Kidney International, Vol. 62 (2002), pp. 78–85

Effects of cyclosporine in osteopontin null mice1

MARILDA MAZZALI, JEREMY HUGHES, MARCIO DANTAS, LUCY LIAW, SUSAN STEITZ,CHARLES E. ALPERS, RAIMUND H. PICHLER, HUI Y. LAN, CECILIA M. GIACHELLI,STUART J. SHANKLAND, WILLIAM G. COUSER, and RICHARD J. JOHNSON

Division of Nephrology and Department of Pathology, University of Washington, Seattle, Washington; Center of MolecularMedicine, Maine Medical Research Institute, Scarborough, Maine; Department of Bioengineering, University of Washington,Seattle, Washington; and Division of Nephrology, Baylor College of Medicine, Houston, Texas, USA

Effects of cyclosporine in osteopontin null mice. tive indicator of tubulointerstitial inflammation and isBackground. Osteopontin (OPN) is a macrophage adhesive up-regulated in both experimental [2–5] and human [6–8]

and cell survival factor that is up-regulated in tubules in tubulo- tubulointerstitial injury. OPN is multifunctional and isinterstitial disease. We have previously reported that rats withinvolved in macrophage recruitment, cell survival, regu-cyclosporine (CsA) nephropathy have increased tubular osteo-lation of inducible nitric oxide synthase activity and inhi-pontin that correlates with the infiltration of macrophages and

interstitial fibrosis. This study tested the hypothesis that the bition of renal stone formation [9–14].absence of OPN would ameliorate CsA nephropathy. Cyclosporine (CsA) nephropathy is a common com-

Methods. OPN knockout (�/�) and wild type (�/�) mice plication in transplantation [15]. CsA nephropathy alsowere fed a low salt diet (Na� 0.01%) for one week and thenhas been generated in rodents and has similar featuresreceived daily CsA injections (30 mg/kg, SC) until sacrifice atto human disease, including afferent arteriolar hyalino-two weeks. Afferent arteriolopathy, tubulointerstitial injury,

macrophage infiltration, collagen III deposition, transforming sis and striped tubulointerstitial fibrosis [16–18]. Pichlergrowth factor-� (TGF-�) expression, and tubular and intersti- et al reported that OPN expression was increased in CsAtial cell proliferation and apoptosis were evaluated.

nephropathy in rats [17]. Furthermore, in individual ratsResults. Wild type mice developed early features of CsAthere was a strong correlation between the degree ofnephropathy, with arteriolar hyalinosis and cortical and tubulo-OPN expression and the severity of the macrophageinterstitial fibrosis. Despite comparable CsA levels, OPN�/�

mice had less arteriolopathy (15 vs. 24%, P � 0.05), a 20% infiltration and fibrosis [17]. This led to the hypothe-reduction in cortical macrophage infiltration (P � 0.05), and sis that OPN might be an important mediator of CsA20% reduction in interstitial collagen deposition (P � 0.05).

nephropathy.OPN�/� mice also showed less cortical interstitial cell pro-Recently the OPN knockout (OPN�/�) mouse hasliferation but no differences in tubular cell proliferation or

been generated, and the phenotype is associated with aapoptosis. OPN�/� mice also developed some neurotoxicity,consisting of ataxia, and this was associated with increased mild defect in wound healing [19]. Importantly, the kid-mortality at two weeks. neys of OPN�/� mice are normal. This observation,

Conclusion. OPN partially mediates arteriolopathy, early coupled with the recent development of a model of CsAmacrophage recruitment and fibrosis in murine CsA nephropa-nephropathy in mice [20], allowed us to directly addressthy. OPN also may be involved in CsA associated neurotoxicity.whether OPN has a functional role in CsA nephropathy.

Osteopontin (OPN) is an Arg-Gly-Asp (RGD-con- METHODStaining) adhesive glycoprotein and is expressed at low

Experimental animalslevels by proximal and distal tubular epithelial cells inDeletion of the OPN gene was performed by targetednormal kidney tissue [1]. However, OPN is a very sensi-

disruption as described previously [19]. Homozygousnull mutant mice (�/�) were viable and fertile, and

1 See Editorial by Hoyer, p. 348. breeding pairs of a (129/SvJx Black Swiss) hybrid geneticbackground were established to generate the animalsKey words: osteopontin, cyclosporine, macrophages, tubulointerstitial

injury. used in the present study. Wild-type breeding pairs ofthe same genetic background (129/SvJ Black Swiss) wereReceived for publication August 15, 2001used as controls. The genotype of animals was confirmedand in revised form February 12, 2002

Accepted for publication February 14, 2002 by polymerase chain reaction (PCR) analysis. Previousstudies indicated that osteopontin null mutant mice ex- 2002 by the International Society of Nephrology

78

Mazzali et al: CsA in osteopontin null mice 79

hibited normal kidney architecture and renal function, Detection of proximal and distal tubular cellproliferation and apoptosisunless exposed to overt renal injury [11, 21].

Prior to immunostaining for BrDU, tissue sectionsStudy design underwent an antigen retrieval step comprising of boiling

in 0.01 mol/L sodium citrate buffer for a total of 10Seventeen mice [7 knockout (�/�) and 10 wild typeminutes. Sections were then incubated with a murine(�/�)], 20 to 24 weeks old, were housed in individualIgG2a monoclonal antibody against BrdU (Cell prolifer-cages in an air, temperature and light controlled environ-ation kit; Amersham) at room temperature for 60 min-ment and received a low salt diet (0.01% Na�; (Zeiglerutes, followed by a peroxidase conjugated goat anti-Bros, Gardners, PA, USA) and water ad libitum. Aftermouse IgG2a (Cell proliferation kit; Amersham) at roomone week on low salt diet, daily subcutaneous injectionstemperature for 30 minutes. A black color was developedof CsA (Novartis) were administered at a dose of 30using DAB with nickel chloride as the chromogen.mg/kg diluted in olive oil for two weeks. These doses of

Apoptotic cells were detected by the TUNEL assayCsA and the use of a low salt diet have been found toas previously described [23]. Briefly, 4 �m formalin-fixed

be necessary for the induction of CsA nephropathy in tissue sections were deparaffinized and rehydrated inmice [20]. At day 13 mice were housed in individual ethanol followed by an antigen retrieval step comprisingmetabolic cages for 24-hour urine collection. On the of boiling in 0.01 mol/L sodium citrate buffer for twofollowing day, mice received IP injections of BrDU label- minutes. Sections were incubated with proteinase K (6.2ing solution (1 mL/100 g body wt; Cell Proliferation kit; �g/mL; Boehringer Mannheim, Mannheim, Germany),Amersham Pharmacia Biotech, Bucks, UK) and were followed by TdT (300 enzyme U/mL; Pharmacia Biotech,sacrificed after two hours. Renal biopsies were obtained Piscataway, NJ, USA) and Bio-14-dATP (0.94 nmol/L;from each animal. Control tissue was obtained from un- Gibco BRL, Grand Island, NY, USA). Biotinylated ATPtreated OPN�/� (N � 6) and OPN�/� (N � 5) mice was detected using the ABC staining method (Vectorfor comparison. The studies were approved by the Uni- Laboratories Inc.) following the manufacturer’s proto-

col. As a positive control, slides were pretreated withversity of Washington Animal Care Committee.DNAase (20 Kunitz U/mL; Sigma Biosciences). Cellswere considered TUNEL positive if their nuclei stainedFunctional datablack and displayed a typical apoptotic morphology withSerum blood urea nitrogen (BUN) was determinedchromatin condensation.colorimetrically using a standard kit (Sigma, St Louis,

In order to identify and accurately quantify proximalMO, USA). Cyclosporine levels were measured by highversus distal tubular epithelial cell proliferation and apo-

performance liquid chromatography in whole blood. ptosis, Fx1A/BrdU and Fx1A/TUNEL double labelingwas performed. The Fx1A antibody was raised against

Renal histology rat proximal tubular brush border antigen as describedRenal biopsies were fixed in either Methyl-Carnoy’s previously and specifically stains the brush border of

or 10% buffered formalin and embedded in paraffin. proximal tubular epithelial cells [24]. Sections were in-Light microscopy was performed in 4 �m sections of cubated with biotinylated sheep Fx1A antibody at 4�C

overnight, followed by horseradish peroxidase conju-Methyl-Carnoy’s fixed tissue stained with periodic acid-gated avidin D (Vector Laboratories) for 20 minutes atSchiff (PAS) reagent. Methyl-Carnoy’s fixed tissue sec-room temperature. Staining was developed using DABtions were analyzed by indirect immunoperoxidase stain-without nickel to produce a brown color. BrdU oring with the following primary antibodies: OP199, a goatTUNEL staining was then performed as indicated pre-polyclonal antibody against osteopontin (OPN) [22];viously using DAB with nickel as the chromogen.F4/80 a rat monoclonal antibody to mouse macrophages

(Caltag, Burlingame, CA, USA); goat antihuman type III In situ hybridization for TGF-�1collagen (Southern Biotechnology Associates Inc, Bir-

A previously characterized pGEM T7 plasmid (Pro-mingham, AL, USA). Sections were incubated with amega, Madison, WI, USA) containing a 974 kb (421

secondary antibody followed by horseradish peroxidase- to 1395 bp) of mouse transforming growth factor-�1conjugated avidin D (Vector Laboratories, Burlingame, (TGF-�1) cDNA was provided by Dr. H.L. Moses (Van-CA, USA), diaminobenzidine (DAB; Sigma) with or derbilt University, Nashville, TN, USA) [25, 15]. Thewithout nickel chloride as a chromogen, and counter- plasmid was linearized with HindIII and Eco RI andstained with methyl green. An irrelevant primary anti- transcribed into antisense and sense (negative control)body of the same isotype was used as a negative control. riboprobes in T7- and SP6-primed reactions, using re-Positive control tissue included sections from diseased agents from Promega, except 35S-UTP, obtained from

New England Nuclear (Boston, MA, USA).mice that expressed these antigens.

Mazzali et al: CsA in osteopontin null mice80

Formalin fixed, paraffin embedded tissues were depar- Macrophage infiltrationaffinized following standard protocols. In situ hybridiza- Macrophage infiltration was measured by determiningtion followed methods previously described [26]. The the amount of cortical and medullary tissue stained forsections were washed with 0.5 standard sodium citrate F4/80 (over background), using computer assisted image(SSC; Gibco) and digested with 5 �g/mL of proteinase K analysis at �200 magnification.(Sigma) in Tris buffer (500 mmol/L NaCl, 10 mmol/L Type III collagen deposition. Type II collagen depositsTris, pH 8.0) for 30 minutes at 37�C. Several 0.5 � SSC were measured by computer image analysis, at �50washes were followed by prehybridization for two hours magnification, and expressed as the percentage of renalin 50 �L of prehybridization buffer [50% formamide, cortex or medulla occupied by type III collagen.0.3 mmol/L NaCl, 20 mmol/L Tris pH 8.0, 5 mmol/L Tubular cell proliferation and apoptosis. These wereethylenediaminetetraacetic acid (EDTA), 1 � Den- measured by counting the number of Fx1A�/BrDU�,hardt’s solution, 10% dextran sulfate, 10 mmol/L dithio- Fx1A�/BrDU�, Fx1A�/TUNEL�and Fx1A�/TUNEL�threitol (DTT), 500 �g/mL yeast tRNA] at 50�C. The cells in 20 sequentially selected 0.1 mm2 grids, in thehybridization was started by adding 500,000 cpm of renal cortex, at �200 magnification, and expressed as35S-labeled riboprobe in 50 �L of prehybridization buffer the number of cells per mm2.and was allowed to proceed overnight at 50�C. Sections Interstitial cell proliferation and apoptosis. Interstitialwere then washed with 2 � SSC, followed by RNAse A cell proliferation and apoptosis were measured by count-(20 �g/mL, 30 min at room temperature), 2 � SSC ing the number of BrDU and TUNEL positive cells inwashes (3 � 1 min), three high-stringency washes with 20 sequentially selected 0.1 mm2 grids in the renal cortex0.1 � SSC/0.1% Tween 20 (Sigma) at 50�C and three and 10 in the renal medulla, at �200 magnification, and2� washes. The slides were dipped in NTB2 nuclear expressed as the number of cells per mm2.emulsion (Kodak, Rochester, NY, USA) and exposed In situ hybridization. Positive cellular labeling was de-in the dark at 4�C for four weeks. After developing, fined as five or more silver grains in a single cell. Forsections were counterstained with hematoxylin and eosin negative controls, simultaneous hybridization with theand were dehydrated and coverslipped. Positive cellular sense riboprobe was performed on replicate tissue.labeling was defined as five or more silver grains in a

Statistical analysissingle cell. For negative controls, simultaneous hybrid-ization with the sense riboprobe was performed on repli- All values are expressed as mean SD, unless other-cate tissue sections. wise stated. Statistical significance (P � 0.05) was evalu-

ated by ANOVA, with appropriate correction for multi-Quantification of morphologic data ple comparisons (Fisher PSLD).

All quantification was performed with the examinerblinded to its origin.

RESULTSCortical tubulointerstitial injury and fibrosis. This wasSurvival was lower in OPN �/� mice treatedscored semiquantitatively by analyzing cortical tubuloin-with CsAterstitial fields on PAS stained tissue sections, and grad-

ing the degree of tubular changes (atrophy/dilation) and The mortality of mice treated with CsA was signifi-interstitial expansion using a score from 0 to 5, as pre- cantly reduced in the OPN�/� mice compared toviously described [17]. OPN�/� mice at two weeks (0% vs. 50%, P � 0.05).

Arteriolar hyalinosis. Arteriolar hyalinosis of the af- Two of five of the OPN�/� mice that died developedferent arteriole was quantified by counting the number ataxia 24 to 48 hours prior to death, whereas this wasof juxtaglomerular afferent arterioles which exhibited not observed in other OPN�/� or OPN�/� mice. Wehyalinosis, characterized by the presence of hyaline ma- cannot exclude that the etiology of the mortality wasterial in the cross sectional vascular area, and was ex- related to worse kidney disease or CsA-induced neuro-pressed as a percentage of the total number of juxtaglom- toxicity, since histologic analyses and BUN and CsAerular afferent arterioles present in the biopsy under levels were not able to be performed in the mice thatexamination with the �40 objective [17]. died. However, in the survivors, both serum BUN and

Tubular expression of osteopontin. This was calcu- CsA blood levels were comparable (Table 1). Food in-lated as the percent of renal cortex occupied by OPN- take was also similar in both groups although OPN�/�positive tubules. Utilizing computer-assisted image anal- mice showed greater weight loss (mean weight loss at 1ysis software (Optimas V6.2; Media Cybernetics, Silver week � 4 g/week, 10% baseline body weight; Table 1).Springs, MD, USA) and digitized images, the percent of

OPN is up-regulated in murine CsA nephropathyarea occupied by OPN positive tubules per 4 mm2 fieldat a magnification of �50 was measured and the mean Osteopontin was up-regulated in tubules of renal cor-

tex of OPN�/� mice after two weeks of CsA treatmentpercent area calculated for each biopsy.

Mazzali et al: CsA in osteopontin null mice 81

Table 1. Cyclosporine levels, body weight and renal function type (Fig. 2 A, B). Interestingly, the interstitial cell prolif-after two weeks of cyclosporine treatment

eration correlated directly with the degree of macro-OPN�/� OPN�/� phage infiltration when both groups were combined (r 2 �

Number of animals 7 5 0.72, P � 0.05; Fig. 2 C, D).CsA ng/dL 3332.0230.8 3247.5496.8 Osteopontin has been reported to be a cell survivalBUN baseline mg/dL 30.31.8 23.42.2

factor, and OPN�/� mice show increased apoptosis inBUN 2 weeks mg/dL 80.915.6 66.716.6BW baseline g 38.21.6 40.70.9 the obstructive uropathy model [10]. However, no differ-BW 1 week g 37.81.4 36.65.9a ences were evident between genotypes in the number ofBW 2 weeks g 35.20.7 39.10.9

TUNEL positive apoptotic cells in the cortex or theHct baseline % 45.31.5 44.91.0Hct 2 weeks % 43.31.6 42.31.0 medulla (Table 2).

Abbreviations are: CsA, cyclosporine levels; OPN, osteopontin; BUN, bloodurea nitrogen; BW, body weight; Hct, hematocrit. OPN�/� mice exhibited comparable TGF-�1

a P � 0.05 vs. baseline body weight mRNA expression

Transforming growth factor-�1 mRNA expressionwas detected in rare scattered intra-glomerular cells andarterial endothelial cells of control kidney tissue from(1.9 0.6% vs. 0.6 0.3% cortical area, CsA 2 weeksboth untreated OPN�/� and OPN �/� mice. In CsA-vs. untreated, N � 5 per group, P � 0.05; Fig. 1 A, B).treated mice TGF-�1 mRNA was expressed in both tu-OPN�/� mice also showed up-regulation of OPN in thebular and interstitial cells in scattered foci, usually butmedulla compared to untreated OPN�/� mice (1.9 not exclusively in regions with an interstitial inflam-0.3 vs. 0.5 0.2% medullary area, P � 0.05, N � 5 permatory infiltrate. No distinct alterations in patterns ofgroup). No expression of osteopontin was detected inTGF-�1 mRNA expression were identified in CsA-either untreated or CsA treated OPN�/� mice.treated OPN�/� versus OPN�/� mice at the two-week

Murine CsA nephropathy is attenuated in time point.OPN�/� mice

After two weeks of CsA injections, both OPN�/� DISCUSSIONand OPN�/� mice developed arteriolopathy and tubu- Recent studies suggest that OPN may have a rolelointerstitial injury consistent with early CsA nephrotox- as a macrophage adhesive factor in renal disease. OPNicity (Fig. 1C). OPN�/� mice, however, displayed sig- binds macrophages with high affinity [27] and OPN willnificantly fewer lesions of arteriolar hyalinosis compared elicit a macrophage-rich infiltration on injection intoto the OPN�/� mice (15 vs. 24%, P � 0.05; Table 2). mice, rats or rabbits [27, 28]. In the kidney, OPN is ex-Interestingly, de novo expression of OPN was found pressed constitutively by distal tubules in the juxta-in the preglomerular arterial vessels in OPN�/� mice medullary cortex and outer medulla [1]. However, OPNtreated with CsA (Fig. 1D). is induced in proximal and distal tubules in both immune

While there was no difference in tubulointerstitial in- and nonimmune models of tubulointerstitial injury [3–5,jury scores, there was significant although mild reduction 29, 30]. In these models the up-regulation of OPN corre-in both cortical macrophage (F4/80) infiltration and corti- lates both spatially and quantitatively with the infiltra-cal type III collagen deposition in the OPN�/� mice tion of macrophages and with the development of fibrosis(Table 2). These changes were present despite similar [3, 10, 17, 31].blood levels of CsA (Table 1). Pichler et al have reported that OPN is up-regulated

In contrast to the difference in cortical injury, there in CsA nephropathy in rats, and that this was associatedwas no difference in macrophage infiltration or type III with increased macrophage infiltration and fibrosis [17].collagen deposition in the medulla between genotypes. Thomas et al have also reported that CsA nephropathy

is associated with tubular cell apoptosis, and this alsoRole of osteopontin in renal cell proliferation correlates with the degree of fibrosis [32]. We thereforeand apoptosis decided to examine the functional role of OPN in a

Cyclosporine A nephropathy in OPN�/� and OPN�/� murine model of CsA nephropathy.mice was associated with both tubular and interstitial The role of OPN in macrophage attraction and pro-cell proliferation, as determined by BrdU incorporation. liferation has been investigated using several differentThere was no difference in proximal (FX1A�) or distal techniques. In the anti-glomerular basement membrane(FX1A�) tubular cell proliferation between genotypes (anti-GBM) model, inhibition of OPN with antibody or(Table 2). However, the number of proliferating BrdU- antisense therapy results in a reduction in macrophage in-positive interstitial cells in the renal cortex was signifi- filtration and fibrosis [31, 33]. In the unilateral ureteral

obstruction (UUO) model, OPN has been shown to me-cantly reduced in OPN�/� mice compared to the wild

Mazzali et al: CsA in osteopontin null mice82

Fig. 1. Osteopontin (OPN) was up-regulatedin cyclosporine (CsA) nephropathy inOPN�/� mice and was expressed in preglo-merular arterial vessels. Tubular expressionof OPN was increased in both cortical andmedullary areas in OPN�/� mice treated withCsA for 2 weeks (B) compared to untreatedOPN�/� mice (A). The histological findingsof CsA nephropathy (tubular atrophy and di-latation, striped fibrosis and afferent arteriolarhyalinosis (C ) was similar in both OPN�/�and OPN�/� mice. However, OPN�/� miceshowed increased incidence of afferent arteri-olar hyalinosis and expression of OPN in affer-ent arterioles (arrow) (D). (A and B, immuno-histochemistry for osteopontin, magnification�50; C, PAS staining, �50; D, osteopontinstaining, �1000).

Table 2. Histological data in control and CsA treated mice

Control 2 weeks of CsA

OPN�/� OPN�/� OPN�/� OPN�/�

N 6 5 7 5TI injury score (PAS) 0.200.12 0.260.18 0.760.43 0.560.23Arteriolar hyalinosis % 0 0 14.9 7.0a 23.8 4.6Type III collagen cortex

% area 1.931.42 2.141.64 8.742.2a 10.453.1F4/80 cortex % area 0.100.11 0.230.21 0.810.13a 1.02 0.09IC proliferationb

BRDU� cells/mm2 3.901.52 5.641.92 20.13.9a 32.0 1.7IC apoptosis

TUNEL� cells/mm2 0.330.30 0.240.20 4.480.64 5.440.74PTC proliferationb

BRDU�/FX1A� cells/mm2 0.640.70 0.350.49 3.141.0 4.31.5PTC apoptosis

TUNEL�/FX1A� cells/mm2 0.110.10 0.050.09 2.130.8 1.560.41DTC proliferationb

BRDU�/FX1A� cells/mm2 0.030.07 0.040.08 1.00.3 0.80.11DTC apoptosis

TUNEL�/FX1A� cells/mm2 0 0.05 0.08 0.110.05 0.200.12MedullaType III collagen % area 5.544.92 5.842.42 22.44.3 23.85.2Macrophage infiltration

% area 0.360.29 0.290.15 11.82.3 11.82.3Medullary cell proliferationb

BRDU� cells/mm2 1.060.38 1.160.40 11.82.3 18.40.8Medullary cell apoptosis

TUNEL� cells/m2 0.230.21 0.680.24 1.110.16 2.801.50

Abbreviations are in Table 1 and: IC, interstitial cell; PTC, proximal tubular cell; DTC, distal tubular cell.a P � 0.05 vs. OPN�/� mice at the same time pointb Proliferation studies with Brdu were performed on only a subset of mice (5 control OPN�/�, 5 OPN�/�, and 4 each of CsA-treated OPN�/� and �/� mice)

Mazzali et al: CsA in osteopontin null mice 83

Fig. 2. The lack of OPN was associated with less interstitial cell proliferation and macrophage infiltration. The number of proliferating BrdUpositive cells (arrows) in interstitial renal cortex (arrows) was reduced in OPN�/� mice (A) compared to OPN�/� (B). The interstitial cellproliferation correlated directly with the degree of macrophage infiltration (C and D). A and B, double staining FX1A/BrdU, �640 (BrdU� cellsstained in black, proximal tubular brush border stained in brown); C shows macrophage infiltration, F4/80 staining, �200). Symbols in D are: (�)OPN�/� control; (�) OPN�/� control; (�) OPN�/� 2 weeks; (�) OPN�/� 2 weeks.

diate the early macrophage infiltration and fibrosis, as OPN knockout mouse. It is also possible that theOPN�/� mice that died may have had worse nephrotox-evidenced by studies using knockout mice [11]. It is im-

portant to realize, however, that blocking OPN only had icity. Nevertheless, our data are consistent with previousstudies that suggest that the role for OPN in macrophagea partial effect on macrophage infiltration in these mod-

els, thereby suggesting the importance of other chemo- recruitment is relatively minor [11, 31, 34].The second interesting finding was that the OPN wastactic (such as, chemokines) [34] and adhesive factors

(such as, ICAM-1) [35] on the renal inflammatory expressed de novo in the preglomerular arterial vesselsof CsA-treated OPN�/� mice, and that OPN�/� miceresponse.

Our primary finding was that OPN mediated the early showed less arteriolar hyalinosis compared to theOPN�/� mice. The afferent arteriolar hyalinosis in CsAmacrophage infiltration and type III collagen deposition

in the cortex in the CsA nephropathy model. The reduc- nephropathy involves endothelial and smooth muscleinjury resulting in the subendothelial accumulation oftion in macrophage infiltration and fibrosis in the

OPN�/� mice were mild (20%), and as such, this did proteinaceous material, and this has been previouslyshown to be mediated by angiotensin II [16, 17] and tonot result in any differences in renal function. It is possi-

ble that the mild differences in macrophage accumula- be ameliorated by vascular endothelial growth factor(VEGF) infusion [36]. Interestingly, OPN has beention in OPN�/� and OPN�/� mice may reflect compen-

satory mechanisms for macrophage recruitment in the shown to be produced by injured vascular smooth muscle

Mazzali et al: CsA in osteopontin null mice84

cells in a variety of conditions [37–40]. In vascular dis- In conclusion, OPN appears to have a modest rolein macrophage recruitment and in the development ofease, OPN has been posited to mediate smooth muscle

cell adhesion and migration [41] as well as endothelial fibrosis in the murine CsA nephropathy model, whichis consistent with studies in other models [2, 11, 31].migration and survival [12]. Indeed, neointimal thick-

ening following coronary or carotid angioplasty can be Interestingly, OPN also may be involved in the inductionreduced by inhibiting OPN with antibody [42] or with of the arteriolopathy in CsA rats. This suggests thatantisense [43] therapy. Therefore, OPN may be partially OPN may have an additional role in mediating renalresponsible for the arteriolar injury that occurs in CsA microvascular injury, and is consistent with emergingtreated mice. roles for OPN in vascular disease [10, 50].

Osteopontin also has been found to be a cell survivalfactor, in part due to its ability to inhibit nitric oxide ACKNOWLEDGMENTSsynthase-2 (NOS2) and the generation of cytotoxic reac- Support for this study was provided by US Public Health Granttive nitrogen metabolites [9, 13, 44]. However, we were DK-52121, HL-68607 and a George O’Brien Kidney Center Grant

(DK47659). Drs. Mazzali and Dantas are recipients of the Postdoctoralnot able to show any difference in tubular or interstitialprogram grant from FAPESP –Fundacao de Amparo a Pesquisa docell survival in the OPN�/� versus OPN�/� mice in Estado de Sao Paulo- SP-Brasil. A portion of these data were presented

this model. This contrasts with the ureteral obstruction in the annual meeting of the American Society of Nephrology 2000,and were published as an abstract (J Am Soc Nephrol 11:605A, 2000).model, in which OPN�/� mice showed significantly

more tubular apoptosis than wild type controls [11]. Reprint requests to Richard J. Johnson, M.D., Nephrology ResearchA third finding was that interstitial cell proliferation Laboratory, Division of Nephrology, Baylor College of Medicine, One

Baylor Plaza-Alkek N530, Houston, Texas 77030, USA.was reduced in the OPN�/� mice. The reduction inE-mail: rjohnson@bcm.tmc.eduproliferating cells correlated with the reduction in macro-

phages present in interstitium. We do not know the exactREFERENCESetiology of the proliferating cells, although they most

1. Kleinman JG, Beshensky A, Worcester EM, Brown D: Expres-likely represent either macrophages or fibroblasts. It ission of osteopontin, a urinary inhibitor of stone mineral crystalof interest that Sahai et al reported that osteopontingrowth, in rat kidney. Kidney Int 47:1585–1596, 1995

mediates the proliferation of a variety of cells in response 2. Yu XQ, Wu LL, Huang XR, et al: Osteopontin expression inprogressive renal injury in remnant kidney: Role of angiotensinto hypoxia [45], as it is known that cyclosporine adminis-II. Kidney Int 58:1469–1480, 2000tration causes vasoconstriction and hypoxia in the rat

3. Pichler R, Giachelli CM, Lombardi D, et al: Tubulointerstitialkidney [46]. disease in glomerulonephritis. Potential role of osteopontin (uro-

pontin). Am J Pathol 144:915–926, 1994We did not observe any difference in TGF-� expres-4. Giachelli CM, Pichler R, Lombardi D, et al: Osteopontin expres-sion. However, only major changes would be noted by

sion in angiotensin II-induced tubulointerstitial nephritis. Kidneythis technique (in situ hybridization). Int 45:515–524, 1994

The relevance of this study to human disease is uncer- 5. Diamond JR, Kees-Folts D, Ricardo SD, et al: Early and persis-tent upregulated expression of renal cortical osteopontin in experi-tain. While several groups have reported a strong associ-mental hydronephrosis. Am J Pathol 146:1455–1466, 1995ation between tubular expression of OPN with interstitial 6. Mezzano SA, Droguett MS, Burgos ME, et al: Overexpression

renal inflammation and fibrosis [4, 11, 17], a recent study of chemokines, fibrogenic cytokines and myofibroblasts in humanmembranous nephropathy. Kidney Int 57:147–158, 2000by Hudkins et al was not able to show this relationship

7. Thomas SE, Lombardi D, Giachelli CM, et al: Osteopontin ex-in CsA nephropathy in humans [47]. This may be due pression, tubulointerstitial disease and essential hypertension. Amto the fact that most studies show that OPN is involved J Hypertens 11:954–961, 1998

8. Hudkins KL, Giachelli CM, Eitner F, et al: Osteopontin expres-in the early macrophage infiltration, and other mediatorssion in human crescentic glomerulonephritis. Kidney Int 57:105–may be responsible for the late inflammatory response 116, 2000

[11]. In the Hudkins study the control kidney tissue also 9. Denhardt DT, Lopez CA, Rollo EE, et al: Osteopontin-inducedmodifications of cellular functions. Ann N Y Acad Sci 760:127–had a high level of OPN expression, which may have142, 1995been due to the effects of cold ischemia [48], and this may 10. Giachelli CM, Steitz S: Osteopontin: a versatile regulator of

have partially confounded the data. Finally, it remains inflammation and biomineralization. Matrix Biol 19:615–622, 200111. Ophascharoensuk V, Giachelli CM, Gordon K, et al: Obstruc-possible that there are significant differences in OPN

tive uropathy in the mouse: Role of osteopontin in interstitialregulation based on species differences [1].fibrosis and apoptosis. Kidney Int 56:571–580, 1999

A final observation was that there appeared to be 12. Scatena M, Almeida M, Chaisson ML, et al: NF-kappaB mediatesalphaVbeta3 integrin induced endothelial cell survival. J Cell Bioldifferences in neurotoxicity of the CsA between the141:1083–1093, 1998OPN�/� and OPN�/� mice. Despite comparable CsA

13. Hwang SM, Lopez CA, Heck DE, et al: Osteopontin inhibitslevels, OPN�/� mice showed ataxia and had a higher induction of nitric oxide synthase gene expression by inflammatory

mediators in mouse kidney epithelial cells. J Biol Chem 269:711–mortality. Indeed, it was not possible to examine later715, 1994time points because of this complication. This finding

14. Hoyer JR: Uropontin in urinary calcium stone formation. Mineremphasizes the multifunctional roles for this complex Electrolyte Metab 269:385–392, 1994

15. Bennett WM, Burdmann EA, Andoh TF, et al: Nephrotoxicitymolecule in disease [49].

Mazzali et al: CsA in osteopontin null mice 85

of immunosuppressive drugs. Nephrol Dial Transplant 9(Suppl 4): osteopontin in experimental crescentic glomerulonephritis in therat. Proc Assoc Am Physicians 110:50–64, 1998141–145, 1994

16. Burdmann EA, Andoh TF, Nast CC, et al: Prevention of experi- 34. Panzer U, Thaiss F, Zahner G, et al: Monocyte chemoattractantprotein-1 and osteopontin differentially regulate monocytes re-mental cyclosporine induced interstitial fibrosis by losartan and

enalapril. Am J Physiol 269:1186–1196, 1995 cruitment in experimental glomerulonephritis. Kidney Int 59:1762–1769, 200117. Pichler RH, Franceschini N, Young BA, et al: Pathogenesis of

cyclosporine nephropathy: Roles of angiotensin II and osteopontin. 35. Nishikawa K, Guo YJ, Myiasaka M, et al: Antibodies to intercellu-lar adhesion molecule-1/lymphocyte function-associated antigen 1J Am Soc Nephrol 6:1186–1196, 1995

18. Young BA, Burdmann EA, Johnson RJ, et al: Cyclosporine A prevent crescent formation in rat autoimmune glomerulonephritis.J Exp Med 177:667–677, 1993induced arteriolopathy in a rat model of chronic cyclosporine ne-

phropathy. Kidney Int 48:431–438, 1995 36. Kang DH: Kim YG, Andoh TF, et al: Post-cyclosporine mediatedhypertension and nephropathy: Amelioration by vascular endothe-19. Liaw L, Birk DE, Ballas CB, et al: Altered wound healing in

mice lacking a functional osteopontin gene (spp1). J Clin Invest lial growth factor. Am J Physiol 280:F727–F736, 200137. Giachelli CM, Steitz S: Osteopontin: a versatile regulator of101:1468–1478, 1998

20. Andoh TF, Lam TT, Lindsley J, et al: Enhancement of chronic inflammation and biomineralization. Matrix Biol 19:615–622, 200038. Malyankar UM, Hanson R, Schwartz SM, et al: Upstream stimu-cyclosporine nephrotocixity by sodium depletion in an experimen-

tal mouse model. Nephrol 3:471–478, 1997 latory factor 1 regulates osteopontin expression in smooth musclecells. Exp Cell Res 250:535–547, 199921. Rittling SR, Denhardt DT: Osteopontin function in pathology:

lessons from osteopontin-deficient mouse. Exp Nephrol 7:103– 39. de Blois D, Lombardi DM, Su EJ, et al: Angiotensin II inductionof osteopontin expression and DNA replication in rat arteries.113, 1999

22. Liaw L, Almeida M, Downey W, et al: Osteopontin promotes Hypertens 28:1055–1063, 199640. Sodhi CP, Phadke SA, Battle D, Sahai A: Hypoxia stimulatesvascular cell adhesion and spreading and is chemotactic for smooth

muscle cells in vitro. Circ Res 74:214–224, 1994 osteopontin expression and proliferation of cultured smooth mus-cle cells: potentiation by high glucose. Diabetes 50:1482–1490, 200123. Baker AJ, Mooney A, Hughes J, et al: Mesangial cell apoptosis:

The major mechanism for resolution of glomerular hypercellularity 41. Liaw L, Lindner V, Schwartz SM, et al: Osteopontin and beta3 integrin are coordinately expressed in regenerating endotheliumin experimental mesangial proliferative nephritis. J Clin Invest 94:

2105–2116, 1994 and stimulate Arg-Gly-Asp-dependent endothelial migration invitro. Circ Res 77:665–672, 199524. Edgington TS, Glassock RJ, Dixon FJ: Autologous immune com-

plex nephritis induced with renal tubular antigen. I. Identification 42. Liaw L, Lombardi DM, Almeida MM, et al: Neutralizing antibod-ies directed against osteopontin inhibit rat carotid neointimal thick-and isolation of the pathogenic antigen. J Exp Med 127:555–

572, 1968 ening after endothelial denudation. Arterioscler Thromb Vasc Biol17:188–193, 199725. Derynck R, Jarrett JA, Chen EY, Goeddel DV: The murine

transforming growth factor beta precursors. J Biol Chem 261:4377– 43. Panda D, Kundu GC, Lee BI, et al: Potential roles of osteopontinand alphaVbeta 3 integrin in the development of coronary artery4379, 1986

26. Alpers CE, Hudkins KL, Ferguson M, et al: Platelet derived restenosis after angioplasty. Proc Natl Acad Sci USA 94:9308–9313, 1997growth factor A chain expression in developing and mature human

kidneys and in Wilm’s tumor. Kidney Int 48:146–154, 1995 44. Noiri E, Dickamn K, Miller F, et al: Reduced tolerance to acuteischemia in mice with a targeted disruption of the osteopontin27. Singh RP, Patarca R, Schwartz J, et al: Definition of a specific

interaction between the early T lymphocyte activation-1 (Eta-1) gene. Kidney Int 56:77–82, 199945. Sahai A, Mei C, Schrier RW, Tannen RL: Mechanisms of chronicprotein and murine macrophage in vitro and its effect upon macro-

phages in vivo. J Exp Med 171:1931–1942, 1990 hypoxia-induced renal cell growth. Kidney Int 56:1277–1281, 199946. Zhong Z, Arteel GE, Connor HD, et al: Cyclosporin A increases28. Giachelli CM, Lombardi D, Johnson RJ, et al: Evidence for

a role of osteopontin in macrophage infiltration in response to hypoxia and free radical production in rat kidneys: Prevention bydietary glycine. Am J Physiol 275(4 Pt 2):F595–F604, 1998pathological stimuli in vivo. Am J Pathol 152:353–358, 1998

29. Magil AB, Pichler RH, Johnson RJ: Osteopontin in chronic 47. Hudkins KL, Le QC, Segerer S, et al: Osteopontin expression inhuman cyclosporine toxicity. Kidney Int 60:635–640, 2001puromycin aminonucleoside nephrosis. J Am Soc Nephrol 8:1383–

1390, 1997 48. Kleinman JG, Worcester EM, Beshensky AM, et al: Upregula-tion of osteopontin expression by ischemia in rat kidney. Ann NY30. Kaneto H, Morrissey J, McCracken R, et al: Osteopontin expres-

sion in the kidney during unilateral ureteral obstruction. Miner Acad Sci 760:321–323, 199549. Denhardt DT, Noda M O’Regan AW, et al: Osteopontin as aElectrolyte Metab 24:227–237, 1998

31. Okada H, Moriwaki K, Kalluri R, et al: Osteopontin expressed means to cope with environmental insults: Regulation of inflam-mation, tissue remodeling and cell survival. J Clin Invest 107:1055–by renal tubular epithelium mediates interstitial monocyte infiltra-

tion in rats. Am J Physiol (Renal Physiol) 278:F110–F121, 2000 1061, 200150. Gadeau AP, Chaulet H, Daret D, et al: Time course of osteopon-32. Thomas SE, Andoh TF, Pichler RH, et al: Accelerated apoptosis

characterizes cyclosporine associated interstitial fibrosis. Kidney tin, osteocalcin and osteonectin accumulation and calcificationafter acute vessel wall injury. J Histochem Cytochem 49:79–86,Int 53:897–908, 1998

33. Yu XQ, Nikolic-Paterson DJ, Mu W, et al: A functional role for 2001