Chronic rejection of mouse kidney allografts

Kidney International, Vol. 55 (1999), pp. 1935–1944

DIALYSIS – TRANSPLANTATION

Chronic rejection of mouse kidney allografts

ROSLYN B. MANNON, JEFFREY B. KOPP, PHILIP RUIZ, ROBERT GRIFFITHS, MATILDE BUSTOS,JEFFREY L. PLATT, PAUL E. KLOTMAN, and THOMAS M. COFFMAN

Departments of Medicine, Pediatrics, and Immunology, Duke and VA Medical Centers, Durham, North Carolina; NIDDK/National Institutes of Health, Bethesda, Maryland; Department of Pathology, University of Miami Medical Center,Miami, Florida; Department of Medicine, Mount Sinai School of Medicine, New York, New York, USA

Chronic rejection of mouse kidney allografts. of acute rejection are a major risk factor for the develop-Background. Chronic renal allograft rejection is the leading ment of chronic rejection [3], the observations that strin-

cause of late graft failure. However, its pathogenesis has not gent HLA-matching is protective, and that inadequatebeen defined.immunosuppression is associated with the developmentMethods. To explore the pathogenesis of chronic rejection,of chronic rejection. Nonimmunological factors suggestedwe studied a mouse model of kidney transplantation and exam-

ined the effects of altering the expression of donor major histo- to play a role include glomerular hyperfiltration, ische-compatibility complex (MHC) antigens on the development of mia-reperfusion injury, and uncontrolled hypertensionchronic rejection. [4]. Although there have been a number of recent ad-Results. We found that long-surviving mouse kidney allografts

vances in the prevention and treatment of acute allograftdevelop pathological abnormalities that resemble chronic re-jection in humans. Furthermore, the absence of MHC class I rejection, little progress has been made in developingor class II antigens did not prevent the loss of graft function therapies for chronic rejection. In kidney transplants,nor alter the pathological characteristics of chronic rejection. chronic rejection is the leading cause of late graft failure.Expression of transforming growth factor-b (TGF-b), a pleio-

The immunological component underlying chronic re-tropic cytokine suggested to play a role in chronic rejection, wasjection must be triggered at some level by the recognitionmarkedly enhanced in control allografts compared with isografts.

However, TGF-b up-regulation was significantly blunted in of major histocompatibility complex (MHC) alloantigensMHC-deficient grafts. Nonetheless, these differences in TGF-b by recipient T cells. In general, allorecognition may occurexpression did not affect the character of chronic rejection, through two pathways. The first is through direct recogni-including intrarenal accumulation of collagens.

tion of MHC alloantigens on donor cells by recipient TConclusions. Reduced expression of either class I or II directcells. The second is through indirect recognition, in whichallorecognition pathways is insufficient to prevent the develop-

ment of chronic rejection, despite a reduction in the levels of donor MHC alloantigens are taken up, processed, andTGF-b expressed in the allograft. This suggests that the severity presented to recipient T cells as peptides in the contextof chronic rejection is independent of the level of MHC dispar-

of recipient MHC. A key role for allorecognition andity between donor and recipient and the level of TGF-b expres-T-cell activation in the pathogenesis of chronic rejectionsion within the allograft.is suggested by studies in which blockade of T-cell co-stimulatory signals with CTLA4Ig ameliorates chronic re-jection [5, 6]. However, the relative contribution of theChronic rejection of a kidney transplant is character-direct versus indirect pathways in the pathogenesis ofized by a progressive loss of renal function associatedchronic rejection has not been defined.with arteriosclerosis, interstitial fibrosis, tubular atrophy,

Regardless of how alloantigen is recognized, severaland glomerular sclerosis. Although the pathogenesis ofchronic rejection is not well characterized, both immuno- immune effector systems have been implicated in chroniclogical and nonimmunological mechanisms may be in- rejection, including antibodies, complement, and variousvolved [1, 2]. A role for immunological factors is sug- cytokines. Because of its ability to regulate immune re-gested by clinical studies showing that previous episodes sponses and extracellular matrix production, several

studies have focused on a possible role for transforminggrowth factor-b (TGF-b) in this process. There are sev-Key words: transplantation, renal graft rejection, TGF-b, fibrosis,

mouse MHC, arteriosclerosis, glomerulosclerosis. eral lines of evidence suggesting that TGF-b may partici-pate in graft rejection. For example, urinary levels ofReceived for publication June 11, 1998TGF-b are increased in kidney transplant recipients [7,and in revised form October 15, 1998

Accepted for publication December 1, 1998 8], and during acute and chronic rejection in humans,TGF-b expression is enhanced in kidney grafts [9]. En- 1999 by the International Society of Nephrology

1935

Mannon et al: Chronic rejection of mouse kidney allografts1936

hanced TGF-b gene expression has also been demon- Measurement of kidney transplant functionstrated within rejecting cardiac allografts [10, 11]. Al- To assess kidney transplant function accurately, clear-though some of its biological effects might promote graft ances of inulin and paraaminohippuric acid (PAH) wereinjury in rejection, TGF-b may also act to suppress im- measured as previously described [16]. Clearances ofmune responses. In murine models, the administration carboxyl-14C-inulin and glycyl-3H-PAH were used as mea-of TGF-b prolonged survival of cardiac allografts [12, sures of glomerular filtration rate (GFR) and renal plasma13]. Thus, although TGF-b expression may be enhanced flow, respectively. Six isografts, four control allografts,in rejecting grafts, its role in the pathogenesis of chronic and three class I2 and four class II2 kidney allograftsallograft rejection remains unclear. were studied.

In this article, we examine the effects of altering ex-Histopathologypression of donor alloantigens in a mouse model of

chronic rejection. We find that in the absence of either Following the clearance study, a portion of each graftwas placed in formalin and stained with hematoxylin-MHC class I or class II antigens, TGF-b production iseosin and Masson’s trichrome. The sections were evalu-reduced in the graft, but the course and severity ofated by a renal pathologist (P.R.) who was masked to thechronic rejection are unchanged. Thus, reduction of anti-experimental groups. Sections were graded for overallgen density on the donor graft does not prevent theseverity of histopathological abnormalities using a semi-development of chronic rejection. Although a reducedquantitative scale as we have previously described [17].level of donor alloantigens results in reduced expressionThe severity of chronic interstitial, tubular, vascular, andof TGF-b, this decrease in intragraft TGF-b is not suffi-glomerular changes was graded separately using a semi-cient to alter the course of chronic rejection in this model.quantitative scale, in which 0 was no abnormality and 1,2, 3, and 4 represented mild, moderate, moderately severe,

METHODS and severe abnormalities, respectively. These changesincluded focal glomerular hypercellularity, membranousAnimalsthickening, focal, segmental mesangial expansion, chronicFor the control allografts, [C57BL/6 J 3129/J]F1 (H-2b)inflammation, tubular dilation and casts, fibrosis, andmice were bred in the Durham VA Animal Facility andperivascular fibrosis. An overall histological score wasused as donors. To examine the effects of donor MHCobtained by summing these individual values.expression on chronic rejection, class II-deficient mice

(class II2; H-2b) [14] or class I-deficient mice (class I2; ImmunohistologyH-2b) [15] were used as kidney donors. [BALB/c 3 DBA/

Immunopathology was performed as previously de-2 J]F1 (H-2d) mice were purchased from Jackson Labora-

scribed [18]. Briefly, kidney tissue was embedded in OCTtory (Bar Harbor, ME, USA) and were used as recipients

compound and snap frozen in precooled 2-methylbutane.of kidney grafts in all of the experiments. As nonrejecting

Tissue sections (4 mm) were cut with a cryostat, air dried,isograft controls, kidneys from [BALB/c 3 DBA/2 J]F1 and fixed in acetone. Fixed sections were washed withmice were transplanted into their littermates. phosphate-buffered saline (pH 7.4), blocked with normal

mouse serum, and stained with appropriately dilutedMouse kidney transplantationprimary monoclonal antibody. Binding of the primary

Vascularized kidney transplants were performed in antibody was detected with affinity-purified fluoresceinmice as previously described [16]. Briefly, animals were isothiocyanate-labeled goat antirat IgG, followed by aanesthetized with isoflurane, and the donor kidney, ureter, secondary layer of affinity-isolated fluorescein isothiocy-and bladder were harvested en bloc, including the renal anate-rabbit antigoat IgG (Cappel, Durham, NC, USA).artery with a small aortic cuff and the renal vein with a The primary rat monoclonal antibodies used includedsmall caval cuff. These vascular cuffs were anastomosed TIB126 (anti-class I), TIB120 (anti-class II), GK1.5 (anti-to the recipient abdominal aorta and vena cava, respec- CD4), and 3.155 (anti-CD8), prepared as hybridoma su-tively, below the level of the native renal vessels. Total pernatants (American Type Culture Collection, Rock-ischemic time averaged 35 to 40 minutes. Donor and ville, MD, USA), 30-H122 (anti-Thy 1.2; Boehringerrecipient bladders were attached dome to dome. The Mannheim, Indianapolis, IN, USA), RA36B2 (anti-right native kidney was removed at the time of trans- B220), and M1/70HL (anti-CD11b/Mac-1; Pharmingen,plant, and the left native kidney was removed through San Diego, CA, USA).a flank incision four days later. Overall surgical mortality

Quantitative immunohistologywas approximately 40%, and there were no significantdifferences in perioperative mortality between the exper- Quantitative immunohistology was performed as pre-

viously described [19, 20]. Cells were quantitated usingimental groups.

Mannon et al: Chronic rejection of mouse kidney allografts 1937

a 1 cm2 grid divided into 100 1 mm2 squares placed in the (DMSO), as previously described [21]. RNA samples weredried briefly, resuspended in 10 ml of glyoxal/DMSO solu-eyepiece of a Leitz DMRB microscope (Leica, Berlin,

Germany). Cells were counted in 10 randomly selected tion (50% DMSO, 1.2 m deionized glyoxal, 20 mm sodiumphosphate, pH 6.8), and incubated at 508C for one hour.squares and were scored with the 340 objective. Final

counts were expressed as the number of cells per square The samples were cooled on ice, loaded with 2.5 ml ofgel loading buffer (36% sucrose, 0.07% xylene cyanol,mm (mean 6 sem). All sections were evaluated by an

individual who was masked to the experimental groups. 0.07% bromphenol blue), and run on a 1.2% agarosegels in recirculating 10 mm sodium phosphate buffer,

Immunohistology of transforming growth pH 6.8. The gels were blotted onto a nylon membranefactor-b1, -b2 and collagens (Hybond-N; Amersham, Arlington Heights, IL, USA),

Tissue was fixed in neutral-buffered formalin or meth- exposed to long-wave ultraviolet light for three minutes,acarn (methanol Carnoy’s solution). At four weeks, five and baked for two hours at 808C. Hybridization wasisografts, five control allografts, and three class I2 and performed at 428C overnight with a full-length rat cDNAthree class II2 allografts were stained and analyzed. For for TGF-b1 [22] and mouse cDNA for TGF-b3 (obtainedTGF-b localization, paraffin sections from tissues fixed in from S.W. Qian, NCI, NIH) labeled with 32P by the ran-formalin or methacarn were digested with hyaluronidase dom primer method (Boehringer Mannheim). Following(Sigma, St. Louis, MO, USA) and subjected to immuno- the overnight hybridization, the filters were washed fourstaining using the Histostain kit (Zymed, Burlingame, times with 2 3 standard saline citrate (SSC)/0.1% sodiumCA, USA) using the directions supplied. The first anti- dodecyl sulfate (SDS) for 20 minutes at room tempera-body, rabbit anti-TGF-b1 peptide [CC antibody; a gift of ture, then washed once with 0.1 3 SSC/0.1% SDS forDr. Kathleen Flanders, National Cancer Institute (NCI), 30 minutes at 558C. Filters were exposed to film (KodakNIH, Bethesda, MD, USA] was applied at 2 mg/ml over- X-OMAT, Rochester, NY, USA) for up to one week atnight at 48C. This antibody recognized both TGF-b1 and 2808C. The filters were later stripped and rehybridizedTGF-b2 within the extracellular matrix. For collagen with a 32P-labeled b-actin cDNA using conditions as de-staining, methacarn-fixed tissue sections were digested scribed earlier here. The intensity of the signals on auto-with pronase. Biotinylated goat antibodies directed radiograms was quantitated by laser densitometry, andagainst collagen I and III (Southern Biotechnology Asso- the results are expressed as arbitrary densitometry unitsciates, Birmingham, AL, USA) were used at a dilution that correspond to the area under the densitometric peak.of 5 mg/ml for 60 minutes at 378C; for these antibodies,

Statistical analysisthe normal serum block and the second antibody wereomitted. Rabbit-anticollagen IV antiserum (a gift of Dr. Data are presented as the mean 6 standard error of

the mean. For the hemodynamic studies, the mean ofHynda Kleinman, NIDR, NIH) was used at a dilutionof 1:100 for 60 minutes at 378C. Negative controls for two clearance periods is given. Statistical significance was

assessed using the Student’s t-test or Wilcoxon rank sumthe first antibody included normal rabbit IgG (Zymed),normal rabbit serum (Life Technologies, Rockville, MD, test as appropriate, except for immunohistochemistry

scores, which were analyzed by analysis of variance usingUSA), and biotinylated goat antihuman Ig antibody(Zymed). These gave negligible staining. the Instat program (GraphPad, San Diego, CA, USA).

The stained sections were evaluated by observersmasked to the identity of the experimental groups. For RESULTSTGF-b, the scale included: 0 5 absent or present only

Glomerular filtration rate is reduced in long-survivingwithin the renal capsule; 0.5 5 trace staining within themouse kidney allograftsinterstitium; 1 5 limited focal interstitial staining; 2 5

We first evaluated whether mouse kidney allograftswidespread focal interstitial staining; and 3 5 widespreadthat survive for prolonged periods relative to skin orstaining. For collagens I, III, and IV, the extent of stainingcardiac allografts [16, 17, 23] develop clinical features ofwas rated in a similar fashion, with 0 5 absent; 0.5 5chronic rejection. To assess the physiological functiontrace staining; 1 5 mild increase in staining; 2 5 moderateof these allografts, we performed renal hemodynamicincrease in staining; and 3 5 considerable increase inmeasurements. Figure 1 shows GFR measured in trans-staining.planted kidneys six weeks following transplantation.

RNA isolation and analysis GFR in isografts (6.1 6 0.7 ml/min/kg) was similar tolevels seen in normal mice following uninephrectomyTotal cellular RNA was isolated from kidneys follow-[24]. Compared with the isografts, GFR was significantlying clearance studies using the guanidinium isothiocya-

nate method [21]. Expression of TGF-b1 and TGF-b3 reduced in allograft kidneys (2.6 6 0.7 ml/min/kg, P 50.008). Thus, six weeks following transplantation, renalmRNA was determined by Northern blot. Fifteen micro-

grams of RNA were sized fractionated on agarose gels, function in mouse kidney allografts was markedly de-pressed compared with nonrejecting isografts.after denaturation by glyoxal and dimethyl sulfoxide


1.5 in allografts (P , 0.005). Thus, by six weeks aftertransplantation, mouse kidney allografts have reducedrenal function and exhibit many of the histological fea-tures seen in chronic rejection of human kidney allografts.

Chronic rejection is also seen in allografts frommultiple histocompatibility complex deficient donors

To determine whether altering the potential for directrecognition of donor alloantigens might affect the char-acteristics of chronic rejection, kidneys from MHC classI2 or MHC class II2 donors (H-2b) were transplantedinto MHC-disparate (H-2d) recipients. GFR measuredin recipients of MHC-deficient grafts six weeks after trans-plantation is shown in Figure 1. The GFR in MHC classI2 (2.6 6 0.4 ml/min/kg) and MHC class II2 (3.8 61.4 ml/min/kg) allografts were not significantly differentfrom that of control allografts (2.6 6 0.7 ml/min/kg).Although the level of GFR was somewhat higher in classII2 allografts compared with the other allograft groups,

Fig. 1. Renal function six weeks following transplantation measured this difference was not statistically significant. Histologi-in isografts (h), allografts (j), MHC class I2 ( ), and MHC class II2

cal abnormalities in MHC class I2 (Fig. 2c) and MHC( ) allografts. GFR in the experimental groups is expressed in ml/min/kg. GFR was significantly reduced in allografts (*P 5 0.008) and MHC class II2 (Fig. 2d) allografts were similar qualitatively toclass I– allografts (**P 5 0.01) compared with nonrejecting isografts. those seen in control allografts. Like the control allo-

grafts, MHC-deficient grafts exhibited extensive intersti-tial fibrosis, tubular atrophy, perivascular and medialfibrosis, and significant glomerular abnormalities. Addi-Kidney allografts have histological features oftionally, the mean histological scores of MHC class I2

chronic rejection(14.6 6 2.0) and MHC class II2 allografts (12.1 6 3.8)

To determine whether the reduction in GFR was ac- were not significantly different from the scores of controlcompanied by histological abnormalities suggesting

allografts (14.4 6 1.5). Thus, the absence of normal ex-chronic rejection, we examined the kidney grafts by light

pression of either MHC class I or class II antigens on themicroscopy. Only minor histological abnormalities werekidney graft did not prevent a loss of graft function norappreciated in kidney isografts (Fig. 2a), principally con-alter the pathological characteristics of chronic rejection.sisting of mild membranous thickening of the peripheral

glomerular capillary loops and focal mesangial hypercel- Immunohistochemical analysis of allograft infiltrateslularity within the glomeruli. No significant interstitial

Significant infiltration by inflammatory cells was pres-fibrosis or vascular abnormalities were evident in theent in all of the allograft groups. To characterize theseisografts. By comparison, the control kidney allograftsimmune cell infiltrates, morphometric analysis was per-revealed marked histopathological alterations, as de-formed on sections of kidney grafts stained with mono-picted in Figure 2b. For example, allografts containedclonal antibodies against T cell and macrophage markers.an increase in glomerular mesangial and endocapillaryThe results of these analyses are shown in Figure 3.cell hyperplasia, with membranous thickening of the cap-Similar numbers of CD41 T cells infiltrating the graftsillary loops and marked periglomerular fibrosis. Exten-were seen in each group. However, the number of CD81

sive interstitial fibrosis, a hallmark of chronic rejectionT cells was significantly reduced in the MHC class I2

in humans, was also present. Control allografts also con-allografts (40 6 4 cells/mm2) compared with controls (59 6tained diffuse and patchy chronic inflammatory cell infil-6 cells/mm2, P 5 0.05) and MHC class II2 allograftstrates accompanied by significant tubular dilation and(82 6 10 cells/mm2, P 5 0.008). Although the numberatrophy. Finally, there were notable vascular alterationsof macrophages tended to be higher in control comparedin the allografts exemplified by perivascular and medialwith MHC-deficient allografts, this difference was notfibrosis, although significant intimal proliferation was notstatistically significant. Thus, despite the similar histolog-seen. Semiquantitative grading of the isografts and allo-ical appearance of chronic rejection in the allograftgrafts was performed to estimate the severity of thesegroups, the absence of class I alloantigens on the donorhistopathological abnormalities, with cumulative scoreswas associated with a significant reduction in the numberbeing generated for the individual organs. In isografts,

the mean score was 4.9 6 0.4 compared with 14.4 6 of CD81-infiltrating cells.


Fig. 2. Masson’s trichrome-stained sections of transplanted kidneys six weeks following transplantation. Relatively normal architecture was seenin nonrejecting isografts (a). Marked histological abnormalities consistent with chronic rejection are seen in allografted kidneys as described inthe text (b). In MHC class I– (c) and class II2 allografts (d), changes characteristic of chronic rejection were seen, and the scope and severity weresimilar to control allografts (magnification 3100). Reproduction of this figure in color was supported by a grant from Roche Pharmaceuticals.

Transforming growth factor-b immunostaining in tively, TGF-b staining was significantly increased in allo-mouse kidney grafts grafts compared with the negligible staining seen in iso-

grafts (2.0 6 0.4 vs. 0.4 6 0.1, P 5 0.006). To determineBecause of its ability to modulate immune responseswhether or not altered expression of donor MHC anti-and extracellular matrix protein production, TGF-b hasgens might affect this response, TGF-b immunostainingbeen suggested to be a causative factor in the pathogene-was also performed in allografts from MHC-deficientsis of chronic transplant rejection [9]. Therefore, we ex-donors (Fig. 4c). Although the distribution of TGF-bamined TGF-b expression in mouse kidney grafts. Bystaining in MHC-deficient grafts was similar to controls,four weeks after transplantation, intense TGF-b stainingthe relative intensity of staining was markedly reducedwas present within the interstitium and particularly(1.3 6 0.3 for class I2 allografts, 0.5 6 0.1 for class II2around small arteries of control allografts (Fig. 4b). Thisallografts, P 5 0.04 compared with allograft controls).staining in allografts was often localized to infiltratingThese differences in intragraft TGF-b were transient,cells rather than surrounding tubules. In contrast, asand by six weeks after transplantation, TGF-b stainingshown in Figure 4a, only trace amounts of staining for

TGF-b were seen in isografts. When analyzed quantita- was negligible in all of the experimental groups.


blood vessels (Fig. 6a). Collagen III staining was in-creased in both control (Fig. 6b) and MHC class II2

allografts (Fig. 6c), especially within the cortical intersti-tium and around small- and medium-sized arterioles. Asshown in Table 2, this was reflected by a 3.8-fold increasein the quantitative analysis of collagen III staining incontrol allografts compared with isografts (P 5 0.01)and a 4.3-fold increase in staining in the MHC class II2

allografts compared with isografts (P , 0.001). CollagenI staining was also significantly increased in both control(ninefold, P 5 0.01) and MHC class II2 allografts (10-fold, P , 0.001) compared with isografts (Table 2). Inthe control and class II2 allografts, increases in the inten-sity of staining in the cortical and medullary interstitiumwere observed (data not shown). Collagen IV stainingwas observed in the basement membranes of the glomer-uli, tubules, blood vessels, and glomerular mesangium inFig. 3. Morphometric analysis of graft-infiltrating lymphocytes. Sec-

tions of kidney grafts were stained and quantitatively analyzed by immu- isografts (data not shown). Similarly, collagen IV in con-nofluorescence as described in the Methods section. Allografts ( ), trol and class II2 allografts was present in normal amountsMHC class I2 allografts ( ), and class II2 allografts ( ) are shown.

in basement membrane, but was modestly increasedNo infiltrating cells were detected in isografts. There were no differencesin the numbers of CD41 T cells infiltrating the allografts. However, the within the glomerular mesangium (data not shown).number of CD81 T cells was significantly reduced in MHC class I2

allografts compared with controls and MHC class II2 allografts (*P 50.05 and P 5 0.008, respectively). The number of macrophages was DISCUSSIONreduced in MHC-deficient grafts compared with controls, but this wasnot significant. Chronic rejection is a common complication of suc-

cessful kidney transplantation in humans. The character-istic clinical picture of chronic renal allograft rejectionconsists of a progressive decline in GFR, often associated

Transforming growth factor-b mRNA expression in with proteinuria. Although this entity is the leading causemouse kidney grafts of late kidney graft failure, its pathogenesis is not well

To determine whether differences in intragraft TGF-b characterized, and there are no effective treatments onceprotein accumulation were associated with alterations in these particular lesions have developed. To better under-intragraft levels of TGF-b mRNA, we assessed mRNA stand the mechanisms contributing to the developmentlevels of TGF-b1 and TGF-b3 by Northern blot analysis of chronic rejection, various animal models have beenfour weeks following transplantation. As shown in Figure used. These include accelerated arteriosclerosis in mouse5, TGF-b mRNA levels generally correlated with protein carotid artery loops transplanted onto the carotid arterylevels. Expression of TGF-b1 and TGF-b3 mRNA in of a histoincompatible recipient [25], coronary arterio-control allografts was increased 3.3-fold and 6.6-fold, sclerosis in vascularized mouse [26] and rat hearts [27],respectively, compared with isografts (Table 1). TGF- and chronic rejection in F344 3 LEW kidney allograftsb1 and TGF-b3 mRNA levels in MHC class II2 allografts [28]. In this report, we have documented the develop-were modestly reduced compared with control allografts ment of chronic rejection in a mouse model of kidney(2.1-fold and 5.8-fold increase in expression, respec- transplantation. These mouse renal allografts develop atively, compared with isografts). Similar results were number of features that are observed in chronic rejectionseen with MHC class I2 grafts (data not shown). Al- in human kidney grafts, including interstitial fibrosis andthough these differences in mRNA levels between trans- glomerulosclerosis. This model therefore provides a sys-plant groups were substantial, they did not reach statisti- tem for studying the mechanisms of kidney injury incal significance. chronic rejection, and we have used it to examine the

role of altering donor MHC antigen expression on theAccumulation of collagen in mouse kidney allografts character of chronic rejection.

Interstitial fibrosis and glomerular sclerosis are two of The triggering stimulus for alloimmune responses, in-the defining features of chronic rejection. To evaluate cluding those that lead to chronic rejection, is the recog-the nature of the abnormal extracellular matrix in these nition of alloantigens from the donor organ by T cellskidneys, we performed immunoperoxidase staining for of the recipient. Direct recognition of alloantigens, incollagens type I, III (Fig. 6), and IV. In isografts, collagen which donor MHC antigens are directly recognized by

the antigen receptor on recipient T cells, has been sug-III was present in trace amounts in areas adjacent to


gested to contribute to the intensity and high precursor with glomerular injury [30], and a number of studieshave suggested that TGF-b may play a similar role infrequency of alloimmune responses and may be mostchronic rejection [7, 9, 10]. In particular, the effects ofimportant in initiating acute allograft rejection [5, 6]. InTGF-b to stimulate extracellular matrix production andcontrast, the indirect pathway for allorecognition hasdeposition may be a key component in the pathogenesisbeen implicated in the pathogenesis of chronic rejectionof chronic kidney injury [30, 31]. In comparing the con-[29]. In previous studies, we demonstrated that kidneytrol and MHC-deficient allografts, we find that the inten-grafts from donor mice lacking class I MHC expressionsity of TGF-b expression within the grafts is dissociatedhave relatively well-preserved renal function during thefrom the development of chronic injury and, specifically,early weeks following transplantation, despite histologi-with abnormal deposition of collagen chains in the renal

cal changes characteristic of acute rejection [16]. Kidney interstitium. Thus, despite the reduced levels of TGF-bgrafts from donors lacking MHC class II antigens also found in MHC-deficient grafts, the severity of kidneymaintained GFR and demonstrated an infiltrating cell injury and the enhancement of collagen deposition inpopulation with a relative deficiency of CD41 T cells these grafts are virtually identical to controls. This could(abstract; Mannon et al, J Am Soc Nephrol 5:984, 1994). be due to altered receptor expression for TGF-b within

In contrast, allografts from MHC-deficient donors de- the experimental groups. Alternatively, factors othervelop pathological changes of chronic rejection of the than TGF-b may contribute to the development intersti-same severity as allografts from mice with normal MHC tial and perivascular fibrosis in this model. These mightantigen expression. The occurrence of chronic rejection, include other fibrogenic growth factors and cytokines,

such as fibroblast growth factor [11, 32] and platelet-despite a dramatic decrease of donor MHC antigen ex-derived growth factor [33, 34], insulin-like growth factorpression, suggests that only a threshold level of antigen[35], interleukin-6 [36], connective tissue growth factorstimulation may be required to initiate these lesions.[37], or vasoactive factors such as inducible nitric oxideAlternatively, the mechanisms leading to chronic rejec-synthase [38, 39]. Finally, the role of TGF-b in this modeltion in this model may be relatively insensitive to a re-may be complex and may not depend on a sustainedduced capacity for allorecognition. As significant andintrarenal elevation of TGF-b production.equivalent numbers of CD41 cells are seen in both MHC-

The mouse model that we describe here has a numberdeficient groups, these cells may also play a role in theof features that are similar to chronic rejection in hu-

pathogenesis of this disorder. mans. However, there are also some important differ-Although altered expression of donor alloantigens did ences. For example, the tempo of the development of

not affect the course and severity of chronic rejection, it renal fibrosis in this model is more rapid than that typi-was clearly associated with altered expression of TGF-b cally seen in humans with chronic rejection. Further-within the grafts. In control allografts four weeks after more, in humans and in a number of animal models, grafttransplant, we found that intragraft TGF-b levels were arteriosclerosis with intimal proliferation is a cardinalstrikingly enhanced. Although there was a modest increase feature of chronic rejection. Although we identify manyin TGF-b protein in four-week MHC-deficient grafts rela- of the histopathological features of chronic rejection intive to isografts, the level of expression was notably less this mouse model, the vascular lesions, although severe,

do not precisely mimic the changes evident in humanthan allograft controls. This difference in TGF-b is espe-renal allograft rejection. There is little evidence of arte-cially interesting given the putative relationship betweenriosclerosis. Intimal proliferation is absent, and medialTGF-b and chronic rejection [7, 9, 10]. The mechanismproliferation predominates. The lack of significant arte-underlying this difference in TGF-b expression in MHC-riosclerosis and fibrointimal hyperplasia in these murinedeficient grafts is unclear. As reduced TGF-b levels werekidney allografts may reflect the duration of the response.seen in both class I- and class II-deficient grafts, the cellOur studies were performed at six weeks after trans-populations responsible for generating TGF-b do notplantation, and the development of florid arteriosclerosisseem to result from a predominant CD41 or CD81-is typically not seen until 12 to 16 weeks in rat kidney

mediated response. Rather, reduced TGF-b productionor heart grafts and as early as eight weeks in mouse

in the MHC-deficient grafts may reflect a reduction in cardiac allografts. In a previous study in mouse kidneythe overall intensity of the alloimmune response. Finally, allografts, vascular changes were not seen until 10 weeks,because we have examined TGF-b expression at only and then only in large sized arteries [23]. Although thetwo time points—four and six weeks—it is also possible reasons for the lack of characteristic vascular involve-that the kinetics of TGF-b expression differ between the ment in this model are unclear, our studies suggest thatMHC-deficient and control animals. arteriosclerosis is not a prerequisite for the development

Transforming growth factor-b appears to have a role in of glomerulosclerosis and interstitial fibrosis in chronicrejection.the development of progressive kidney injury associated

Fig. 4. Immunoperoxidase staining of transplanted kidneys for TGF-b1 and TGF-b2 four weeks following transplantation. In isografts (a,N 5 5), there was minimal staining. In allografts (b, N 5 5), there wasintense staining, particularly within the interstitium and around smallarteries, as shown. In class II2 allografts (c, N 5 3), this pattern wassimilar, but staining was less intense (magnification 350). Reproduc-tion of this figure in color was supported by a grant from Roche Pharma-ceuticals.

Fig. 6. Immunoperoxidase staining for collagen type III in isografts(a), control allografts (b), and MHC class II2 allografts (c) four weeksfollowing transplantation. Only trace amounts of collagen III are seenin isografts compared with the increase in staining within the interstitiumof control allografts. In class II2 kidneys, collagen III staining was similarto and nearly indistinguishable from control allografts (magnification3100). Reproduction of this figure in color was supported by a grantfrom Roche Pharmaceuticals.


Fig. 5. Northern blot analysis for TGF-b in transplanted kidneysfour weeks following transplantation. Results were analyzed bylaser densitometry and normalized for b-actin expression. Com-pared with isografts, enhanced expression for TGF-b1 and TGF-b3was seen in control allografts and was modestly reduced in MHCclass II2 allografts.

Table 1. Quantitation of TGF-b mRNA expression on ACKNOWLEDGMENTSNorthern blot by laser densitometry

These studies were supported by National Institutes of Health grantTransplant type TGF-b3 TGF-b3 PO1-DK38103 and by grants from the Research Service of the Veterans

Administration. Roslyn B. Mannon is supported by a National Insti-Isografts 179616 27611tutes of Health Mentored Clinical Scientist Development Award (1Allografts 5836111 177688K08 AI01389-01). A preliminary version of this report was publishedClass II allografts 367 630 156643in abstract form (J Am Soc Nephrol 4:914, 1993). Reproduction of

All values are mean arbitary densitometry units 6 sem and normalized to color figures was made possible by a grant from Roche Pharmaceuticals.b-actin content. The authors thank Ms. Norma Turner for secretarial assistance and

Ms. Karen Ferri Roberts for expert technical assistance.

Reprint requests to Roslyn B. Mannon, M.D., Nephrology SectionTable 2. Quantitation of collagen expression in (111I), VA Medical Center, 508 Fulton Street, Durham, North Carolinatransplanted kidneys 27705, USA.E-mail: [email protected] type Collagen I Collagen III

Isografts 0.260.1 0.660.2REFERENCESAllografts 1.860.5a 2.360.5a

Class II allografts 2.0 60.2b 2.660.2b

1. Hostetter TH: Chronic transplant rejection. Kidney Int 46:266–Staining intensity was graded on a semiquantitative scale as described in the 279, 1994

Methods section. 2. Matas AJ, Burke JF, DeVault GA Jr, Monaco A, Pircsh JD:a P 5 0.01 compared to isografts Chronic Rejection. J Am Soc Nephrol 4:S23–S29, 1994b P , 0.001 compared to isografts3. Almond PS, Matas A, Gillingham K, Dunn DL, Payne WD,

Gores P, Gruessner R, Najarian JS: Risk factors for chronicrejection in renal allograft recipients. Transplantation 55:752–757,1993

In summary, long-term surviving mouse kidney allo- 4. Tilney NL, Paul LC: Antigen-Independent Events Leading tografts demonstrate reduced renal function and histopath- Chronic Graft Dysfunction. Philadelphia, Lippincott-Raven, 1996

5. Azuma H, Chandraker A, Nadeau K, Hancock WW, Carpenterological changes resembling chronic rejection of humanCB, Tilney NL, Sayegh MH: Blockade of T-cell costimulationkidney grafts. In kidney grafts with reduced expressionprevents development of experimental chronic renal allograft re-

of MHC class I or II antigens, intragraft TGF-b expres- jection. Proc Nat Acad Sci USA 93:12439–12444, 1996sion was reduced, but the severity of the functional and 6. Russell ME, Hancock WW, Akalin E, Wallace AF, Glysing-

Jensen T, Willett TA, Sayegh MH: Chronic rejection in thepathological abnormalities of chronic rejection was notLEW to F344 rat model: Blockade of CD28-B7 costimulation byaltered. Thus, abrogation of either class I or II expressionCTLA4Ig modulates T cell and macrophage activation and attenu-

is by itself insufficient to prevent the development of ates arteriosclerosis. J Clin Invest 97:833–838, 19967. Gaciong A, Koziak K, Religa P, Lisiecka A, Morzycka-Micha-chronic rejection.


lik M, Rell K, Kozlowska-Boszko B, Lao M: Increased expres- 23. Russell PS, Chase CM, Colvin RB, Plate JMD: Kidney trans-plants in mice: An analysis of the immune status of mice bearingsion of growth factors during chronic rejection of human kidney

allograft. Transplant Proc 27:928–929, 1995 long term, H-2 incompatible transplants. J Exp Med 147:1449–1468,19788. Coupes BM, Newstead CG, Short CD, Brenchley PEC: Trans-

forming growth factor b1 in renal allograft recipients. Transplanta- 24. Nesbitt T, Coffman TM, Griffiths R, Drezner MK: Cross-trans-plantation of kidneys in normal and hyp-mice: Evidence that thetion 57:1727–1731, 1994

9. Shihab FS, Yamamoto T, Nast CC, Cohen AH, Noble NA, Gold hyp-mouse phenotype is unrelated to an intrinsic renal defect.J Clin Invest 89:1453–1459, 1992LI, Border WA: Transforming growth factor-b and matrix protein

expression in acute and chronic rejection of human renal allografts. 25. Shi C, Russell ME, Bianchi B, Newell JB, Haber E: Murinemodel of accelerated transplant arteriosclerosis. Circ Res 75:199–J Am Soc Nephrol 6:286–294, 1995207, 199410. Waltenberger J, Wanders A, Fellstrom B, Miyazono K, Heldin

26. Russell PS, Chase CM, Winn HJ, Colvin RB: Coronary athero-C-H, Funa K: Induction of transforming growth factor-b duringsclerosis in transplanted mouse hearts. Am J Pathol 144:260–274,cardiac allograft rejection. J Immunol 151:1147–1157, 1993199411. Zhao XM, Frist WH, Yeoh TK, Miller GG: Expression of cyto-

27. Adams DH, Tilney NL, Collins JJ, Karnovsky MJ: Experimentalkine genes in human cardiac allografts: Correlation of IL-6 andgraft arteriosclerosis. I. The Lewis-to-F344 allograft model. Trans-transforming growth factor-beta (TGF-b) with histological rejec-plantation 53:1115–1119, 1992tion. Clin Exp Immunol 93:448–451, 1993

28. White E, Hildemann WH, Mullen Y: Chronic kidney allograft12. Raju GP, Belland SE, Eisen HJ: Prolongation of cardiac allograftreactions in rats. Transplantation 8:602–617, 1969survival with transforming growth factor-b1 in rats. Transplantation

29. Muluk SC, Clerici M, Via CS, Weir MR, Kimmel PL, Shearer58:392–396, 1994GM: Correlation of in vitro CD41 T helper cell function with13. Wallick SC, Figari IS, Morris RE, Levinson AD, Palladinoclinical graft status in immunosuppressed kidney transplant recipi-MA: Immunoregulatory role of transforming growth factor bents. Transplantation 52:284–291, 1991(TGF-b) in development of killer cells: Comparison of active and

30. Sharma K, Ziyadeh FN: The emerging role of transforming growthlatent TGF-b1. J Exp Med 172:1777–1784, 1990factor-b in kidney diseases. Am J Physiol 266:F829–F842, 199414. Cosgrove D, Gray D, Dierich A, Kaufman J, Lemeur M, Benoist

31. Bruijn JA, Roos A, de Geus B, de Heer E: Transforming growthC, Mathis D: Mice lacking MHC class II molecules. Cell 66:1051–factor-b and the glomerular extracellular matrix in renal pathology.1066, 1991J Lab Clin Med 123:34–45, 199315. Koller BH, Marrrack P, Kappler JW, Smithies O: Normal devel-

32. Zhao XM, Citrin BS, Miller GG, Frist WH, Merrill WH,opment of mice deficient in b2-m, MHC class I proteins, and CD81

Fischell TA, Atkinson JB, Yeoh TK: Association of acidic fibro-T cells. Science 248:1227–1230, 1990blast growth factor and untreated low grade rejection with cardiac16. Coffman T, Geier S, Ibrahim S, Griffiths R, Spurney R, Smithies allograft vasculopathy. Transplantation 56:1005–1010, 1995O, Koller B, Sanfilippo F: Improved renal function in mouse 33. Azuma H, Heemann UW, Tullius SG, Tilney NL: Cytokines andkidney allografts lacking MHC class I antigens. J Immunol 151:425– adhesion molecules in chronic rejection. Clin Transplant 8:168–180,

435, 1993 199417. Mannon RB, Nataraj C, Kotzin BL, Griffiths R, Geier S, 34. Abboud HE: Role of platelet-derived growth factor in renal injury.

Ibrahim S, Sanfilippo FP, Platt JL, Kurlander R, Coffman TM: Am J Physiol 57:297–309, 1995Rejection of kidney allografts by MHC class I deficient mice. Trans- 35. Motomura N, Lou H, Maurice P, Foegh ML: Acceleration ofplantation 59:746–755, 1995 arteriosclerosis of the rat aorta allograft by insulin growth factor-I.

18. Wrenshall LE, Cerra FB, Rubinstein P, Platt JL: Regulation Transplantation 63:932–936, 1997by heparan sulfate and interleukin-1a of the ontogenic expression 36. Russell ME, Wallace AF, Hancock WW, Sayegh MH, Adamsof T-cell receptor, CD4, and CD8 in developing thymus. Hum DH, Sibinga NES, Wyner LR, Karnovsky MJ: Upregulation ofImmunol 38:165–171, 1993 cytokines associated with macrophage activation in the Lewis-

19. Gomez-Morales M, Bustos M, Montes A, Andujar M, Medina- to-F344 rat transplantation model of chronic cardiac rejection.Cano MT, Ramirez C, O’Valle F, Aguilar D, Aneiros J, Garcia Transplantation 59:572–578, 1995del Moral R: Influence of intrarenal deposits of cyclosporin A 37. Ito Y, Aten J, Bende RJ, Oemar BS, Rabelink TJ, Weening JJ,on acute renal transplant rejection. Nephron 70:402–409, 1995 Goldschmeding R: Expression of connective tissue growth factor

20. Platt JL, Lebien TW, Michael AF: Interstitial mononuclear cell in human renal fibrosis. Kidney Int 53:853–861, 1998populations in renal graft rejection. J Exp Med 155:17–30, 1982 38. Russell ME, Wallace AF, Wyner LR, Newell JB, Karnovsky

21. Sambrook J, Fritsch E, Maniatis T: Molecular Cloning: A Labora- MJ: Upregulation and modulation of inducible nitric oxide syn-tory Manual (2nd ed). Cold Spring Harbor, Cold Spring Harbor thase in rat cardiac allografts with chronic rejection and transplantPress, 1989 arteriosclerosis. Circulation 92:457–464, 1995

22. Qian SW, Kondiah P, Roberts AB, Sporn MB: cDNA cloning 39. Koglin J, Glysing-Jensen T, Mudgett JS, Russell ME: Exacer-by PCR of rat transforming growth factor beta-1. Nucleic Acids bated transplant arteriosclerosis in inducible nitric oxide-deficient

mice. Circulation 97:2059–2065, 1998Res 18:3059, 1990

Documents

Chronic rejection of mouse kidney allografts