close

Вход

Забыли?

вход по аккаунту

?

Colloidal gold localization of type IV collagen in the extracellular matrix of rat gastric mucosaInfluence of alcohol and prostaglandin.

код для вставкиСкачать
THE ANATOMICAL RECORD 230:235-242 (1991)
Colloidal Gold Localization of Type IV Collagen in
the Extracellular Matrix of Rat Gastric Mucosa:
Influence of Alcohol and Prostaglandin
KATHRYN S. RIGHTOR, PHILIP A. MITCHELL, THOMAS A. MILLER, AND
KARMEN L. SCHMIDT
Departments of Pathology and Laboratory Medicine (K.S.R., P.A.M., K.L.S.) and Surgery
(T.A.M.), University of Texas Medical School, Houston, Texas
ABSTRACT
The effect of acute alcohol exposure on the gastric mucosal basal
lamina, and its major structural protein type IV collagen, was assessed by transmission electron microscopy (TEM) and immunogold (IG) labeling of this collagenous material. Fasted rats orally received either 50% or 100% ethanol. Five or 60
minutes later animals were sacrificed and mucosal samples were obtained from the
glandular epithelium for TEM or IG localization of type IV collagen. For IG studies, the number of gold particleslarea lamina densa was quantified in interfoveolar, pit, and gland regions as a n index of the molecular integrity of type IV collagen. Both ethanol concentrations induced epithelial exfoliation with pleating of
the denuded lamina densa. Absolute ethanol, and to a lesser extent 50% ethanol,
caused frequent rupture of a thickened, precipitated lamina densa. Immunolabeling of type IV collagen varied with the experimental protocol. In control tissues
exposed to oral saline, binding was greatest in the interfoveolar zone. Low binding
occurred with 100% ethanol in all regions when compared with controls, but 50%
ethanol evoked significantly higher binding in interfoveolar regions, in a similar
fashion to controls. In additional studies in which 16,16 dimethyl prostaglandin E,
(PGE,) (10 pgikg) was injected subcutaneously prior to oral ethanol exposure,
PGE, pretreatment prevented the large decrease in IG binding induced by absolute
ethanol, but the level still remained significantly less than with corresponding
controls. In contrast, pretreatment with PGE, prior to 50% ethanol exposure restored type IV collagen immunolabeling to control levels. These results indicate
that ethanol induces a concentration-dependent lowering of IG binding to type IV
collagen which also effects its reversibility by PGE,.
Although alcohol1 is a potent necrotizing agent when
applied to the gastric mucosa, its site and mechanism
of action remain elusive. The epithelium (Lacy, 1985)
and vasculature (Szabo, 1987) have been widely studied as targets of ethanol insult. The extracellular matrix, composed of basal laminae and interstitial stroma,
has also been suggested as a possible focus of alcoholmediated injury (Lacy, 1987; Lacy and Ito, 198413; Rutten and Ito, 1983; Morris and Wallace, 1981; Svanes et
al., 1982) but its individual components have not been
examined in detail. Since the basal lamina serves as
the primary interface between the lamina propria and
the gastric epithelium and endothelium, damage to the
basal lamina may have a vital impact on the ability of
the mucosa to recover from injury. Support for this contention is based on the fact that in other tissues, the
presence of a n intact basal lamina has been shown to
be a prerequisite to epithelial cell migration and its
ability to resurface a denuded connective tissue surface
(Thorning and Vracko, 1977).
Acute exposure of the gastric mucosa to ethanol induces surface erosions and hemorrhagic, necrotic le'Alcohol and ethanol are used interchangeably
0 1991 WILEY-LISS. INC.
sions which are repaired, respectively, by restitution
and healing. Restitution, a well-studied phenomenon,
is presently defined as the rapid repair of superficial
gastric injury by the migration of viable cells from the
gland isthmus to resurface a denuded basal lamina
(Lacy, 1988, 1985; Schmidt et al., 1986). Healing of
hemorrhagic and necrotic gastric lesions is a n extremely slow process which requires cell proliferation,
differentiation, and migration into the wound site (Ito
and Lacy, 1984; Silen and Ito, 1985; Yoemans e t al.,
1973).
Although ethanol alters the ultrastructure of the
gastric mucosal basal lamina (Lacy and Ito, 1984a,b;
Ito and Lacy, 1985; Schmidt et al., 1986; Schmidt and
Miller, 1988b), its effects upon individual components
of this structure remain unexplored. Among these components are proteins and glycoproteins including type
Received October 2, 1989; accepted October 2, 1990.
Address reprint requests to Karmen L. Schmidt, Ph.D., Department
of Pathology and Laboratory Medicine, University of Texas Medical
School, 6431 Fannin Street, Houston, T X 77030.
236
K.S. RIGHTOR ET AL.
IV collagen, laminin, entactin, and several proteoglycans (Timpl and Martin, 1982). If alcohol alters one or
more of these components, the ability of the gastric
mucosa to repair itself by restitution or healing may be
greatly compromised. Such a presumption is supported
by the findings of Black and colleagues (1985), who
demonstrated that anionic sites corresponding to glycosaminoglycans in the basal lamina, a s depicted by
ruthenium red staining, are reduced when gastric tissue is insulted by acid.
The present study focused upon the effects of alcohol
on type IV collagen because i t is the major protein
forming the structural backbone of epithelial and endothelial cell basal laminae (Fitch et al., 19821, and has
been identified as the extracellular matrix scaffold for
branching morphogenesis (Chen and Little, 1987). Accordingly, we have assessed the effects of alcohol upon
the ultrastructure of gastric mucosal basal lamina and
quantitatively evaluated immunogold labeling of type
IV collagen during restitution. Alterations in the conformation or number of immunogold binding sites as
reflected by the number of gold particles bound per
unit area of lamina densa were used a s a n index of the
molecular integrity of type IV collagen molecules following exposure to alcohol (Novotny et al., 1986).
Lacy and Ito (1982) and Schmidt et al. (1985) have
shown that, following ethanol exposure, prostaglandins prevent necrotic lesion formation and reduce the
depth of injury in gastric mucosa such that damage
occurs primarily at the surface. Prostaglandin pretreatment prior to alcohol is still accompanied by severe injury in the interfoveolar and pit regions of the
epithelium while the deeper gland region is spared.
Whether prostaglandins can alter the effect of alcohol
upon the basal lamina and its components is unknown,
but preservation of basal laminae should permit rapid
restitution and long-term healing to proceed more effectively. Therefore, the ability of 16,16 dimethyl prostaglandin E, to modulate alcohol-induced alterations
of immunogold binding to type IV collagen was also
assessed.
MATERIALS AND METHODS
Female Sprague-Dawley rats, weighing 180-200 g
(TIMCO, Houston, TX), were fasted overnight in wirebottom cages to prevent ingestion of hair and feces. At
least three animals were employed in each treatment
group. On the following morning, they were injected
subcutaneously with 10 pglkg body weight of 16,16
dimethyl prostaglandin E, (PGE,) or a n equal volume
of saline (S). The route of administration and dose of
PGE, used were chosen because Schmidt et al. (1985,
1986) have previously shown that subcutaneous injection of PGE, in this fashion decreases the depth of gastric mucosal injury induced by absolute alcohol. In addition, this technique has been widely used by other
investigators (Faust et al., 1989; Rowe e t al., 1987, Ruwart and Rush, 1984; Takeuchi e t al., 1986, 1988;
Wang et al., 1989). Thirty minutes after PGE, or S
pretreatment, 1 ml of 50% or 100% ethanol or a n equal
volume of S was administered orally to the animals via
a n oro-esophageal tube. These concentrations of alcohol were employed because they reproducibly induce a
consistent pattern of injury to the rat gastric epithelium t h a t is not commonly seen with lower concentra-
tions. At 5 or 60 min post-ethanol exposure, animals
were sacrificed, and tissue samples from the glandular
portion of the stomach were removed from the greater
curvature in a plane perpendicular to the long axis of
the stomach (Schmidt et al., 1985) because alcohol-induced lesions consistently form in this region within 5
min of exposing the stomach to these concentrations of
ethanol. The 60 min sacrifice time was selected because
restitution actively engages and proceeds in gastric
mucosa within 1h after alcohol exposure (Lacy and Ito,
198413; Ito and Lacy, 1985; Schmidt et al., 1985, 1986).
To retrieve gastric samples, animals underwent laparotomy and were perfused with fixative via the abdominal aorta while under ether anesthesia. Tissue
blocks measured approximately 1 mm x 2 mm. Specimens were subsequently prepared for either transmission electron microscopy (TEM) or immunogold (IG)
localization of type IV collagen. Half-strength Karnovsky’s fixative (Karnovsky, 1965) was used for standard TEM preparation while 1% paraformaldehyde
was employed as the fixative for IG localization of type
IV collagen. Paraformaldehyde was selected for IG
studies because it preserves antibody binding to type
IV collagen in muscle cell external laminae (Stephens
et al., 1982), which we have subsequently confirmed in
studies with gastric mucosa (Smith and Schmidt,
1987). Once perfusion was initiated, the stomach was
filled with 2 ml of the perfusion fixative and perfusion
was carried out 10 min longer.
Gastric tissues prepared for TEM were post-fixed in
osmium tetroxide which also contained 1% potassium
ferrocyanide (Karnovsky, 1971). Samples were embedded in Spurr’s resin (Spurr, 19691, thin-sectioned,
stained with uranyl acetate and lead citrate, and photographed in a JEOL-100CX electron microscope at 60
kV. Tissues destined for IG studies were fixed for 1% h
in paraformaldehyde and then embedded in Lowicryl
K4M (Carlemalm et al., 1982) using a procedure modified from Bendayan (1984). Postfixation with osmium
tetroxide was omitted since i t has been previously
shown that the antigenicity of collagen is sensitive to
fixation and embedding procedures (Laurie e t al.,
1980). Tissue blocks were placed in resin-filled gelatin
capsules and the resin was polymerized for 3 days under indirect ultraviolet light a t -30°C and for a n additional 2 days at room temperature. Ultrathin sections were cut, placed on Butvar-coated nickel grids
(Handley and Olsen, 1979), and labeled for IG within
48 h after sectioning. IG analysis was determined simultaneously on sections from all treatments within a
single time point (either 5 or 60 min). Since Lowicryl
K4M is a hydrophilic, polar resin, and etching is a technique used to either dissolve resin or decrease its hydrophobicity, etching was not necessary to unmask antigens (Carlemalm et al., 1982; Martinez-Hernandez,
1987). For immunogold labeling, grids were first incubated with 1%bovine serum albumin (BSA) to reduce
non-specific binding, then placed directly on a drop of a
1:20 dilution of type IV collagen antisera (courtesy of
Dr. H. Kleinman, NIH, Bethesda, MD) for 60 min. After washing several times with phosphate buffered saline (PBS), grids were incubated in a 1:4 dilution of 20
nm protein A-colloidal gold (EY Laboratories, San Mateo, CA) for 30 min, then washed thoroughly. IG controls included replacing the primary antibody with ei-
ALCOHOL, PG, AND GASTRIC EXTRACELLULAR MATRIX
Fig. 1. Transmission electron micrograph (TEM) showing the interfoveolar region of normal gastric mucosa. The finely particulate lamina densa (arrowheads) closely follows the epithelial contours (E). Bar
= 0.5 pm.
Fig. 2. TEM showing the surface epithelium (E) in gastric tissue
after 60 min of treatment with 100%ethanol. The lamina densa (ar-
ther saline or pre-immune serum. Subsequently,
sections were stained with uranyl acetate for 3-5 min
and viewed on a JEOL 1200-EX transmission electron
microscope at 60 kV.
For each animal, a minimum of 5 photographstregion
were taken at 6,000 x in each of the interfoveolar, midpit, and glandular isthmus regions. Thus, in each
group studied, including corresponding controls, at
least 15 photographstregion were available for evaluation. Micrographs were enlarged to 30,000 x , the epithelial lamina densa was measured using a digitizing
tablet and a trapezoidal algorithm area program, and
the number of gold particles per area lamina densa was
determined. The basal lamina was clearly discernable
at this magnification, and since the lamina densa widened after alcohol exposure, area provided a more accurate depiction of any changes than did length. A minimum of 400 gold counts was obtained per region of
each treatment group in order to limit error (Loud,
1987). Since the groups of animals sacrificed at 5 and
60 min following ethanol exposure were processed on
different days, data were normalized such t h a t SIS controls within a region were equivalent. The mean gold
countslarea of lamina densa were statisticalIy analyzed
using analysis of variance and Student-NewmanKeuls’ test (Keuls, 1952; Newman, 1939) for multiple
comparisons. A P value < .05 was taken to represent
statistical significance.
RESULTS
Transmission Electron Microscopy
In control tissues, TEM revealed a n intact gastric
epithelium resting upon a thin and continuous basal
lamina (Fig. 1).Following exposure to either 50% eth-
237
rowheads) has become flocculent and thickened, and the remaining
cells are necrotic. Bar = 0.5 km.
Fig.3.TEM of the interfoveolar gastric mucosa after treatment with
100% ethanol following prostaglandin pretreatment. Note the pleated
lamina densa (arrowheads) and exfoliated but otherwise normal-appearing surface mucous cells (El. Bar = 0.5 pm.
anol or 100% ethanol, superficial mucosal injury was
characterized by epithelial cell death and exfoliation.
The lamina rara could no Ionger be identified in regions of epithelial cell exfoliation and the lamina densa
often appeared thickened and flocculent (Fig. 2). As
reported previously, tissue damage was time and alcohol concentration-dependent such that the most severe
alcohol damage was observed in animals sacrificed 60
min after 100% ethanol exposure, and the least severe
at 5 min after 50% ethanol exposure (Schmidt and
Miller, 1988a).
Tissues pretreated with PGE, prior to oral saline appeared normal in every way when compared to controls. Exposure to PGE, prior to ethanol also resulted
in morphology relatively comparable to controls, except that damaged surface mucous cells were exfoliated, as noted in the salinelethano1 combinations. Cell
sloughing left behind a pleated but predominantly normal basal lamina, which maintained its finely particulate TEM appearance (Fig. 3).
lmmunogold Electron Microscopy
Under control conditions, paraformaIdehyde fixation
provided satisfactory IG binding to type IV collagen
although a loss of ultrastructural detail was noted.
Nevertheless, general morphological characteristics of
the various groups were similar to those noted in the
TEM study. Binding was non-specific and minimal
when the primary antibody was omitted or preimmune
serum substituted. In control animals, binding to type
IV collagen, shown by the presence of electron-dense
gold particles, was specifically localized to the basal
lamina, mainly the lamina densa (Fig. 4).Following
238
K.S. RIGHTOR ET AL.
Figs. 4-8
239
ALCOHOL, PG, AND GASTRIC EXTRACELLULAR MATRIX
treatment with 50% ethanol or 100% ethanol, gold particles were localized to a thickened and flocculent lamina densa (Figs. 5, 6, respectively). Gold label was
clearly evident in the lamina densa from tissues exposed to PGE, prior to either concentration of ethanol
(Figs. 7, 8). Background measurements of non-specifically bound gold particles among treatment groups
were not significantly different ( P < .05).
No significant differences were found in IG binding
between groups receiving the same treatment at either
5 or 60 min post alcohol exposure. For this reason, the
data were combined and are shown in subsequent figures. Control tissues exhibited significantly higher
binding in the interfoveolar zone compared with pit
and gland regions (Fig. 9a). For the S/50% ethanol
group, IG binding was also higher in the interfoveolar
regions than in either pit or gland zones. However,
compared with controls, S/50%ethanol treated animals
demonstrated decreased IG binding in all three regions
(Fig. 9a-c). Irnmunogold binding was further decreased
in S/lOO% ethanol treated tissues, being significantly
lower in all regions when compared with S/Sand S/50%
ethanol treated groups (Fig. 9a-c).
Regardless of the mucosal region evaluated, IG binding in PGE, pretreated controls was comparable to that
found in saline-exposed controls (Fig. 10a). Moreover,
in PGE,/ethanol-treated tissues, IG binding was increased significantly over IG binding in comparable
ethanol-treated tissues without PGE, pretreatment
(Fig. 10b,c). Similar to S-pretreated tissues, PGE,-pretreated tissue revealed significantly higher binding in
the interfoveolar zone as compared to the pit and gland
zones for S/50% ethanol-treated tissues, although no
significant differences among these three regions were
noted in 100% ethanol tissues. Prostaglandin in combination with 50% ethanol elevated the degree of gold
binding to or above control levels. In PGE2/100% ethanol-treated tissues, however, the mucosa was sufficiently damaged such that even though IG binding levels were increased over S/lOO% ethanol levels, they
nevertheless remained a t a significantly low level
when compared with corresponding controls.
Fig. 4.TEM of normal (5 min salineisaline treated) gastric tissue
after immunogold localization of type IV collagen. The protein-A gold
has localized specifically to the epithelial basal lamina, primarily the
lamina densa (arrowheads). The basal lamina of the vessel (V) also
shows specificity to type IV collagen. Bar = 0.5 pm.
Fig. 5. Type IV collagen binding in the mid-pit region of 5 rnin
saline/50% ethanol-treated tissue. Binding per area basal lamina is
reduced compared with control levels (Fig. 4);the lamina densa (arrowheads) is now thickened. Bar = 0.5 pm.
Fig. 6.Extensively diminished type IV collagen immunolabeling in
the mid-pit region of 5 min saline/100% ethanol-treated gastric mucosa. The lamina densa (arrowheads) is flocculent. Bar = 0.5 pm.
Fig. 7.Five minute prostaglandini50% ethanol-treated mid-pit tissue showing an elevated degree of type IV collagen binding per area
basal lamina (arrowheads) when compared with controls (Fig. 4). Bar
= 0.5 pm.
Fig. 8.This 5 rnin prostaglandin/lOO% ethanol-treated gastric mucosa shows increased type IV immunogold binding (arrowheads) when
compared to saline/lOO% tissues (Fig. 5) but has not increased to the
levels of salineisaline mucosa (Fig. 4). Bar = 0.5 pm.
71
I
T
Region: a. lnterfoveolar
Treatment:
b. Pit
c. Gland
0 Saline 13 50% Ethanol I
3100% Ethanol
Fig. 9. Quantitative analysis of immunogold binding to type IV
collagen in the gastric mucosal interfoveolar (a),pit (b) and gland (c)
regions. Tissues were pretreated with saline followed by treatment
with either saline, 50% ethanol, or 100% ethanol. *Groups within a
region are significantly different from controls ( P < ,051. Data represents combined results of mucosa retrieved at 5 and 60 min post alcohol or saline exposure. N represents at least 30 observationsigroup.
DISCUSSION
The present study has shown that acute alcohol exposure decreases IG binding to type IV collagen in epithelial basal lamina of rat gastric mucosa. In addition,
we found that this diminution in IG binding became
more pronounced as the concentration of ethanol increased and that, regardless of the alcohol concentration employed, the reduction in IG binding occurred
within 5 min after alcohol administration and remained unchanged 1h later. Finally, although the interfoveolar region, as compared to the pit and gland
isthmus zones, showed substantial IG binding after saline exposure that was not significantly different following 50% ethanol exposure, all three zones exhibited
very low binding after exposure to absolute alcohol.
The decrease in binding of immunogold to type IV
collagen following exposure to alcohol is a finding
which may be influenced by a number of factors. Degree of binding may reflect changes either in the number of antibody binding sites available or the binding
affinity of these sites. The number of binding sites
available is affected by the overall number of type IV
collagen molecules present, the number of binding sites
occupied by other associated molecules, and the availability of sites due to the molecular conformation of
type IV collagen. That ethanol may have directly compromised type IV collagen molecules and thereby reduced binding capacity is supported by the fact that
alcohol is known to denature proteins. However, basal
laminae and most of the collagen IV in tissues are
known to be relatively insoluble due to stabilization by
disulfide bridges and nonreducible crosslinking (Timpl
and Dziadek, 1986).Other unknown factors could also
contribute to the observed loss in binding capacity. For
example, Nakanashi and coworkers (1987) noted t h a t
other damaging agents such a s acetic acid or taurocholic acid induce the gastric mucosa to release a metalloproteinase which denatures type IV collagen.
Whether alcohol can render a similar effect remains to
be determined. Yet another possibility is that antigenic
sites on type IV collagen molecules are influenced by
neighboring matrix macromolecules and that exposure
240
K.S. RIGHTOR ET AL.
10
a Oral Saline Treatment
b. 50%EK)H Treatment
7
0
Region: I n t m a r
Pit
Gland
Intetfoveolar
Pt
Gland
c. 100% EtOH Treatment
lnterfowolar
pit
Gland
SubcutaneousPretreatment: rSaline mPGE,
Fig. 10. Quantitative analysis comparing pretreatment with saline and 16,16dimethyl prostaglandin
E, (PGE,). Oral treatment was either saline (a),50% ethanol (b), or 100% ethanol ( c ) .*Groups within
a particular regiodtreatment combination are significantly different from controls (P < .05). Data
represent combined results of mucosa retrieved a t 5 and 60 min post alcohol exposure, as well as
corresponding controls. N represents a t least 30 observationsigroup.
to ethanol may confound such relationships resulting
in a loss in binding capability. Little and Chen (1982),
in a study using embryonic chicken heart fibroblasts,
found a loss in collagedprocollagen antigen as the extracellular matrix matured. After gentle proteolysis,
the antigenicity increased suggesting that during development antigenic sites were gradually masked by
other matrix components. Whether alcohol can inf luence neighboring matrix molecules in ways which reduce the binding capacity of type IV collagen remains
to be explored.
In addition to the impact of ethanol exposure, both
fixation and processing are known to affect the antigenicity of type IV collagen (Bendayan, 1985; Priestly,
1984). However, because tissue preparation and processing were standardized for all specimens in the
present study, such techniques should have no influence on interpreting our data since we were more concerned with actual changes in binding properties due
to experimental manipulation rather than absolute
numbers of binding sites. In support of this assumption, Grant and Leblond (1988) have utilized type IV
collagen’s immunolabeling densities to perform quantitative comparisons of thick and thin basement membranes, and found fixation and processing techniques
not to be a problem in interpreting results.
Although the cause of the concentration-dependent
decrease in binding of gold to type IV collagen following alcohol exposure is unknown, the higher concentration may either have destroyed the entire molecule
or changed its structural conformation enough to induce further loss of antibody binding. Schmidt and
Miller (1988a) have shown that even though severe
surface damage occurs with both 50% ethanol and
100% ethanol, damage which includes the gastric
gland cells is significantly higher after treatment with
absolute alcohol. Such effects upon the depth of epithelial injury may also appear in alterations of the molecular components of the extracellular matrix.
The finding that the degree of IG binding was higher
in the interfoveolar zone in control stomachs and in
those following 50% ethanol exposure but not after
100% ethanol warrants consideration. Increased binding in the interfoveolar region when compared with the
pit and gland regions suggests the possibility of a physiologically important difference in the type IV collagen
of this region of the gastric mucosa. Concentration gradients of other basal lamina molecules, including laminin, have been shown to regulate the migratory behavior of cells in culture (Lacovara et al., 1984; McCarthy
and Furcht, 1984). It is conceivable that the migration
behavior of surface mucous cells toward the luminal
surface may be influenced by concentration gradients
of basal lamina molecules. The lack of regional differences in IG binding following absolute alcohol compared with the interfoveolar differences observed in saline and 50% ethanol groups appears to be accounted
for by the devastating effect of this highly potent damaging agent. Thus, following 100% ethanol, the collagen binding sites of the interfoveolar region presumably were rendered indistinguishable from type IV
collagen lying deeper in the pit and gland.
The lowered binding affinity exhibited by type IV
collagen following exposure to either concentration of
ethanol indicates that type IV collagen has been altered, possibly making the basal lamina less viable a s
a substrate for restitution. In 1974, Vracko examined
epithelial cell migration, and found that if the basal
lamina is destroyed or disrupted, orderly cell replacement does not occur. The importance of a n intact basal
lamina in restitution of gastric mucosa was first suggested by Morris and Wallace (1981). Using a n ex vivo
chamber preparation of rat gastric mucosa, these investigators exposed their tissue preparation to 40%
ethanol with and without 50 mM HCl. Damage induced
by ethanol administered in a n acid medium brought
about the rapid conversion of ethanol-induced injury
into hemorrhagic erosions, severely destroying the
ALCOHOL, PG, AND GASTRIC EXTRACELLULAR MATRIX
basal lamina in hyperemic areas, whereas ethanol
given alone caused widespread exfoliation of the surface epithelium while sparing the basal lamina. These
investigators concluded that migration of cells from the
gastric pit to re-establish epithelial integrity, now
known as restitution, might require the presence of a n
intact basal lamina, and that its destruction is likely to
impair this response. Subsequently Lacy and Ito
(1984a; Ito and Lacy, 1985) reported that exposing gastric mucosa to absolute alcohol for 30-40 s left a
pleated but otherwise normal basal lamina as a scaffold for rapid repair. Recently, Schmidt and associates
(1986; Schmidt and Miller, 1988b) studied restitution
between 5 min and 24 h after acute alcohol injury by
100% ethanol and reported that the ultrastructure of
basal laminae was markedly changed in tissues with
necrotic and hemorrhagic injury. The basal lamina appeared to be directly altered by alcohol because it was
flocculent and thickened when compared with controls.
Furthermore, hemorrhage confined to the lamina propria was associated with rupture of the vascular basal
lamina. However, extravasation of plasma and cellular
elements into the gastric lumen also required breakage
of the basal lamina underlying the gastric epithelium.
Lastly, several investigators (Ito and Lacy, 1985; Morris and Wallace, 1981; Schmidt and Miller, 1988b;
Svanes et al., 1982) reported that restitution was not
observed in regions of gastric mucosa where the basal
lamina had been destroyed following exposure to ethanol. These authors postulated that a damaged basal
lamina was a primary defect in preventing restitution.
In recent years, the use of prostaglandins as protective substances against the effects of various noxious
agents on gastric mucosa has been observed and has
generated considerable controversy with respect to the
extent of this protection. Several thorough histologic
studies have helped define the limits of such protection.
For example, when tissues are evaluated macroscopically, the hemorrhagic and necrotic lesions induced by
certain necrotizing chemicals such as ethanol are virtually eliminated in gastric mucosa in which prostaglandin pretreatment has been rendered (Robert et al.,
1979). When these tissues are assessed microscopically,
however, Lacy and Ito (1982) and others (Schmidt et
al., 1985, 1986; Tarnawski et al., 1985) have shown
t h a t although prostaglandins reduce the depth of gastric mucosal injury which accounts for the elimination
of the hemorrhagic and necrotic lesions following exposure to high concentrations of ethanol, tissue injury
still occurs, but is limited to the superficial layers of
the epithelium. In addition, we have shown that in
ethanol-exposed mucosa prostaglandin pretreatment
also preserves much of the ultrastructure of the connective tissue of the lamina propria including the basal
lamina (Schmidt and Miller, 1988a,b). In the present
study, PGE, was effective in protecting the molecular
integrity of type IV collagen, as measured by IG binding, if administered prior to ethanol. This effect provided complete protection from 50% ethanol and partial protection from 100% ethanol. In fact, IG binding
in the interfoveolar region of mucosa exposed to 50%
ethanol and pretreated with PGE, was actually increased compared with PGE,/saline controls (see Fig.
10). One possible explanation for this finding might be
a differential sensitivity to ethanol of molecules asso-
241
ciated with the basal lamina. As cited earlier, matrix
components other than collagen have been shown to
mask collagen/procollagen antigenic binding sites in a
study of embryonic chicken heart fibroblasts (Little
and Chen, 1982). If molecules other than type IV collagen occupy some of the IG binding sites and are more
sensitive to the effects of ethanol than is the collagen
molecule, then a low concentration of ethanol could
cause a release of these molecules before the binding
sites are disrupted. This release would provide more
binding sites and result in higher IG binding. Such a
circumstance could have occurred in mucosa exposed to
50% ethanol and pretreated with PGE,. If this explanation is correct, i t would also explain why PGE, did
not increase IG binding in those animals receiving oral
saline instead of alcohol.
The mechanism underlying the ability of PGE, to
prevent or markedly attenuate the adverse effects of
ethanol on IG binding is unknown. One possibility is
that the PGE, analogue may act directly upon type IV
collagen, rendering i t impervious to the effects of ethanol. Another possibility is that PGE, may retard the
amount of alcohol which reaches the basal lamina by
barrier effects, or may dilute the concentration of the
alcohol by vascular permeability, and thereby lessen
its toxic effects. Yet a third possibility is that alcohol
induced the formation of oxygen radicals which in t u r n
caused the injury to the basal lamina, and PGE, either
prevented their formation or adequately scavenged
them so that they were no longer deleterious. Evidence
does exist suggesting that ethanol injury to the gastric
mucosa may involve oxygen radicals (Pihan et al.,
19871, and prostaglandins have been shown to inhibit
radical formation (Simpkins et al., 1986). Whether
such a mechanism was responsible for our findings will
require further study. Regardless of the underlying
mechanism(s1, PGE, appears to influence not only epithelial cells but extracellular molecules as well.
In summary, we have shown that ethanol in a concentration-dependent fashion significantly inhibits the
ability of type IV collagen to bind to IG particles, indicating loss of molecular integrity. Such ethanol-induced alterations of a major component of basal laminae may be a n important basis for inhibited restitution
following alcohol injury. Interestingly, PGE, was able
to prevent or markedly attenuate these adverse effects
on IG binding which correlated with its ability to render histologic protection to the deep epithelium. To further clarify the role of the basal lamina in injury and
repair, future studies need to be directed toward the
effects of ethanol on other component macromolecules
of the gastric basal lamina during periods of restitution
and healing and what influence prostaglandins have
on such effects.
ACKNOWLEDGMENTS
Tfiis work was supported by NIAAA grant AA06887
to Dr. Schmidt, and NIH grant DK25838 to Dr. Miller.
LITERATURE CITED
Bendayan, M. 1984 Protein A-gold electron microscopic immunocytochemistry: Methods, Applications, and Limitations. J. Electron
Microsc. Tech., lt243-270.
Bendayan, M. 1985 The enzyme-gold technique: a new cytochemical
approach for the ultrastructural localization of macromolecules.
242
K.S. RIGHTOR ET AL.
genic determinants in proteins coincide with surface region acIn: Techniques in Immunocytochemistry, Vol. 3. G.R. Bullock and
cessible to large probes (antibody domains). PNAS, 83:226-230.
P. Petrusz, eds. Academic Press, Inc., London, pp. 179-236.
Pihan, G., C. Regillo, and S. Szabo 1987 Free radicals and lipid perBlack, B.A., G.P. Morris, and J.L. Wallace 1985 Effects of acid on the
oxidation in ethanol- or aspirin-induced gastric mucosal injury.
basal lamina of the rat stomach and duodenum. Virchows Archiv
Dig. Dis. Sci., 32:1395-1401.
[Cell Pathol.], 50:109-118.
Carlemalm, E., R.M. Garavito, and W. Villiger 1982 Resin develop- Priestly, J.V. 1984 Pre-embedding ultrastructural immunocytochemistry: immunoenzyme techniques. In: Immunolabelling for Elecment for electron microscopy and an analysis of embedding at low
tron Microscopy. J.M. Polak and M. Varndell, eds. Elsevier Scitemperature. J . Microsc., 126t123-143.
ence Publishers, Amsterdam, pp. 37-52.
Chen, J.M., and C.D. Little 1987 Cellular events associated with lung
Robert, A., J.E. Nezamis, C. Lancaster, and A.J. Hanchar 1979 Cytobranching morphogenesis including the deposition of collagen
protection by prostaglandins in rats. Gastroenterology, 77:761type IV. Dev. Biol., 12Ot311-321.
767.
Faust, T.W., E. Lee, J.S. Redfern, and M. Feldman 1989 Effects of
Rowe, P.H., M.J. Starlinger, E. Kasdon, M.J. Hollands, and W. Silen
prostaglandin F, alpha on gastric mucosal injury by ethanol in
1987 Parenteral aspirin and sodium salicylate are equally injurats: comparison with prostaglandin F, alpha. Prostaglandins,
rious to the rat gastric mucosa. Gastroenterology, 93:863-871.
37:493-504.
Fitch, J.M., E. Gibney, R.D. Sanderson, R. Mayne, and T.F. Linsen- Rutten, M.J., and S. Ito 1983 Morphology and electrophysiology of
guinea pig gastric mucosal repair in vitro. Am. J. Physiol., 244:
mayer 1982 Domain and basement membrane specificity of a
G1714182.
monoclonal antibody against chicken type IV collagen. J . Cell
Ruwart, M.J., and B.D. Rush 1984 Prostacyclin inhibits gastric empBiol., 95t641-647.
tying and small-intestinal transit in rats and dogs. GastroenterGrant, D.S., and C.P. Leblond 1988 Immunogold quantitation of lamiology, 87t392-395.
nin, type IV collagen, and heparan sulfate proteoglycan in a vaSchmidt, K.L., R.L. Bellard, G.S. Smith, J.M. Henagan, and T.A.
riety of basement membranes. J. Histochem. Cytochem., 36t271Miller 1986 Influence of prostaglandin on repair of rat stomach
283.
damaged by absolute ethanol. J . Surg. Res., 41r367-377.
Handley, D.A., and B.R. Olsen 1979 Butvar B-98 as a thin support
Schmidt,
K.L., J.M. Henagan, G.S. Smith, P.J. Hilburn, and T.A.
film. Ultramicroscopy, 4:479-480.
Miller 1985 Prostaglandin cytoprotection against ethanol-inIto, S., and E.R. Lacy 1984 Characteristics of ethanol-induced lesions
duced gastric injury in the rat: a histologic and cytologic study.
in rat gastric mucosa. In: Mechanisms of Mucosal Protection in
Gastroenterology, 88:649-659.
the Upper Gastrointestinal Tract. A. Allen et al., eds. Raven
Schmidt, K.L., and T.A. Miller 1988a Morphological characteristics of
Press, New York, pp. 57-63.
prostaglandin cytoprotection. Toxicol. Pathol., 16r223-236.
Ito, S., and E.R. Lacy 1985 Morphology of rat gastric mucosal damage
and restitution in the presence of luminal ethanol. Gastroenter- Schmidt, K.L., and T.A. Miller 1988b Ultrastructural aspects of prostaglandin cytoprotection in an alcohol injury model. J . Clin. Gasology, 88t250-260.
troenterol., 10.584-592.
Karnovsky, M.J. 1965 A formaldehyde-glutaraldehyde fixative of
Silen, W., and S. Ito 1985 Mechanisms for rapid re-epithelialization of
high osmolality for use in electron microscopy. J . Cell Biol., 27:
the gastric mucosal surface. Annu. Rev. Physiol., 47t217-229.
137A,138A.
Karnovsky, M.J. 1971 Use of ferrocyanide-reduced osmium tetroxide Simpkins, C.O., S.T. Alailima, E.A. Tate, and M. Johnson 1986 The
effect of enkaphalins and prostaglandins on 0, release by neuin electron microscopy. J . Cell Biol., 51:146A.
trophils. J . Surg. Res., 41r645-652.
Keuls, M. 1952 The use of the “studentized range” in connection with
Smith, K.B., and K.L. Schmidt 1987 Protein A-gold localization of
a n analysis of variance. Euphytica, 1:112-122.
type IV collagen in gastric mucosa. Tex. SOC.Elec. Micros. J., 18r2
Lacovara, J., E.B. Cramer, and J.P. Quigley 1984 Fibronectin en(Abstract).
hancement of directed migration of B16 melanoma cells. Cancer
Spurr, A.R. 1969 A low-viscosity epoxy resin embedding medium for
Res., 44t1657-1663.
electron microscopy. J . Ultrastruct. Res., 26r31-43.
Lacy, E.R. 1985 Prostaglandins and histological changes in the gasStephens, H., M. Bendayan, and M. Silver 1982 Immunocytochemical
tric mucosa. Dig. Dis. Sci., 30t83S94S.
localization of collagen types and laminin in skeletal muscle with
Lacy, E.R. 1987 Gastric mucosal defense after superficial injury. Clin.
the protein A-gold technique. Biol. Cell, 44:81-84.
Invest. Med., IOt189-200.
Lacy, E.R. 1988 Epithelial restitution in the gastrointestinal tract. J. Svanes, K., S. Ito, K. Takeuchi, and W. Silen 1982 Restitution of the
surface epithelium of in vitro frog gastric mucosa after damage
Clin. Gastroenterol., IO(Supp1. l):S72S77.
with hyperosmolar sodium chloride: morphologic and physiologic
Lacy, E.R., and S. Ito 1982 Microscopic analysis of ethanol damage to
characteristics. Gastroenterology, 82t1409-1426.
rat gastric mucosa after treatment with a prostaglandin. GastroSzabo, S. 1987 Mechanisms of mucosal injury in the stomach and
enterology, 83t619-625.
duodenum: time-sequence analysis of morphologic, functional,
Lacy, E.R., and S. Ito 1984a Ethanol-induced insult to the superficial
biochemical and histochemical studies. Scand. J . Gastroenterol.,
rat gastric epithelium: A study of damage and rapid repair. In:
127St21-28.
Mechanisms of Mucosal Protection in the Upper Gastrointestinal
Takeuchi, K., 0. Furukawa, and S. Okabe 1986 Induction of duodenal
Tract. A. Allen, et al., eds. Raven Press, New York, pp. 49-56.
ulcers in rats under water-immersion stress conditions. Influence
Lacy, E.R., and S. It0 1984b Rapid epithelial restitution of the rat
of stress on gastric acid and duodenal alkaline secretion. Gastrogastric mucosa after ethanol injury. Lab. Invest., 51.573-583.
enterology, 91:554-563.
Laurie, G.W., C.P. Leblond, I. Cournil, and G.R. Martin 1980 Immunohistochemical evidence for the intracellular formation of base- Takeuchi, K., H. Nishiwaki, N. Hara, and S. Okabe 1988 Effects of
gastric distension and prostaglandin on acid ethanol-induced mument membrane collagen (type IV) in developing tissues. J . Hiscosal lesions in the rat. Dig. Dis. Sci., 33r1569-1577.
tochem. Cytochem., 28t1267-1274.
Tarnawski, A,, D. Holander, and J . Stachura 1985 Ultrastructural
Little, C.D., and W.-T. Chen 1982 Masking of extracellular collagen
changes in the gastric mucosal microvessels after ethanol. Gasand the co-distribution of collagen and fibronectin during matrix
troenterol. Clin. Biol., 9t93-97.
formation by cultured fibroblasts. J . Cell Sci., 55:35-50.
Loud, A.V. 1987 Electron Microscopic Morphometry. Anal. Quant. Thorning, D., and R. Vracko 1977 Renal glomerular basal lamina
scaffold: embryologic development, anatomy, and role in cellular
Cytol. Histol., 9:7-12.
reconstruction of rat glomeruli injured by freezing and thawing.
Martinez-Hernandez, A. 1987 Methods for electron immunohistoLab. Invest., 37t105-119.
chemistry. Methods Enzymol., 145t103-133.
McCarthy, J.B., and L.T. Furcht 1984 Laminin and fibronectin pro- Timpl, R., and M. Dziadek 1986 Structure, development, and molecular pathology
mote the haptotactic migration of B16 mouse melanoma cells in
-. of basement membranes. Int. Rev. Exp. Pathol.,
29:l-112.
uitro. J. Cell Biol., 98r1474-1480.
Timpl, R., and G. Martin 1982 Components of basement membranes.
Morris, G.P., and J.L. Wallace 1981 The roles of ethanol and of acid in
In: Immunochemistry of the Extracellular Matrix, Vol. 11. H.
the production of gastric mucosal erosions in rats. Virchows Arch.
Furthmayr, ed. CRC Press, Boca Raton, Florida, pp. 119-151.
[Cell Pathol.], 38r23-28.
Vracko, R. 1974 Basal lamina scaffold: anatomy and significance for
Nakanashi, J., S. Murakami, Y. Ishikura, S. Tsuchiya, and Y. Mori
maintenance of orderly tissue structure: a review. Am. J. Pathol.,
1987 Production of metalloproteinase in gastric tissue with acetic
77,314-338.
acid-induced ulcer or taurocholic acid-induced gastritis. J . PharWang, J.Y., S. Yamasaki, K. Takeuchi, and S. Okabe 1989 Delayed
macobiodyn., IOt92-97.
healing of acetic acid-induced gastric ulcers in rats by indomethNewman, D. 1939 The distribution of range in samples from a normal
acin. Gastroenterology, 96t393-402.
population, expressed in terms of an independent estimate of
Yoemans, N.D., D.J.B. St. John, and W.G.R.M. deBoer 1973 Regenstandard deviation. Biometrika, 31t20-30.
eration of gastric mucosa after aspirin-induced injury in the rat.
Novotny, J., M. Hanaschumacher, E. Haber, R.E. Bruccoleri, W.B.
Am. J. Dig. Dis., 18t773-780.
Carlson, D.W. Fanning, J.A. Smith, and G.D. Rose 1986 Anti-
Документ
Категория
Без категории
Просмотров
0
Размер файла
1 074 Кб
Теги
matrix, colloidal, prostaglandin, extracellular, gold, localization, typed, rat, gastric, alcohol, mucosainfluence, collagen
1/--страниц
Пожаловаться на содержимое документа