urine OPN were measured onays 0, 4, 7, 11, 15, and 20 after adriamycin treatment. Miceere killed on the same sampling days (n � 7 for each day),

nd the kidneys were subjected to histopathologic examina-ions and processed for immunohistochemistry (IHC) forPN expression and macrophage infiltration. Glomeruli were

solated with a sieving technique to analyze the OPN expres-ion by reverse transcription-polymerase chain reaction (RT-CR). All animal experiments were performed with the ap-

roval of the Institutional Animal Care and Use Committee of v

he National Defense Medical Center, Taiwan, and wereonsistent with the NIH Guide for the Care and Use ofaboratory Animals.Measurement of urine protein, serum urea nitrogen, and

erum creatinine. Samples of urine or serum were mi-rofuged and stored at �70°C until analyzed. Urinary proteinmg/dL) was measured by the bicinchoninic acid methodBCA protein assay kit; Pierce, Rockford, Ill) with bovineerum albumin (BSA) as the standard and normalized byrine creatinine. Serum urea nitrogen was measured with arease assay kit (Sigma 640–5; Sigma Chemical Co., St.ouis, Mo), and serum creatinine was measured with a picriccid colorimetric kit (Sigma 555–1) as described previously.20

ll samples were tested in duplicate.Histopathology and IHC. Renal tissues were fixed in 10%

uffered formalin and embedded in paraffin for routine his-opathologic evaluation. Sections of the formalin-fixed renalissue were immersed in xylene to remove paraffin, rehy-rated in graded ethanol, stained with hematoxylin and eosinHE), and examined with a light microscope (Olympus, To-yo, Japan).For histopathologic evaluations of EPHL and sclerosis, at

east 50 glomeruli in renal tissue sections for each case werexamined. The number of glomeruli with EPHLs was ex-ressed as a percentage of the total number of evaluatedlomeruli as described previously.12 The severity of sclerosisas semi-quantified by morphological changes on a scale of–4, according to a previous published criterion.19

For detection of OPN by IHC, paraffin-embedded tissuesere cut into 5-�m sections. The sections were pretreatedith saponin (0.5%, Sigma) for 5 min and then incubatedith Tris-buffered saline, pH 7.4, containing 0.05% Tween0 (TBST) and 2% BSA for the blocking step (30 min atoom temperature). We used TBST for all of the followingashing steps. Sections were incubated overnight at 4°C with

abbit anti-mouse OPN antibody (1:200 dilution; Assay De-igns, Ann Arbor, Mich), washed with TBST, and incubatedor 1 h at room temperature with horseradish peroxidase-onjugated protein G (Pierce) as described previously.19

ound protein G was visualized with 3,3=-diaminobenzidineDAKO, Carpinteria, Calif), and the slides were lightly coun-erstained with hematoxylin. Intensity of IHC was scored asescribed previously.19

For staining of F4/80, a mouse macrophage marker, we usedethanol Carnoy’s solution-fixed and paraffin-embedded tissue

locks. The sections were subjected to a microwave heatingrocedure before the application of rat anti-F4/80 antibodies1:50 dilution) (Serotec, Raleigh, NC) at 4°C overnight, fol-owed by washing with TBST and incubation with horserad-sh peroxidase-conjugated rabbit anti-rat IgG (DAKO) forh. The sections were then visualized with 3,3=-diaminoben-

idine (DAKO) and counterstained with methyl green. Theeri-glomerular macrophage number was counted by calcu-ating F4/80 positive cells per glomerular cross-section (c/cs) as published previously.21

Isolation of glomeruli. The glomeruli were extractedrom the kidneys with a sieving technique as described pre-

iously.22 Briefly, the kidneys were removed, washed with

saci1tsTm

fisp8(Rot(ACGKiacwpl57sbtCG

1o3b(wr2rbnMrt

ermis

R

oa(cPl(petw0

FFsftc

Translational Research218 Shui et al October 2007

terile phosphate-buffered saline (PBS), and decapsulated,nd the cortex were separated from the medulla. A sample ofortex was cut into 1–2 mm3 blocks, which were ground upn PBS and the suspension passed sequentially through 60,00, and 200 mesh nets and the glomeruli were collected fromhe surface of the 200 mesh net. The glomerular samples wereubjected to total RNA extraction with Trizol reagent (Lifeechnologies, Rockville, Md) for RT-PCR according to theanufacturer’s instruction.Measurement of glomerular OPN mRNA by RT-PCR. For

rst-strand cDNA synthesis, we used 3 �g of total RNA in aingle-round RT reaction, with 2.5 �g of oligo (dT)12–18

rimer, 1 mM dNTPs, 1� first-strand buffer, 0.4 mM DTT,0 units of RNase-Out recombinant ribonuclease inhibitorInvitrogen, Carlsbad, Calif), and 300 units of REFerscript IINase H–reverse transcriptase (Invitrogen) in a total volumef 25 �L. PCR was performed with 0.9 �L of the RT reac-ion mixture, 0.4 �M of primers for either mouse OPNF:5=-CTCGTGCAGGAAGAACAGAAGC-3=; R:5=-GAGTC-AGTCAGCTGGATGAACC-3=) or GAPDH (F:5=-TCCGC-CCTTCTGCCGATG-3=; R:5=-CACGGAAGGCCATGCCA-TGA-3=), 1� PCR buffer, 0.25 mM dNTPs, and 1.5 units oflenTag DNA polymerase (Ab Peptides Inc., St. Louis, Mo)

n a total volume of 15 �L. The suitable cycles for PCRmplification were determined to prevent overamplifying theDNA; optimal numbers of PCR cycle for OPN and GAPDHere 30 and 20, respectively. Amplification of cDNA waserformed by an initial incubation at 94°C for 10 min, fol-owed by the suitable amplication cycles of 94°C for 45 s,8°C for 1 min, and 72°C for 45 s, and a final extension at2°C for 10 min. The PCR products were electrophoreticallyeparated on a 1.5% agarose gel and visualized with ethidiumromide, and then they were quantified with a gel documen-ation system (Bio-CAPT; Vilber-Lourmat, Marne-la-Valléeedex, France). OPN mRNA levels were normalized toAPDH mRNA levels.Western blot analysis. The urine samples were run on a

0% SDS–polyacrylamide gel. The gel was electroblottednto a PVDF membrane, which was incubated for 2 h in0 mL of blocking buffer (5% skim milk dissolved in TBSTuffer) and incubated with rabbit anti-mouse OPN antibody1:1000 dilution, Assay Designs) for 1 h. Blots were thenashed 3 times with TBST buffer and incubated with horse-

adish peroxidase-conjugated goat anti-rabbit antibodies (1:000 dilution; JacksonImmuno, West Grove, Pa) for 1 h atoom temperature. After washing 3 times, the membrane-ound antibody was detected by incubation with chemilumi-escent reagent plus (PerkinElmer Life Sciences, Boston,ass) and captured on X-ray film. Data were presented as the

atio of the density of OPN protein to the creatinine concen-ration of urine.

Statistics. All data are presented as the mean � standardrror of the mean. Statistical analysis was performed withepeated measures analysis of variance (ANOVA). A New-an–Keuls test was carried out when the ANOVA compar-

sons gave a significant result. Differences were considered

ignificant at P � 0.05. o

ESULTS

Clinical manifestations. To confirm the establishmentf the FSGS model, proteinuria was evaluated in thedriamycin-injected mice. Compared with basal levels0.25 � 0.03), a significant increase of creatinine-orrected urine protein levels (1.13 � 0.21) (F � 10.76,

� 0.05) was observed on day 11, and the proteinevels remained high throughout the experimentFig 1, A). Serum urea nitrogen and serum creatinine, 2arameters for evaluating renal function, were alsoxamined. The former showed a significant and persis-ent increase from day 7 (44.1 � 6.3 mg/dL comparedith basal levels of 22.2 � 2.1 mg/dL; F � 10.11, P �.05), whereas the latter showed a significant increase

ig 1. Clinical manifestations in mice with adriamycin-inducedSGS. Three typical clinical parameters of (A) urine protein, (B)erum urea nitrogen, and (C) serum creatinine for evaluating renalunction were measured at various time points after adriamycinreatment. n � 7, *P� 0.05 compared with day 0; **P � 0.01ompared with day 0.

n day 11 compared with basal levels (1.33 � 0.16

m(t

eaw

tec(10

with day

Translational ResearchVolume 150, Number 4 Shui et al 219

g/dL vs 0.13 � 0.04 mg/dL; F � 18.02, P � 0.05)Fig 1, B and C). These clinical manifestations showedhat FSGS nephropathy was successfully established.

Glomerular EPHLs and sclerosis of the FSGS model. Tovaluate glomerular EPHLs and sclerosis in thedriamycin-treated mice, histopathologic examination

Fig 2. Glomerular histopathology in adriamycin-in15, or 20 showing a gradual increase of collapsesegmental sclerosis in the glomeruli. (G) Semi-quanof glomeruli containing EPHLs in tissue sections. Tplots the percentage of the cellular and fibrocellulscores of glomeruli in tissue sections. (I–N) IHC ofinduced FSGS mice on day 0, 4, 7, 11, 15, or 20.glomeruli. (O) Scores for OPN staining in the EPHstages of FSGS. The scores were calculated as descperi-glomerular regions of adriamycin-treated mimacrophage infiltration in and around the glomeglomerular region on day 0, 4, 7, 11, 15, or 20. Or0.05 compared with day 0; **P � 0.01 compared

as performed on kidney sections obtained at different fi

ime points. As shown in Fig 2, A–G, the FSGS modelxhibited a significant and steady increase in the per-entage of glomeruli containing EPHLs from day 11day 11, 19.5 � 7.6; day 15, 29.1 � 5.1; day 20, 39.2 �2.8 compared with 0 � 0 on day 0; F � 10.03, P �.05), with the cellular type predominating over the

GS mice. (A–F) Kidney tissue on day 0, 4, 7, 11,erular tufts, together with increased EPHLs andlot showing the change with time in the percentageow the total percentage of hyperplasia, and the line(H) Semi-quantitative plot showing the sclerosisressions of EPHLs in the glomeruli of adriamycin-s indicate lesions corresponding to EPHLs in the

ngial cells, and podocytes in glomeruli at differentthe Methods section. (P–U) F4/80 positive cells iny 0, 4, 7, 11, 15, and 20. The arrows indicateted. (V) Counts of F4/80 positive cells in peri-gnification of tissue sections, 400�. n � 7, *P �0.

duced FSof glom

titative phe bars shar types.OPN expThe arrowLs, mesaribed in

ce on daruli affeciginal ma

brocellular type on day 11 (14.0 � 2.9 vs 5.5 � 3.2)

(c(2lctmmsnol2lp

frets11bcmt

tSutcle

lPloe1bTts

atiso

92

D

EmcoEtoiOaiOkEuwqaoa

FuiC0

Translational Research220 Shui et al October 2007

P � 0.05, cellular type vs fibrocellular type), and theonverse of the histologic types of EPHL on days 159.9 � 2.0 vs 19.1 � 4.1) and 20 (11.2 � 2.8 and7.2 � 5.5) (each P � 0.05, cellular type vs fibrocel-ular type). This gradual shift in predominance fromellular type to fibrocellular type hyperplasia suggestshat EPHLs could be involved in the progress of glo-erular sclerosis by gradual deposition of extracellularatrix proteins.10 Early segmental sclerosis was occa-

ionally noticed on day 7; however, a statistically sig-ificant increase of segmental sclerosis was observedn day 15 (F � 29.41, P � 0.01) (Fig 2, H), which isater than the onset of increase of EPHL on day 11 (FigG). Besides, mononuclear cell infiltration and focalobular atrophy were occasionally identified in theeri-glomerular regions (Fig 2, E and F).Colocalization of OPN expression and macrophage in-

iltration. Within glomeruli, the expression of OPN wasestricted to the EPHLs (Fig 2, I–N), despite that OPNxpression was also observed in tubular cells aroundhe glomeruli. As shown by Fig 2, O, the stainingcores for OPN in EPHL increased with time on days1, 15, and 20 (day 11, 64.20 � 7.90; day 15, 97.70 �9.50; day 20, 166.30 � 22.40), compared with theasal levels of 1.8 � 0.1 (F � 22.9, P � 0.01 each). Inontrast, OPN expression scores for both podocytes andesangial cells were extremely low as compared with

hat of EPHLs at each time point (Fig 2, O).As demonstrated, F4/80 positive cells were noticed in

he peri-glomerular region on days 11, 15, and 20 (Fig 2,–U). The number of F4/80 positive cells per glomer-lar cross-section (c/gcs) also increased significantly onhe days (day 11, 9.9 � 1.6 c/gcs; day 15, 11.5 � 2.4/gcs; day 20, 15.6 � 3.5 c/gcs) compared with basalevels (day 0, 0.1 � 0.02 c/gcs) (F � 15.07, P � 0.01ach) (Fig 2, V).

Glomerular expressions of OPN mRNA. OPN mRNAevels in the isolated glomeruli were evaluated by RT-CR. As shown in Fig 3, glomerular OPN mRNA

evels were significantly increased in the FSGS modeln day 11 and remained persistently increased to thend of the experiment (day 11, 1.15 � 0.12; day 15,.16 � 0.24; day 20, 1.39 � 0.14), compared with theasal levels of 0.37 � 0.06 (F � 9.77, P � 0.01 each).he increased OPN mRNA levels correlated well with

he OPN protein expression in glomerular EPHLs ashown by IHC in Fig 2, O.

Dynamic changes in urine OPN levels. Western blotnalysis of urine OPN levels at various time points ofhe FSGS model are shown in Fig 4. Unlike the earlyncrease of OPN levels in tissue (Figs. 2, O and 3), theignificant increase of urine OPN levels can only be

bserved at the late stage of FSGS on day 20 (6034 � F

13), compared with basal levels (401 � 119) (F �3.65, P � 0.01).

ISCUSSION

OPN expression was restrictedly associated withPHLs within glomeruli and not obvious in other glo-erular components such as podocytes and mesangial

ells, despite that OPN also expressed in tubular cellsutside the glomeruli (Fig 2, I–O). Both glomerularPHL percentage (Fig 2, G) and glomerular OPN pro-

ein levels (Fig 2, O) increased significantly as early asn day 11 and continuously increased until the day 20n the FSGS animal model. Likewise, an increase ofPN mRNA expression in the isolated glomeruli was

lso detected on day 11 of the FSGS model (Fig 3), andts increase coincided with the increase tendency ofPN protein shown by IHC (Fig 2, O). As EPHL is aey determinant of progression of FSGS,3,4,6 thePHL-associated OPN could serve as an injury markerseful for indicating the presence of EPHLs (day 11),hich in turn might be helpful in predicting the subse-uent appearance of sclerotic lesions (day 15). Besides,s the significant increase of urinary OPN can only bebserved on day 20 (Fig 4), urinary OPN could act asnoninvasive indicator for entering the late stage of

ig 3. Glomerular OPN mRNA expression. (A) The RT-PCR prod-cts for OPN and GAPDH at different time points after adriamycinnjection are shown in the upper and lower panels, respectively. (B)hange in the OPN/GAPDH mRNA ratio with time. n � 7, **P �.01 compared with day 0.

SGS. However, it should be noticed that part of uri-

nuntdtn

1cabisit(iicsoAacpf

evsrsF

tuEt(oosTfisEmsEpiau

uOoEccmrpfpcmsuimcmgcbn1g2

Fil*v

Translational ResearchVolume 150, Number 4 Shui et al 221

ary OPN might come from tubular cells showing OPNpregulation (Fig 2, I–N); the urine OPN representedot only OPN levels of EPHLs, but also those ofubular cells.13 The increase of OPN in urine could helpifferentiate FSGS from glomerular diseases of otherypes, as urinary OPN was reported to decrease in IgAephropathy.23

Although cellular type EPHL appeared early on day1 (Fig 2, G) and this characteristic histologic featureould be helpful for the prognosis of FSGS, the lesionsre easily overlooked by light-microscope examinationecause of their focalized distribution and morpholog-cal similarity to normal parietal epithelial cells. Ourtudy showed that IHC staining of OPN that expressesn the EPHLs of glomeruli can accentuate the outline ofhe abnormally aggregated hyperplastic epithelial cellsie, EPHL), the contour of which is usually insignificantn HE tissue sections for routine histopathologic exam-nation (Fig 2, I–N vs Fig 2, A–F); therefore, OPNould be helpful in identifying the EPHLs at an earlytage of FSGS when they contain only a few cells andccupy a very small portion of the whole tissue section.lthough OPN has been a potential biomarker for di-

gnosis and prognosis of some diseases such as cres-ent nephropathies and myocardial infarction,15–17 it isremature to conclude that OPN is a specific biomarker

ig 4. Urine OPN levels. (A) Western blot analysis of OPNn samples taken at the indicated times. (B) Changes in the OPNevels normalized to the corresponding urine creatinine levels. n � 9,*P � 0.01 compared with day 0. The dashed lines show the meanalue for the control group.

or FSGS as OPN upregulation is a common phenom- r

non observed in different types of renal cells under aariety of kidney injuries.13–15 However, OPN mighttill be a useful sensor for detecting the early EPHL-elated injuries of glomeruli in FSGS; additional humantudies are required to confirm our result, especially forSGS caused by suspected environmental toxins.11

Even if some EPHLs in FSGS have histologic fea-ures similar to crescents of typical crescentic glomer-lonephritis (Fig 2, I–N),2–7 the pathogenic roles ofPHLs in FSGS remain unclear. In the FSGS model,

he time point for the initial development of EPHLsFig 2, G) was earlier than that of the developmentf sclerotic lesions (Fig 2, H), ie, typical EPHLs devel-ped significantly on day 11, whereas remarkableclerotic lesions developed in glomeruli on day 15.he shift of the EPHLs from the cellular type to thebrocellular type (Fig 2, G) indicated an active depo-ition of extracellular matrix proteins within thePHL-containing glomeruli and implied that most glo-eruli would eventually develop sclerosis and fibro-

is.10 One underlying mechanism for the fibrosis ofFHL and sclerosis of glomeruli might be the overex-ression of OPN, as OPN is a well-known factor tonduce fibrosis for tissue remodeling after injuries,17

nd modulation of OPN expression can inhibit glomer-lar sclerosis.24

In the current study, OPN expressed in both glomer-lar and tubular cells; the main sources of glomerularPN were the parietal epithelial cells that were locatedn the inside of Bowman capture and formed thePHLs (Fig 2, I–N). Glomerular OPN is involved inrescentic glomerulonephritis by attracting mononu-lear cells to the lesion sites and contributes to inflam-ation, crescentic development, and glomerular scle-

osis and fibrosis.8,10,25 Tubular OPN plays bothrotective and pathogenic roles; that is, OPN derivedrom tubular cells can inhibit renal stone formation,revent tubular cell apoptosis, and participate in tubularell regeneration and repair,13 but it can also recruitacrophages that contribute to renal interstitial fibro-

is.24 However, the roles of OPN in FSGS are stillnclear. Although the existence of OPN in the EPHLsn the FSGS animal model we used did not evokeassive macrophage infiltration as shown by light mi-

roscopy (Fig 2, A–F), we observed focal characteristicacrophage infiltration around the glomeruli (peri-

lomerular in distribution) by IHC (Fig 2, P–U). It isonceivable that these macrophages might be attractedy OPN secreted from the EPHLs and/or tubutar cellsearby the EPHLs, as the expression of OPN (days1–20, Fig 2, O) was in accordance with the peri-lomerular infiltration of macrophage (days 11–20, Fig, V). We suggest that OPN could play a role in

ecruiting macrophage from outside of the glomeruli,

wtaehteEa

ecioIEFftOt

R

1

1

1

1

1

1

1

1

1

1

2

2

2

2

2

2

2

Translational Research222 Shui et al October 2007

hich, in FSGS, could perpetuate the development ofhe EPHLs, compromize the glomerular structure, andccelerate glomerular sclerosis and fibrosis.8–10 How-ver, besides OPN, other monocyte chemoattractantsave been shown to be involved in macrophage infil-ration in other glomerulonephropathies.26 Thus, thexpression of other monocyte chemoattractants inPHLs of FSGS should also be considered and calls fordditional investigation.In conclusion, the FSGS model showed an increased

xpression of OPN in early glomerular EPHLs, whichoincided well with the peri-glomerular macrophagenfiltration and correlated with the following increasesf glomerular sclerosis and urine OPN protein levels.mportantly, IHC staining of OPN helps accentuate thePHL contour for microscopic inspection at earlySGS stages. We propose that OPN is an injury markeror EPHLs and plays a pathogenic role in the persis-ence and progression of EPHLs in FSGS. Detection ofPN levels in glomeruli and urine could have prognos-

ic values for FSGS.

EFERENCES

1. Benchimol C. Focal segmental glomerulosclerosis: pathogenesisand treatment. Curr Opin Pediatr 2003;15:171–80.

2. D’Agati V. Pathologic classification of focal segmental glomer-ulosclerosis. Semin Nephrol 2003;23:117–34.

3. Nagata M, Horita S, Shu Y, Shibata S, Hattori M, Ito K, et al.Phenotypic characteristics and cyclin-dependent kinase inhibi-tors repression in hyperplastic epithelial pathology in idiopathicfocal segmental glomerulosclerosis. Lab Invest 2000;80:869–80.

4. Schwartz MM, Lewis EJ. Focal segmental glomerular sclerosis:the cellular lesion. Kidney Int 1985;28:968–74.

5. Ramirez F, Travis LB, Cunningham RJ, Cavallo T, Rajarman S,Brouhard BH, et al. Focal segmental glomerulosclerosis, cres-cent, and rapidly progressive renal failure. Int J Pediatr Nephrol1982;3:175–8.

6. Shu KH, Lien JD, Chou G, Yang YF, Lu YS, Wang JY. Focalsegmental glomerulosclerosis with crescent—a case report. Tai-wan Yi Xue Hui Za Zhi 1986;85:643–8.

7. Verani RR, Hawkins EP. Recurrent focal segmental glomerulo-sclerosis. A pathological study of the early lesion. Am J Nephrol1986;6:263–70.

8. Hudkins KL, Giachelli CM, Eitner F, Couser WG, Johnson RJ,Alpers CE. Osteopontin expression in human crescentic glomer-ulonephritis. Kidney Int 2000;57:105–16.

9. Lan HY, Yu XQ, Yang N, Nikolic-Paterson DJ, Mu W, PichlerR, et al. De novo glomerular osteopontin expression in ratcrescentic glomerulonephritis. Kidney Int 1998;53:136–45.

0. Atkins RC, Nikolic-Paterson DJ, Song Q, Lan HY. Modulatorsof crescentic glomerulonephritis. J Am Soc Nephrol 1996;7:

2271–8.

1. Conlon PJ, Butterly D, Albers F, Rodby R, Gunnells JC, HowellDN. Clinical and pathologic features of familial focal segmentalglomerulosclerosis. Am J Kidney Dis 1995;26:34–40.

2. Ka SM, Rifai A, Chen JH, Cheng CW, Shui HA, Lee HS, et al.Glomerular crescent-related biomarkers in a murine model ofchronic graft versus host disease. Nephrol Dial Transplant 2006;21:288–98.

3. Xie Y, Sakatsume M, Nishi S, Narita I, Arakawa M, Gejyo F.Expression, roles, receptors, and regulation of osteopontin in thekidney. Kidney Int 2001;60:1645–57.

4. Endlich N, Sunohara M, Nietfeld W, Wolski EW, Schiwek D,Kranzlin B, et al. Analysis of differential gene expression instretched podocytes: osteopontin enhances adaptation of podo-cytes to mechanical stress. FASEB J 2002;16:1850–2.

5. Masutani K, Tokumoto M, Nakashima H, Tsuruya K, KashiwagiM, Kudoh Y, et al. Strong polarization toward Th1 immuneresponse in ANCA-associated glomerulonephritis. Clin Nephrol2003;59:395–405.

6. Suezawa C, Kusachi S, Murakami T, Toeda K, Hirohata S,Nakamura K, et al. Time-dependent changes in plasma osteopon-tin levels in patients with anterior-wall acute myocardial infarc-tion after successful reperfusion: correlation with left-ventricularvolume and function. J Lab Clin Med 2005;145:33–40.

7. Chandrashekhar Y. Multifunctional role of osteopontin in ven-tricular remodeling. J Lab Clin Med 2005;145:5–8.

8. Sadlier DM, Connolly SB, Kieran NE, Roxburgh S, Brazil DP,Kairaitis L, et al. Sequential extracellular matrix-focused andbaited-global cluster analysis of serial transcriptomic profilesidentifies candidate modulators of renal tubulointerstitial fibrosisin murine adriamycin-induced nephropathy. J Biol Chem 2004;279:29670–80.

9. Shui HA, Ka SM, Lin JC, Lee JH, Jin JS, Lin YF, et al.Fibronectin in blood invokes the development of focal segmentalglomerulosclerosis in mouse model. Nephrol Dial Transplant2006;21:1794–802.

0. Montinaro V, Hevey K, Aventaggiato L, Fadden K, Esparza A,Chen A, et al. Extrarenal cytokines modulate the glomerularresponse to IgA immune complexes. Kidney Int 1992;42:341–53.

1. Chao TK, Rifai A, Ka SM, Yang SM, Shui HA, Lin YF, et al.The endogenous immune response modulates the course of IgA-immune complex mediated nephropathy. Kidney Int 2006;70:283–97.

2. Kreisberg JI, Karnovsky MJ. Glomerular cells in culture. KidneyInt 1983;23:439–47.

3. Gang X, Ueki K, Kon S, Maeda M, Naruse T, Nojima Y.Reduced urinary excretion of intact osteopontin in patients withIgA nephropathy. Am J Kidney Dis 2001;37:374–9.

4. Kramer AB, Ricardo SD, Kelly DJ, Waanders F, van Goor H,Navis G. Modulation of osteopontin in proteinuria-induced renalinterstitial fibrosis. J Pathol 2005;207:483–92.

5. Nakamura H, Kitazawa K, Honda H, Sugisaki T. Roles of andcorrelation between alpha-smooth muscle actin, CD44, hyal-uronic acid and osteopontin in crescent formation in humanglomerulonephritis. Clin Nephrol 2005;64:401–11.

6. Chen S, Liu Z, Chen H, Zhou H, Wang J, Li L. Glomerularchemokine expression and the effect of steroid and cyclophos-phamide pulse therapy in human crescentic glomerulonephritis.

Chin Med J (Engl) 2002;115:1301–7.