| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Regular Articles |
From MediCity Research Laboratories, University of Turku, and the National Public Health Institute, Turku, Finland
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Human vascular adhesion protein 1 (hVAP-1) is a homodimeric endothelial cell molecule composed of two 90-kd subunits. It mediates subtype-specific, selectin-independent lymphocyte binding to endothelial cells.4-6 hVAP-1 is mainly expressed on high endothelial venules (HEVs) in PLN-type lymphatic tissues, but immunoreactive hVAP-1 can also be found in endothelial cells of other tissues as well as in smooth muscle cells and follicular dendritic cells.4 The expression of hVAP-1 is up-regulated during inflammation in the vessels of the skin, gut, and synovium.5,7,8 The sialic acids decorating VAP-1 are essential for its function, inasmuch as hVAP-1 has been shown to be nonfunctional in lymphocyte binding assays if these oligosaccharide modifications are removed.9 Under physiologically relevant shear stress VAP-1 has been shown by intravital microscopy to mediate the formation of the initial contacts of labeled human lymphocytes with inflamed rabbit mesenterial venules,5 suggesting that VAP-1 would function at an early step of the multistep adhesion cascade.
To obtain more information on the importance of VAP-1 in lymphocyte homing in vivo and to be able to manipulate genetically the expression of VAP-1, we have recently isolated the cDNA and gene encoding mouse VAP-1 (mVAP-1)10,11 and produced a mAb against it. Antibody stainings of frozen tissue sections from PLN and gut have shown that mVAP-1 is expressed on PLN HEVs, in lamina propria vessels, and in smooth muscle cells of the mouse gut. The analysis of the predicted mVAP-1 protein core revealed that it is a novel type II transmembrane molecule with an 83% identity to hVAP-1. Moreover, mVAP-1 displays significant identity to the semicarbazide-sensitive Cu-containing amine oxidase (SSAO) enzyme family. The members of this superfamily are enzymes that catalyze the oxidative deamination of different amines and have widely differing substrate specificities.12,13 Based on the expression and presence of a quinone cofactor, enzyme-bound copper, and enzyme activity only against primary amines or monoamines, the Cu-containing amine oxidases are clearly distinct from the flavinyl adenosine diamine (FAD)-containing intracellular (mitochondrial) monoamine oxidases.14,15 The true biological role of these enzymes has remained unclear, although they have been reported to be involved in the pathogenesis of different vasculopathies.16,17
Although hVAP-1 has been shown to be inducible in clinical samples,8 it has not been possible to study the characteristics of VAP-1 in controlled animal models. Nonobese diabetic (NOD) mice are a good model system for a lymphocyte-dependent inflammatory reaction, because these mice spontaneously develop insulitis and thereafter a syndrome with clinical findings resembling those of insulin dependent diabetes mellitus.18
In the present study we have for the first time in any species been able to examine the distribution of VAP-1 in formalin-fixed paraffin-embedded sections with good resolution and have analyzed many tissues for which no information on VAP-1 synthesis had been available earlier. We have also followed the expression of mVAP-1 during the development of insulitis in the pancreata of NOD mice and shown that the expression of VAP-1 is induced in islet vessels during the formation of islet infiltrates and that this up-regulation of expression correlates with the amount of lymphocyte infiltration, suggesting a novel biological role for VAP-1 and SSAOs in the development of diabetes. Analyses of biochemical characteristics and enzyme activity of mVAP-1 in several tissues revealed tissue-specific differences in glycosylation of VAP-1 and the existence of naturally occurring substrates for this enzyme. The results show that there are important differences between human and mouse VAP-1. Finally, these data not only represent the first thorough cellular analyses of mVAP-1 but, in fact, any single molecular species of a whole large family of semicarbazide-sensitive monoamine oxidases.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
TK10-79 is a rat anti-mVAP-1 mAb, and its production has been described elsewhere.10 Briefly, a 117-bp C-terminal fragment of mVAP-1 was cloned into glutathione S-transferase (GST) gene fusion vector (Pharmacia Biotech, Uppsala, Sweden) to produce a soluble GST-mVAP-1 fusion protein, which was used to immunize rats. A new anti-hVAP-1 mAb TK8-18, which detects both the monomeric and dimeric forms of hVAP-1, was produced by immunizing BALB/c mice with immunoaffinity-purified hVAP-1 as described earlier.19 Hermes-1, a rat mAb against human CD44,20 and 3G6, a mouse mAb against an antigen expressed on chicken peripheral T cells,4 were used as negative control antibodies in immunostainings and in immunoblottings. Anti-MAdCAM-1 mAb MECA-36721 was a kind gift from Dr. E. C. Butcher. Fluorescein isothiocyanate-conjugated goat anti-rat IgG and peroxidase-conjugated goat anti-rat IgG were from Sigma Chemical Co. (St. Louis, MO). Peroxidase-conjugated sheep anti-mouse IgG was from Dakopatts (Glostrup, Denmark).
Cell Culture and Expression of mVAP-1 cDNA
CHO (Chinese hamster ovary) cells were from the American Type
Culture Collection and grown in 
-MEM (Gibco BRL, Paisley, UK) plus
CHO nucleosides supplemented with 20% fetal calf serum (FCS), 2 mmol/L
glutamine (Biological Industries, Beit Haemek, Israel), 128 U/ml
penicillin, and 128 µg/ml streptomycin. Electroporation was used
(Bio-Rad Gene Pulser apparatus; 0.3 kV, 960 µF, 0.4 cm cuvette in
RPMI plus 1 mM Na-pyruvate, 2 mM L-glutamine without serum)
to transfect these cells (5 x 106/transfection) with
20 µg of expression vector pcDNA3 alone or with the vector containing
full-length mVAP-1 cDNA. After transfection these stably transfected
cells were selected in the presence of 0.5 mg/ml geneticin (Gibco BRL).
Fluorescence-Activated cell sorting (FACS) Analyses
Stably transfected CHO cells were detached from the culture flasks by trypsin-EDTA treatment. For the immunofluorescence stainings the cells were either left intact or permeabilized and/or fixed by incubation either in 100% methanol or acetone (2 minutes, -20°C) or in 2% glutaraldehyde in phosphate-buffered saline (PBS) (20 minutes, 20°C). After washing and blocking (20 minutes in RPMI1640 containing 10% FCS) the primary mAbs were added at 10 µg/ml (diluted in PBS containing 1% FCS and 1 mM sodium azide). Fluorescein isothiocyanate-conjugated anti-rat IgG with 5% normal mouse serum was used as a second-stage reagent. The cells were finally fixed in PBS containing 1% formalin. Approximately 10,000 cells were analyzed with a FACScan cytometer (Becton Dickinson, Mountain View, CA).
Tissue Specimens and Immunohistochemistry
Tissue samples were freshly prepared from different organs of mice. BALB/c and NOD mice were obtained from the local colony in Turku, and they were fed with normal mouse chow and housed under specific pathogen-free conditions. Wild mice were caught with traps from local farms. The tissue samples were either snap-frozen or fixed in formalin. The formalin-fixed, paraffin-embedded BALB/c tissue sections were deparaffinized with xylene and rehydrated in a series of decreasing concentrations of ethanol before microwave treatments (3 x 2 minutes, 900 W (50% power) in 10 mM citric acid, pH 6.0). The endogenous peroxidase activity was removed by incubating the sections for 30 minutes in 1% H2O2 in PBS. The staining was done according to the standard protocol with the avidin-biotin complex technique (Vectastain ABC kit (rat IgG Elite), Vector Laboratories, Burlingame, CA). The primary mAbs TK10-79 (anti-mVAP-1) and Hermes-1 (control mAb) were used at a concentration of 2 µg/ml at +7°C overnight. The sections were slightly counterstained with hematoxylin.
Because MECA-367 (against MAdCAM-1) does not work on paraffin-embedded
tissue sections (data not shown), acetone-fixed serial cryostat
sections of pancreata from 3- and 12-week NOD mice were stained by
immunoperoxidase staining, using a protocol described
earlier.10
Primary mAbs TK10-79, MECA-367, and Hermes-1
were used at a concentration of 100 µg/ml. The stainings with TK10-79
and MECA-367 were read blindly by counting all islets in the sections
and every vessel around and within the islets. To avoid counting the
expression of the smooth muscle VAP-1, larger arteries and veins were
omitted from the analysis of both mVAP-1 and MAdCAM-1 stainings. The
staining intensity was scored semiquantitatively as follows: (-)
negative staining, (+) weak staining, (++) moderate staining, (+++)
strong staining. The number of infiltrating lymphocytes present in the
islets of 12-week mice was scored as follows: (0) no lymphocytes
present in the islet, (1) mild periislet insulitis, (2) insulitis
occupying 
50% of the islet area, (3) islet filled with lymphocytes
(>50% of the area) as described previously.22
Because of
the sample size, certain categories were finally combined as shown in
the tables. The statistical analyses were performed with a
2
test.
Immunoblotting and Glycosidase Digestions
Samples from different BALB/c mouse tissues were minced and lysed
in a lysis buffer (150 mM NaCl, 10 mM Tris-base, pH 7.2, 1.5 mM
MgCl2, 1% NP-40, 1% aprotinin, and 1 mM
phenylmethylsulfonyl fluoride), and after removal of the insoluble
material by centrifugation, all of the supernatants, except the
supernatant from adipose tissue, were depleted of endogenous
immunoglobulins by incubation with protein G-Sepharose beads
(Pharmacia, Sweden). Mouse smooth muscle cell lysate was prepared from
aorta and the white adipose tissue lysate from abdominal fat. The
precleared samples were then mixed with Laemmli sample buffer with or
without reduction (5% 2-mercaptoethanol) before loading onto 5–12.5%
sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE)
gradient gels. The human smooth muscle lysate was prepared from tissue
samples obtained from surgical operations and run in a parallel lane on
the same gel as the mouse lysates. For glycosidase treatments the
clarified lysate supernatants (mAb TK10-79 does not work in
immunoprecipitation) were treated with different enzymes, using a
protocol previously described for hVAP-1.9
In brief, for
sialidase treatments 50-µl aliquots of the lysate were incubated with
Vibrio cholerae sialidase (Behring, 5 mU, 2 h,
+37°C), which removes 2,3, 2,6, and 2,8 linked sialic
acids.23
Endo--N-acetylgalactosaminidase
(Genzyme, 16 mU, +37°C, overnight) was used to remove O-glycans from
the sialidase-treated samples, and peptide: N-glycosidase F treatment
(Genzyme, 1.2 U, +37°C, overnight) was used to digest N-linked
oligosaccharides.
Amine Oxidase Assays
Stably transfected CHO cells (5–10 x 106) were detached with trypsin-EDTA, washed with PBS, resuspended in 1 ml PBS, and finally lysed by using a Braun sonicator (Melsungen, Germany). The enzyme activities were measured according to a radioisotopic method,24 using 25 µl of the lysate in all reactions. Unlabeled substrates (benzylamine, ß-phenylethylamine, tyramine, tryptamine, methylamine, and histamine; obtained from Sigma Chemical Co.) were added to the reaction at 1 mM as cold competitors before the addition of 14C-labeled benzylamine. The reactions were incubated at 37°C for 60 minutes before extraction of the labeled aldehyde reaction products in toluene. Finally, the extracts were measured with a liquid scintillation counter.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
mAb TK10-79 has been raised against a bacterial expressed 39 amino
acid peptide of mVAP-1.10
To confirm the specificity of
the anti-mVAP-1 mAb TK10-79, mVAP-1 and mock transfected CHO cells were
stained with this mAb and a control mAb (Figure 1)
. Neither of the untreated
transfectants stained positively with mAb TK10-79 (Figure 1
, untreated
samples). No VAP-1 reactivity was observed with glutaraldehyde fixation
or acetone permeabilization of the transfectants (Figure 1
and data not
shown). However, when the cells were fixed with methanol, a positive
signal was detected from mVAP-1 CHO transfectants, but not from mock
cells, with TK10-79, while the control mAb stainings remained negative,
showing that the mAb TK10-79 is immunoreactive against mVAP-1. These
data indicate that mAb TK10-79 does not detect the native mVAP-1 in
these transfectants, but rather a certain epitope of the mVAP-1 protein
core after fixation of the cells and denaturation of the protein.
|
Previously mVAP-1 has been shown to be expressed in gut and PLN by staining of frozen tissue sections, but the staining intensity was relatively weak.10 Expression of mVAP-1 in all other tissues and cells is unknown. However, during the course of this work we found out that mAb TK10-79 stained mVAP-1 well in formalin-fixed, paraffin-embedded tissue sections. In fact, mAb TK10-79 is the first anti-VAP-1 mAb (of 15) in any species that recognizes VAP-1 in these histological specimens of superior resolution. Therefore paraffin sections were prepared from different mouse organs to determine the distribution of mVAP-1. In all organs examined, mVAP-1 is expressed at least on a few endothelial cells. In addition, smooth muscle cells and fat cells are positive in all tissues studied. mVAP-1 is always absent from all types of epithelial cells, leukocytes, and fibroblasts.
The staining of paraffin sections of mouse gut revealed exactly the
same staining pattern as previously detected from frozen sections. The
smooth muscle cells and lamina propria endothelial cells were mVAP-1
positive, as shown in Figure 2A
. Some,
but not all, HEVs in Peyer’s patches were also VAP-1 positive (Figure 2B)
. In PLN HEVs mVAP-1 was expressed on both the endothelial cells and
pericytes, whereas the lymphocytes were mVAP-1 negative. In germinal
centers of the lymph nodes mVAP-1-positive dendritic cells were
detected (data not shown). In spleen mVAP-1 is expressed in many
vessels but not on splenocytes or marginal zone macrophages (Figure 2C)
.
|
, strong
mVAP-1 expression was seen on the sinusoidal vessels in bone marrow,
but mVAP-1 was absent from osteocytes and hematopoetic cells. We found
mVAP-1 on the endothelial cells of thymic vessels in both the cortex
and medulla in a conspicuously luminal pattern (Figure 2E)
.
mVAP-1 is present in multiple nonlymphoid tissues. In lung the
capillary endothelial cells and alveolar epithelial cells were mVAP-1
negative, whereas the smooth muscle cells of the bronchii and
endothelium of larger veins were mVAP-1 positive (Figure 2F)
. In
heart, the cardiac muscle cells were mVAP-1 negative, but the
endocardium and myocardial capillary endothelium expressed this antigen
(Figure 2G)
. The endothelium of large arteries (aorta) was faintly
mVAP-1 positive. In kidney, the glomerular endothelium, Bowman’s
capsule, peritubular capillaries, and renal tubuli were mVAP-1
negative, and the only reactivity was found in the muscle cell layer of
the incoming glomerular artery (Figure 2H)
. In liver the sinusoidal
endothelial cells lacked or had only a marginal mVAP-1 reactivity,
whereas the central veins displayed endothelial VAP-1 positivity
(Figure 2I)
. mVAP-1 was absent from hepatocytes and bile duct cells. In
pancreas, exocrine and endocrine cells lacked mVAP-1, which was only
faintly detectable in blood vessels. In brain, neurons and glial cells
were mVAP-1 negative, and only occasional positive vessels were found.
In skin, mVAP-1 was found only at low levels on a few dermal vessels.
In both white (Figure 2J)
and brown fat tissue the adipocytes were
strongly mVAP-1 positive. mVAP-1 was absent from striated skeletal
muscle cells (data not shown).
Our conclusion is that mVAP-1, like all other known adhesion molecules (reviewed in ref. 1 ), is present not only on endothelial cells but also on certain other cell types in different organs. The endothelial staining pattern is compatible with the possibility that VAP-1 functions at the sites of lymphopoesis (thymus and bone marrow), in the physiological recirculation of lymphocytes (secondary lymphoid organs) and in leukocyte influx into sites of inflammation (nonlymphoid tissues).
mVAP-1 Is Expressed in Islet Vessels of Nonobese Diabetic Mice
To investigate the expression of mVAP-1 in inflammation we
analyzed the well-established model of insulitis in diabetic mice. In
our colony, none of the animals has insulitis at 3 weeks, whereas by 12
weeks 80% of the islets are inflamed in females. In this disease
MAdCAM-1 serves as a model of an inducible, functionally important
adhesion molecule.22,25
To compare the distribution of
mVAP-1- and MAdCAM-1-positive vessels in the endocrine pancreas of NOD
mice, frozen tissue sections from 12-week-old animals with emerging
insulitis were stained with mAbs TK10-79 and MECA-367 (an anti-MAdCAM-1
mAb). The expression of both molecules could be found
in the vessels of endocrine and exocrine pancreas, and a similar
staining pattern of mVAP-1 was seen on paraffin sections (data not
shown). Because mAb MECA-367 does not work on paraffin sections, usage
of frozen tissue sections was mandatory to compare the localization of
these vascular addressins on islet vessels. The staining of serial
sections revealed that mVAP-1 and MAdCAM-1 colocalize in the same
vessels in many islets (two representative islets are shown in Figure 3
), although vessels that were positive
only for VAP-1 or MAdCAM-1 were also found.
|
The observation that mVAP-1- and MAdCAM-1-positive vessels
colocalize led us to investigate whether the expression of mVAP-1, like
that of MAdCAM-1,25
is up-regulated during the
pathogenesis of diabetes. To that end we compared the expression of
mVAP-1 in the pancreata of NOD mice at the age when most of the islets
are devoid of lymphocyte infiltration (3–4 weeks) and at the age when
advanced insulitis occurs (12 weeks). As a positive control for the
scoring, we also stained and analyzed the expression of MAdCAM-1 in the
same pancreata. When the number of MAdCAM-1-positive vessels was
compared at the ages of 3–4 and 12 weeks, more vessels having moderate
or strong MAdCAM-1 expression were found at 12 weeks (Table 1)
, confirming the earlier observations
and hence our scoring system with the frozen sections. When the
stainings for mVAP-1 were analyzed, there were statistically
significantly (P < 0.001) more vessels modestly
or strongly VAP-1 positive in 12-week-old than in 3–4-week-old animals
(Table 1)
. These results indicate that during the development of
insulitis the expression of both mVAP-1 and MAdCAM-1 is induced (Table 1)
.
|
and Table 2
). Thus the up-regulation of mVAP-1 in
islet vessels of 12-week NOD mice correlates statistically
significantly with a high degree of lymphocyte influx.
|
|
To study the biochemical properties of mVAP-1 in different
tissues, lysates from several mouse tissues were run in parallel on
SDS-PAGE gel, transferred to a filter, and stained with the anti-mVAP-1
mAb. Under nonreducing conditions a specific 220-kd antigen was
detected from most examined tissues (Figure 5A
, lanes 1–13). In prolonged exposures
(data not shown) the 110-kd antigen was also detectable in all tissues
except PLN/MLN lym-phocytes, which remained constantly negative. Under
reducing conditions a prominent ~110-kd antigen was detected from the
smooth muscle lysate (Figure 5B)
, but the ~220-kd antigen
disappeared, and a large smear at higher molecular weights became
reproducibly detectable. The probing with the control mAb remained
negative. A similar result under reducing conditions was also seen from
CHO-mVAP-1 transfectants (data not shown). Moreover, in all examined
tissues mVAP-1 was roughly of the same size, and the intensity of the
signal corresponded well with the staining pattern. The comparison of
the apparent molecular weight of mVAP-1 in nonreduced and reduced
SDS-PAGE analyses to the predicted mass of the amino acid sequence and
to the structure of hVAP-1 shows that most, if not all, mVAP-1 is in a
dimeric form composed of two 110-kd subunits in the examined tissues.
|
, lane 14) and
detected with an anti-hVAP-1 mAb TK8-18. A predominant 170-kd molecule
was seen together with larger forms typical of smooth muscle VAP-1
(unpublished results). Thus mVAP-1 has a clearly slower electrophoretic
mobility than hVAP-1. mVAP-1 Is a Sialoglycoprotein
There is a marked size discrepancy between the size of mVAP-1
monomer seen in immunoblotting (~110 kd) and the predicted molecular
mass from the cDNA sequence (84.5 kd). Because mVAP-1 contains six
potential N-glycosylation sites and six putative
O-glycosylation sites,10
we next analyzed
whether the difference in the two molecular mass estimates is due to
oligosaccharide modifications. Cell lysates from peripheral lymph
nodes, heart, adipose tissue, and gut smooth muscle were digested with
different glycosidases. Sialidase treatment decreased, although quite
modestly, the molecular weight of mVAP-1 in all examined tissues
(Figure 6)
, indicating that mVAP-1
protein core is decorated with sialic acids. Cleavage of the
lysates with N-glycanase also increased the
electrophoretic mobility of mVAP-1, showing that mVAP-1 contains
N-linked oligosaccharide decorations. However, mVAP-1 may not have
major O-linked side chains, because further digestions of
sialidase-treated mVAP-1 with O-glycanase did not have any
detectable effect on the mobility of mVAP-1 when compared to the
samples treated with sialidase only. The change in the electrophoretic
mobility of desialylated or de-N-glycosylated mVAP-1 varied
to some extent between different tissues and was greatest in the
adipose tissue. Thus mVAP-1 is a sialoglycoprotein that is differently
glycosylated in distinct tissues.
|
Because mVAP-1 has previously been shown to possess
activity only against benzylamine, a synthetic amine not found in
vivo,15
we tested whether any endogenously found
amines could be oxidized by mVAP-1. Lysates from CHO-mVAP-1
transfectants were used in a radioisotopic assay as the source of the
enzyme, and [14C]benzylamine was used as the substrate.
The effect of different amines was tested by adding them unlabeled in
the reaction and by analyzing the effect of possible competition on the
oxidation of the labeled benzylamine. As expected, an excess of cold
benzylamine almost completely inhibited the oxidation of the labeled
benzylamine. Interestingly, mVAP-1 also binds ß-phenylethylamine,
tyramine, tryptamine, and methylamine, because these amines also
inhibited the oxidation of labeled benzylamine (Figure 7)
. In contrast, histamine is not a
substrate for mVAP-1. These data indicate that mVAP-1 can oxidatively
deaminate naturally occurring primary amines.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The appearance of mononuclear cells (mainly lymphocytes) in the islets of Langerhans in the pancreas is a key event in the pathogenesis of insulin-dependent diabetes mellitus.26,27 This inflammatory response, called diabetic insulitis, leads to the destruction of the insulin-secreting ß cells and later to the onset of the disease. Previously it has been shown with NOD mice that MAdCAM-1, a mucosa-specific vascular addressin,21,28 is induced on islet vessels of these mice during the development of insulitis and that it mediates the binding of lymphocytes to the islet endothelium.25 The expression of MAdCAM-1 correlates with the degree of lymphocyte infiltration,29 and furthermore, the incidence of diabetes can be reduced by blocking the function of MAdCAM-1 by an anti-MAdCAM-1 mAb,22 indicating that MAdCAM-1 is required for the development of diabetes. Here we have shown that mVAP-1 is expressed on islet vessels in NOD mice, its expression is induced on islet endothelium during the development of the insulitis, and mVAP-1 is induced in islet vessels with a high degree of lymphocyte infiltration. Our findings show that VAP-1 may have a role similar to that of MAdCAM-1 in the pathogenesis of diabetes. Intriguingly, in humans inhibitory mAbs against inflammation-induced VAP-1 have been shown to block the function of VAP-1 in vitro.8 However, the functional role of mVAP-1 remains to be studied, because the anti-mVAP-1 MAb TK10-79 is noninhibitory. After development of a function blocking anti-mVAP-1 mAb, it will be of interest to use this model to investigate whether mVAP-1 is causally involved in the development of the disease.
In this study the expression of VAP-1 was investigated for the first
time in formalin-fixed paraffin-embedded sections in any species. It
allowed us to define the expression of mVAP-1 precisely and in multiple
tissues not earlier amenable for analyses. In most tissues and cell
types, the expressions of mVAP-1 and hVAP-1 were the same. There are,
nevertheless, some clear distinctions. Immunoperoxidase stainings of
human liver tissue sections have previously revealed strong VAP-1
positivity in the vascular endothelium of portal vessels and sinusoidal
endothelium of the liver.8,30
In contrast, the sinusoids
are mostly mVAP-1 negative in BALB/c mice. This is notable, because in
humans, liver VAP-1 has been shown to mediate T-cell binding to hepatic
endothelium and to support the binding of tumor-infiltrating
lymphocytes to tumor endothelium in hepatocellular
carcinomas.30,31
Because previous Northern blot analyses
have revealed that the expression of mVAP-1 and hVAP-1 in liver does
not differ at the mRNA level,6,10
one explanation for the
lack of mVAP-1 staining is that in mouse liver, VAP-1 may have
different posttranslational modifications and therefore be poorly
detectable by the antibody. Another possible explanation for the almost
complete lack of mVAP-1 expression in sinusoidal endothelium is that
the mice have been housed under specific pathogen-free conditions.
Because the liver is a major lymphatic organ (reviewed in ref. 32
), the
lack of immunostimulation may explain the low level of mVAP-1
expression. This latter possibility is also supported by the
observation that mVAP-1 seems to be more faintly expressed on PLN HEVs
of these animals than hVAP-1 is in human PLN (data not shown). Another
tissue where the expression of mVAP-1 differs from that of hVAP-1 is
kidney. In humans the peritubular small vessels (through which
lymphocytes infiltrate into a rejecting kidney) display strong VAP-1
positivity, but in mouse these capillaries are negative. Although it is
possible that the modifications of renal VAP-1 may also be different,
it should be kept in mind that genuine species-specific differences can
also lie behind the detected differences in the expression of mouse and
human VAP-1 (Table 3)
.
|
The SDS-PAGE separation and immunoblotting analysis of mVAP-1 in different tissues revealed that the apparent molecular weight of mVAP-1 is higher than that of hVAP-1. This is in accordance with previous reports stating that the molecular weight of different SSAOs varies between species.33 The glycosidase digestions of lysates from different mouse tissues revealed that mVAP-1 contains at least some sialic acids and N-linked oligosaccharides. Digestions with the same glycosidases lead to different changes in mobility in mVAP-1 and hVAP-1 on an SDS-PAGE gel analysis. Treatment of hVAP-1 with sialidase paradoxically increases the apparent molecular weight of hVAP-1, indicating that hVAP-1 is decorated with abundant sialic acid residues affecting the net charge of the molecule or that the removal of them causes a differential conformational change leading to an altered mobility.9 N-Glycanase treatment leads to a detectable change in the mobility of only smooth muscle-derived hVAP-1 but not that of HEV-originating hVAP-1. Moreover, smooth muscle and endothelial hVAP-1 consistently display different mobilities in SDS-PAGE. Interestingly, the glycosylation of mVAP-1 seems to differ not only between hVAP-1 and mVAP-1, but also between various mouse tissues. This indicates, together with our previous findings of tissue-specific mVAP-1 mRNA transcription start sites,11 that there may be heterogeneity in the structure and presumably in the function of mVAP-1 in different cell types.
The substrate specificity and affinities of different SSAOs have been
reported to vary unpredictably between different
species.34,35
For example, certain amines like tyramine
and tryptamine are good substrates for rat but poor substrates for
human SSAO. Previously both human and mouse VAP-1 have been shown to
have activity against benzylamine, but the activity of mVAP-1 toward
amines present in vivo has remained unknown. This study
reveals that mVAP-1 is active against methylamine, tyramine,
tryptamine, and ß-phenylethylamine. Thus the substrate specificity of
mVAP-1 is different from that of hVAP-1 because only methylamine of the
amines tested here has been shown to be oxidized by
hVAP-16
(Table 3)
. Furthermore, hVAP-1 has higher affinity
toward methylamine than does mVAP-1 (unpublished observations). This
study reveals for the first time physiological substrates for a mouse
SSAO because, eg, tyramine is ingested in food and methylamine is
formed endogenously from metabolic pathways involved in the degradation
of sarcosine, creatinine, and adrenaline.13
Based on our
observations the substrate specificity of mVAP-1 resembles the
specificity of rat SSAO more than it does the specificity of human SSAO
(hVAP-1). However, regardless of the substrate specificity, the
aldehydes produced in the reaction have been shown to be important in
vascular injury seen in pancreas, kidney, and arteries during
inflammatory disorders.16,17
Moreover, the oxidation
reaction of these amines results in the production of
H2O2, which is a potent modulator of the local
microenvironment. Therefore, it will be of interest to study the
effects of mVAP-1-dependent H2O2 production
because it has been shown that this oxygen radical affects the
expression of other adhesion molecules36,37
and leukocyte
rolling in vivo.38
A prominent VAP-1 staining can be found in both mouse and human white
adipose tissue, indicating that adipocytes widely express VAP-1 (Figure 2
and data not shown). This is in accordance with a recent (partial)
cloning report of a rat membrane-bound adipocyte amine oxidase, which
probably represents the rat homologue of mouse and human VAP-1, and the
identity of which to mVAP-1 is 95%.10,39
The very
prominent mVAP-1 staining in the adipose tissue is of interest because
some leukocyte adhesion molecules have been suggested to participate in
the regulation of adipose tissue mass. For instance, intercellular
adhesion molecule 1-deficient mice spontaneously become obese, and
Mac-1 (CD11b/CD18)-deficient mice are susceptible to diet-induced
obesity, indicating that leukocyte adhesion and fat metabolism may be
linked phenomena.40
Based on a recent finding that
benzylamine has effects on glucose transport of cultured
adipocytes,41
VAP-1 may also affect glucose metabolism,
because we have now shown that mVAP-1 can deaminate not only
benzylamine but also naturally found amines. Moreover, while the
manuscript was being revised, a mouse SSAO cDNA sequence from
adipocytes was reported,42
which was 100% identical to
the mVAP-1 sequence we had published earlier. SSAO activity was shown
to be involved in body mass regulation in rats. Therefore, it will be
interesting to examine whether the VAP-1-deficient mice gain weight
normally or if they are susceptible to obesity.
mVAP-1 is expressed in endothelial, smooth muscle, and fat cells. In addition to the restricted expression pattern, another level of functional control exists. In experiments with dogs and pigs, a cross-reacting anti-hVAP-1 mAb has been given iv to animals with skin inflammations, and the localization of the in vivo bound mAb has been visualized from tissue sections (Jaakkola et al, manuscript in preparation). The experiments showed that luminal VAP-1 is only found at sites of inflammation but not in resting vessels, which nevertheless express abundant cytoplasmic VAP-1.
Mouse and human VAP-1’s are 83% identical proteins that share many
common features in expression, biochemical structure, and enzyme
activity. Nevertheless, there are several distinct parameters
(summarized in Table 3
) that should be taken into account when
comparing studies between animals and humans. The expression of mVAP-1
in primary and secondary lymphatic tissues and its inducibility at
sites of inflammation (eg, in insulitis) make it a potential adhesion
molecule that mediates both physiological lymphocyte recirculation as
well as extravasation into sites of inflammation. On the other hand,
the analyses described here represent the first detailed cellular and
molecular study of any member of the large, but so far not well
understood, class of semicarbazide sensitive monoamine oxidase family.
Moreover, the activity of mVAP-1 against naturally occurring amines
provides new clues to the biological role of these enzymes.
|
Acknowledgements |
|---|
|
Footnotes |
|---|
Supported by the Finnish Academy, the Finnish Cancer Union, the Finnish Cultural Foundation, the Finnish Medical Foundation, the Emil Aaltonen Foundation, the Paulo Foundation, and the Sigrid Juselius Foundation.
Accepted for publication July 25, 1999.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  S. Jalkanen and M. Salmi VAP-1 and CD73, Endothelial Cell Surface Enzymes in Leukocyte Extravasation Arterioscler. Thromb. Vasc. Biol., January 1, 2008; 28(1): 18 - 26. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Salmi and S. Jalkanen Developmental regulation of the adhesive and enzymatic activity of vascular adhesion protein-1 (VAP-1) in humans Blood, September 1, 2006; 108(5): 1555 - 1561. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. H. Yu, L.-X. Lu, H. Fan, M. Kazachkov, Z.-J. Jiang, S. Jalkanen, and C. Stolen Involvement of Semicarbazide-Sensitive Amine Oxidase-Mediated Deamination in Lipopolysaccharide-Induced Pulmonary Inflammation Am. J. Pathol., March 1, 2006; 168(3): 718 - 726. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Merinen, H. Irjala, M. Salmi, I. Jaakkola, A. Hanninen, and S. Jalkanen Vascular Adhesion Protein-1 Is Involved in Both Acute and Chronic Inflammation in the Mouse Am. J. Pathol., March 1, 2005; 166(3): 793 - 800. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Abella, L. Marti, M. Camps, M. Claret, J. Fernandez-Alvarez, R. Gomis, A. Guma, N. Viguerie, C. Carpene, M. Palacin, et al. Semicarbazide-Sensitive Amine Oxidase/Vascular Adhesion Protein-1 Activity Exerts an Antidiabetic Action in Goto-Kakizaki Rats Diabetes, April 1, 2003; 52(4): 1004 - 1013. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Salmi, C. Stolen, P. Jousilahti, G. G. Yegutkin, P. Tapanainen, T. Janatuinen, M. Knip, S. Jalkanen, and V. Salomaa Insulin-Regulated Increase of Soluble Vascular Adhesion Protein-1 in Diabetes Am. J. Pathol., December 1, 2002; 161(6): 2255 - 2262. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Marti, A. Abella, C. Carpene, M. Palacin, X. Testar, and A. Zorzano Combined Treatment With Benzylamine and Low Dosages of Vanadate Enhances Glucose Tolerance and Reduces Hyperglycemia in Streptozotocin-Induced Diabetic Rats Diabetes, September 1, 2001; 50(9): 2061 - 2068. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Morin, J.-M. Lizcano, E. Fontana, L. Marti, F. Smih, P. Rouet, D. Prévot, A. Zorzano, M. Unzeta, and C. Carpéné Semicarbazide-Sensitive Amine Oxidase Substrates Stimulate Glucose Transport and Inhibit Lipolysis in Human Adipocytes J. Pharmacol. Exp. Ther., April 12, 2001; 297(2): 563 - 572. [Abstract] [Full Text]
|
|
|
|
 |
|
 N. Andrés, J. M. Lizcano, M. J. Rodríguez, M. Romera, M. Unzeta, and N. Mahy Tissue Activity and Cellular Localization of Human Semicarbazide-sensitive Amine Oxidase J. Histochem. Cytochem., February 1, 2001; 49(2): 209 - 218. [Abstract] [Full Text]
|
|
|
|
|
|
T. Martelius, M. Salmi, H. Wu, C. Bruggeman, K. Hockerstedt, S. Jalkanen, and I. Lautenschlager Induction of Vascular Adhesion Protein-1 during Liver Allograft Rejection and Concomitant Cytomegalovirus Infection in Rats Am. J. Pathol., October 1, 2000; 157(4): 1229 - 1237. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. S. Alexander and D. N. Granger Lymphocyte Trafficking Mediated by Vascular Adhesion Protein-1 : Implications for Immune Targeting and Cardiovascular Disease Circ. Res., June 23, 2000; 86(12): 1190 - 1192. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||
