| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
* German Cancer Research Center, Department of Experimental Pathology, Germany;
# Internal Medicine University of Utrecht, Netherlands;
Institute for Prophylaxis and Epidemiology, Ludwig-Maximilians-University of Munich, Germany;
§ Medical Policlinic, Ludwig-Maximilians-University of Munich, Germany; and
|| Serono Pharmaceutical Institute, Geneva, Switzerland
1Correspondence: Medizinische Poliklinik der Ludwig-Maximilians-Universität München, AG Klinische Biochemie, Schillerstrasse 42, D-80336, Munich, Germany. E-mail: nelson{at}medpoli.med.uni-muenchen.de
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: chemokine receptors • inflammation • monocyte • endothelium
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
2
3)
.
The recruitment of these leukocytes from the peripheral circulation
into the transplanted organ involves a complex interplay between a
series of molecules expressed on the leukocyte and endothelial surface
(4
5
6)
. Chemokines, a large superfamily of structurally related
cytokines, have been shown to selectively promote the rapid adhesion,
chemotaxis, and activation of specific leukocyte effector
subpopulations (6
7
8
9)
.
Chemokines are characterized by a series of shared structural
elements including the conserved cysteine residues used to define the
C, C-C, C-X-C, and C-X3-C chemokine subgroups
(where X represents an intervening amino acid residue between the first
two amino-terminal proximal cysteines). All of the various biological
actions of chemokines appear to be directed through their interaction
with a large family of seven-transmembrane spanning, G-protein-coupled
receptors (7
8
9)
. The cell type-specific expression of these receptors
appears to control to a significant degree the leukocyte specificity of
chemokine action (7
8
9)
.
The chemokine RANTES (regulated on activation, normal T cell expressed
and
secreted),2
a member of the C-C chemokine subfamily, is a potent chemoattractant
for T cells, monocytes, natural killer cells, basophils, and
eosinophils (10)
. RANTES is a ligand for a number of chemokine
receptors including CCR1, CCR3, CCR5, CCR9, and DARC (Duffy antigen
receptor for chemokines) in humans (7
8
9
10
11)
. Most tissues can produce
RANTES in response to proinflammatory stimuli such as interleukin 1ß
(IL-1ß) or tumor necrosis factor 
(TNF-) (9
, 11)
. Platelets
sequester RANTES protein in their -granules and release it during
acute stages of inflammation (9
, 11
12
13)
.
Chemokines such as RANTES are thought to play pivotal roles in the
cellular infiltrates that underlie various disease processes. For
example, RANTES is expressed in vivo in diseases
characterized by a mononuclear cell infiltrate, including delayed-type
hypersensitivity, necrotizing glomerulonephritis, inflammatory lung
disease, and renal allograft rejection (9
, 11
, 14
15
16
17
18
). In studies of
human kidneys undergoing acute cellular rejection, RANTES protein was
found localized to mononuclear infiltrating cells, renal tubular
epithelial cells, and the endothelium of peritubular capillaries (17
,
18
). Since acute cellular rejection is characterized by an
intravascular and interstitial cellular infiltrate consisting of
monocytes/macrophages, T lymphocytes, and occasional eosinophils,
RANTES is potentially a key player in the pathogenesis of acute
rejection (9
, 11
, 17
, 18
).
Based on these observations, a model for the role of RANTES in
renal allograft rejection was proposed (11
, 17
, 18
). Early in
rejection, the microvascular endothelium becomes inflamed and platelets
degranulate, releasing RANTES protein that binds to the endothelial
surface. The inflamed renal tubules and endothelial cells produce
additional chemokines, including RANTES. The accumulated surface-bound
chemokines then provide directional signals to circulating leukocytes
as they roll across the endothelial surface (4
5
6
, 11
, 17
, 18
).
Leukocytes recognize the surface-bound protein, up-regulate integrins,
firmly adhere to the endothelial surface, and undergo diapedesis and
extravasation. As the leukocytes become activated, they produce
additional cytokines and chemokines, thus amplifying and propagating
the inflammatory response (11
, 17
, 18
).
We investigated the functional role of RANTES and its receptors in rat
models of acute renal allograft rejection. Modification of the amino
terminus of the RANTES protein can dramatically alter its properties
(19
20
21)
. The addition of a single methionine residue changes the
agonist protein into a RANTES receptor antagonist with nanomolar
potency (19)
. This antagonist, Met-RANTES, is bioactive in mouse and
rat (A. E. Proudfoot, unpublished results), and has been shown to
suppress inflammation in murine models of allergic skin and rheumatoid
arthritis and to partially inhibit necrotizing glomerulonephritis
(22
23
24)
. We describe here the effect of Met-RANTES on development of
the cellular infiltrate in acute kidney transplant rejection.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. Primary human dermal microvascular
endothelial cells (DMVEC) from human neonatal foreskin were obtained
from Dr. K. Degitz (Dermatology, LMU, Munich, Germany). The cells were
carried in MCDB 131 media (Life Technologies, Gaithersburg, Md.)
supplemented with 10% fetal calf serum (Boehringer Mannheim, Mannheim,
Germany), 1 x 10-6M hydrocortisone acetate
(Sigma, St. Louis, Mo.), 0.5 µg/ml dibutyryladenosinemonophosphate
(Sigma), 2 mM glutamine (Seromed, Berlin, Germany), 100 U/ml
penicillin, 100 µg/ml streptomycin (Life Technologies GMBH,
Karlsruhe, Germany), EGF 10 µg/ml (Collaborative Biomedical Products,
Bedford, Mass.), and incubated at 37°C and 5%
CO2. The cells were grown on flasks, petri plates
(Costar, Cambridge, Mass.), or 96-well, flat-bottom plates (Nunc,
Roskilde, Denmark) precoated with 0.5% gelatin (Sigma). Cells were
characterized through morphological appearance and immunofluorescence
flow cytometry for CD31 surface expression.
Histology reagents
Materials for histological studies were obtained from Merck
(Rahway, N.J.) and for immunological measurements from Sigma. IL-1ß
and TNF- were purchased from Sigma. Generation of recombinant human
RANTES and the human RANTES-specific monoclonal antibody VL1 was
described previously (26)
. Met-RANTES was produced for in
vivo studies as described (19
, 22
23
24
).
Animals and renal transplantation
Inbred male rats were used in all experiments. Lewis (LEW,
RT1l) rats served as recipients of Fisher 344
(F344 RT1lvl) or Brown Norway (BN
RT1n) kidneys. The animals, purchased from
Charles River GmbH (Sulzfeld, Germany), weighed 190 to 250 g (Lew
and F344) and 140 to 170 g (BN) to adjust for ureter diameter.
Transplantation was performed using a modification of the technique
originally described by Fisher and Lee (27)
. Briefly, animals were
anesthetized by ether-drop anesthesia and the donor kidney was flushed
with 5 ml of ice-cold 0.9% NaCl with or without 100 µg Met-RANTES.
The kidney and ureter were removed en bloc including the renal artery
with a 5 mm aortic cuff and the renal vein with a 3 mm vena cava patch.
The kidneys were stored in 0.9% NaCl at 4°C.
The donor kidney was transplanted to the abdominal aorta and inferior vena cava of the recipient animal, below the left renal artery, by end-to-side anastomoses with 8–0 nonabsorbable monofilament nylon suture. Ureter anastomosis was performed end-to-end with 11–0 nonabsorbable monofilament nylon suture. Total ischemic time of the donor kidney varied between 30 and 40 min. Hydronephrosis was evaluated both macroscopically at time of death and by light microscopy. All animals with hydronephrosis were excluded from the experimental groups. The left kidney of the recipient was always removed at the time of transplantation. In the Fisher to Lewis transplantation, the right kidney was left in place as an internal control for the effects of Met-RANTES. In Brown Norway to Lewis transplantations, a bilateral nephrectomy was performed at the time of transplantation.
Experimental groups
Cyclosporin A (CyA) (kindly provided by Sandoz, Basel,
Switzerland) was dissolved in olive oil and administered subcutaneously
(s.c.) in a concentration of 2.5 mg/kg body weight per day for 12 days,
starting 4 h posttransplantation. Met-RANTES was dissolved in
water, adjusted to 0.9% sodium chloride, and injected once daily
intravenously (i.v.) at a dose of 200 µg per day in Fisher/Lewis and
50 µg per day in Brown Norway/Lewis transplantation experiments.
The experimental groups were as follows:
Group 1: Fisher 344 kidney into Lewis rat with one endogenous kidney.
Group 1a: with Met-RANTES 200 µg/day for 7 days (n=9).
Group 1b: without Met-RANTES for 7 days (n=9).
Group 2: Brown Norway kidney into bilaterally nephrectomized Lewis rat with CyA 2.5 mg/kg body weight administered per day.
Group 2a: with Met-RANTES, 50 µg/day for 12 days (n=4).
Group 2b: without Met-RANTES for 12 days (n=4).
Serum analysis
Blood taken from the aorta at the time of death was analyzed for
creatinine, urea, using an automated serum analyzer. These measurements
were only relevant in the Brown Norway to Lewis transplant model
(bilateral nephrectomy).
Histology
Organs (lung, liver, kidney, spleen) were removed under deep
anesthesia, quickly blotted free of blood, weighed, and processed as
needed for histology, immunohistochemistry, or in situ
hybridization. The organs were cut into 1 mm slices and either
immersion-fixed in 4% formaldehyde in phosphate-buffered saline (PBS)
pH 7.35 (PBS: 99 mM
NaH2PO4, 108 mM
NaH2PO4, and 248 mM NaCl)
for 24 h or fixed in methacarn for 8 h, and embedded in
paraffin or frozen in liquid nitrogen and stored at -80°C until used
for immunohistochemistry. Light microscopy was performed on 3 µM
sections stained by periodic acid-Schiff or Goldner-Elastica.
Immunohistochemistry
The monoclonal antibody ED1 (Serotec/Camon) was used on
methacarn fixed, paraffin-embedded tissue (3 µM) to demonstrate
monocytes/macrophages. An alkaline phosphatase anti-alkaline
phosphatase detection system was applied for visualization (Dako,
Bucks, U.K.). Controls, omitting the first or second antibody for each
section tested were negative.
Morphometry
The vascular injury score for preglomerular vessels with
endothelial damage, thrombus, and endothelialitis were assessed as
showing no injury [0], a mild [1], moderate [2], and severe [3]
degree of injury and evaluated in whole kidney sections including
cortex, outer, and inner medulla. A degree-specific vascular injury
index was defined as the percentage of vessels with the respective
degree of injury encountered in a whole kidney section. The total
vascular injury score was calculated as the sum of all vessels, with
all degrees of vascular injury, whereby the number of vessels with
degree one was multiplied by one, that of degree two, by a factor of
two, and that of degree three by a factor of three (28)
. Tubular damage
was evaluated as nonexistent [0], mild [1], moderate [2], and
severe [3] as judged in 20 high-power fields of cortex and outer
stripe of outer medulla. The total tubular damage score was calculated
as described for the total vascular injury score. The extent of
interstitial infiltration by mononuclear cells was judged as
nonexistent [0], mild [1], moderate [2], or severe [3] and the
total interstitial inflammation score was calculated as described for
the total vascular injury score. The number of monocytes/macrophages
and T cells within capillary convolutes of glomeruli was calculated as
the mean of the respective numbers in all glomeruli in one kidney
section.
In situ hybridization
Single-stranded RNA probes were generated by in vitro
transcription of a cDNA clone of rat RANTES (Dr. H. Sprenger, Marburg,
Germany). In vitro transcription was carried out using a
Trans-Probe-T kit (Pharmacia) and digoxigenin-labeled uridine
triphosphate (Boehringer). The subsequent hybridization, washing, and
development of color reaction were performed essentially as described
(28
, 29)
.
RNase protection assay
Total RNA was isolated from whole rat kidney as described
previously (29)
. RNase protection experiments were performed using a
commercial RPA kit (PharMingen, probe rCK-1). This kit allowed the
simultaneous measurement of mRNA species for rat: IL-1, IL-1-ß,
TNF-, TNF-ß, IL-2, IL-3, IL-4, IL-5, IL-6, IL-10, and interferon
(IFN-), and the housekeeping genes GAPDH and L32. Total RNA (20
µg) was used for each determination. The protected samples were run
out on a precast gel (Quickpoint Rapid Nucleic Acid Separation System,
Novex). The intensity of the specific bands was quantitated using a
Molecular Dynamics Storm 840 PhosphorImager, normalized to L32
gene expression, and averaged over the three animals analyzed.
In vitro binding assay
DMVEC were grown to confluence on coated 96-well, flat-bottom
plates. The resultant endothelial monolayer was either left untreated
or treated with IL-1ß (0.1 to 5 ng/ml) for 12 h. The cells were
then washed twice with 1xPBS at 4°C and fixed for 15 min at 4°C
with 2% paraformaldehyde, 0.2% glutaraldehyde in 1xPBS. The
monolayer was then washed four times with cold 1xPBS. The RANTES
binding assay was a modification of a previously described procedure
(17
, 18
). Horseradish peroxidase (HRP) -conjugated, anti-human-RANTES
monoclonal antibody VL1 (0.5 µg/ml) was preincubated at 25°C for 30
min with an excess of recombinant human RANTES (10 µg/ml) in 1xPBS.
The chemokine–antibody complex (used to assay the relative chemokine
binding capacity of the microvascular endothelium) was added at 50 µl
per well to the fixed cells and the plates were incubated at 25°C for
45 min, followed by four rounds of washing with cold 1xPBS. The HRP
reaction was developed for 1–4 min and the optical density at 405 nm
of the plate was determined using an enzyme-linked immunoassay (ELISA)
plate reader. A standard curve for relative `fold' binding of the
complex was generated by determining the signal from a serial dilution
of the RANTES/VL1 monoclonal antibody complex bound to activated (5
ng/ml IL-1ß) microvascular endothelium. The results are normalized
such that the relative binding to unstimulated microvascular
endothelial cells represents a value of one. All experiments were
performed in quadruplicate and the results displayed are representative
of four separate experiments.
Fluorescence-activated cell sorting (FACS) analysis
Flow cytometry analysis of DMVEC was performed essentially as
described (30)
. Briefly, confluent DMVEC stimulated with or without
IL-1ß (5 ng/ml for 12 h) were trypsinized, reacted with
saturating concentrations of ICAM-1 monoclonal antibody (mAb) RR1/1
(kindly provided by Dr. R. Rothlein), E-selectin mAb, VCAM-1 mAb (both
Serotec) or isotype control for 30 min on ice, stained with fluorescein
isothiocyanate-conjugated goat anti-mouse IgG (Boehringer Mannheim),
and analyzed in a FACScan (Becton Dickinson, Rutherford, N.J.). After
correction for unspecific binding, data were expressed as specific mean
log fluorescence intensity in channels.
In vitro model system of monocyte recruitment on
microvascular endothelium under physiological flow conditions
The interaction of monocytes with DMVEC was studied in laminar
flow assays performed essentially as described (31
32
33)
. Briefly, DMVEC
were grown to confluence in 35 mm petri dishes and stimulated with
IL-1ß (5 ng/ml) for 12 h or left untreated. The plates were
assembled as the lower wall in a parallel wall flow chamber and mounted
on the stage of an Olympus IMT-2 inverted microscope with 20x and 40x
phase contrast objectives. Monocytes (MonoMac 6 cells) were cultured as
described (25
, 34
) and resuspended at 106
cells/ml in assay buffer (HBSS) containing 10 mM HEPES pH 7.4 and 0.5%
HSA. Shortly before assay, 1 mM Mg2+ and 1 mM
Ca2+ was added. The cell suspensions were kept in
a heating block at 37°C during the assay and perfused into the flow
chamber at a rate of 1.5 dyn/cm2 for 5 min. For
inhibition experiments, monocytes were preincubated with Met-RANTES at
different concentrations (0.01–1 µg/ml) for 30 min on ice. The
number of firmly adherent cells after 5 min was quantitated in multiple
fields (at least five per experiment) by analysis of images recorded
with a long integration JVC 3CCD video camera and a JVC SR L 900 E
video recorder and expressed as cells/mm2. The
type of adhesion analyzed was restricted to primary, i.e., direct
interactions of monocytes with endothelium. As an inverse measure of
firm arrest, the number of cells rolling at reduced velocity on
endothelium was determined within the last 30 s of the 5 min
intervals and assessed as the percentage of all interactions in the
field. The number of cells spreading or transmigrating after 5 min
intervals was determined in high-power fields as described (35)
and
expressed as percentage of cells firmly attached.
Statistical analysis
Values are given as mean ± SE. Statistical
analysis was performed using the Mann-Whitney U-Wilcoxon rank sum test.
A P value <0.05 was considered to show a significant
difference between two groups.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The effect of Met-RANTES on this process was examined by treating transplanted animals with daily i.v. injections of Met-RANTES at 200 µg per animal. The initial injection of Met-RANTES was given within 1 h after vascular anastomosis during transplantation surgery. No additional immune suppressive agent was given during the course of the experiment. Light microscopy and immunohistology showed no obvious effect of Met-RANTES treatment on the endogenous kidney (data not shown).
During organ transplant rejection, the transplanted organ generally
increases in weight due to inflammation. The results summarized in
Table 1
show that Met-RANTES-treated animals had a statistically significant
reduction in transplanted organ weight relative to the untreated
animals. The results also suggested a reduction in T cell and monocyte
infiltration of glomeruli; however, this reduction could not be
considered statistically significant (Mann-Whitney U-Wilcoxon rank sum
test). The most profound effects of Met-RANTES treatment are summarized
in Table 2
. The data demonstrate a significant reduction in the vascular injury
and tubular rejection score of the Met-RANTES-treated animals relative
to that seen in the untreated animals. Although the general trend
regarding interstitial rejection score showed an apparent reduction in
the Met-RANTES-treated animals, this could not be considered
statistically significant (Mann-Whitney U-Wilcoxon rank sum test).
|
|
Histological and immunohistochemical examples showing the
effects of Met-RANTES on the rejection process (day 7) are displayed in
Fig. 1
and Fig.
2. Figure 1a
shows the vascular damage, with mononuclear cells
present within the lumen and the wall of an artery characteristic of
untreated kidneys. In contrast, a corresponding section from a
Met-RANTES-treated animal (Fig. 1b
) shows no vascular
rejection. The immunohistochemical stain shown in Fig. 2a
is
an example of vascular rejection from an untreated rat showing
ED1-positive cells (monocyte/macrophage cells) within the lumen,
arterial wall, and periarterial tissue. A corresponding
Met-RANTES-treated rat section seen in Fig. 2b
shows few
infiltrating ED1-positive cells. The interstitial region of an
untreated animal shown in Fig. 1c
demonstrates infiltration
of a large number of red staining mononuclear cells within the
interstitium and tubules. By contrast, the panel shown in Fig. 1d
from a Met-RANTES-treated animal demonstrates reduced
mononuclear infiltration and less tubular damage, with a well-developed
red brush border of proximal tubules. Figure 2c
, taken from
an untreated rat, shows pronounced interstitial rejection with
significant ED1-positive cells within the interstitium. A corresponding
section from a Met-RANTES-treated animal (Fig. 2d
) shows a
reduction in ED1-positive cells. Figure 1e
is a high
magnification showing the interstitial rejection characteristic of the
untreated animals with infiltrating mononuclear cells within the
tubules and the tubular epithelium (e.g., tubulitis). A corresponding
tissue section from a Met-RANTES-treated rat (Fig. 1f
) shows a reduction of mononuclear cells in the
tubules. This phenomenon is also demonstrated in Fig. 2e
, a
blowup of a section taken from an untreated animal. It shows
ED1-positive cells within peritubular capillaries and the interstitium.
Figure 2f
, a section taken from a Met-RANTES-treated rat,
demonstrates a reduced infiltration of ED1-positive cells within these
tissue spaces.
|
|
Localization of rat RANTES mRNA by in situ hybridization
Tissue sections taken from rejecting Fisher rat kidneys were
used in in situ hybridization studies to demonstrate
cell-specific expression of RANTES mRNA in the rejecting kidney. The
results were similar to those previously described for RANTES
expression during rejection of human renal allografts (17
, 18
, 26
).
Strong expression by infiltrating mononuclear cells and renal tubules
and limited but identifiable expression by some endothelial
cells were seen. Figure 3
a shows endothelial expression of RANTES. Mononuclear cells
attached to and beneath the endothelium are also positive for RANTES
expression. Figure 3b
is a corresponding `sense'-negative
control probe. Epithelial cell expression of RANTES mRNA is shown in
Fig. 3c
. In the lower middle section of this figure is an
erythrocyte that is negative, with RANTES expressing mononuclear cells
seen just to the left and below the erythrocyte.
|
Met-RANTES-treated animals show a reduction in the expression of
proinflammatory cytokine mRNA as determined by RNase protection assays
The increased expression of proinflammatory cytokines such
as IL-1, IL-1ß, IL-2, IL-3, IL-6, TNF-ß, TNF-, and IFN-
are characteristic of renal transplant rejection and suggest an ongoing
inflammatory process (36
37
38
39)
. We examined the effect of Met-RANTES on
the expression of cytokines in transplanted Fisher rat kidneys.
Whole-organ RNA samples were isolated from control kidneys, untreated
transplanted kidneys, and Met-RANTES-treated transplanted kidneys. The
mRNA levels of IL-1, IL-1ß, TNF-ß, TNF-, IL-2, IL-3, IL-4,
IL-5, IL-6, IL-10, and IFN- were determined by RNase protection
assay relative to the internal standards L32 and GAPDH. The results
show that 7 days after transplantation, untreated rejecting kidneys
up-regulated (relative to control kidneys) mRNAs coding for IL-1
(2.4-fold), TNF-ß (3.2-fold) and IFN- (1.7), with the most
pronounced increase seen in IL-1ß (8.4-fold) and TNF- (4.6-fold)
(Fig. 4
). No mRNA expression of IL-2, IL-4, or IL-5 was detected. The
corresponding Met-RANTES-treated animals showed a reduced
average expression of IL-1 (25%), IL-1ß (48%), TNF-ß (34%),
TNF- (24%), and IFN- (24%) relative to the untreated animals.
|
Transplantation of Brown Norway rat kidneys into Lewis rats: effect
of Met-RANTES in conjunction with low-dose cyclosporin A
We sought to determine whether Met-RANTES could complement
low-dose CyA treatment. For this procedure, we chose a renal transplant
model that would yield a more vigorous rejection episode: the
transplantation of Brown Norway kidneys into the Lewis rat. A bilateral
nephrectomy was performed at the time of transplantation. The level of
CyA used (2.5 mg/kg of body weight administered s.c. per day) had
previously been shown not to significantly block renal rejection in
this model (28; H.-J. Gröne, unpublished results). Finally, to
better detect any additive action, a reduced dose of Met-RANTES (50
µg per animal per day) was used, and the experiments were run for 12
days. The results summarized in Table 3
show a statistically significant reduction in vascular and tubular
damage in the Met-RANTES/low-dose CyA-treated animals as compared with
the animals treated with low-dose CyA alone. In addition, a significant
reduction in mononuclear cell infiltration of interstitial spaces was
seen. These histological observations were confirmed by functional
measurements showing a reduction in serum creatinine in the
Met-RANTES/CyA-treated animals relative to the CyA controls (0.98 ± 0.12 vs. 1.42 ± 0.17 mg% (n=3)).
|
Histological examples are shown in Fig. 5
(a–f). Figure 5a
shows the typical
vascular damage seen in the low-dose CyA-treated animals, while a
section from a corresponding Met-RANTES/low-dose CyA-treated animal in
Fig. 5b
shows no vascular damage. A low-dose CyA-treated
animal (Fig. 5c
) displays a glomerulus with pronounced
mesangiolysis and aneurysmatic dilation of capillaries, a typical
glomerular manifestation of rejection in this model. By contrast, a
corresponding Met-RANTES-treated animal (Fig. 5d
)
demonstrates reduction of this phenomenon and reduced interstitial
mononuclear cell infiltration. Figure 5e
is a high
magnification showing the characteristic interstitial rejection seen in
the low-dose CyA-treated animals. Figure 5f
shows a marked
reduction in the number of mononuclear cells within the interstitial
spaces and a pronounced improvement of tubular morphology in a
Met-RANTES-treated animal.
|
Direct RANTES binding and adhesion molecule expression on activated
microvascular endothelium
Since a reduction of monocyte infiltration into vascular
luminal spaces represented a prominent feature of Met-RANTES treatment
in both transplantation models (Fig. 1a
, 1b
, Fig. 2a
, 2b
, and Fig. 5a
, 5b
), we set out to study potential
mechanisms for this effect. It has been suggested that RANTES protein,
released by activated platelets or secreted by locally inflamed tissue,
accumulates on the surface of inflamed endothelium where it may support
monocyte recruitment (11
, 17
, 18
). We examined the capacity of human
microvascular endothelium (DMVEC) to sequester RANTES protein using a
modification of an assay previously used to detect endothelial surface
binding of RANTES in tissue sections (17
, 18
). An mAb specific for
RANTES was incubated with an excess of recombinant RANTES protein and
the resulting complex was added to fixed microvascular endothelium that
had been-treated for 12 h prior to fixation with various
concentrations of IL-1ß. Using an ELISA-like format, the capacity of
DMVEC to bind the antigen–mAb complex as opposed to mAb alone was
determined. Whereas the microvascular endothelium could bind RANTES
protein without prestimulation, the binding was significantly increased
after prestimulation with IL-1ß (Fig. 6
).
|
To further characterize microvascular endothelial activation, the
surface expression of molecules involved in monocyte adhesion (i.e.,
E-selectin and the Ig superfamily members ICAM-1 and VCAM-1) was
studied (30)
. Analysis revealed that resting DMVEC expressed
constitutive surface levels of ICAM-1, but little VCAM-1 or E-selectin
was detected (Table 4
). Activation of DMVEC with IL-1ß for 12 h resulted in an
up-regulation of ICAM-1 expression and a marked induction of VCAM-1 and
E-selectin surface expression (Table 4)
.
|
Met-RANTES blocks the firm adhesion of monocytes to inflamed
microvascular endothelium but does not affect subsequent events in
diapedesis
To gain insight into potential mechanisms of action of Met-RANTES,
we studied whether a blockade of RANTES receptors could inhibit the
firm arrest and diapedesis of monocytes on microvascular endothelium.
To this end, we used human monocytic MonoMac 6 cells that show the
adhesive characteristics and integrin repertoire of mature monocytes
and express several chemokine receptors, including CCR1 (40)
(data not
shown). DMVEC were grown to confluence on petri dishes and were either
left unstimulated or activated with IL-1ß (5 ng/ml) for 12 h.
The microvascular endothelium was then tested in a parallel wall flow
chamber where the MonoMac 6 cells were perfused through the chamber at
a shear rate of 1.5 dyn/cm2 for 5 min. Under such
physiological flow conditions, MonoMac 6 cells undergo short periods of
rolling, and the attachment of a proportion of cells can be readily
converted into shear-resistant arrest. After 5 min of accumulation, the
number of MonoMac 6 cells that had undergone firm adhesion to the
endothelium was determined (Fig. 7
a).
|
Few monocytic cells firmly adhered to unstimulated microvascular
endothelium and the preexposure of the endothelial cells to RANTES
protein showed no significant effect. Prestimulation of the
microvascular endothelium with IL-1ß resulted in an increase in
shear-resistant adhesion of monocytes. Inhibition with mAb (data not
shown) confirmed previous findings that this process is mediated by
monocyte 4 and ß2 integrins that interact with ICAM-1 and VCAM-1
expressed on activated endothelium, respectively (32
, 35
). Consistent
with the immobilization of RANTES in direct binding assays, preexposure
of IL-1ß-activated microvascular endothelium to RANTES protein
markedly enhanced the firm arrest and accumulation of monocytes within
5 min (Fig. 7a
). Notably, preincubation of monocytes with
Met-RANTES at various concentrations (0.01–1 µg/ml) completely
blocked RANTES-mediated shear resistant adhesion of monocytes on
IL-1ß-activated DMVEC (Fig. 7a
and data not shown). In
parallel, the fraction of monocytes rolling on the activated
microvascular endothelium, which can be used as an inverse measure of
firm arrest, was reduced after preexposure to RANTES but was restored
by Met-RANTES (data not shown), indicating that the number of initial
interactions with the activated endothelium was unaffected. After firm
arrest, a fraction of monocytes underwent shape change or spreading,
and some ultimately migrated in between or under the endothelial cells.
However, RANTES or Met-RANTES did not alter spreading or transmigration
(Fig. 7b
), inferring the involvement of other signals. Thus,
these results indicate that Met-RANTES may reduce monocyte recruitment
during renal transplant rejection by blocking monocyte arrest to
inflamed microvascular endothelium. However, Met-RANTES does not appear
to effect the subsequent events involved in monocyte diapedesis.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
42
43)
.
Chemokines are thought to play pivotal roles in leukocyte recruitment
and, thus, the mononuclear cell infiltrate characteristic of allograft
rejection (9
, 17
, 18
, 44
45
46
). The chemokine RANTES, a chemoattractant
for T cells and monocytes, is highly expressed by human and rat kidneys
undergoing transplant rejection (17
, 18
, 44
, 45
). In a study of human
renal allograft rejection, RANTES protein was identified in 17/20
biopsies from kidneys undergoing acute cellular rejection (17)
. RANTES
mRNA was found expressed by infiltrating mononuclear cells and renal
tubular epithelial cells, but expression of RANTES mRNA by endothelial
cells was almost undetectable (17
, 18
). In contrast, RANTES protein
prominently localized to the endothelium of peritubular capillaries.
This suggested that much of the RANTES protein found on the endothelium
was not produced by the inflamed endothelium, but rather had been
deposited there presumably through platelet degranulation (17
, 18
).
Thus situated, the chemokine would be ideally positioned to enhance the
recruitment of monocytes and T cells into the graft.
The kinetics of RANTES mRNA expression during renal allograft rejection
in the Fisher into Lewis transplant model has been described (44
, 45
).
We found a similar pattern of expression where up-regulation of RANTES
mRNA in whole kidney was seen by day 4 after transplantation (data not
shown). The tissues expressing RANTES mRNA on day 7 included the
epithelium, infiltrating mononuclear cells, and some endothelial cells
(Fig. 3)
.
A blockade of RANTES receptors using the RANTES analog Met-RANTES was used to study the functional role of RANTES and its receptors in the early inflammatory events associated with rejection. In the Fisher kidney into Lewis rat transplantation model, treatment with Met-RANTES led to a significant reduction in damage to the preglomerular vessels and tubules of the transplanted kidney and reduced mononuclear cell infiltration of the vascular lumen (endotheliitis). Morphologically, the untreated kidneys showed tubules with collapsed lumina and flattened epithelia with microvesicular cytoplasm whereas the Met-RANTES-treated kidneys showed excellent preservation of normal epithelial differentiation.
The expression of specific cytokines can be an indication of ongoing
inflammation (36
37
38
, 46
). Kidneys taken from untreated animals (Fisher
into Lewis) showed up-regulation of mRNAs coding for IL-1, IL-1ß,
TNF-, TNF-ß, and IFN, with IL-1ß and TNF- showing the most
pronounced increase (Fig. 4)
. This is interesting given that IL-1ß
and TNF- are capable of inducing the expression of a functional
program related to thrombosis and inflammation (47
, 48
). Through an
autocrine process, damaged endothelium produces additional quantities
of these cytokines, leading to amplification of inflammation (47
48
49)
.
Animals treated with Met-RANTES showed a reduced expression of each of
the proinflammatory cytokines studied. This effect may be mediated
through the suppression of monocytic infiltration and thus a reduction
in monocyte-derived proinflammatory factors or, alternatively,
Met-RANTES could act directly on the inflamed endothelium. We have
recently demonstrated that functional chemokine receptor (CCR2) can be
induced on endothelial cells (50)
. Thus, Met-RANTES may also directly
block chemokine-mediated signals to the endothelium.
Though Met-RANTES did not completely suppress transplant rejection, the arrest of vascular and tubular injury seen in the Met-RANTES-treated animals demonstrates that it can have a significant effect on the damage associated with vascular inflammation. The interaction between RANTES, leukocytes, and inflamed endothelium was analyzed in vitro. An increase in the direct binding of RANTES to microvascular endothelial cells was seen after pretreatment of the endothelium with IL-1ß, suggesting an induction of surface `chemokine presentation' molecules on the activated/inflamed endothelial cells. This `priming' effect makes intuitive sense as the direct binding of chemokines to `normal' microvascular endothelium could promote immune effector cell binding and activation, leading to damage of the vascular tissue.
We then examined the expression of various adhesion molecules on
microvascular endothelium known to be involved in the rolling and firm
adhesion of monocytes. ICAM-1 was highly expressed, and some VCAM-1 and
E selectin was also detected on resting DMVEC. Surface expression of
each of these proteins increased significantly after treatment with
IL-1ß (Table 4)
. The increased expression of these molecules has been
described in rejecting human renal allografts (51
, 52
). Thus, inflamed
microvascular endothelium appears to increase its capacity to bind
RANTES protein and up-regulate the expression of molecules required for
the efficient rolling and firm adhesion of leukocytes.
The RANTES-induced activation of monocytes and the blockade of RANTES
receptors with Met-RANTES in the context of microvascular endothelium
were then studied. In these experiments, the accumulation of monocytic
cells on microvascular endothelium under physiological flow conditions
was assayed. In accordance with our findings in the direct RANTES
binding assays, we found that preexposure of microvascular endothelium
to RANTES could enhance the firm arrest of monocytes only after
preactivation of the endothelium with IL-1ß. The preincubation of
monocytes with Met-RANTES completely blocked the RANTES-mediated,
shear-resistant accumulation on the `inflamed' microvascular
endothelium. In contrast, the subsequent spreading or migration of the
monocytes was not affected by RANTES or Met-RANTES, implicating other
agents/chemokines in these later events in diapedesis. Our results
expand on recent reports that the CXC chemokines Mig and IP-10, which
are induced and immobilized on stimulated endothelium, can induce the
firm adhesion of activated effector T cells under flow conditions via
CXCR3. In this study, a blockade of CXCR3 with an antireceptor
monoclonal antibody did not alter subsequent transmigration (33)
.
Moreover, soluble eotaxin has been shown to augment eosinophil adhesion
in static assays. A blockade of its receptor, CCR3, with a specific
monoclonal antibody revealed a contribution to firm eosinophil arrest
on activated endothelium in shear flow, but a substantial inhibition
was seen only in combination with an 4 monoclonal antibody (53)
.
It has been proposed that chemokines bound to endothelial proteoglycans
are protected from being washed away in vascular flow and may thereby
more efficiently recruit leukocytes (18
, 54
, 55
). Met-RANTES may act in
part by preventing the shear-resistant arrest and subsequent
recruitment of mononuclear cells induced by RANTES and potentially
other RANTES receptor cross-reactive chemokines immobilized on the
inflamed microvascular endothelium. This may be a mode of action by
which Met-RANTES reduces the vascular and tubular damage during acute
renal transplant rejection.
CyA is a potent immunosuppressive agent often used in transplantation.
Its immunosuppressive action is mediated, at least in part, through the
modulation of cytokine gene expression (56)
. The ability of Met-RANTES
to enhance the effects of low-dose CyA treatment was studied in a
second rat model of renal transplant rejection. Kidneys from Brown
Norway rats were transplanted into Lewis rats. This protocol resulted
in a more aggressive clinical course of rejection than that seen in the
Fisher kidney into Lewis rat model. These experiments show that
treatment with Met-RANTES and low-dose CyA results in a dramatic
reduction in inflammatory events compared with treatment with CyA
alone. Met-RANTES significantly reduced damage to vessels and tubules
and caused a pronounced reduction in interstitial rejection (Table 3
,
Fig. 5
). These results have clinical relevance given the dose-dependent
nephrotoxicity associated with cyclosporin therapy (56)
. A reduction in
the amount of cyclosporin A required to achieve sufficient
immunosuppression would be of considerable benefit to the renal
transplant.
Even though the clinical outcome of the renal transplant studies and
the results of the in vitro experiments are clear,
additional, so far undefined modes of action of Met-RANTES may also
play a role in vivo. Although Met-RANTES is less potent at
antagonizing CCR3-mediated effects, it is still more effective at
preventing eosinophil recruitment in vivo in an allergic
skin model than neutralizing the CCR3-specific ligand eotaxin (22
, 57
).
The efficiency of Met-RANTES in limiting the inflammation seen in our
rat allograft models may be due to its ability to bind to multiple
receptors and thereby block action of additional chemokines. Met-RANTES
is a potent antagonist of human CCR1 (ligands: MIP-1, RANTES, MCP-3,
HCC-1, MPIF-1, MIP-5, LKN-1, MIP-1
) and retains its ability to bind
to CCR5 (ligands: MIP-1, RANTES, MIP-1ß, MCP-2, HCC-1) (7
, 8
, 19
,
21
). Both of these receptors are expressed by monocytes and may play a
role in the early recruitment seen in allograft rejection.
The clinical advantage of antiinflammatory agents such as Met-RANTES
may lie in their acute mode of action. As demonstrated here, Met-RANTES
appears to slow the mononuclear cell recruitment and vascular damage
seen early in rejection. Limiting damage at these early stages may
lower the inclination toward development of the subsequent inflammatory
events involved in the process of transplant rejection (41
42
43
, 58
,
59
). These results strongly suggest that therapies directed toward
blocking of chemokines and chemokine receptors may have a positive
effect on solid allograft survival.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
, tumor necrosis factor . 
Received for publication December 4, 1998.
Revision received February 22, 1999.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
IIbß3 and stimulated by platelet-activating factor. J. Clin. Invest. 100,2085-2093[Medline]
