The plasminogen/plasmin system has the potential to affect the outcome of inflammatory diseases by regulating accumulation of fibrin and other matrix proteins. In human and experimental crescentic glomerulonephritis (GN), fibrin is an important mediator of glomerular injury and renal impairment. Glomerular deposition of matrix proteins is a feature of progressive disease. To study the role of plasminogen and plasminogen activators in the development of inflammatory glomerular injury, GN was induced in mice in which the genes for these proteins had been disrupted by homologous recombination. Deficiency of plasminogen or combined deficiency of tissue type plasminogen activator (tPA) and urokinase type plasminogen activator (uPA) was associated with severe functional and histological exacerbation of glomerular injury. Deficiency of tPA, the predominant plasminogen activator expressed in glomeruli, also exacerbated disease. uPA deficiency reduced glomerular macrophage infiltration and did not significantly exacerbate disease. uPA receptor deficiency did not effect the expression of GN. These studies demonstrate that plasminogen plays an important role in protecting the glomerulus from acute inflammatory injury and that tPA is the major protective plasminogen activator.

The plasminogen/plasmin system has important functions in the dissolution of fibrin and in the breakdown of other extracellular matrix proteins (1). It has pivotal roles in clot lysis and wound healing (2, 3), but also has the potential to influence the outcome of inflammatory processes, particularly where fibrin and extracellular matrix deposition play an important role (4).

Glomerulonephritis (GN) results from immunologically mediated glomerular inflammation and is the most common cause of renal failure worldwide. Glomerular fibrin deposits are observed in severe proliferative and in crescentic forms of human GN, in which the outcome for renal function is often poor (5). These forms of GN are associated with recruitment of T cells and monocytes (6) and cross-linking of fibrin both in the glomerular tuft and within Bowman's space. Fibrin in Bowman's space is chemotactic for macrophages (7). This accumulation of macrophages and fibrin forms the histological appearance of a crescent (8, 9).

The glomerulus, being a complex filter, is particularly susceptible to injury induced by local deposition of fibrin and other matrix proteins. Defibrination studies have demonstrated that fibrin is an important mediator of glomerular injury and crescent formation in experimental GN (8, 9). Local dysregulation of coagulation (10) and fibrinolysis (11) may contribute to glomerular fibrin deposition. Upregulation of tissue factor has been demonstrated in experimental (12) and human crescentic GN (13). Dysregulation of the plasminogen/plasmin system, with decreased glomerular fibrinolytic activity, decreased tissue type plasminogen activator (tPA) and increased plasminogen activator inhibitor-1 (PAI-1) has also been demonstrated in experimental crescentic GN (11) and in human GN (14).

The extent to which local generation of plasmin plays a protective role in preventing glomerular fibrin deposition in crescentic GN is not known. Plasmin can cleave crosslinked fibrin, directly degrade matrix proteins (1) and can activate matrix metalloproteinases (15), suggesting a protective function for plasmin in the glomerulus. Conversely, plasmin may activate latent transforming growth factor-β (16), which promotes matrix deposition (17), enhances PAI-1 production (17) and down regulates plasminogen activator production in the kidney (18).

To evaluate the plasminogen/plasmin system in immune initiated glomerular injury, proliferative GN was induced in mice with genetic deficiencies of key molecules of this system. These included mice deficient in plasminogen (Plg−/−) (19), mice with combined deficiencies of tPA and uPA (tPA−/−:uPA−/−) (2), mice with single deficiencies of either tPA (tPA−/−) or urokinase-type plasminogen activator (uPA−/−) (2) and mice deficient in the cell surface receptor for uPA, (uPAR−/−) (20). Mice were compared with their appropriate genetically unmodified (or wild-type [WT]) mice in each case. GN was induced by planting an exogenous Ag, sheep anti–mouse GBM globulin, in the glomeruli of mice previously sensitized to sheep globulin (SG) which induced a rapidly progressive, proliferative GN with renal impairment in WT mice.

Materials And Methods

Study Design.

Mice deficient in plasminogen, tPA, uPA, both tPA and uPA, and uPAR were created and characterized as previously described (2, 19, 20). Mice were 48–69 d old and clinically free of disease at induction of GN. Studies of Plg−/−, tPA−/−, uPA−/−, uPAR−/− mice and their WT mice used male mice weighing between 20 and 30 g, except Plg−/− mice, which weighed between 10.7 and 21.3 g. tPA−/−:uPA−/− mice and their WT controls were female. WT female mice weighed 16.9 to 20.5 g and tPA−/−:uPA−/− mice weighed 13.7 to 18.8 g.

Sheep anti–mouse GBM globulin was prepared as previously described (21). Mice were sensitized with 2 mg SG in 200 μl of Freund's Complete Adjuvant (Sigma Chemical Co., Castle Hill, Australia) injected s.c. at two sites. 5 mg of sheep anti–mouse GBM globulin was injected i.v. after 10 d to initiate glomerular disease. The dose of anti-mouse GBM globulin was reduced to 4 mg for mice weighing between 15.0 to 19.9 g and to 3 mg for mice less than 15.0 g in weight.

Animals were studied 10 days after anti-GBM globulin injection. tPA−/− mice (n = 10), uPA−/− mice (n = 9), and uPAR−/− mice (n = 8) were matched with WT mice (n = 9), with equal fractions of each WT genotype. Female tPA−/−:uPA−/− mice (n = 5) were matched with female WT mice (n = 7). Because of severity of disease in the Plg−/− mice (n = 8), they were studied 8 days after induction of disease, together with WT mice (n = 4). Two Plg−/− mice, which were killed because of severe disease (at days 6 and 7), and another Plg−/− animal that died on day 7 of disease were included in the analysis.

Assessment of Glomerular Injury.

Renal tissue was fixed in Bouin's fixative, embedded in paraffin, cut in 3 μm sections and stained with periodic acid Schiff's stain (PAS) for histological analysis. Glomerular cellularity was determined by counting cell nuclei in at least 20 glomeruli per animal and results expressed as cells per glomerular cross section (cells/gcs). The presence of PAS+ material within glomeruli was scored semiquantitatively in a minimum of 50 glomeruli per mouse as follows: 0 = no deposition of PAS+ material, 1 = up to one third of the cross-sectional area of the glomerulus staining PAS+, 2 = one third to two thirds involvement, 3 = greater than two thirds involvement. Crescent formation (defined as the presence of 3 or more layers of cells in Bowman's space) was assessed (minimum of 50 glomeruli per mouse) and results expressed as a percentage incidence.

Tissue sections (6 μm) were stained to demonstrate macrophages using a three-layer immunoperoxidase technique, as previously described (22). The primary Ab was M1/70 (monoclonal anti-mouse Mac-1; American Type Culture Collection [ATCC], Rockville, MD). Sections of spleen provided a positive control for each animal and protein G–purified rat Ig was substituted for the primary Ab to provide a negative control. A minimum of 20 glomeruli were assessed per mouse and results expressed as cells/gcs.

Fibrin immunofluorescence was performed on sections (4 μm) cut from frozen tissue. An IgG fraction of goat anti–mouse fibrin/ fibrinogen serum (Nordic Immunological Laboratories, Drawer, CA) was conjugated to digoxigenin (Boehringer Mannheim, Castle Hill, Australia) and used at a concentration of 35 μg/ml. Binding of the primary Ab was detected with FITC-conjugated sheep anti-digoxigenin Fab fragments (Boehringer Mannheim) at a dilution of 1 in 50. Glomerular fibrin deposition was assessed semiquantitatively (minimum of 30 glomeruli per mouse) using the scoring system described for the assessment of glomerular PAS+ material.

Serum creatinine (sCr) was measured by the alkaline picric acid method by autoanalyzer (Cobas Bio, Roche Diagnostics, Basel, Switzerland). As sCr is proportional to weight, values for Plg−/− and tPA−/−:uPA−/− mice were adjusted to the mean weight of their WT group at the end of the experiment.

Measurement of Circulating Mouse Anti–Sheep Globulin Antibody.

Mouse anti–sheep globulin titers were assessed by ELISA on serum collected at the end of the experiment. Microtiter plates were coated with 10 μg/ml SG in carbonate/bicarbonate buffer (pH 9.5), washed, blocked with 1% BSA, then incubated with dilutions of mouse serum. Bound mouse Ig was detected with horseradish peroxidase conjugated rabbit anti–mouse Ig (Sigma Chemical Co.) at a dilution of 1 in 2,000 using 0.1 M 2,2′-azinodi-3-ethylbenzthiazoline sulphonate (ABTS; Boehringer Mannheim, Castle Hill, Australia) in 0.02% H2O2 as the substrate. The absorbance at 405 nm was read on a microtiter plate reader. Serum from each mouse was tested in serial dilutions from 1 in 100 to 1 in 126,000, with serum from six non-immunized mice used as negative controls.

Statistics.

Results are expressed as the mean ± SEM. Significances of differences were determined by the unpaired t test for the Plg−/− and tPA−/−:uPA−/− mice and their respective WT. Multiple group comparisons (i.e., tPA−/−, uPA−/−, uPAR−/− and WT mice) were analyzed by ANOVA, followed by Fisher's protected least significant differences (PLSD) test.

Results

Effects of Plasminogen and Plasminogen Activator Deficiencies on Histological Features of GN.

All WT mice developed proliferative GN (Fig. 1,A) with occasional glomeruli exhibiting crescent formation. Areas of basement membrane thickening were demonstrated by silver methenamine staining (not shown) and segmental fibrin deposition was observed within the glomerular tuft by direct immunofluorescence (Fig. 1,B). Glomeruli of both Plg−/− and tPA−/−:uPA−/− mice were more severely affected, with tuft necrosis and relatively hypocellular glomeruli (Fig. 1, C and E). Despite this relative hypocellularity, there were more macrophages within glomeruli and significantly increased numbers of glomeruli exhibiting crescent formation compared to their respective WT controls (Table 1). Crescents in the Plg−/− and tPA−/−: uPA−/− animals were larger than those found in WT mice and contained both cellular elements (Fig. 1, C and E, arrowed), and fibrin (Fig. 1, D and F, arrowed). Deposition of PAS+ material within glomeruli was significantly increased (Table 1) indicating more extra cellular matrix deposition in nephritic Plg−/− and tPA−/−:uPA−/− mice, compared to WT animals.

Of mice deficient in a single plasminogen activator or uPAR, only tPA−/− mice exhibited histologically more severe disease with increased glomerular hypercellularity and increased numbers of glomerular macrophages. Deposition of PAS+ material was increased compared to WT mice (Table 1) but did not reach the extent of the Plg−/− or tPA−/−:uPA−/− mice. Mice deficient in uPA had fewer macrophages infiltrating glomeruli compared to WT mice, but no other differences in the histological appearances or quantitative indices of renal injury (Table 1). Nephritic uPAR−/− mice showed similar histological appearances to WT mice (Table 1).

Effect of Plasminogen and Plasminogen Activator Deficiencies on Glomerular Fibrin Deposition in GN.

In contrast to the segmental fibrin deposition observed only within the glomerular tuft in WT mice (Fig. 1,B), Plg−/− and tPA−/−: uPA−/− mice developed extensive glomerular fibrin deposition within the glomerular tuft and in Bowman's space (Fig. 1, D and F). Semi-quantitative scoring of the extent of glomerular fibrin deposition confirmed that these differences were statistically significant compared to their WT controls (Fig. 2). Mice deficient in tPA also showed significant increases in glomerular fibrin deposition over WT (Fig. 3), though the extent of glomerular fibrin deposition did not reach that of the Plg−/− or the tPA−/−:uPA−/− mice. There were no significant changes in glomerular fibrin deposition in nephritic uPA−/− and uPAR−/− mice compared to WT mice.

Effect of Plasminogen and Plasminogen Activator Deficiencies on Renal Function in GN.

The importance of the increased glomerular deposition of fibrin and other matrix proteins induced by deficiency of plasminogen activation was demonstrated by the increased impairment of renal function observed in plasminogen or plasminogen activator-deficient mice. Plg−/− and tPA−/−:uPA−/− mice developed more severe renal failure, measured by sCr compared to their appropriate WT groups (Fig. 2). Mice deficient in tPA alone also developed more renal impairment compared to their WT control. uPA−/− mice showed a trend towards more severe renal impairment (sCr P = 0.08 compared to WT). The severity of renal impairment of nephritic uPAR−/− mice was unchanged from that of WT mice with GN (Fig. 3).

Assessment of Systemic Immune Responses.

In addition to the activity of the coagulation and fibrinolytic systems, the extent of immune response to the disease initiating antigen is important in determining the severity of injury in this model of GN. Although deficiencies of the plasminogen/ plasmin system are not known to modulate immune responses, the specific immune response to the nephritogenic Ag (SG) was quantified in the various groups of mice with GN by measurement of circulating levels of mouse anti–sheep globulin antibodies. Ab titers, measured by ELISA at dilutions of sera from 1 in 100 to 1 in 102,400, were not significantly different for any genetically deficient mice compared with their WT control group, except in Plg−/− mice that showed slightly lower absorbance values at the two highest antibody dilutions. These minor differences are unlikely to significantly modulate the expression of GN and could not contribute to the more severe disease seen in these Plg−/− mice.

Discussion

These studies demonstrate that generation of plasmin plays a critical role in protecting glomerular structure and function in rapidly progressive GN. A similar degree of severe glomerular damage was observed with plasminogen deficiency and combined deficiency of tPA and uPA, suggesting that these two plasminogen activators contribute the major plasminogen activating activity in GN. This supports the earlier observation that tPA and uPA are the only two physiologically significant plasminogen activators in mice (19). Mice deficient in tPA alone also developed more severe renal injury than WT mice, but not to the extent observed in combined PA or the Plg-deficient mice. More severe GN did not develop in uPA−/− or uPAR−/− mice.

In the initial characterization of mice with combined deficiencies of tPA and uPA, minor spontaneous renal fibrin deposition was reported in a minority of animals beyond the age of 2 to 3 mo (2). However, in this study disease was induced in tPA−/−:uPA−/− mice that were younger than 2 mo old to avoid any possibility that spontaneous glomerular fibrin deposition may significantly influence the results. In the original description of Plg−/− mice, dilated renal pelvises had been noted in four mice, without fibrin deposition (19). None of the Plg−/− mice in the current study had macroscopic or microscopic evidence of pelvicalyceal dilatation or obstruction.

There was a significantly increased propensity to form crescents in glomeruli of mice with plasminogen or combined plasminogen activator deficiency. Crescents are the result of accumulation of fibrin and macrophages, together with proliferation of glomerular epithelial cells in Bowman's space. They are indicative of severe glomerular injury and a poor outcome for renal function. Fibrin has previously been shown to be an important chemotactic stimulus for accumulation of macrophages in Bowman's space and crescent formation (7). The enhanced crescent formation and macrophage infiltration in Plg−/− and tPA−/−: uPA−/− mice is consistent with the observation that enhanced glomerular fibrin accumulation occurring in the absence of local plasmin generation provides a greater stimulus for inflammatory cell recruitment and crescent formation. The more intense inflammatory reaction resulted in necrosis of the glomerular tuft and hypocellular glomeruli.

tPA is the predominant PA within mouse (23), rabbit (11) and human (24, 25) glomeruli and is produced mainly by glomerular endothelial cells. The major site of uPA production within the murine kidney is the tubular epithelium (23), although human glomerular epithelial cells in culture also produce uPA (24, 26) and uPA immunofunctional activity has been demonstrated in rabbit (11) and human glomeruli (25). uPA has been proposed to have a role in maintenance of tubular patency (23).

Mice deficient in tPA alone showed significantly increased renal failure, glomerular hypercellularity, glomerular macrophage numbers, glomerular fibrin deposition and deposition of PAS+ material within glomeruli. They also demonstrated a trend towards increased crescent formation. Glomerular capillary thrombosis was minimal in nephritic tPA−/− mice and their glomeruli were hypercellular. However, the severity of injury in these mice was not sufficient to cause necrosis of the glomerular tuft.

Mice deficient in uPA alone did not show significant differences in histological indices of disease from WT mice, but did show a trend towards increased renal impairment. The significance of this result is unclear, but the extensive expression of uPA in renal tubular epithelial cells of normal mice suggests that tubular dysfunction may have contributed to any increased renal impairment in uPA−/− mice. The demonstration of significantly reduced glomerular macrophage infiltration in uPA−/− mice supports the hypothesis that uPA assists migration of macrophages to sites of inflammatory injury.

The increased renal impairment, fibrin deposition and other histological indices of disease in the tPA−/− mice suggest a more important role for tPA in protecting glomeruli against inflammatory injury, but do not exclude a role for uPA. However, the more severe glomerular injury in mice deficient in both plasminogen activators implies that the effects of uPA deficiency in glomerular inflammation become evident in the absence of tPA.

Deficiency of the uPA receptor did not significantly modify the histological or functional manifestations of GN. The observation that glomeruli of uPAR-deficient mice contained similar numbers of macrophages to WT mice suggests that unlike uPA, uPAR is not important in macrophage infiltration into nephritic glomeruli. In normal mice, uPAR mRNA is expressed only at low levels in the kidney, in contrast to the abundant renal expression of uPA mRNA (27). Sites of uPA expression in the kidney do not correspond well to its receptor (27), suggesting that in the kidney uPA can act independently of uPAR. Although a prominent role for uPA and uPAR could not be demonstrated in this short term model of renal injury, this does not preclude a significant role for these proteins in chronic renal injury, where prolonged accumulation of matrix proteins leads to renal scarring and glomerulosclerosis.

In summary, these studies demonstrate an important protective role for plasminogen and plasminogen activators in acute inflammatory renal injury associated with crescentic GN. They are consistent with previous observations that tPA and uPA appear to be the major plasminogen activators in vivo but also suggest that the major protective role in acute inflammatory glomerular injury is attributable to tPA. uPA facilitates recruitment of macrophages into nephritic glomeruli. The uPA receptor does not appear to play any significant role in protecting the glomerulus in GN. Therapies aimed at enhancing or replacing fibrinolytic activity in crescentic GN may be of benefit in human disease.

Acknowledgments

This work was supported by grants from the National Health and Medical Research Council (NH&MRC) of Australia, the Flemish Interuniversity Institute of Flanders and Human Frontiers in Science Program.

References

References
1
Vassalli
J-D
,
Sappino
A-P
,
Belin
D
The plasminogen activator/plasmin system
J Clin Invest
1991
88
1067
1072
[PubMed]
2
Carmeliet
P
,
Schoonjans
L
,
Kiekens
L
,
Ream
B
,
Degan
J
,
Bronson
R
,
De Vos
R
,
van den Oord
JJ
,
Collen
D
,
Mulligan
RC
Physiological consequences of loss of plasminogen activator gene function in mice
Nature (Lond)
1994
368
419
424
[PubMed]
3
Rømer
J
,
Bugge
TH
,
Pyke
C
,
Lund
LR
,
Flick
MJ
,
Degen
JL
,
Dano
K
Impaired wound healing in mice with a disrupted plasminogen gene
Nat Med
1996
2
287
292
[PubMed]
4
Carmeliet
P
,
Collen
D
Gene targeting and gene transfer studies of the biological role of the plasminogen/plasmin system
Thromb Haemost
1995
74
429
436
[PubMed]
5
Kincaid-Smith
P
Coagulation and renal disease
Kidney Int
1972
2
183
190
[PubMed]
6
Huang
XR
,
Holdsworth
SR
,
Tipping
PG
Evidence of a role for delayed type hypersensitivity in glomerular crescent formation
Kidney Int
1994
46
69
78
[PubMed]
7
Holdsworth
SR
,
Thomson
NM
,
Glasgow
EF
,
Atkins
RC
The effect of defibrination on macrophage participation in rabbit nephrotoxic nephritis: studies using glomerular culture and electron microscopy
Clin Exp Immunol
1979
37
38
43
[PubMed]
8
Thomson
NM
,
Moran
J
,
Simpson
IJ
,
Peters
DK
Defibrination with ancrod in experimental immune complex nephritis
Clin Exp Immunol
1975
20
527
557
9
Naish
P
,
Penn
GB
,
Evans
DJ
,
Peters
DK
The effect of defibrination on nephrotoxic serum nephritis in rabbits
Clin Sci
1972
42
643
646
[PubMed]
10
Holdsworth
SR
,
Tipping
PG
Macrophage induced glomerular fibrin deposition in experimental glomerulonephritis in the rabbit
J Clin Invest
1985
76
1367
1374
[PubMed]
11
Malliaros
J
,
Holdsworth
SR
,
Wojta
J
,
Erlich
J
,
Tipping
PG
Glomerular fibrinolytic activity in anti-GBM glomerulonephritis in rabbits
Kidney Int
1993
44
557
564
[PubMed]
12
Tipping
PG
,
Erlich
JH
,
Apostolopoulos
J
,
Mackman
N
,
Loskutoff
D
,
Holdsworth
SR
Glomerular tissue factor expression. Correlations between antigen, activity and mRNA in crescentic glomerulonephritis
Am J Path
1995
147
1736
1748
[PubMed]
13
Tipping
PG
,
Dowling
JP
,
Holdsworth
SR
Glomerular procoagulant activity in human proliferative glomerulonephritis
J Clin Invest
1988
81
119
125
[PubMed]
14
Rondeau
E
,
Mougenot
B
,
Lacave
R
,
Peraldi
MN
,
Kruithof
EKO
,
Sraer
JD
Plasminogen activator inhibitor 1 in renal fibrin deposits of human nephropathies
Clin Nephrol
1990
33
55
60
[PubMed]
15
Lee
E
,
Vaughan
DE
,
Parikh
SH
,
Grodinsky
AJ
,
Libby
P
,
Lark
MW
,
Lee
RT
Regulation of matrix metalloproteinases and plasminogen activator inhibitor-1 synthesis by plasminogen in cultured human vascular smooth muscle cells
Circ Res
1996
78
44
49
[PubMed]
16
Nunes
I
,
Shapiro
RL
,
Rifkin
DB
Characterization of latent TGF-beta activation by murine peritoneal macrophages
J Immunol
1995
153
1450
1459
[PubMed]
17
Tomooka
S
,
Border
WA
,
Marshall
BA
,
Noble
NA
Glomerular matrix expression is linked to inhibition of the plasmin protease system
Kidney Int
1992
42
1462
1469
[PubMed]
18
Wilson
HM
,
Reid
FJ
,
Brown
PA
,
Haites
NE
,
Booth
NA
Effect of transforming growth factor beta 1 on plasminogen activators and plasminogen activator inhibitor-1 in renal glomerular cells
Exp Nephrol
1993
6
343
350
[PubMed]
19
Ploplis
VA
,
Carmeliet
P
,
Vazirzadeh
S
,
Vlaenderen
I V
,
Moons
L
,
Plow
E F
,
Collen
D
Effects of disruption of the plasminogen gene in mice on thrombosis, growth and health
Circulation
1995
92
2585
2593
[PubMed]
20
Dewerchin
M
,
Van Nuffelen
A
,
Wallays
G
,
Bouché
A
,
Moons
L
,
Carmeliet
P
,
Mulligan
RC
,
Collen
D
Generation and characterization of urokinase receptor-deficient mice
J Clin Invest
1996
97
870
878
[PubMed]
21
Tipping
PG
,
Huang
XR
,
Berndt
MC
,
Holdsworth
SR
P-selectin directs T lymphocyte mediated injury in delayed-type hypersensitivity responses: studies in glomerulonephritis and cutaneous delayed-type hypersensitivity
Eur J Immunol
1996
26
454
460
[PubMed]
22
Tipping
PG
,
Huang
XR
,
Berndt
MC
,
Holdsworth
SR
A role for P selectin in complement-independent neutrophil-mediated glomerular injury
Kidney Int
1994
46
79
88
[PubMed]
23
Sappino
A-P
,
Huarte
J
,
Vassalli
J-D
,
Belin
D
Sites of synthesis of urokinase and tissue -type plasminogen activators in the murine kidney
J Clin Invest
1991
87
962
970
[PubMed]
24
Angles-Cano
E
,
Rondeau
E
,
Delarue
F
,
Hagege
J
,
Sultan
Y
,
Sraer
JD
Plasminogen activators of human glomeruli: identification and cellular localization
Thromb Haemost
1985
54
688
692
[PubMed]
25
Bergstein
JM
,
Riley
M
,
Bang
NU
Analysis of the plasminogen activator activity of the human glomerulus
Kidney Int
1988
33
868
874
[PubMed]
26
Brown
PAJ
,
Wilson
HM
,
Reid
FJ
,
Booth
NA
,
Simpson
JG
,
Morrison
L
,
Power
DA
,
Haites
NE
Urokinase-plasminogen activator is synthesized in vitro by human glomerular epithelial cells but not by mesangial cells
Kidney Int
1994
45
43
47
[PubMed]
27
Almus-Jacobs
F
,
Varki
N
,
Sawdey
MS
,
Loskutoff
DJ
Endotoxin stimulates expression of the murine urokinase receptor gene in vivo
Am J Pathol
1995
147
688
698
[PubMed]

The expert assistance of L. Moons, E. Simic, J. Tsoupas, P. Davenport, B. Hermans, I. Cornelissen and S. Jansens is acknowledged.

Author notes

Address correspondence to Dr. A. Richard Kitching, Monash University Department of Medicine, Monash Medical Centre, 246 Clayton Road, Clayton 3168, Victoria, Australia.