![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
|
|
|
|
|
||
Vol. 290, Issue 3, 1458-1466, September 1999
Laboratoire de Recherche en Biologie Vasculaire et Cellulaire, EA 1557, Hôpital Lariboisière, Paris Cedex, France (C.R., O.C., M.P.W., J.L.W.); Institut National de Transfusion Sanguine, Paris Cedex, France (J.L.W.); Institut National de la Santé et de la Recherche Médicale Unité 26, Hôpital Fernand Widal, Paris Cedex, France (C.R., J.M.S.); and Berlex Biosciences, Richmond, California (M.N., J.M.)
 |
Abstract |
|---|
 Top
 Abstract
 Introduction
Materials and Methods
Results
Discussion
References
|
|---|
The accelerated formation of advanced glycation end products (AGEs) is implicated in diabetic microvascular and macrovascular complications. The binding of AGEs to their cellular surface receptor (RAGE) induces vascular dysfunction and in particular an increase in vascular permeability. We previously demonstrated that rat recombinant RAGE (rR-RAGE) produced in insect cells corrected the hyperpermeability due to RAGE-AGE interaction and that pharmacokinetic properties of rR-RAGE after i.v. administration in rats were compatible with a potential therapeutic use. In the present study, we showed that recombinant human RAGE (rH-RAGE) had a similar efficacy in inhibiting AGE-induced endothelial alteration and in reducing the hyperpermeability observed in streptozotocin-induced diabetic rats. 125I-rH-RAGE elimination half-life after i.v. administration was similar in diabetic and normal rats (53.7 ± 7.6 and 45.3 ± 4.0 h, respectively). The presence of AGEs is responsible for a higher distribution volume in diabetic rats compared with normal rats (15.3 ± 2.7 and 7.7 ± 0.7 l/kg, respectively). Immunoreactive 125I-rH-RAGE decreased more rapidly than did immunoreactive 125I-rR-RAGE. The differences between 125I-rH-RAGE and 125I-rR-RAGE pharmacokinetics in rat may be related to differences in potential O-glycosylation and protease cleavage sites between the two RAGE molecules.
| |
Introduction |
|---|
|
Top
Abstract
Introduction
Materials and Methods
Results
Discussion
References
|
|---|
Advanced
glycation endproducts (AGEs) are a heterogeneous class of molecules
found in plasma, cells, and tissues. They accumulate in the vessel wall
and the kidney during aging and at an accelerated rate in diabetes
(Brownlee, 1995
). AGEs are formed by nonenzymatic glycation of primary
amino groups on proteins or lipids. The best-characterized AGE receptor
(RAGE) (Schmidt et al., 1999) has been purified and cloned in insect
cells. RAGE is a member of the Ig superfamily. It is composed of an
extracellular region with one V-type domain and two C-type domains, a
single transmembrane domain, and a highly charged intracellular domain
(Neeper et al., 1992; Schmidt et al., 1994). RAGE is present in various
species, and the molecules have a high degree of homology (Neeper et
al., 1992; Schmidt et al., 1992). It is expressed on several cell
types: endothelial cells (ECs), smooth muscle cells,
monocytes/macrophages, cardiac myocytes, neural tissue, and hepatocytes
(Brett et al., 1993; Schmidt et al., 1993).
RAGE-AGEs interactions are thought to contribute to the development of
diabetic complications, including vascular dysfunction (Wautier et al.,
1996; Bierhaus et al., 1997). Our previous studies demonstrated that
antibodies directed against RAGE and soluble RAGE purified from bovine
lung, a truncated form of the receptor, inhibited the binding of
diabetic erythrocytes bearing AGEs to ECs (Wautier et al., 1994).
Rat-soluble RAGE or recombinant rat RAGE (rR-RAGE) produced in insect
cells reduced the early vascular hyperpermeability observed in diabetic
rats (Wautier et al., 1996; Renard et al., 1997). After i.v.
administration, RAGE elimination half-life indicated that daily
administration was possible (Renard et al., 1997). In fact, in diabetic
apolipoprotein E-null mice, i.p. administration of murine-soluble RAGE
(40 µg/day for 6 weeks) suppressed accelerated atherosclerosis (Park
et al., 1998).
In the present study, we show that recombinant human RAGE (rH-RAGE) is as efficient as rR-RAGE in preventing the increased 125I-albumin transfer through a bovine aortic EC monolayer and in correcting the hyperpermeability observed in diabetic rats. We also compared the pharmacokinetics of 125I-rR-RAGE and 125I-rH-RAGE.
| |
Materials and Methods |
|---|
|
Top
Abstract
Introduction
Materials and Methods
Results
Discussion
References
|
|---|
Cloning, Expression, and Purification of rH-RAGE and Vascular Cell Adhesion Molecule 1(VCAM-1)
Human full-length RAGE cDNA was cloned by screening a lung cDNA
library (Neeper et al., 1992). Baculovirus expression of rH-RAGE in
Spodoptera frugiperda (Sf9) cells was obtained as described for rR-RAGE (Renard et al., 1997). The resulting protein obtained corresponded to that expected for the extracellular domain (Schmidt et
al., 1992) and migrated as a homogenous sample of 
35 kDa on SDS-polyacrylamide gel electrophoresis (PAGE).
A DNA fragment coding for the first three Ig-like domains of rat VCAM-1
was obtained from a lung cDNA library (Clontech, Palo Alto, CA) by
using a polymerase chain reaction technique. The primers used were
5'-GAAATGCCTGTGAAGATGGTCG-3' and 5'-CTCATTGAACAACTAATTCCACTTC-3', based
on the known sequence of mouse VCAM-1. A polymerase chain reaction
fragment was subcloned into a pCRII vector (Invitrogen, San Diego, CA)
for DNA sequencing by the dideoxynucleotide chain-termination method.
The deduced amino acid sequence of rat-soluble VCAM-1 has 85% identity
with that of murine-soluble VCAM-1. An EcoRI fragment
of the resultant plasmid was then cloned into a pBacPAK8 vector
(Clontech) for the baculovirus expression. The expression and the
purification of recombinant-soluble VCAM-1 were essentially the same as
for rR-RAGE (Renard et al., 1997) except a Q-Sepharose fast-flow column
(Amersham Pharmacia Biotech, Uppsala, Sweden) was used with 20 mM Tris,
pH 8.0. The purified recombinant rat soluble VCAM-1 has a molecular
mass of around 35 kDa and cross-reacts with antibody against
human VCAM-1 (data not shown). The N-terminal sequence of the purified
protein is
Phe-Lys-Ile-Glu-Ile-Ser-Pro-Glu-Tyr-Lys-Thr-Leu-Ala-Gln-Ile, and
it matches with the sequence predicted from that of cDNA.
Database Search
Amino acid sequences of human, rat, and bovine RAGE are
available on the Swiss-Prot database. They can be obtained on the World
Wide Web (www.expasy.ch). Alignments of the amino acid sequences of
human, rat, and bovine RAGE were performed with the SIM program and
comparison matrix BLOSUM 62 (Swiss-Prot database).
O-glycosylation sites were predicted by using the NetOGlyc
server (www.cbs.dtu.dk/services/NetOGlyc) (Hansen et al., 1997, 1998).
Radiolabeling of Proteins
rH-RAGE, albumin, and murine Fab Ig fragment were labeled with
Na125I by the Iodo-Gen method (Fraker and Speck,
1978; Renard et al., 1997). Precipitation of the iodinated protein by
trichloroacetic acid (TCA; 10%) at +4°C was used to determine
protein concentrations. Specific activities were in the range of 1 µCi/µg for rH-RAGE, 2.5 µCi/µg for albumin, and 1.15 µCi/µg
for murine Fab. Analysis of 125I-rH-RAGE,
125I-albumin, and 125I-Fab
preparations by SDS-PAGE and autoradiography indicated no major
contamination by label or degradation products for any of the proteins.
In Vitro Permeability Assay
In accordance with provisions of the Declaration of Helsinki and
with the rules of our institution, human red blood cells (RBCs) used in
our experiments came from normal volunteers and diabetic patients. In
vitro permeability assays were performed as previously described
(Wautier et al., 1996). ECs were incubated with medium alone or with
normal or diabetic RBCs (2.5 × 109
cells/ml) for 24 h. To determine the effect of rH-RAGE on EC permeability, rH-RAGE (60 µg/ml) was added to diabetic RBCs. VCAM-1 (60 µg/ml) was tested in the same experimental conditions as rH-RAGE. To study the permeability, 125I-albumin was added
to the upper chamber and the permeability coefficient (P) was
determined by the following equation: P = J × 1/A × 1/(CT 
CB), where J is
the flux of molecules across the filter; A is the surface area; and C
is the concentration of tracer in the top (CT)
and bottom (CB) chambers (Albelda et al., 1988;
Wautier et al., 1996).
In Vivo Permeability Studies
In vivo permeability studies were carried out as previously
described (Wautier et al., 1996) in normal and streptozotocin-induced diabetic male Wistar rats. Streptozotocin (45 mg/kg i.v.; Sigma-Aldrich Chimie, Saint-Quentin Fallavier, France) was injected 7 to 9 weeks before experiments. The blood glucose levels of normal and diabetic rats were 90 to 130 mg/dl and above 250 mg/dl, respectively. Normal rats received an i.v. bolus of normal RBCs, diabetic RBCs, diabetic RBCs plus rH-RAGE (5.15 mg/kg), or diabetic RBCs plus VCAM-1 (5.15 mg/kg), and diabetic rats received the same dose of rH-RAGE or VCAM-1
also as an i.v. bolus. In vivo permeability in normal and diabetic rats
was determined by using the tissue-blood isotope ratio (TBIR) method
(Williamson et al., 1987; Wautier et al., 1996).
Pharmacokinetics of 125I-rH-RAGE and Murine 125I-Fab
Plasma kinetics of 125I-rH-RAGE (250 µg/kg, volume 
4 ml/kg) were studied in normal and
streptozotocin-induced diabetic male Wistar rats (200-250 g; CERJ,
Laval, France) after i.v. and s.c. administration. Plasma
concentrations of 125I-rH-RAGE were determined
after precipitation with TCA and also after immunoprecipitation with a
specific monoclonal antibody (antibody 9D9). Animals had free access to
food and water before the experiments. They were anesthetized with
ether before i.v. injection, whereas s.c. injections were given to
conscious animals. After the injection, rats were placed in metabolic
cages for urine and feces collection.
125I-rH-RAGE was administered by i.v. bolus via
the femoral vein. Blood samples (50 µl) were collected in
heparin-containing tubes from the tail vein at 2, 5, 10, 30, and 45 min
and 1, 1.5, 2, 3, 6, 8, 24, 30, 48, 54, 72, and 96 h and
centrifuged for 10 min at 3000g, and plasma was isolated.
After s.c. administration, blood samples were collected at the same
times except that the first sample was taken at 10 min and additional
blood samples were collected at 144, 168, 192, and 216 h. The
blood hematocrit measured at 6, 24, 54, and 96 h after
125I-rH-RAGE injection did not differ from
physiological values (46%; Davies and Morris, 1993).
After i.v. administration of 125I-rH-RAGE, rats
were decapitated at 96 h and radioactivity in organs (intestine,
skin, kidney, vena cava, aorta, liver, spleen, lung, heart, and
thyroid) was determined. Furthermore, tissue distribution of
125I-rH-RAGE was studied in normal and diabetic
rats at the time corresponding to 87.5% of the
125I-rH-RAGE distribution (i.e., 2 and 7.5 h
for normal and diabetic rats, respectively). The amount of
125I-rH-RAGE determined in organs was corrected
for the presence of radioactivity in the residual blood remaining in
the tissue (Ebling et al., 1994).
The pharmacokinetics of a molecule belonging to the Ig superfamily, a monoclonal murine 125I-Fab (250 µg/kg), were also studied in normal rats (n = 4) to verify whether the pharmacokinetic profile determined after s.c. injection was influenced by the rH-RAGE characteristics or by the route of injection. The experimental protocol was the same as that for 125I-rH-RAGE.
Identification of 125I-rH-RAGE
Immunoprecipitation of 125I-rH-RAGE.
A
monoclonal anti-rH-RAGE antibody (antibody 9D9) (Brett et al., 1993)
was used in an immunoprecipitation assay to determine the
immunoreactive fraction of 125I-rH-RAGE. Samples
consisting of 125I-rH-RAGE before injection
diluted in RAGE-free plasma and rat plasma samples taken after
injection of 125I-rH-RAGE were analyzed. Samples
(50 µl) were incubated with the anti-RAGE antibodies (100 µl), PBS
(300 µl), and RAGE-free rat plasma (50 µl) for 1 h at 37°C
and overnight at +4°C. A solution of 14% polyethylene glycol 8000 in
borate buffer (500 µl; Sigma-Aldrich Chimie) was added, and the mix
was incubated overnight. The precipitate obtained by centrifugation (15 min at +4°C and 1800g) was counted in a gamma counter
(Packard Instruments).
SDS-PAGE Analysis.
After i.v. administration of
125I-rH-RAGE, plasma samples were analyzed by
SDS-PAGE. Equal amounts of TCA-precipitable radioactive material
(approximately 7000 cpm) from plasma samples were loaded onto a 15%
acrylamide gel and analyzed under nonreducing conditions. Radiolabeled
proteins and metabolites were autoradiographed with X-ray film
(Amersham Pharmacia Biotech) and intensifying screens for 8 weeks. The
migration zones of labeled proteins or metabolites were compared with
prestained standards (myosin, 202 kDa; 
-galactosidase, 137 kDa; BSA,
42.3 kDa; soybean trypsin inhibitor, 31.6 kDa; lysozyme, 18 kDa;
aprotinin, 7.6 kDa) (Bio-Rad, Paris, France).
20°C until analysis by SDS-PAGE and autoradiography for 1 month.
Pharmacokinetic Analysis
Plasma concentration-time data of
125I-rH-RAGE administered i.v. were analyzed by
using the Siphar Software (SIMED, Créteil, France) with a
noncompartmental method. The terminal disposition rate constant
(
z) was determined by linear regression
analysis and the corresponding half-life
(T1/2z) was
calculated as 0.693/z. The area under the
plasma 125I-rH-RAGE concentration-time curve from
zero to infinity (AUC) was determined by linear trapezoidal estimation
from 0 to the last measured time, with extrapolation to infinity by
adding the value of the last measured plasma concentration divided by
the terminal rate constant (Gibaldi and Perrier, 1982). Total body clearance (CL), distribution volume of the terminal elimination half-life (Vz), extrapolated
distribution volume (Ve), and mean residence time (MRT) were calculated by using standard equations (Gibaldi and Perrier, 1982).
Subcutaneous plasma pharmacokinetics of
125I-rH-RAGE and 125I-Fab
were also analyzed by a noncompartmental approach to determine T1/2z and
AUC. The maximal concentration (Cmax)
and the corresponding experimental time
(Tmax) after s.c. administration of
125I-rH-RAGE are the experimental observed values.
Statistical Analysis
Results are presented as mean ± S.E.. One-way ANOVA followed by the parametric Dunnett's test in the event of significant differences was used to compare permeability of ECs in the presence of normal or diabetic RBCs and to analyze the results of in vivo permeability studies. Mean values of pharmacokinetic parameters were compared with the nonparametric Mann-Whitney two-sample test.
| |
Results |
|---|
|
Top
Abstract
Introduction
Materials and Methods
Results
Discussion
References
|
|---|
In Vitro Permeability
Transfer of 125I-albumin through the EC
barrier was similar in culture medium alone and in the presence of
normal RBCs. Addition of diabetic RBCs to the medium significantly
increased (2.0-fold, P .001) the permeability to
125I-albumin. Preincubation of RBCs with rH-RAGE
prevented the effect of diabetic RBCs on permeability of the EC
monolayer. This effect was dependent on the rH-RAGE concentration, as
the permeability was decreased 1.6- and 2.0-fold at rH-RAGE
concentrations of 30 and 60 µg/ml, respectively (Fig.
1). VCAM-1 preincubation with diabetic
RBCs did not significantly alter the increased passage of
125I-albumin through the EC monolayer.
|
In Vivo Permeability
In diabetic rats the vascular permeability to
125I-albumin was increased in diabetic compared
with normal rats, especially in skin (×2.75), intestine (×2.6),
kidney (×2.36), vena cava (×2.2), and heart (×2.1). After a bolus
injection of rH-RAGE (5.15 mg/kg) in diabetic rats, the
hyperpermeability was corrected. This effect was observed 1 h and
40 min after rH-RAGE injection and was more pronounced in skin and
intestine (Fig. 2A). Infusion of diabetic RBCs in normal rats increased the permeability to
125I-albumin compared with normal RBCs infused in
normal rats (Fig. 2B). The permeability increase was similar to that
observed in diabetic and normal rats infused with rH-RAGE. After
diabetic RBC infusion, the vascular permeability to
125I-albumin was 2.5, 2.2, 2.2, 2.1, and 2.0-fold
higher in heart, skin, kidney, vena cava, and intestine, respectively.
Coadministration of rH-RAGE and diabetic RBCs inhibited the
hyperpermeability to 125I-albumin in each organ
and particularly in skin (Fig. 2B). VCAM-1 did not reduce the
hyperpermeability in organs, indicating that the effect of rH-RAGE was
specific.
|
Plasma Pharmacokinetics after i.v. Administration
Pharmacokinetics of 125I-rH-RAGE Proteins Precipitated
by TCA.
After i.v. bolus (Fig. 3),
125I-rH-RAGE plasma concentrations decreased more
rapidly in diabetic rats than in normal rats and resulted in an AUC
1.4-fold higher in normal than in diabetic rats and in a clearance
1.7-fold higher in diabetic than in normal rats. Distribution
clearances (CLD1 and CLD2,
Table 1) determined by using a
three-compartment model (Veng-Pedersen and Gillespie, 1986) for plasma
concentration-time curve analysis were not significantly different in
normal and diabetic rats, indicating that differences observed in
clearance of normal and diabetic rats were not due to a different
distribution process. The elimination half-life was not significantly
different in diabetic and normal rats. The distribution volume was
2-fold higher in diabetic than in normal rats, which is particularly
high for a 35-kDa protein, because the total body water of rats is 0.7 l/kg (Davies and Morris, 1993). Because
Vz could be influenced by differences
in clearance (Jusko and Gibaldi, 1972), distribution volume at steady
state (Vss) and extrapolated volume
(Ve) were also determined by using a
three-compartment model for plasma concentration-time curve analysis
(Table 1). Results confirmed the high distribution volume of
125I-rH-RAGE in rat. Volume of the central
compartment (Vc) of
125I-rH-RAGE being similar in normal and diabetic
rats (Table 1), the difference in the distribution volume was not due
to a different 125I-rH-RAGE central compartment
in normal and diabetic rats. Vc was
approximately 2-fold higher than the blood volume (56-86 ml/kg in
normal and diabetic rats) and indicates that
125I-rH-RAGE was easily distributed out of the
vascular compartment. Because AGEs are extensively formed in diabetic
animals they could create an additional distribution compartment,
induce a distribution of the protein out of the plasma compartment, and
consequently explain the larger distribution volume of the protein in
diabetic rats.
|
). After i.v.
injection of 125I-rR-RAGE, we observed a
different elimination half-life in diabetic and normal rats but not
with 125I-rH-RAGE. In diabetic rats,
125I-rR-RAGE and
125I-rH-RAGE had a similar elimination half-life
(53.74 ± 7.58 and 57.17 ± 11.62 h, respectively,
P = .740), but in normal rats
125I-rH-RAGE had a higher elimination half-life
than did 125I-rR-RAGE (45.27 ± 4.02 and
26.02 ± 2.36 h, respectively, P = .044).
Distribution volume and total clearance were higher for 125I-rH-RAGE than for
125I-rR-RAGE as shown by a 2.2- and 2.4-fold
increase of Vz in diabetic and normal
rats and by a 1.9- and 1.3-fold increase of CL in diabetic and normal
rats, respectively. AGE concentrations, which are higher in diabetic
than in normal rats, might account for the degradation of free
125I-rH-RAGE.
Immunoprecipitation of 125I-rH-RAGE.
To further
demonstrate that the plasma radioactivity corresponded to intact
rH-RAGE, we performed immunoprecipitation studies. Before
125I-rH-RAGE administration in rats, more than
95% of the radioactivity was recovered after TCA precipitation, and
66.2 ± 1.5% was recovered after immunoprecipitation with a
monoclonal antibody directed against rH-RAGE. One hour after the i.v.
administration, immunoprecipitated 125I-rH-RAGE
represented only 46.7 ± 4.7 and 73.0 ± 1.7% of the
TCA-precipitable radioactivity in normal and diabetic rats,
respectively. Two hours after the administration, immunoprecipitated
percentages of the TCA-precipitable radioactivity decreased to
20.7 ± 2.9 and 46.7 ± 0.9%, for normal and diabetic rats,
respectively. These results suggest that the monoclonal antibody used
in our study recognized 125I-rH-RAGE before its
administration in rats, and that after its administration
125I-rH-RAGE was extensively metabolized,
especially in normal rats.
|
SDS-PAGE.
This extensive metabolism of
125I-rH-RAGE was also observed by SDS-PAGE (Fig.
4). One hour after
125I-rH-RAGE administration in normal and
diabetic rats, autoradiography of plasma samples showed one band of
approximately 35 kDa and another band of lower molecular mass in normal
and diabetic rats. The 35-kDa band could correspond to native
125I-rH-RAGE and the other to
125I-rH-RAGE metabolites. In addition, in plasma
samples of diabetic rats, we also observed a third band corresponding
to products of higher molecular mass that could correspond to complexes
formed with 125I-rH-RAGE. Identification of
125I-rH-RAGE by SDS-PAGE indicated a rapid
metabolism of 125I-rH-RAGE, as did the
immunoprecipitation studies.
|
Plasma Pharmacokinetics after s.c. Administration
Pharmacokinetics of 125I-rH-RAGE Proteins Precipitated
by TCA.
After s.c. administration, we observed similar elimination
half-lives in diabetic and normal rats
(T1/2z = 64.9 ± 8.5 and 57.9 ± 3.9, respectively, P = .499). Two Cmax values characterized the
profile of the 125I-rH-RAGE plasma
concentration-time curve. The first around 2 h after the injection
and the second 40 to 50 h after the injection (Fig.
5).
|
Immunoprecipitation of 125I-rH-RAGE.
To better
understand the two absorption peaks that characterized the s.c.
pharmacokinetics of 125I-rH-RAGE, we
immunoprecipitated plasma samples at the two
Cmax times (1.8 and 50 h) after
s.c. administration of 125I-rH-RAGE in normal
rats. Only 16.1 ± 0.6 and 2.3 ± 1.3% of the TCA-precipitable radioactivity were immunoprecipitated at 1.8 and
50 h, respectively. In normal rats,
125I-Fab, which was used as a control, was
characterized by only one absorption peak, indicating that the mode of
administration was not responsible for the two observed
Cmax values with
125I-rH-RAGE. These results could indicate that
125I-rH-RAGE was extensively metabolized after
s.c. administration and that the second
Cmax could result from the
distribution of radiolabeled metabolites not recognized by the
monoclonal antibody used. Although this pharmacokinetic profile is
surprising, metabolism of proteins after s.c. injection is well known
and has been described for other proteins such as platelet-derived
growth factor (PDGF; Abdiu et al., 1998), insulin (Okumura et al.,
1985), and parathyroid hormone and calcitonin (Parsons et al., 1979).
Biodistribution of 125I-rH-RAGE
We studied the distribution of 125I-rH-RAGE
at the end of the distribution phase after i.v. administration in
diabetic (n = 3) and normal rats (n = 3) (Fig. 6A). The distribution profile of 125I-rH-RAGE in all organs tested was similar in
both types of rat. In normal rats, its distribution was 24.3 ± 5.9, 22.0 ± 3.6, 15.3 ± 1.3, and 12.0 ± 3.4% in
skin, aorta, kidney, and vena cava, respectively, whereas these
percentages were 19.2 ± 3.4, 15.8 ± 2.0, 11.72 ± 3.3, 11.5 ± 3.0, and 10.7 ± 0.8% in aorta, skin, lung, vena
cava, and kidney of diabetic rats, respectively. In both types of rat,
at the end of the experiment (96 h, Fig. 6B) most of the radioactivity
was found in the kidney and aorta at the end of the distribution phase,
but it was also elevated in the liver, suggesting hepatic trapping of
the protein. It has been reported that the tissue distribution of other
proteins such as humanized anti-IgE antibody (Fox et al., 1996) and
recombinant human interleukin-11 (Takagi et al., 1995) is higher in the
liver and kidney. These findings may reflect residual organ blood and possible metabolism and clearance of the studied protein and its metabolites by these organs (Fox et al., 1996).
|
Database Search Results
Alignment of the amino acid sequences of human, rat, and bovine
RAGE indicated 80.8% identity between human and bovine RAGE, 78%
identity between human and rat RAGE, and 73.3% identity between rat
and bovine RAGE (Fig. 7). Cysteine
residues and potential N-glycosylation sites are conserved
in the three molecules. They have several potential
O-glycosylation sites (T304 and
S307 in human RAGE, T194,
S260, S305, and
S311 in rat RAGE, and T101
and S317 in bovine RAGE), and only one is
conserved in all three molecules (S307 of human
RAGE, S305 of rat RAGE, and
S317 of bovine RAGE). Several potential sites of
proteolysis by trypsin-like enzymes are present in RAGE from different
species, but the only major difference is that human RAGE has an
arginine (R221), which may be responsible for an
increased susceptibility of human RAGE to trypsin-like enzymes, and may
affect the half-life of rH-RAGE in rat plasma.
|
| |
Discussion |
|---|
|
Top
Abstract
Introduction
Materials and Methods
Results
Discussion
References
|
|---|
After i.v. injection, pharmacokinetic parameters of
125I-rR-RAGE are different in normal and diabetic
rats. This finding is probably related to the presence in diabetic
animals of AGEs, which create an additional distribution compartment
for 125I-rR-RAGE (Renard et al., 1997). After
i.v. injection, 125I-rH-RAGE had a higher
distribution volume and a higher clearance in diabetic than in normal
rats, but the 125I-rH-RAGE elimination half-life
was similar. It is unlikely that differences were due to the alteration
of proteins during iodination (Bauer et al., 1996), because
rR-RAGE and rH-RAGE were radiolabeled by the same protocol and have
no major difference in their phenylalanine and tyrosine contents.
Immunoprecipitation by anti-RAGE antibody is more specific than TCA
precipitation, and plasma concentrations determined by TCA
precipitation might be overestimated, while those determined by
immunoprecipitation might be underestimated, since a small conformational change may result in nonreactivity with the monoclonal antibody used. Despite this discrepancy between the two methods, immunoreactive 125I-rH-RAGE plasma concentrations
decrease rapidly after i.v. injection, whereas 93% of
125I-rR-RAGE is immunoprecipitable 2 h after
i.v. injection (Renard et al., 1997). The low percentage of
immunoprecipitated 125I-rH-RAGE in the present
study might be due to several factors: 125I-rH-RAGE may be more catabolized by
proteases, a different glycosylation between rH-RAGE and rR-RAGE can
lead to a different metabolism, or 125I-rH-RAGE
may form complexes with AGEs that modify its conformation.
Pharmacokinetic profiles after s.c. injection of
125I-rH-RAGE also suggested a rapid metabolism of
the human protein in rat. Human recombinant
125I-tumor necrosis factor 
(TNF-) has a
longer elimination half-life in monkeys than in rabbits (Bocci et al.,
1987), which further demonstrates the importance of the experimental
model. In our in vitro degradation experiments we did not detect a
difference between rR-RAGE and rH-RAGE stability in rat plasma, which
might have explained the difference observed in the animal experiments. However, the sensitivity of the technique, even after densitometry analysis, is low. A possible complex mechanism involving cell surface
associated proteases or surface components could be responsible for the
difference between rR-RAGE and rH-RAGE pharmacokinetics in the rat. In
the same rat model, we previously showed that purified bovine
125I-RAGE pharmacokinetics are similar to those
of 125I-rR-RAGE (Wautier et al., 1996; Renard et
al., 1997), which indicates that differences between human and rat RAGE
cannot simply be explained by the heterologous nature of the proteins.
To investigate possible explanations, we compared the amino acid
sequences of rat, human, and bovine RAGE and found some differences
between the three molecules. The arginine-221 in the human protein
could be an additional cleavage site for trypsin-like enzymes,
resulting in fragments of 24 and 13 kDa that both have tyrosine
residues and remain radiolabeled. This could explain the extensive
catabolism of 125I-rH-RAGE and the presence of
radiolabeled low molecular weight products on SDS-PAGE. Further studies
using a mutated rH-RAGE with a substitution of arginine-221 would be
useful to test this hypothesis. Prediction of different potential
O-glycosylation sites suggests an alternative hypothesis,
because human, rat, and bovine RAGE have different potential
O-glycosylation sites that could lead to a different
glycosylation pattern. Previous studies on protein pharmacokinetics
indicate that carbohydrate moieties modulate catabolism of proteins, as
observed with tissue plasminogen activator (Lucore et al.,
1988) and interferon (Bocci et al., 1982).
In vitro and in vivo, rR-RAGE (Renard et al., 1997) and rH-RAGE reverse
vascular permeability produced by AGEs. Despite the catabolism of
125I-rH-RAGE in rat, the biological efficacy
indicates that rH-RAGE fragments are still active or, as observed in
genetically modified mice, that recombinant RAGE is active at a low
concentration (Park et al., 1998). Delineation of the active peptide of
the rH-RAGE molecule might allow better understanding of the
discrepancy between the immunoreactive rH-RAGE blood level and its
biological activity. Development of rH-RAGE as a treatment for
reversing alterations due to AGEs will require further studies to
improve stability and bioavailability of the protein. Protection of
rH-RAGE from degradation in the bloodstream or the use of injectable
depot formulations, in which rH-RAGE would be embedded in a polymeric matrix or vesicles (e.g., liposomes) (Putney and Burke, 1998) and
released slowly, could be methods for overcoming degradation after i.v.
or s.c. administration.
| |
Acknowledgments |
|---|
We thank Dr. Lei Zhao (Berlex Biosciences) for his assistance with recombinant RAGE and VCAM-1 purification and Clotilde Zoukourian (Laboratoire de Recherche en Biologie Vasculaire et Cellulaire) for assistance in the in vivo permeability experiments.
| |
Footnotes |
|---|
Accepted for publication May 24, 1999.
Received for publication February 15, 1999.
1 This work was supported by a North Atlantic Treaty Organization grant (to J.L.W.).
Send reprint requests to: Prof. Jean-Luc Wautier, INTS, 6 rue Alexandre Cabanel, 757 39 Paris Cedex 15, France. E-mail: wautier{at}ints.fr
| |
Abbreviations |
|---|
AGE, advanced glycation end products;
RAGE, receptor for AGEs;
rH-RAGE, recombinant human RAGE;
EC, endothelial
cell;
rR-RAGE, recombinant rat RAGE;
VCAM-1, vascular cell adhesion
molecule 1;
PAGE, polyacrylamide gel electrophoresis;
TCA, trichloroacetic acid;
Fab, fragment having antigen-binding site;
RBC, red blood cell;
TBIR, tissue-blood isotope ratio;
z, terminal disposition rate constant;
T1/2z, elimination half-life;
AUC, area under the curve;
CL, total clearance;
Vz, volume of distribution of the
elimination phase;
Vc, volume of central
compartment;
Ve, extrapolated volume;
Vss, distribution volume at steady state;
MRT, mean residence time;
Cmax, maximal
concentration;
Tmax, time of
Cmax.
| |
References |
|---|
|
Top
Abstract
Introduction
Materials and Methods
Results
Discussion
References
|
|---|
This article has been cited by other articles:
|
|
 |
|
  M.-P. Wautier, W. El Nemer, P. Gane, J.-D. Rain, J.-P. Cartron, Y. Colin, C. Le Van Kim, and J.-L. Wautier Increased adhesion to endothelial cells of erythrocytes from patients with polycythemia vera is mediated by laminin {alpha}5 chain and Lu/BCAM Blood, August 1, 2007; 110(3): 894 - 901. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Basta, A. M. Sironi, G. Lazzerini, S. Del Turco, E. Buzzigoli, A. Casolaro, A. Natali, E. Ferrannini, and A. Gastaldelli Circulating Soluble Receptor for Advanced Glycation End Products Is Inversely Associated with Glycemic Control and S100A12 Protein J. Clin. Endocrinol. Metab., November 1, 2006; 91(11): 4628 - 4634. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Aragno, R. Mastrocola, C. Medana, F. Restivo, M. G. Catalano, N. Pons, O. Danni, and G. Boccuzzi Up-Regulation of Advanced Glycated Products Receptors in the Brain of Diabetic Rats Is Prevented by Antioxidant Treatment Endocrinology, December 1, 2005; 146(12): 5561 - 5567. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J.-L. Wautier and A. M. Schmidt Protein Glycation: A Firm Link to Endothelial Cell Dysfunction Circ. Res., August 6, 2004; 95(3): 233 - 238. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
 |
 |
|
 |
 |
 |
1, where
125I/51Cr is the ratio of radioactivity in
tissue to that in an arterial blood sample harvested before excising
the heart. B, effect of a transfusion of diabetic RBCs on
vascular permeability. Normal RBCs (n = 7),
diabetic RBCs (n = 7), or diabetic RBCs plus
rH-RAGE (60 µg/ml; n = 6) were injected into
normal rats and TBIR was determined 1 h after the injection. The
results are presented as mean values. Bars, mean ± S.E..
*P < .05; **P < .01; and
***P < .001.
