Renal urokinase-type plasminogen activator (uPA) receptor but not uPA deficiency strongly attenuates ischemia reperfusion injury and acute kidney allograft rejection.
ABSTRACT Central mechanisms leading to ischemia induced allograft rejection are apoptosis and inflammation, processes highly regulated by the urokinase-type plasminogen activator (uPA) and its specific receptor (uPAR). Recently, up-regulation of uPA and uPAR has been shown to correlate with allograft rejection in human biopsies. However, the causal connection of uPA/uPAR in mediating transplant rejection and underlying molecular mechanisms remain poorly understood. In this study, we evaluated the role of uPA/uPAR in a mice model for kidney ischemia reperfusion (IR) injury and for acute kidney allograft rejection. uPAR but not uPA deficiency protected from IR injury. In the allogenic kidney transplant model, uPAR but not uPA deficiency of the allograft caused superior recipient survival and strongly attenuated loss of renal function. uPAR-deficient allografts showed reduced generation of reactive oxygen species and apoptosis. Moreover, neutrophil and monocyte/macrophage infiltration was strongly attenuated and up-regulation of the adhesion molecule ICAM-1 was completely abrogated in uPAR-deficient allografts. Inadequate ICAM-1 up-regulation in uPAR(-/-) primary aortic endothelial cells after C5a and TNF-alpha stimulation was confirmed by in vitro experiments. Our results demonstrate that the local renal uPAR plays an important role in the apoptotic and inflammatory responses mediating IR-injury and transplant rejection.
- [Show abstract] [Hide abstract]
ABSTRACT: PURPOSE: The plasminogen system plays many roles in normal epithelial cell function, and components are elevated in diseases, such as cancer and asthma. The relative contribution of each component to epithelial function is unclear. We characterized normal cell function in airway epithelial cells with increased expression of selected pathway components. METHODS: BEAS-2B R1 bronchial epithelial cells stably overexpressing membrane urokinase plasminogen activator receptor (muPAR), soluble spliced uPAR (ssuPAR), the ligand (uPA) or inhibitors (PAI1 or PAI2), were characterized for pathway expression. Cell function was examined using proliferation, apoptosis, and scratch wound assays. A549 alveolar epithelial cells overexpressing muPAR were similarly characterized and downstream plasmin activity, MMP-1, and MMP-9 measured. RESULTS: Elevated expression of individual components led to changes in the plasminogen system expression profile, indicating coordinated regulation of the pathway. Increased muPAR expression augmented wound healing rate in BEAS-2B R1 and attenuated repair in A549 cells. Elevated expression of other system components had no effect on cell function in BEAS-2B R1 cells. This is the first study to investigate activity of the splice variant ssuPAR, with results suggesting that this variant plays a limited role in epithelial cell function in this model. CONCLUSIONS: Our data highlight muPAR as the critical molecule orchestrating effects of the plasminogen system on airway epithelial cell function. These data have implications for diseases, such as cancer and asthma, and suggest uPAR as the key therapeutic target for the pathway in approaches to alter epithelial cell function.Beiträge zur Klinik der Tuberkulose 02/2013; · 2.06 Impact Factor
- [Show abstract] [Hide abstract]
ABSTRACT: Renal ischemia reperfusion injury leads to acute kidney injury (AKI) and is associated with tissue edema, inflammatory cell infiltration, and subsequent development of interstitial renal fibrosis and tubular atrophy. The purpose of this study was to investigate the value of the functional magnetic resonance imaging (MRI) techniques, T2 mapping, and diffusion-weighted imaging (DWI) in characterizing acute and chronic pathology after unilateral AKI in mice. Moderate or severe AKIs were induced in C57Bl/6 mice through transient unilateral clamping of the renal pedicle for 35 minutes (moderate AKI) or 45 minutes (severe AKI), respectively. Magnetic resonance imaging was performed in 10 animals with moderate AKI and 7 animals with severe AKI before surgery and at 5 time points thereafter (days 1, 7, 14, 21, 28) using a 7-T magnet. Fat-saturated T2-weighted images, multiecho turbo spin echo, and diffusion-weighed sequences (7 b values) were acquired in matching coronal planes. Parameter maps of T2 relaxation time and apparent diffusion coefficient (ADC) were calculated, and mean values were determined for the renal cortex, the outer medulla, and the inner medulla. Inflammatory cell infiltration with monocytes/macrophages (F4/80), T-lymphocytes (CD4, CD8), and dendritic cells (CD11c) as well as the degree of interstitial fibrosis 4 weeks after AKI were determined through renal histology and immunohistochemistry. Statistical analysis comprised unpaired t tests for group comparisons and correlation analysis between MRI parameters and kidney volume loss. Increase of T2 relaxation time, indicating tissue edema, was most pronounced in the outer medulla and reached maximum values at d7 after AKI. At this time point, T2 values in the outer medulla were significantly increased to 53.8 ± 2.5 milliseconds after the severe AKI and to 46.3 ± 2.3 milliseconds after the moderate AKI when compared with the respective contralateral normal kidneys (40.9 ± 0.9 and 36.4 ± 1.2 milliseconds, respectively; P < 0.01). The T2 values reached baseline by d28. Medullary ADC was significantly reduced at all time points after AKI; restriction of diffusion was significantly more pronounced after the severe AKI than after the moderate AKI at d14 and d28. Changes of renal T2 and ADC values were associated with the severity of AKI as well as the degree of inflammatory cell infiltration and interstitial renal fibrosis 4 weeks after AKI. Furthermore, relative changes of both MRI parameters significantly correlated with kidney volume loss 4 weeks after AKI. Measuring T2 and ADC values through MRI is a noninvasive way to determine the presence and severity of acute and chronic renal changes after AKI in mice. Thus, the method should prove useful in animal and human clinical studies.Investigative radiology 07/2013; · 4.85 Impact Factor
- The Korean Journal of Internal Medicine 03/2014; 29(2):166-9.
Renal Urokinase-Type Plasminogen Activator (uPA) Receptor
but not uPA Deficiency Strongly Attenuates Ischemia
Reperfusion Injury and Acute Kidney Allograft Rejection
Faikah Gueler,1* Song Rong,1* Michael Mengel,†Joon-Keun Park,* Julia Kiyan,*
Torsten Kirsch,* Inna Dumler,* Hermann Haller,* and Nelli Shushakova2*‡
Central mechanisms leading to ischemia induced allograft rejection are apoptosis and inflammation, processes highly regulated
by the urokinase-type plasminogen activator (uPA) and its specific receptor (uPAR). Recently, up-regulation of uPA and uPAR
has been shown to correlate with allograft rejection in human biopsies. However, the causal connection of uPA/uPAR in mediating
transplant rejection and underlying molecular mechanisms remain poorly understood. In this study, we evaluated the role of
uPA/uPAR in a mice model for kidney ischemia reperfusion (IR) injury and for acute kidney allograft rejection. uPAR but not
uPA deficiency protected from IR injury. In the allogenic kidney transplant model, uPAR but not uPA deficiency of the allograft
caused superior recipient survival and strongly attenuated loss of renal function. uPAR-deficient allografts showed reduced
generation of reactive oxygen species and apoptosis. Moreover, neutrophil and monocyte/macrophage infiltration was strongly
attenuated and up-regulation of the adhesion molecule ICAM-1 was completely abrogated in uPAR-deficient allografts. Inade-
quate ICAM-1 up-regulation in uPAR?/?primary aortic endothelial cells after C5a and TNF-? stimulation was confirmed by in
vitro experiments. Our results demonstrate that the local renal uPAR plays an important role in the apoptotic and inflammatory
responses mediating IR-injury and transplant rejection. The Journal of Immunology, 2008, 181: 1179–1189.
receptor (uPAR; CD87) was originally identified as a proteinase
receptor for uPA, directing pericellular proteolysis. However, ac-
cumulating data clearly demonstrate that uPAR can also activate a
variety of intracellular signal pathways via its lateral interaction
with different cell surface proteins such as integrins, growth factor
receptors, and G-protein-coupled membrane proteins. These inter-
actions enable the uPA/uPAR system not only to control pericel-
lular fibrinolytic and proteolytic activities, but also to modulate
cell adhesion, migration, proliferation, and differentiation (1, 2).
Moreover, the important role of the uPA/uPAR system has re-
cently been demonstrated in both innate and adaptive immune-
mediated responses (3). The uPA/uPAR system modulates Ag pro-
cessing and presentation (4), lymphocyte activation (5), generation
of pro- and anti-inflammatory signals (6), activation of intracellu-
rokinase-type plasminogen activator (uPA)3is a multi-
functional molecule that serves either as a proteolytic
enzyme or as a signal-inducing ligand. The urokinase
lar signaling pathways (7), cytotoxicity (8), cell adhesion (9), and
migration (10–12), all of which are critical steps in cell-mediated
immune responses. Furthermore, uPA potentiates neutrophil acti-
vation (7) and superoxid production (13). Recently, we were also
able to demonstrate an important link between the uPA/uPAR and
complement system in the regulation of immunological responses
in kidney (14) and lung tissues (15).
Renal ischemia reperfusion (IR) injury after transplantation
leads to acute renal failure, which profoundly affects both early
and long-term allograft function (16). Accumulating evidence
demonstrates that prolonged cold ischemia time is a strong risk
factor for unfavourable outcome after allogenic kidney transplan-
tation (17, 18) and suggests that the severity of IR injury to the
allograft determines its immunogenicity and subsequent graft fate
(19). The mechanisms underlying IR damage of kidney tissue
seem to be multifactorial and interdependent. The oxygen supply
to the tissue by reperfusion leads to the generation of reactive
oxygen species (ROS) exceeding the protective anti-oxidative ca-
pacity of kidney cells. Oxidative stress is known to be a major
apoptotic stimulus in allograft nephropathy (20). Furthermore, IR
injury stimulates the components of innate immune response, such
as complement activation and up-regulation of multiple proinflam-
matory genes including chemokines, cytokines, cytokine receptors,
and adhesion molecules. This inflammatory response induced by
innate mechanisms early after transplantation is markedly ampli-
fied by the subsequent adaptive response (16). Therefore, IR injury
initiates and induces the alloimmune response leading to acute and
chronic allograft rejection (21, 22).
Recently, up-regulation of the uPA/uPAR system has been dem-
onstrated under hypoxia/reoxygenation conditions in vitro (23,
24). Moreover, uPA/uPAR activation has been shown to correlate
with allograft rejection in a human biopsy study (25, 26) implying
a probable involvement of this system in acute and chronic allo-
graft rejection. However, the causal connection of uPA/uPAR in
*Department of Nephrology and†Department of Pathology, Medical School Han-
nover, Hannover, Germany; and‡Phenos GmbH, Hannover, Germany
Received for publication January 11, 2007. Accepted for publication May 13, 2008.
The costs of publication of this article were defrayed in part by the payment of page
charges. This article must therefore be hereby marked advertisement in accordance
with 18 U.S.C. Section 1734 solely to indicate this fact.
1F.G. and S.R. contributed equally to this work.
2Address correspondence and reprint requests to Dr. Nelli Shushakova, Nephrology
Department, Hannover Medical School, Carl-Neuberg Strasse 1, D-30625 Hannover,
Germany. E-mail address: firstname.lastname@example.org
3Abbreviations used in this paper: uPA, urokinase-type plasminogen activator;
uPAR, urokinase-type plasminogen activator receptor; IR, ischemia reperfusion;
ROS, reactive oxygen species; RT, room temperature; DHE, dihydroethidium; PFA,
paraformaldehyde; MAEC, mouse aortic endothelial cell; PBMC, peripheral blood
mononuclear cell; MO, monocytes/macrophages; PMN, polymorphonuclear leuko-
cyte; ATN, acute tubular necrosis; C5aR, receptor for C5a anaphylatoxin; MO, mac-
rophage/monocyte; WT, wild type.
Copyright © 2008 by The American Association of Immunologists, Inc. 0022-1767/08/$2.00
The Journal of Immunology
mediating of IR injury and transplant rejection as well as the pos-
sible molecular mechanisms for uPA/uPAR mediated initiation
and perpetuation of inflammatory reaction after transplantation re-
The aim of our study was to assess the role of the renal uPA/
uPAR system in renal IR injury and in allogenic kidney transplant
rejection. We demonstrate that uPAR but not uPA deficiency of the
allograft protected from IR injury and acute allogenic transplant
rejection. As underlying mechanisms we could elucidate impaired
susceptibility of the uPAR-deficient allografts to IR-induced ROS-
mediated apoptosis and reduced expression of adhesion molecules,
leading to impaired migration of host monocytes/macrophages and
neutrophils into uPAR?/?allografts.
Materials and Methods
The uPA- and uPAR-deficient mice (uPA?/?and uPAR?/?), generated as
previously described (27), were a gift from Peter Carmeliet and Mieke
Dewerchin (Center for Transgene Technology and Gene Therapy, Univer-
sity of Leuven, Belgium). uPA/uPAR deficiency was verified by PCR
genotyping. The uPA?/?mice on a C57BL/6J background and the corre-
sponding WT controls (WT1), as well as uPAR?/?mice on a mixed
C57BL/6J (75%) ? 129/Sv (25%) background and their WT littermate
controls (WT2) served as kidney donors (H2b). BALB/c (H2d) mice served
as recipients in kidney transplantation experiments and were supplied by
Charles River Laboratories. The animals were bred under pathogen-free
conditions in the animal facility of Phenos GmbH (Hannover, Germany)
and cared for in accordance with our institution’s guidelines for exper-
imental animals. All experiments were approved by the animal protec-
tion committee of the local authorities. Mice weighing 25–30g were
used for all experiments. For IR injury 7–12 mice were used in each
group (8 uPAR?/?mice, 7 WT2 mice, 10 uPA?/?mice, and 12 WT1
mice). For transplant experiments 6 mice for each group were used for
the survival study and additional 6 mice for each group and each time
point were sacrificed at 4 h, at day 1, and at day 6 after transplantation
for histological and molecular analysis.
Renal ischemia reperfusion injury
Renal IR-injury was induced in homozygous male uPA?/?, uPAR?/?,
WT1, and WT2 mice by bilateral clamping of both renal pedicals. The
animals were anesthetized with isoflurane. After median laparatomy, renal
pedicals were bluntly dissected and a nontraumatic vascular clamp was
applied for 35 min. At 24 h, postischemia kidney function was estimated by
serum creatinine measurement using an automated method (Beckman
For transplant experiments, homozygous female uPA?/?, uPAR?/?, WT1,
and WT2 mice were used as kidney donors and female BALB/c (H2d) mice
served as recipients. Vascularized kidney transplantation was performed as
described previously (28). In brief, the animals were anesthetized with
isoflurane, and the left donor kidney attached to a cuff of the aorta and the
renal vein with a small caval cuff and the ureter were removed en bloc.
After left nephrectomy of the recipient, the vascular cuffs were anasto-
mosed to the recipient abdominal aorta and vena cava, respectively, below
the level of the native renal vessels. The ureter was directly anastomosed
to the bladder (29). The times of cold and warm ischemia of allografts were
60 and 30 min, respectively. The right native kidney was removed through
flank incision either on the day of transplantation or four days later. After
transplantation, kidney function was estimated at designated time points
(ranging from 24 h to several weeks) by serum creatinine level measure-
ment. The general physical condition of the animals was monitored for any
evidence of rejection.
Kidney grafts were harvested 24 h and six days after transplantation and
one half of each allograft was immediately fixed in buffered formalin and
embedded in paraffin. Sections of 3 ?m were stained with hematoxylin-
eosin and periodic acid-Schiff stain, and evaluated according to the updated
Banff classification (30) by a nephropathologist, who was masked to the
experimental groups. Immunohistochemistry was performed using the fol-
lowing primary Abs: rat anti-mouse monocytes/macrophages (F4/80; Se-
rotec), polyclonal rabbit anti-mouse active caspase 3 (BD Pharmingen),
monoclonal rat anti-mouse neutrophils (Gr-1, a gift from Prof. Hoffmann,
Hannover Medical School, Hannover, Germany), monoclonal rat anti-
mouse T lymphocytes (CD4 and CD8; BD Pharmingen), rabbit polyclonal
anti-mouse CD25 (Santa Cruz Biotechnology), rabbit polyclonal anti-
mouse uPAR (Santa Cruz Biotechnology). For indirect immunofluores-
cence, nonspecific binding sites were blocked with 10% normal donkey
serum (Jackson ImmunoResearch Laboratories) for 30 min. Thereafter,
sections were incubated with the primary Ab for 1 h. All incubations were
performed in a humid chamber at room temperature (RT). For fluorescent
visualization of bound primary Abs, sections were further incubated with
Cy3 conjugated secondary Abs (Jackson Immuno Research Laboratory) for
1 h. Specimens were analyzed using a Zeiss Axioplan-2 imaging micro-
scope with the computer program AxioVision 3.0 (Zeiss). Analysis of in-
flammation was done by semiquantitative scoring of the infiltrating cells in
10 randomly chosen, nonoverlapping fields of cortex and outer medulla
(original magnification, ?200). Score: 0, no; 1, weak; 2, moderate; 3, high;
and 4, very high numbers of infiltrating cells. For CD4 and CD25 expres-
sion, absolute cell numbers were counted in 20 nonoverlapping view fields
each specimen. Active caspase-3 expression in the outer strip of the outer
medulla was scored as follows: 0 ? 10%, 1 ? 10–30%, 30–50%, 4 ? 50%
of the tubuli affected. ICAM-1 expression in the cortex was scored as
follows: 0 ? 10%, 1 ? 10–30%, 30–50%, 4 ? 50% of the glomeruli
affected. The analysis was done without knowledge of the animal
Mixed lymphocyte reactivity (MLR)
Priming of alloantigen-specific T cells from kidney graft recipients was
investigated by performing MLR assay based on the measurement of BrdU
incorporation during DNA synthesis. The responder spleen cells obtained
from naive BALB/c mice or from WT2 or uPAR?/?allograft recipients at
day 6 after transplantation were treated with ammonium chloride solution
(Cell Systems) to lyse erythrocytes, washed three times, and resuspended at
3 ? 106cells/ml in complete RPMI 1640 medium (Life Technologies) sup-
plemented with 10% FCS (Sigma-Aldrich), 2 mM L-glutamine, 100 U/ml
penicillin, 100 mg/ml streptomycin, and 100-?l aliquots were delivered in
triplicate to the wells of a 96-well, flat-bottom tissue culture plate. Stimulator
cells were prepared from the spleens of syngeneic (i.e., BALB/c) and allograft
donors (i.e.WT2, uPAR?/?). After lysis of erythrocytes the stimulator cells
were treated with 50 ?g/ml mitomycin C for 30 min at 37°C, washed and
resuspended in culture medium at 3 ? 106cells/ml, and 100-?l aliquots per
added and 18h later responder cell proliferation was quantified using colori-
metric Cell Proliferation ELISA kit (Roche Diagnostics GmbH) in accordance
to the manufacturer’s instructions. Syngeneic stimulator cells were used as
background controls and were subtracted from alloresponses.
Priming of alloantigen-specific T cells from kidney allograft recipients was
also tested by enumerating of IFN-?-producing T cells using a mouse
IFN-? ELISpot kit (BD Biosciences) in accordance to the manufacturer’s
instructions. Spleen cell suspensions were prepared from naive BALB/c
mice or from allograft recipients at day 6 after transplantation and used as
responder cells. Spleen cells from WT2 or syngeneic BALB/c mice were
prepared and treated with mitomycin C for use as stimulator cells in the
assay as described above. Responder and stimulator cells were cultured
together for 24 h at 37°C in 5% CO2. The resulting spots were counted
using the A.EL.VIS 4-Plate ELISPOT reader with the A.EL.VIS ELISPOT
ROS generation by oxidation of dihydroethidium
The redox-sensitive fluorophore dihydroethidium (DHE) was used to eval-
uate O2-production in the kidney in situ (31). Frozen tissue Cryosections of
6 ?M were incubated with 0.1 mM DHE dissolved in HEPES-Tyrode
buffer solution (132 mM NaCl, 4 mM KCl, 1 mM CaCl2, 0.5 mM MgCl2,
9.5 mM HEPES, and 5 mM glucose) for 12 min RT. After incubation,
images were obtained using the Leica imaging system IM 500 (Ex, 520 nm;
Em, 605 nm). Semiquantitative scoring was performed as follows: score: 0,
no; 1, weak; 2, moderate; 3, high; and 4, very high intensity.
For TUNEL-assay (terminal deoxynucleotidyl transferase-mediated dUTP
nick-end labeling) 2 ?m sections of 4% paraformaldehyde (PFA)-fixed
paraffin-embedded tissues were deparaffinized, treated with the terminal
deoxynucleotidyl transferase enzyme and incubated in a humidified cham-
ber at 37°C for 1 h. After washing, the tissue was treated with FITC-labeled
anti-digoxygenin, incubated for 60 min, and washed. Negative controls
1180 uPAR AND KIDNEY TRANSPLANTATION
were prepared under the same conditions, with the omission of the terminal
deoxynucleotidyl transferase enzyme. TUNEL pos. cell numbers were
counted in 20 nonoverlapping view fields each specimen without knowl-
edge of the animal assignment.
RNA extraction and real time quantitative PCR
Frozen kidneys were grinded in liquid nitrogen and total RNA was ex-
tracted using Trizol reagent (Invitrogen). For real-time quantitative PCR
(qPCR), 1 ?g of DNase-treated total RNA was reverse transcribed using
Superscript II Reverse transcriptase (Invitrogen) and qPCR was performed
on an SDS 7700 system (Applied Biosystems) using Rox dye (Invitrogen),
FastStart taq Polymerase (Roche Diagnostics) and gene specific primers
and FAM-Tamra-labeled probes (BioTez). PCR amplification was con-
ducted at 10 min 96°C and 40 cycles at 10 s 95°C and 1 min at 60°C.
?-actin served as the reference gene for normalization. Primer sequences
are available on request. Quantification was conducted using qgene soft-
Mouse aortic endothelial cell (MAEC) culture
Isolation of mouse aortic endothelial cells from 6- to 8-wk-old WT or
uPAR?/?mice was performed as described previously (33). The cells were
cultivated in medium consisting of endothelial cell growth medium 2 (Clo-
netics/Cambrex) and DMEM (1:1) supplemented with 20% FCS, 100
?g/ml endothelial cell growth supplement (Sigma-Aldrich), 100 U/ml pen-
icillin, 100 mg/ml streptomycin, 2 mM L-glutamine, 0.5% nonessential
amino acids, and 0.1 mg/ml heparin. MAEC at passage 3 were used for
endothelial cell characterization and for all experiments. The endothelial
nature of cells was confirmed by the typical cobblestone morphology of
confluent monolayers, by Dil-Ac-LDL uptake and by surface expression of
CD31 and CD106 analyzed by immunnocytochemistry (data not shown).
WT and uPAR?/?MAEC were seeded and cultured on glass coverslips.
Serum-starved cells were fixed with 4% PFA in PBS for 20 min at RT.
Nonspecific binding was blocked by 2 h incubation at RT with 3% BSA in
PBS; the preparations were washed three times with PBS. Incubations with
primary Ab (rat anti-mouse receptor for C5a anaphylatoxin (C5aR) clone
20/70 from Hycult Biotechnology or polyclonal rabbit anti-TNFR1 from
Santa Cruz Biotechnology) were performed for 2 h at RT. Incubations with
Alexa 488-conjugated secondary Abs (Molecular Probes) were performed
for 1 h. After staining, cells were embedded in Poly-Mount mounting me-
dium (Polysciences). The fluorescence cell images were captured using a
Leica TCS-SP2 AOBS confocal microscope (Leica Microsystems). All the
images were taken with oil-immersed ? 63 objective, NA ? 1.4.
ICAM-1 cell ELISA
Cell surface expression of ICAM-1 on MAEC was measured by cell
ELISA as described previously (34). In brief, MAEC were grown in 96-
well plate until 80% confluent, starved for 4 h and then stimulated for 16 h
with 100 ng/ml recombinant murine C5a (mrC5a; Sigma-Aldrich) or with
5 ng/ml recombinant murine TNF-? (mrTNF?; R&D Systems). These con-
centrations were chosen in preliminary experiments ranging from 0.5 ng/ml
to 200 ng/ml for each stimulus. For inhibition experiments MAEC were
preincubated for 2 h with 100 ng/ml pertussis toxin (Sigma-Aldrich) or
with 20 ?g/ml hamster anti-mouse-TNFR1 mAb (R&D Systems). MAEC
incubated in medium without stimuli served as a control. After fixation
with 3% PFA and blocking with 3% BSA to prevent nonspecific binding
the cells were incubated for 2 h with polyclonal rabbit-anti-mouse ICAM-1
Ab (Santa Cruz Biotechnology).The specific binding of Abs was then eval-
uated by incubation of cells for 1 h with secondary peroxidase conjugated
goat anti-rabbit IgG (Santa Cruz Biotechnology) followed by addition of
in uPAR?/?mice. A, Renal function was investigated by measurement of
serum creatinine level before (control) and at 24 h after IR. IR-injury
caused severe renal dysfunction with elevation of serum creatinine level in
WT1, WT2, and uPA-deficient mice. In contrast, uPAR deficiency led to
markedly attenuated loss of renal function after IR injury (???, p ? 0.001
vs WT2, 7–12 mice were investigated for each group). B, Representative
specimens after TUNEL assay 24-h postischemia are shown (original mag-
nification, ?200). Because WT1 and WT2 mice demonstrated comparable
results, representative data from WT2 kidney is shown. WT kidneys (a)
and uPA-deficient kidneys (b) had increased numbers of apoptotic cells
mainly in the tubular epithelium. In contrast, postischemic uPAR-deficient
kidneys (C) showed almost no TUNEL positive cells. The results of quan-
tification are presented as mean ? SEM in the lower panel (???, p ? 0.001
vs WT2, six mice were investigated for each group).
Attenuation of IR-induced renal dysfunction and apoptosis
allografts. uPAR (A) and uPA (B) mRNA expression in WT2 kidney al-
lografts was investigated by qPCR before (control) and at 4 and 24 h after
allogenic kidney transplantation. Both uPAR and uPA mRNA expression
was time-dependently up-regulated after transplantation (?, p ? 0.05; ??,
p ? 0.01 vs control, six mice were investigated for each group). C, The
uPAR protein expression in WT2 allografts was investigated by immuno-
cytochemistry. The upper row, uPAR expression in the cortex area; the
lower row, uPAR expression in the outer strip of the outer medulla. Low
baseline uPAR protein expression seen in normal WT2 kidney (control)
was increased within 24 h and even more at 6 days after transplantation.
(Original magnification ?200; six mice were investigated for each group).
uPA and uPAR expression is up-regulated in WT kidney
1181 The Journal of Immunology
tetramethylbenzidine substrate solution (R&D Systems), stopping the re-
action with 0.5 M H2SO4and measuring the OD at 450 nm. The substrate
was then washed away with deionized water, the plate allowed to dry and
0.5% trypan blue was added to stain for the number of cells/well. Excess
of trypan blue was washed away and 1% SDS was added to solubilize the
trypan blue stained cells. Each well was then read at 595 nm. The OD of
ICAM-1 staining was divided by the OD of trypan blue staining to yield
ICAM-1 expression for each well.
Peripheral blood mononuclear cell (PBMC) adhesion assay
The adhesion of WT2 PBMC isolated by Ficoll-Paque separation to the
endothelial surface of aortas obtained from WT2 or uPAR?/?mice was
determined by the counting of adherent cells fluorescently labeled with the
acetyloxymethyl ester of calcein (Calbiochem). A total of 1–2 mm pieces
of aortas cleaned carefully of periadventitial fat and connective tissue and
opened longitudinally were placed adventitia-side down on collagen I
coated 96-well plates containing 10 ?l of endothelial cell basal medium 2
(Clonetics/Cambrex) supplemented with 5% FCS to allow adherence of the
aortic pieces to the substratum. When the pieces were well-attached (after
4 h), 200 ?l of EBM-2 medium containing 50 ng/ml mrC5a, or 50 ng/ml
mrTNF? was added. These concentrations were chosen in preliminary ex-
periments ranging from 0.5 ng/ml to 200 ng/ml for each stimulus. For
inhibitory experiments the pieces were preincubated for 2 h with 100 ng/ml
pertussis toxin or with 20 ?g/ml neutralizing hamster anti-mouse-TNFR1
mAb. Aorta pieces incubated in medium alone served as a control. After
16 h, the aorta pieces were washed twice and 100?103fluorescently labeled
WT2 PBMC in 200 ?l medium were added. The cells were allowed to
adhere for 45 min at 37°C. Unbound cells were removed by washing three
times. Photographs of aorta pieces were then made using the Leica imaging
microscope with the digital image-processing program. The bound leuko-
cytes were counted without knowledge of the group assignment in four
different view-fields per aorta piece.
Data are shown as mean ? SEM. Normal distribution was analyzed by
Klomogorov-Smirnov-test and statistical significance was calculated by
Student’s t test for independent groups. SPSS 12.01 software was used.
uPA/uPAR deficiency in IR injury
The uPA/uPAR system is involved in a variety of signaling cas-
cades which mediate IR injury. We first tested whether uPA or
uPAR are mediators or effectors of renal IR injury. We performed
Survival analysis of mice receiving uPA- and uPAR-deficient allografts as compared with the animals receiving corresponding WT allografts is shown.
WT1 and WT2 allografts revealed identical results and are presented as WT. Survival of recipients of uPA?/?allografts was comparable to WT allograft
controls. The majority of all these transplant recipients died within 4 wk after transplantation. In contrast, recipients of uPAR?/?allografts showed
prolonged allograft survival over 20 wk after transplantation. B, Renal function was estimated in recipient mice before (day 0) and at designated time points
after transplantation by serum creatinine measurement. Recipients of uPA?/?and both WT1 and WT2 allografts had a similar steep rise of serum creatinine
at 24 h and at 6 days after transplantation, correlating with acute rejection. In contrast, recipients of uPAR-deficient allografts had only a moderate initial
rise in creatinine within 24 h after transplantation and remained stable at slightly increased serum creatinine levels thereafter. The results are presented as
mean ? SEM (???, p ? 0.001 vs WT2, six mice were investigated for each group). C, Morphological changes one day after kidney transplantation are
shown. All WT (a) and the uPA-deficient allografts (b) showed diffuse severe ATN. In contrast, the uPAR-deficient allograft recipients (c) had no signs
of ATN. (PAS stain, original magnification ?200; six mice were investigated for each group).
Effect of allograft uPA and uPAR deficiency on survival, renal function and renal morphology after allogenic kidney transplantation. A,
1182uPAR AND KIDNEY TRANSPLANTATION
bilateral renal pedical clamping in uPA?/?and uPAR?/?mice
and their corresponding wild-type controls (WT1 and WT2, re-
spectively). Both WT1 and WT2 controls and uPA?/?mice dem-
onstrated severe loss of renal function within 24 h postischemia, as
reflected in elevation of serum creatinine levels. In contrast,
uPAR-deficient animals were protected from IR injury and dem-
onstrated statistically significant attenuated loss of renal function
(Fig. 1A). Because apoptosis is a hallmark of IR injury, we per-
formed a TUNEL assay in normal kidney tissue and in kidneys
24 h after ischemia (Fig. 1B). No TUNEL-positive cells were de-
tected in normal kidneys in any of the four groups of animals (data
not shown). We detected increased numbers of TUNEL-positive
nuclei in WT1, WT2 and uPA?/?kidneys. Because WT1 and
WT2 mice showed almost identical results, only the data from
WT2 mice are shown. However, in uPAR?/?kidneys the amount
of TUNEL-positive cells was significantly reduced. These results
suggest that uPAR contributes to hypoxia-induced apoptosis inde-
pendently from uPA.
uPA/uPAR expression is up-regulated in rejecting kidney
Because activation of uPA/uPAR system has been shown to cor-
relate with allograft rejection in a human biopsy study (25), we
tested whether or not a similar phenomenon may be observed in
kidney allografts in mice. We performed allogenic kidney trans-
plantation using WT2 mice (H2b) as donors, and BALB/c mice
(H2d) as recipients. The expression of uPA and uPAR was then
investigated in normal WT2 kidneys and in WT2 kidney allografts
at 4 and 24 h after transplantation. We detected strong up-regula-
tion of uPAR (Fig. 2A) mRNA 4 h after transplantation with fur-
ther increase after 24 h. The up-regulation of uPA mRNA was less
pronounced but also could be observed at 24 h after transplantation
(Fig. 2B). At protein level (Fig. 2C), the weak expression of uPAR
seen in normal WT2 kidney was strongly up-regulated within 24 h
after transplantation and further increased at day 6. As expected,
no positive signal was detected in uPAR?/?allografts (data not
shown). These data demonstrate that transplantation results in local
activation of the uPA/uPAR system in rejecting allografts from
uPAR but not uPA deficiency improved kidney allograft survival
and attenuated loss of renal function
Further, we analyzed an influence of uPA or uPAR deficiency of
the allograft on recipient survival (Fig. 3A). We performed allo-
genic kidney transplantation using uPA- and uPAR-deficient mice
(H2b) and their corresponding WT controls as donors, and BALB/c
mice (H2d) as recipients. All recipients of WT as well as uPA?/?
allografts died shortly after allogenic kidney transplantation. In
contrast, recipients of uPAR-deficient allografts showed superior
survival for more than 20 wk after transplantation.
Next, we studied the effect of uPA/uPAR deficiency on renal
function after transplantation by measuring serum creatinine levels
(Fig. 3B). To monitor the effects of rejection on renal function we
removed both recipient kidneys and transplanted the donor kidney
in the same operation. An increase in serum creatinine was de-
tected in all groups 24 h after transplantation; however this effect
transplantation-induced apoptosis. TUNEL assay was performed in WT2
and uPAR?/?allografts at 4 (upper row) and 24 h (middle row) after
transplantation. WT2 allografts (a and c) had many TUNEL positive nuclei
at 4 h (a) with further increase at 24 h (c) after transplantation. In contrast,
uPAR-deficient allografts had almost no TUNEL positive cells at (b) and
24 h (d) after transplantation. (TUNEL assay, original magnification
?200). The quantification results are presented as mean ? SEM in the
right-hand panel (??, p ? 0.01; ???, p ? 0.001 vs WT2, six mice were
investigated for each group). The immunohistochemistry for cleaved
caspase-3 showed in WT2 allografts an intense up-regulation of cleaved
caspase-3 level in the tubuli of the outer stripe of the outer medulla, the
area most sensitive to hypoxic damage (e). uPAR-deficient allografts in
contrast showed only few cleaved caspase-3 positive tubuli (f).
Allograft uPAR-deficiency strongly protected kidneys from
generation after transplantation. Generation of ROS was investigated by
DHE stain. The upper row shows low basal level of DHE staining in WT2
(a) and uPAR?/?(b) kidneys before transplantation. An increase in gen-
eration of ROS in the tubulo-interstitial compartment was detected in WT2
allografts as early as 4 h (c) after transplantation; this effect was still de-
tectable 24 h after transplantation (e). In contrast, uPAR-deficient allografts
showed impaired generation of ROS at 4 (d) and 24 h (f) after transplan-
tation. The results of the semiquantitative analysis are presented as mean ?
SEM in the right-hand panel (???, p ? 0.001 vs WT2, six mice were
investigated for each group).
uPAR-deficiency of the allograft strongly reduces ROS
1183 The Journal of Immunology
was strongly attenuated in mice which received uPAR?/?allo-
grafts. Six days after transplantation, recipients of uPA-deficient
allografts and WT allografts developed deleterious loss of renal
function as reflected in severe creatinine elevation. In contrast, the
renal function of recipients of uPAR-deficient allografts remained
stable without any significant increase in serum creatinine. Histo-
logical analysis revealed severe diffuse acute tubular necrosis
(ATN) in WT and uPA?/?but not in uPAR?/?allografts at d1
after transplantation (Fig. 3C). These results indicate that the
uPAR but not uPA of allograft origin contributes significantly to
the loss of renal function and allograft rejection in this model of
allogenic kidney transplantation. Because uPA deficiency had no
protective effects in either the IR model or in kidney transplanta-
tion, the following experiments were performed using only uPAR-
deficient animals and their respective WT2 controls.
Apoptosis was markedly reduced in uPAR-deficient allografts
Because hypoxia-induced apoptosis is the pathophysiological cor-
relate for ATN, we performed a TUNEL assay to investigate the
cell death rate in uPAR?/?and WT2 kidney allografts (Fig. 4).
WT2 allografts ubiquitously developed TUNEL positive nuclei as
early as 4 h (a) and more pronouncedly 24 h (c) after transplan-
tation. In contrast, uPAR?/?allografts demonstrated almost no
signs of apoptosis at these time points (b and d). To elucidate
whether the differences in cell death between WT2 and uPAR?/?
allografts were due to necrosis or to apoptosis we performed im-
munohistochemistry for cleaved caspase-3. Especially in the outer
stripe of the outer medulla, the area which is most sensitive to
hypoxic damage, and in the medium of the vessels, cleaved
caspase-3 level was increased in WT2 allografts at 24 h after trans-
plantation (data not shown) and even more so at day 6 after kidney
transplantation (e). In comparison, uPAR?/?allografts (f) showed
markedly lower levels of cleaved caspase-3. These results were in
line with our previous finding in the IR injury model and clearly
demonstrate that local uPAR expression is also pivotal in ischemia
triggered apoptosis in the allogenic kidney transplantation model.
Generation of ROS was reduced in uPAR-deficient allografts
Hypoxia leads to tissue damage via generation of ROS contribut-
ing significantly to the mitochondrial apoptotic pathway (35).
Scavengers for ROS, H2O2, and superoxide have been shown to
inhibit apoptosis induced by ischemia-reperfusion (36). Therefore,
we tested the hypothesis that strongly reduced apoptosis in
uPAR?/?allografts might be due to decreased production of ROS.
DHE staining (Fig. 5) performed in normal WT2 (a) and uPAR?/?
(b) kidneys revealed similar basal levels of ROS. However, in
contrast to WT allografts demonstrating severe up-regulation of
ROS at 4 (c) and 24 h (e) after transplantation, ROS generation
was markedly reduced in uPAR-deficient allografts (d and f). This
result suggests that uPAR contributes to the initial apoptosis trig-
gering step - the generation of ROS.
uPAR deficiency of the allograft and inflammatory cell
The inflammatory cell infiltration of the allograft is a hallmark of
allograft rejection (37, 38). Therefore, we performed immunohis-
tochemistry for different cell subsets to elucidate the composition
of the cell infiltrates. No differences were observed between the
groups in CD8 and CD4 infiltrates. Moreover, the CD25 expres-
sion on CD4 T cells was comparable in WT2 and uPAR?/?allo-
grafts (Fig. 6A). In line with this observation the MLR assay did
not reveal any differences between splenocytes obtained from
WT2 and uPAR?/?allograft recipients at day 6 after transplanta-
tion (Fig. 6B). The splenocytes from WT2 and uPAR?/?allograft
recipients showed a similar increase of proliferative response to
WT2 stimulator cells compared with splenocytes from naive
BALB/c mice. The frequency of alloantigen-specific IFN-?-pro-
ducing T cells tested by ELISPOT assay was also comparable in
WT2 and uPAR?/?allograft recipients (Fig. 6C). This result
shows that uPAR deficiency did not decrease the level of donor-
reactive T cell priming in this model.
Furthermore, we analyzed infiltration of neutrophils (Fig. 7A,
upper panel) and monocytes/macrophages (Fig. 7A, middle panel).
We found that uPAR deficiency of the allograft decreased the
number of infiltrating neutrophils and monocytes/macrophages as
compared with WT2 control allografts. These results underline the
cell infiltration and alloantigen-specific T cell priming after transplantation.
A, At day 6 after transplantation, WT2 allografts (a) and uPAR-deficient
allografts (b) showed similar CD8 positive T cell infiltration. Furthermore,
CD4 positive T cell infiltrates of similar density were seen in both WT2
and uPAR?/?allografts (c and d). No differences were observed concern-
ing CD25 positive T cells between WT2 and uPAR?/?allografts (e and f).
The results of the quantification analysis are presented as mean ? SEM in
the right-hand panel six mice were investigated for each group). B, In vitro
MLR assay was performed with splenocytes isolated from naive BALB/c
mice and from recipients of WT2 and uPAR?/?kidney allografts at day 6
after transplantation. No differences between WT2 and uPAR?/?allograft
recipients were observed. Experiments were performed in triplicate, n ? 5
mice for each group, data are presented as mean ? SEM. C, IFN-?-pro-
ducing cells at day 6 after transplantation in spleen cells of naive BALB/c
mice and recipients of WT2 and uPAR?/?kidney allografts were deter-
mined by ELISPOT analysis. Numbers of IFN-?-producing cells were sim-
ilar in uPAR?/?and WT2 allograft recipients. Experiments were per-
formed in triplicate, n ? 5 mice for each group, data are presented as
mean ? SEM.
uPAR-deficiency of the allograft does not change host T
1184 uPAR AND KIDNEY TRANSPLANTATION
importance of local renal uPAR for monocyte/macrophage and
Adhesive interaction between up-regulated adhesion molecules
on activated endothelium with blood leukocytes is a crucial step
for leukocyte infiltration into the site of inflammation. Therefore,
we analyzed the expression of ICAM-1 in WT2 and uPAR?/?
allografts after kidney transplantation. Baseline ICAM-1 expres-
sion of uPAR?/?and WT kidneys was comparable (data not
shown). However, a strong ICAM-1 up-regulation seen in WT2
allografts after transplantation was significantly impaired in
uPAR?/?allografts (Fig. 7A, lower panel).
Members of the chemokine family play a central role in in-
flammatory cell infiltration into extravascular sites by attracting
and stimulating specific subsets of leukocytes (39). MCP-1 is an
important mediator for monocyte recruitment (40) whereas
MIP-2 is necessary for an adequate neutrophil infiltration (39).
Therefore, we compared the transplantation-induced changes of
MCP-1 and MIP-2 mRNA expression in WT2 and uPAR?/?
allografts. As expected, we observed a strong up-regulation of
these proinflammatory mediators in WT2 allografts. Surpris-
ingly, uPAR?/?allografts demonstrated a practically identical
pattern of the transient overexpression of MIP-2 mRNA, with
a maximum at 24 h after transplantation. The kinetics of
MCP-1 mRNA up-regulation was also similar in WT and
uPAR?/?allografts, and moreover, the grade of MCP-1 mRNA
up-regulation was even significantly higher in uPAR?/?kidney
(Fig. 7B). These results suggest that attenuated leukocyte infil-
tration into uPAR-deficient allografts may be due to impaired
up-regulation of ICAM-1 but not to altered chemokine
uPAR deficiency of blood vessel decreases TNF?/C5a-induced
ICAM-1 up-regulation and reduces WT leukocyte adhesion
Recently, the interaction of complement activation product C5a
with its receptor (C5aR) has been shown to induce a strong in-
crease in gene expression for cell adhesion molecules in endothe-
lial cells similar to those induced by TNF-? (41). Because both
TNF-? and activated complement are important mediators of IR
injury and transplant rejection (42, 43), we hypothesized that
TNF-? and/or C5a-dependent up-regulation of endothelial adhe-
sion molecules may be impaired in uPAR?/?allografts and may
be an explanation for the reduced monocyte/macrophage and gran-
ulocyte infiltration. To test this hypothesis, we performed addi-
tional in vitro experiments with the primary culture of MAEC from
WT2 and uPAR?/?mice. After demonstration of similar cell sur-
face expression of both TNFR1 and C5aR on WT2 and uPAR?/?
MAEC by immunocytochemistry (Fig. 8A) we investigated the
expression of ICAM-1 after stimulation of cells with TNF-? and
C5a for 16 h by cell ELISA. The treatment with 5 ng/ml mrTNF?
resulted in a strong up-regulation of ICAM-1 expression in both
WT2 and uPAR?/?MAEC, however, this effect was significantly
decreased in uPAR?/?MAEC compared with WT2. The pretreat-
ment with TNFR1 blocking Ab strongly decreased the TNF-?-
induced up-regulation of ICAM-1 suggesting the involvement of
TNFR1 (Fig. 8B). The treatment of MAEC with 100 ng/ml mrC5a
resulted in a moderate but significant up-regulation of ICAM-1 in
WT2 MAEC, however, this effect was completely abrogated in
uPAR?/?MAEC. The pretreatment with the C5aR blocking per-
tussis toxin prevented the C5a-induced up-regulation of ICAM-1
in WT2 MAEC verifying the role of C5aR (Fig. 8C).
allografts (a) showed more tubulo-interstitial infiltrates of Gr-1 positive neutrophils as compared with uPAR-deficient allografts (b). Dense infiltration of
F4/80 positive monocytes/macrophages seen in the tubulo-interstitium of WT2 allografts (c) was strongly reduced in uPAR-deficient allografts (d). ICAM-1
expression was studied by immunohistochemistry. Twenty-four hours after transplantation ICAM-1 was heavily up-regulated in the glomeruli of WT2
allografts (e) whereas uPAR?/?allografts (f) showed almost no ICAM-1 up-regulation. The results of the semiquantification analysis are presented as
mean ? SEM in the right-hand panel (??, p ? 0.01; ???, p ? 0.001 vs WT2, original magnification ?200, six mice were investigated for each group).
B, The expression of MCP-1 (upper panel) and MIP-2 (low panel) mRNA was analyzed by TaqMan RT-PCR in normal kidneys (control) and in WT2 and
uPAR?/?allografts obtained at 24 h and day 6 after transplantation. Data are expressed as the mean ? SEM (??, p ? 0.01 vs WT2, six mice were
investigated for each group).
uPAR-deficiency of the allograft protects from host leukocyte infiltration after transplantation. A, At day six after transplantation, WT2
1185The Journal of Immunology
These results were confirmed by a PBMC adhesion assay ex
vivo. In these experiments, adhesion of normal WT2 PBMC to the
endothelial surface of aortas obtained from WT2 or uPAR?/?
mice and stimulated in vitro with mrC5a or mrTNF? was inves-
tigated. As demonstrated in Fig. 8D, 16-h stimulation of WT2
aorta with TNF-? or C5a induced significant increase in adhesion
of WT2 PBMC. These effects could be strongly reduced or
completely abolished when aortas were pretreated with TNFR1
blocking Ab and pertussis toxin, respectively. In contrast, the
TNF-? and C5a-dependent up-regulation of endothelial adhesion
was completely abolished in uPAR?/?aortas. These results
clearly demonstrate that uPAR expression on endothelial cells is
necessary for adequate TNF-? and C5a signaling leading to up-
regulation of adhesion molecules and mediating leukocyte
Our data provide the first evidence that local renal uPAR but not
uPA expression plays a pivotal role in the pathogenesis of IR in-
jury and allogenic transplant rejection. We ruled out that uPAR
deficiency protected kidney tissue from generation of ROS and
consecutively from severe apoptosis in IR injury and acute kidney
allograft rejection. In the transplantation model, the resistance of
the allograft against hypoxia-induced apoptosis due to uPAR de-
ficiency was linked to better renal function and increased allograft
survival. The inadequate up-regulation of ICAM-1 in the blood
vessels of uPAR?/?allografts resulted in attenuated monocyte/
macrophage and neutrophil infiltration.
Allogenic transplant rejection is triggered by several stimuli
such as allograft hypoxia and subsequent apoptosis and inflam-
matory response. Tubular cell apoptosis is considered to be an
important pathway leading to tubular atrophy in progressive
renal disease. Recently, it has been shown that uPAR may mod-
ify the rate of apoptotic renal cell death (44). In vitro, human
glioma cells exposed to uPAR antisense have been reported to
undergo more apoptotic cell death (45). In contrast to these
observations, the major finding of this study was that uPAR
deficiency strongly protected renal cells from apoptosis via re-
duced ROS formation in both models, IR injury and kidney
transplantation. One possible explanation for this discrepancy
might be that uPAR exposes pro- or anti-apoptotic action and
modulates the cell survival/apoptosis ratio depending on the
cell type and/or the apoptosis-triggering pathways. This hypoth-
esis is supported, for example, by demonstration of the
?-induced increase of WT leukocyte adhesion in vitro. A, Nonstimulated serum-starved MAEC obtained from WT and uPAR?/?mice were immunostained
with monoclonal anti-C5aR (upper panel) or polyclonal anti-TNFR1 (lower panel) Abs and Alexa 488-conjugated secondary Abs. Basal expression of
C5aR and TNFR1 was comparable in uPAR?/?and WT MAEC. (The frame size of images is 240 ? 240 ?m.) B, mrTNF? (left histogram) stimulation
resulted in a strong up-regulation of ICAM-1 expression in WT2 and to a lesser extent also in uPAR?/?MAEC. Pretreatment with TNFR1 blocking Ab
strongly decreased the TNF-?-induced up-regulation of ICAM-1. Nonstimulated cells served as controls. mrC5a stimulation (right histogram) resulted in
a moderate but significant up-regulation of ICAM-1 in WT2 MAEC, however, this effect was completely abrogated in uPAR?/?MAEC. Pertussis toxin
prevented the C5a-induced up-regulation of ICAM-1 in WT MAEC. (Data are expressed as the mean ? SEM from three independent experiments,
performed in four parallels for each condition. Significant differences were determined by Student’s t test for C5a/TNF-?-stimulated vs nonstimulated cells
(??, p ? 0.01; ???, p ? 0.001) and for WT2 vs uPAR?/?cells (??, p ? 0.01; ???, p ? 0.001). C, Adhesion of WT2 PBMC to the endothelial surface
of aortas was studied in vitro. Aortas from WT2 or uPAR?/?mice were preincubated or not with 100 ng/ml pertussis toxin or with 20 ?g/ml anti-TNFR1
Abs for 2 h and then stimulated for 16 h with 50 ng/ml C5a or TNF-?. Unstimulated aortas served as a control. C5a- and TNF-?-dependent increase of
WT2 PBMC adhesion to uPAR?/?aorta was completely abrogated as compared with WT2 aortas. The data are presented as mean ? SEM for five
individual experiments performed in triplicate (??, p ? 0.01 between C5a/TNF-?-stimulated and nonstimulated aortas and ??, p ? 0.01; ???, p ? 0.001
between WT2 and uPAR?/?aortas under the same stimulation conditions).
uPAR-deficiency strongly reduces C5a/TNF-?-induced up-regulation of ICAM-1 in endothelial cells and completely abrogates C5a/TNF-
1186 uPAR AND KIDNEY TRANSPLANTATION
anti-apoptotic and uPA-independent action of uPAR in endo-
thelial cells in which anti-uPAR Abs as well as soluble recom-
binant uPAR blocked the apoptotic effect of cleaved high mo-
lecular mass kininogen by preventing of the binding of high
molecular mass kininogen to these cells (46).
Recently, it has been shown that uPA stimulates ROS produc-
tion in VSMC in vitro (47). Because hypoxia triggers ROS gen-
eration and mediates apoptosis, we investigated the possible role of
uPA/uPAR interaction in ROS generation and subsequent ROS-
dependent apoptosis using both uPA- and uPAR-deficient mice.
Our results clearly demonstrate that in both models uPAR defi-
ciency protected renal tissue from ROS generation independently
from uPA. The molecular mechanisms underlying the attenuated
ROS formation in uPAR?/?allografts remain to be investigated.
uPAR can interact laterally with a wide variety of membrane pro-
teins including integrins, endocytic receptors, caveolin, the gp130
cytokine receptor, the epidermal growth factor receptor, chemoat-
tractant receptors (1) and platelet-derived growth factor receptor
(2). This interaction might activate both pro- and anti-apoptotic
downstream signaling pathways. For example, recently ROS-trig-
gered apoptosis in polymorphonuclear leukocytes (PMNs) cells
has been shown to be ?2 integrin-dependent (48). Therefore,
uPAR may stimulate ROS production via its lateral interaction
with other molecules.
The early inflammatory response during reperfusion of allo-
grafts is initiated by the infiltration of PMNs into the graft (38)
followed by infiltration of monocytes/macrophages (MO). Mac-
rophages constitute 38 to 60% of infiltrating cells during acute
allograft rejection and increased influx of MO has been strongly
correlated to complement activation and acute rejection in a
human protocol biopsy study (49). Furthermore, it has been
shown that MO infiltration 3 mo after transplantation correlated
inversely with graft survival (50). It is well known that uPAR
deficiency on the surface of leukocytes strongly reduces their
migratory capacity both in vivo and in vitro (15, 51, 52). How-
ever, it should be noted that in our transplant experiments, the
host leukocytes expressed uPAR normally. Despite the normal
expression of uPAR on host leukocytes, we found strongly re-
duced PMN/MO infiltration in kidney allografts from uPAR?/?
as compared with those from WT2 mice. This finding stresses
the importance of the local uPAR expression on resident renal
cells in the allograft for adequate cell adhesion and subsequent
PMN/MO accumulation after kidney transplantation, indepen-
dently from the uPAR expression on infiltrating cells. Recently
it has been shown that inhibition of PMN infiltration into car-
diac allografts may have a significant downstream impact on the
efficacy of recipient T cell responses to the allograft (38). In
contrast, in our model of acute renal allograft rejection we did
not find any differences in CD4 and CD8 positive cell infiltra-
tion between WT2 and uPAR?/?allografts after transplanta-
tion. Moreover, the expression of T cell activation Ag CD25
used as a marker for T cell activation (53) was also similar in
WT2 and uPAR?/?allografts. Furthermore, by investigating of
priming of alloreactive T cells by MLR and IFN-? ELISPOT
assays we did not found differences of T cell function of WT2
and uPAR?/?allografts recipients. These results suggest that
superior survival of uPAR?/?deficient allografts is not medi-
ated by T cell-dependent alloimmune response.
The migration of inflammatory cells into an extravascular site
requires a series of coordinated signals including the generation of
a chemotactic gradient by the cells of the extravascular compart-
ment and up-regulation of adhesion molecules on activated endo-
thelium. The strong up-regulation of MCP-1 and MIP-2 has been
demonstrated in animal models during renal ischemia as well as in
renal biopsies from patients with acute and chronic renal rejection
(54). In line with this report, we observed a strong up-regulation of
these proinflammatory mediators which was similar in WT2 and
uPAR?/?allografts. Therefore, it could be concluded that the level
of these chemoattractants was not responsible for the diminished
PMN/MO infiltration in uPAR?/?allografts.
Adhesion molecules are rapidly up-regulated early after trans-
plantation (55) and after induction of IR injury (56). Therefore, we
analyzed expression of ICAM-1 in WT2 and uPAR?/?allografts
given the facts that, being the major counterreceptor for leukocyte
?2-integrins, this adhesion molecule is rapidly up-regulated within
the first 2–3 h after transplantation (57) and the blockade of
ICAM-1 strongly attenuated PMN infiltration in some experimen-
tal models of renal ischemia (58–60). Indeed, we could demon-
strate that the transplantation-induced up-regulation of ICAM-1
observed in WT2 allografts was strongly attenuated in uPAR?/?
allografts. However, the reduced expression of ICAM-1 in
uPAR?/?allografts was not sufficient to prevent T cell infiltration.
This result coincides with observation of Zhang and coworkers
(61) demonstrating that primed alloreative T cells do not require
allograft expression of ICAM-1 to infiltrate heart allografts. The
interaction of integrin leukocyte function Ag-1 with its alternative
ligand ICAM-2, or interaction of other T cell homing receptors
with their ligands on endothelium, such as CD44/E-selectin inter-
action, may be involved in lymphocyte adhesion to the vascular
endothelium of the graft and its subsequent infiltration bypassing
the requirement for ICAM-1 (62).
By performing in vitro experiments with primary aortic endo-
thelial cells we could elucidate impaired TNF-? and C5a signaling
in uPAR?/?cells leading to strongly decreased TNF-?-induced
and completely abrogated C5a-induced up-regulation of ICAM-1.
In line with a prior report that only the TNFR1 receptor is involved
in TNF-?-induced ICAM-1 up-regulation (63) we could demon-
strate that the up-regulation of ICAM-1 in MAEC after TNF-?
stimulation was mediated predominantly via TNFR1 because the
neutralizing anti-mouse TNFR1 Ab almost completely blocked
this effect. These results coincide with strongly reduced adhesion
of WT2 PBMC to isolated aortas from uPAR?/?mice compared
with WT2 aortas after stimulation with TNF-? and C5a. This result
is in line with a prior report that platelet uPAR is critical for the
response to TNF-? (64) and coincides also with our previous ob-
servation that uPAR is necessary for C5a/C5aR-mediated re-
sponses in mouse alveolar macrophages (15) and in human mes-
angial cells (14). Collectively, these data suggest that uPAR plays
a major role in mediating IR injury and subsequently influences
early inflammation and allograft survival in renal transplantation
rather than directly affecting T cell mediated alloimmune
In summary, this study shows that the local expression of uPAR
on resident renal cells of the allograft contributes to the develop-
ment of acute allograft rejection via at least two different path-
ways, ischemia-induced apoptosis and the infiltration of host leu-
kocytes. These results suggest that uPAR, whose role was
presumed to be involved in the migratory behavior of infiltrating
cells, has a broader critical function as an early regulator of isch-
emia-triggered initial generation of ROS, which, in turn, induces
the apoptosis in intrinsic renal cells. uPAR is also necessary for
adequate TNF-? and C5a signaling, leading to up-regulation of
ICAM-1 on endothelial cells of the allograft which is a crucial step
for adequate leukocyte recruitment to the inflamed tissue. These
observations underscore a new role of uPAR in acute allograft
rejection and highlights uPAR as a target for prevention of organ
dysfunction and damage in IR injury.
1187 The Journal of Immunology
We thank Yvonne Nikolai, Herle Chlebusch, and Kerstin Bankes for ex-
cellent technical assistance.
The authors have no financial conflict of interest.
1. Blasi, F., and P. Carmeliet. 2002. uPAR: a versatile signalling orchestrator. Nat.
Rev. Mol. Cell Biol. 3: 932–943.
2. Kiyan, J., R. Kiyan, H. Haller, and I. Dumler. 2005. Urokinase-induced signaling
in human vascular smooth muscle cells is mediated by PDGFR-?. EMBO J. 24:
3. Mondino, A., and F. Blasi. 2004. uPA and uPAR in fibrinolysis, immunity and
pathology. Trends Immunol. 25: 450–455.
4. Woodhead, V. E., T. J. Stonehouse, M. H. Binks, K. Speidel, D. A. Fox, A. Gaya,
D. Hardie, A. J. Henniker, V. Horejsi, K. Sagawa, et al. 2000. Novel molecular
mechanisms of dendritic cell-induced T cell activation. Int. Immunol. 12:
5. Nykjaer, A., B. Moller, R. F. Todd 3rd, T. Christensen, P. A. Andreasen,
J. Gliemann, and C. M. Petersen. 1994. Urokinase receptor: an activation antigen
in human T lymphocytes. J. Immunol. 152: 505–516.
6. Sitrin, R. G., S. B. Shollenberger, R. M. Strieter, and M. R. Gyetko. 1996. En-
dogenously produced urokinase amplifies tumor necrosis factor-? secretion by
THP-1 mononuclear phagocytes. J. Leukocyte Biol. 59: 302–311.
7. Abraham, E., M. R. Gyetko, K. Kuhn, J. Arcaroli, D. Strassheim, J. S. Park,
S. Shetty, and S. Idell. 2003. Urokinase-type plasminogen activator potenti-
ates lipopolysaccharide-induced neutrophil activation. J. Immunol. 170:
8. Goldfarb, R. H., T. Timonen, and R. B. Herberman. 1984. Production of plas-
minogen activator by human natural killer cells: large granular lymphocytes.
J. Exp. Med. 159: 935–951.
9. Chapman, H. A. 1997. Plasminogen activators, integrins, and the coordinated
regulation of cell adhesion and migration. Curr. Opin. Cell Biol. 9: 714–724.
10. May, A. E., S. M. Kanse, L. R. Lund, R. H. Gisler, B. A. Imhof, and
K. T. Preissner. 1998. Urokinase receptor (CD87) regulates leukocyte recruitment
via ?2 integrins in vivo. J. Exp. Med. 188: 1029–1037.
11. Gyetko, M. R., G. H. Chen, R. A. McDonald, R. Goodman, G. B. Huffnagle,
C. C. Wilkinson, J. A. Fuller, and G. B. Toews. 1996. Urokinase is required for
the pulmonary inflammatory response to Cryptococcus neoformans: a murine
transgenic model. J. Clin. Invest. 97: 1818–1826.
12. Gyetko, M. R., S. Sud, T. Kendall, J. A. Fuller, M. W. Newstead, and
T. J. Standiford. 2000. Urokinase receptor-deficient mice have impaired neutro-
phil recruitment in response to pulmonary Pseudomonas aeruginosa infection.
J. Immunol. 165: 1513–1519.
13. Cao, D., I. F. Mizukami, B. A. Garni-Wagner, A. L. Kindzelskii, R. F. Todd 3rd,
L. A. Boxer, and H. R. Petty. 1995. Human urokinase-type plasminogen activator
primes neutrophils for superoxide anion release: possible roles of complement
receptor type 3 and calcium. J. Immunol. 154: 1817–1829.
14. Shushakova, N., N. Tkachuk, M. Dangers, S. Tkachuk, J. K. Park, J. Zwirner,
K. Hashimoto, H. Haller, and I. Dumler. 2005. Urokinase-induced activation of
the gp130/Tyk2/Stat3 pathway mediates a pro-inflammatory effect in human mes-
angial cells via expression of the anaphylatoxin C5a receptor. J. Cell Sci. 118:
15. Shushakova, N., G. Eden, M. Dangers, J. Zwirner, J. Menne, F. Gueler,
F. C. Luft, H. Haller, and I. Dumler. 2005. The urokinase/urokinase receptor
system mediates the IgG immune complex-induced inflammation in lung. J. Im-
munol. 175: 4060–4068.
16. Boros, P., and J. S. Bromberg. 2006. New cellular and molecular immune path-
ways in ischemia/reperfusion injury. Am. J. Transplant. 6: 652–658.
17. Schwarz, A., M. Mengel, W. Gwinner, J. Radermacher, M. Hiss, H. Kreipe, and
H. Haller. 2005. Risk factors for chronic allograft nephropathy after renal trans-
plantation: a protocol biopsy study. Kidney Int. 67: 341–348.
18. Gueler, F., W. Gwinner, A. Schwarz, and H. Haller. 2004. Long-term effects of
acute ischemia and reperfusion injury. Kidney Int. 66: 523–527.
19. Land, W. 2002. Postischemic reperfusion injury to allografts: a case for “innate
immunity”? Eur. Surg. Res. 34: 160–169.
20. Ha, H., J. Park, Y. S. Kim, and H. Endou. 2004. Oxidative stress and chronic
allograft nephropathy. Yonsei Med. J. 45: 1049–1052.
21. Land, W. G. 2005. The role of postischemic reperfusion injury and other non-
antigen-dependent inflammatory pathways in transplantation. Transplantation
22. Dragun, D., U. Hoff, J. K. Park, Y. Qun, W. Schneider, F. C. Luft, and H. Haller.
2000. Ischemia-reperfusion injury in renal transplantation is independent of the
immunologic background. Kidney Int. 58: 2166–2177.
23. Tjarnstrom, J., L. Holmdahl, P. Falk, M. Falkenberg, P. Arnell, and B. Risberg.
2001. Effects of hyperbaric oxygen on expression of fibrinolytic factors of human
endothelium in a simulated ischaemia/reperfusion situation. Scand. J. Clin. Lab.
Invest. 61: 539–545.
24. Yoon, S. Y., Y. J. Lee, J. H. Seo, H. J. Sung, K. H. Park, I. K. Choi, S. J. Kim,
S. C. Oh, C. W. Choi, B. S. Kim, et al. 2006. uPAR expression under hypoxic
conditions depends on iNOS modulated ERK phosphorylation in the MDA-MB-
231 breast carcinoma cell line. Cell Res. 16: 75–81.
25. Tang, W. H., H. Friess, F. F. di Mola, M. Schilling, C. Maurer, H. U. Graber,
C. Dervenis, A. Zimmermann, and M. W. Buchler. 1998. Activation of the
serine proteinase system in chronic kidney rejection. Transplantation 65:
26. Roelofs, J. J., A. T. Rowshani, J. G. van den Berg, N. Claessen, J. Aten,
I. J. ten Berge, J. J. Weening, and S. Florquin. 2003. Expression of urokinase
plasminogen activator and its receptor during acute renal allograft rejection. Kid-
ney Int. 64: 1845–1853.
27. Carmeliet, P., L. Schoonjans, L. Kieckens, B. Ream, J. Degen, R. Bronson,
R. De Vos, J. J. van den Oord, D. Collen, and R. C. Mulligan. 1994. Physiological
consequences of loss of plasminogen activator gene function in mice. Nature 368:
28. Mannon, R. B., C. Doyle, R. Griffiths, M. Bustos, J. L. Platt, and
T. M. Coffman. 2000. Altered intragraft immune responses and improved
renal function in MHC class II-deficient mouse kidney allografts. Transplan-
tation 69: 2137–2143.
29. Han, W. R., L. J. Murray-Segal, and P. L. Mottram. 1999. Modified technique for
kidney transplantation in mice. Microsurgery 19: 272–274.
30. Racusen, L. C. 2004. The Banff schema and differential diagnosis of allograft
dysfunction. Transplant. Proc. 36: 753–754.
31. Fleming, I. 2001. Cytochrome p450 and vascular homeostasis. Circ. Res. 89:
32. Muller, P. Y., H. Janovjak, A. R. Miserez, and Z. Dobbie. 2002. Processing of
gene expression data generated by quantitative real-time RT-PCR. BioTechniques
33. Kreisel, D., A. S. Krupnick, W. Y. Szeto, S. H. Popma, D. Sankaran,
A. M. Krasinskas, K. M. Amin, and B. R. Rosengard. 2001. A simple method for
culturing mouse vascular endothelium. J. Immunol. Methods 254: 31–45.
34. Hubbard, A. K., and C. Giardina. 2000. Regulation of ICAM-1 expression in
mouse macrophages. Inflammation 24: 115–125.
35. Jonassen, J. A., L. C. Cao, T. Honeyman, and C. R. Scheid. 2003. Mechanisms
mediating oxalate-induced alterations in renal cell functions. Crit. Rev. Eukary-
otic Gene Expression 13: 55–72.
36. Wang, X., N. Takahashi, H. Uramoto, and Y. Okada. 2005. Chloride channel
inhibition prevents ROS-dependent apoptosis induced by ischemia-reperfusion in
mouse cardiomyocytes. Cell. Physiol. Biochem. 16: 147–154.
37. Tinckam, K. J., O. Djurdjev, and A. B. Magil. 2005. Glomerular monocytes
predict worse outcomes after acute renal allograft rejection independent of C4d
status. Kidney Int. 68: 1866–1874.
38. El-Sawy, T., J. A. Belperio, R. M. Strieter, D. G. Remick, and R. L. Fairchild.
2005. Inhibition of polymorphonuclear leukocyte-mediated graft damage syner-
gizes with short-term costimulatory blockade to prevent cardiac allograft rejec-
tion. Circulation 112: 320–331.
39. Perez de Lema, G., H. Maier, E. Nieto, V. Vielhauer, B. Luckow, F. Mampaso,
and D. Schlondorff. 2001. Chemokine expression precedes inflammatory cell
infiltration and chemokine receptor and cytokine expression during the initiation
of murine lupus nephritis. J. Am. Soc. Nephrol. 12: 1369–1382.
40. Wenzel, U., A. Schneider, A. J. Valente, H. E. Abboud, F. Thaiss,
U. M. Helmchen, and R. A. Stahl. 1997. Monocyte chemoattractant protein-1
mediates monocyte/macrophage influx in anti-thymocyte antibody-induced glo-
merulonephritis. Kidney Int. 51: 770–776.
41. Albrecht, E. A., A. M. Chinnaiyan, S. Varambally, C. Kumar-Sinha,
T. R. Barrette, J. V. Sarma, and P. A. Ward. 2004. C5a-induced gene expression
in human umbilical vein endothelial cells. Am. J. Pathol. 164: 849–859.
42. Meldrum, K. K., D. R. Meldrum, X. Meng, L. Ao, and A. H. Harken. 2002.
TNF-?-dependent bilateral renal injury is induced by unilateral renal ischemia-
reperfusion. Am. J. Physiol. 282: H540–H546.
43. de Vries, B., J. Kohl, W. K. Leclercq, T. G. Wolfs, A. A. van Bijnen,
P. Heeringa, and W. A. Buurman. 2003. Complement factor C5a mediates
renal ischemia-reperfusion injury independent from neutrophils. J. Immunol.
44. Zhang, G., H. Kim, X. Cai, J. M. Lopez-Guisa, P. Carmeliet, and A. A. Eddy.
2003. Urokinase receptor modulates cellular and angiogenic responses in ob-
structive nephropathy. J. Am. Soc. Nephrol. 14: 1234–1353.
45. Kin, Y., S. K. Chintala, Y. Go, R. Sawaya, S. Mohanam, A. P. Kyritsis, and
J. S. Rao. 2000. A novel role for the urokinase-type plasminogen activator re-
ceptor in apoptosis of malignant gliomas. Int. J. Oncol. 17: 61–65.
46. Cao, D. J., Y. L. Guo, and R. W. Colman. 2004. Urokinase-type plasminogen
activator receptor is involved in mediating the apoptotic effect of cleaved high
molecular weight kininogen in human endothelial cells. Circ. Res. 94:
47. Menshikov, M., O. Plekhanova, H. Cai, K. Chalupsky, Y. Parfyonova,
P. Bashtrikov, V. Tkachuk, and B. C. Berk. 2006. Urokinase plasminogen acti-
vator stimulates vascular smooth muscle cell proliferation via redox-dependent
pathways. Arterioscler. Thromb. Vasc. Biol. 26: 801–807.
48. Zhang, B., J. Hirahashi, X. Cullere, and T. N. Mayadas. 2003. Elucidation of
molecular events leading to neutrophil apoptosis following phagocytosis: cross-
talk between caspase 8, reactive oxygen species, and MAPK/ERK activation.
J. Biol. Chem. 278: 28443–28454.
49. Sund, S., A. V. Reisaeter, H. Scott, T. E. Mollnes, and T. Hovig. 2004. Glomer-
ular monocyte/macrophage influx correlates strongly with complement activation
in 1-week protocol kidney allograft biopsies. Clin. Nephrol. 62: 121–130.
50. Srinivas, T. R., P. S. Kubilis, and B. P. Croker. 2004. Macrophage index
predicts short-term renal allograft function and graft survival. Transpl. Int.
51. Gyetko, M. R., R. F. Todd 3rd, C. C. Wilkinson, and R. G. Sitrin. 1994. The
urokinase receptor is required for human monocyte chemotaxis in vitro. J. Clin.
Invest. 93: 1380–1387.
1188uPAR AND KIDNEY TRANSPLANTATION
52. Waltz, D. A., R. M. Fujita, X. Yang, L. Natkin, S. Zhuo, C. J. Gerard,
S. Rosenberg, and H. A. Chapman. 2000. Nonproteolytic role for the uroki-
nase receptor in cellular migration in vivo. Am. J. Respir. Cell Mol. Biol. 22:
53. Wildner, O., and J. C. Morris. 2000. Therapy of peritoneal carcinomatosis from
colon cancer with oncolytic adenoviruses. J. Gene Med. 2: 353–360.
54. Segerer, S., P. J. Nelson, and D. Schlondorff. 2000. Chemokines, chemokine
receptors, and renal disease: from basic science to pathophysiologic and thera-
peutic studies. J. Am. Soc. Nephrol. 11: 152–176.
55. Nagano, H., K. C. Nadeau, M. Takada, M. Kusaka, and N. L. Tilney. 1997.
Sequential cellular and molecular kinetics in acutely rejecting renal allografts in
rats. Transplantation 63: 1101–1108.
56. De Greef, K. E., D. K. Ysebaert, V. Persy, S. R. Vercauteren, and M. E. De Broe.
2003. ICAM-1 expression and leukocyte accumulation in inner stripe of outer
medulla in early phase of ischemic compared to HgCl2-induced ARF. Kidney Int.
57. Neto, J. S., A. Nakao, K. Kimizuka, A. J. Romanosky, D. B. Stolz, T. Uchiyama,
M. A. Nalesnik, L. E. Otterbein, and N. Murase. 2004. Protection of transplant-
induced renal ischemia-reperfusion injury with carbon monoxide. Am. J. Physiol.
58. Rabb, H., Y. M. O’Meara, P. Maderna, P. Coleman, and H. R. Brady. 1997.
Leukocytes, cell adhesion molecules and ischemic acute renal failure. Kidney Int.
59. Patel, N. S., S. Cuzzocrea, P. K. Chatterjee, R. Di Paola, L. Sautebin, D. Britti,
and C. Thiemermann. 2004. Reduction of renal ischemia-reperfusion injury in
5-lipoxygenase knockout mice and by the 5-lipoxygenase inhibitor zileuton. Mol.
Pharmacol. 66: 220–227.
60. Haller, H., D. Dragun, A. Miethke, J. K. Park, A. Weis, A. Lippoldt, V. Gross,
and F. C. Luft. 1996. Antisense oligonucleotides for ICAM-1 attenuate reperfu-
sion injury and renal failure in the rat. Kidney Int. 50: 473–480.
61. Zhang, Q. W., D. D. Kish, and R. L. Fairchild. 2003. Absence of allograft
ICAM-1 attenuates alloantigen-specific T cell priming, but not primed T cell
trafficking into the graft, to mediate acute rejection. J. Immunol. 170:
62. Sackstein, R. 2005. The lymphocyte homing receptors: gatekeepers of the mul-
tistep paradigm. Curr. Opin. Hematol. 12: 444–450.
63. De Cesaris, P., D. Starace, G. Starace, A. Filippini, M. Stefanini, and E. Ziparo.
1999. Activation of Jun N-terminal kinase/stress-activated protein kinase path-
way by tumor necrosis factor ? leads to intercellular adhesion molecule-1 ex-
pression. J. Biol. Chem. 274: 28978–28982.
64. Piguet, P. F., C. Vesin, Y. Donati, F. Tacchini-Cottier, D. Belin, and
C. Barazzone. 1999. Urokinase receptor (uPAR, CD87) is a platelet receptor
important for kinetics and TNF-induced endothelial adhesion in mice. Circula-
tion 99: 3315–3321.
1189 The Journal of Immunology