 |
INTRODUCTION |
Clostridium perfringens is a Gram-positive,
endospore-forming, anaerobic bacterium that produces a plethora of
protein toxins, including a 35-kDa single polypeptide named C. perfringens enterotoxin (CPE).1 Considerable
experimental and epidemiologic evidence (1-3) now implicates CPE as
the virulence factor responsible for the diarrheal and cramping
symptoms of several important human gastrointestinal illnesses, which
include C. perfringens type A food poisoning, the second
most commonly reported foodborne disease in the United States, and
non-foodborne diarrheal illnesses, such as antibiotic-associated diarrhea and sporadic diarrhea.
CPE appears to induce its gastrointestinal effects (at least in large
part) through a multi-step, cytotoxic action that initiates when the
toxin binds to one or more protein receptors (4-7). Binding of CPE to
its receptor(s) results in the formation of a small (~90 kDa)
CPE-containing complex in the plasma membrane of sensitive mammalian
cells (4). The small complex then apparently associates with one or
more additional eucaryotic plasma membrane proteins, forming a large
(~160 kDa) complex (4, 8-10). Formation of large complex causes the
development of massive plasma membrane permeability alterations (5, 6,
11, 13), which collapse the cellular-osmotic equilibrium (14) and
trigger cell death (15, 16). In the intestines, CPE-induced death of
enterocytes produces histologic damage (17, 18), which appears to be
largely responsible for the onset of the intestinal fluid and
electrolyte transport disturbances that clinically manifest as diarrhea
during CPE-associated gastrointestinal illness (17, 18).
Several independent observations have strongly suggested that large
complex formation plays a central role in the CPE cytotoxic pathway.
For example, the failure of CPE to induce cytotoxic effects at 4 °C
has been ascribed to a specific blockage in large complex formation at
low temperatures (10). Second, a close correlation has been
demonstrated between the cytotoxic activity and large complex forming
ability of various CPE deletion fragments (19). The importance of large
complex formation for CPE action has also received support from recent
random mutagenesis studies (20), which showed that introduction of
several point mutations into the region of native CPE containing amino
acids 45-116 produced a strong inhibition or elimination of cytotoxic
activity. The reduced (or absent) cytotoxic activity of those CPE point
mutants was found to specifically correlate with a sharp reduction, if not total inhibition, in the large complex forming ability of each CPE mutant.
The availability of strong evidence supporting the importance of large
complex in CPE-induced pathophysiology implies that elucidating large
complex formation is necessary for fully understanding CPE action.
Although clarifying the process of large complex formation requires the
identification of all eucaryotic protein constituents of CPE-containing
large complex, only limited information is currently available
concerning the composition of large complex. Affinity chromatography
experiments (9, 21, 22) have suggested that eucaryotic plasma membrane
proteins of ~45-50 and ~65-70 kDa are present in large complex.
More recently, expression cloning experiments have suggested that
certain claudins that can function as CPE receptors (8, 23-25) may
also be present in large complex.
The recent association of claudin(s) with CPE-containing large complex
potentially provides some new insights into the eucaryotic protein
constituents of large complex. For example, it has been reported (26)
that claudins, which are believed to primarily localize to epithelial
TJs, can interact with occludin, which is an ~65-kDa protein that
also localizes to TJs. Reports of associations between claudins and
occludin, coupled with affinity chromatography results suggesting that
an ~65-70 kDa protein is present in large complex, led us to
hypothesize that occludin may be an eucaryotic component of
CPE-containing large complex. However, when that hypothesis was tested
in the current study, surprising results were obtained; several
heterogeneous species of CPE-containing large complex were identified,
only one of which contains occludin.
| |
EXPERIMENTAL PROCEDURES |
Materials
Previously described methods (27) were used to purify native CPE
to homogeneity from C. perfringens strain NCTC 8239 and to
assay the biological activity of the resultant purified toxin. Aliquots
(2 mg) of the purified native CPE were radioiodinated as described
previously (5), using lactoperoxidase-glucose oxidase (Bio-Rad) and 2 mCi of Na125I (17 mCi/mg; ICN radiochemicals). Using
previously described assays (5), the radiolabeled CPE preparation was
determined to retain binding and cytotoxic activity.
Rabbit polyclonal antibody raised against a fusion protein consisting
of the C-terminal 150 amino acids of human occludin fused to
glutathionine S-transferase was purchased from
Zymed Laboratories Inc. NRS IgG was purchased from
Sigma. Acrylic beads containing immobilized protein A were purchased
from Sigma.
Cell Cultures
CaCo-2 human intestinal carcinoma cells were routinely
maintained at 37 °C in minimal essential Eagle's medium (Sigma)
containing 10% fetal bovine serum (Life Technologies), 1% minimal
essential medium nonessential amino acids (Sigma), 100 units/ml
penicillin, and 100 µg/ml streptomycin.
Rat-1/R12 cells used in these experiments were obtained from the
American Type Culture Collection (CRL-2210, Manassas, VA). Rat-1/R12
cells are Rat-1 fibroblast cells stably transfected with pUHD15-1neo,
which contains a tetracycline transactivator gene and a neomycin
resistance gene. These cells are tet-off cells, i.e.
expression of the transactivator gene is repressed by tetracycline or
doxycycline. Rat-1/R12 cells do not naturally express occludin, as
determined by Western blot analysis (see "Results").
To prepare Rat-1/R12 cells expressing occludin, the occludin cDNA
insert from the pCB6 occludin vector, prepared previously (28), was
subcloned into the EcoRI/XbaI sites on the pTRE
vector (CLONTECH). That pTRE construct, named
pTRE-occ, also encodes the addition of an 11-amino acid VSV-G tag to
the cytoplasmic C terminus of occludin. The pTRE-occ vector, along with
the pTKhyg hygromycin resistance vector (CLONTECH),
was then transfected into Rat-1/R12 cells (at a ratio of 20:1 pTRE-occ
to pTKhyg) using LipofectAMINE (Life Technologies, Inc.). Stable
transfectants were selected in 200 µg/ml G418 and 200 µg/ml
hygromycin. A clonal cell line, named FL-22, which showed lower basal
levels of occludin expression in the presence of doxycycline (40 ng/ml)
but greater occludin expression in the absence of doxycycline, was
selected for further culture. Both the parental Rat-1/R12 cells and the FL-22 cells were routinely grown in Dulbecco's modified Eagle's medium (Sigma), with 10% tet-off certified system fetal bovine serum
(CLONTECH).
Immunofluoresence Localization of Occludin in FL-22 Cells
Immunofluorescence analysis was performed to determine where
occludin localizes in FL-22 cells. Taking advantage of the fact that
the occludin expressed by FL-22 cells is tagged at its C terminus with
a VSV-G tag, occludin localization in FL-22 cells was evaluated using
monoclonal antibody P5D4, an anti-VSV-G monoclonal antibody described
previously (29), which was kindly supplied by Dr. Thomas Kreis
(University of Geneva, Switzerland). After treatment with monoclonal
antibody P5D4, FL-22 cells were reacted with affinity-purified,
species-specific Texas Red anti-mouse IgG secondary antibodies (diluted
1:100), which were purchased from Jackson Immunoresearch. Fixation,
permeabilization, quenching, and image capture was performed as
described previously (29).
ZO-1 localization in FL-22 cells was used as a marker for identifying
cell contact regions (30). FL-22 cells were subjected to
immunofluorescence analysis using a ZO-1 primary antibody (1:350 dilution) purchased from Zymed Laboratories Inc. The
secondary antibody used to detect binding of ZO-1 antibodies was an
affinity-purified, species-specific, fluroscein
isothiocyanate-conjugated donkey anti-mouse IgG (Jackson
Immunoresearch), used at a 1:100 dilution.
Large Complex Western Immunoblots
Confluent (5-8 days old) CaCo-2 cells were removed from tissue
culture flasks by gentle scraping. After washing with warm PBS (140 mM NaCl, 9 mM Na2HPO4,
1.5 mM KH2PO4, pH 7.4) containing 0.9 mM CaCl2 and 0.5 mM
MgCl2, an aliquot (2 × 106) of the washed
CaCo-2 cells was suspended in 200 µl of PBS containing Ca2+ and Mg2+. CPE (0.5 µg) was then added,
and the resultant mixtures were incubated (with gentle shaking) at
37 °C for desired time periods. After incubation, the cells were
washed with PBS containing Ca2+ and Mg2+,
treated with 10 µl of DNase (1 mg/ml stock; Roche Molecular Biochemicals), extracted with 100 µl of 2× SDS sample buffer without 
-mercaptoethanol (4) at room temperature for 10 min, and analyzed by
SDS-PAGE Western blots (see below). In some experiments (see Fig. 3),
CaCo-2 cells were prepared and treated with CPE as described above,
except that unbound CPE was removed from these cultures after 10 min.
Those CPE-treated cultures were then washed with warm PBS containing
Ca2+ and Mg2+ and further incubated in CPE-free
minimal essential Eagle's medium at 37 °C for an additional 10-80 min.
Samples containing extracted large complex were electrophoresed
(without sample boiling) at either 30 mA for 3 h on 6%
polyacrylamide gels containing SDS or at 4 mA overnight on 4%
polyacrylamide gels containing SDS. The separated proteins on those
gels were then electrotransferred onto a nitrocellulose membrane,
blocked with Blotto (5% powdered milk in Tris-buffered saline, pH
7.5), and incubated with either affinity-purified anti-CPE rabbit
polyclonal IgG, prepared as described (20), or with affinity-purified
polyclonal rabbit antibodies raised against the C-terminal region of
human occludin (Zymed Laboratories Inc.). The
immunoblots were then treated with a 1:10,000 dilution of goat
anti-rabbit IgG-horseradish peroxidase conjugate (Sigma) and developed
with chemiluminescent substrate (Pierce), as described previously
(20).
Isolation of Large Complexes by Preparative Electrophoresis
Scraped CaCo-2 cells (~5 × 107) in 5 ml of
PBS containing Ca2+ and Mg2+ were incubated
with 12.5 µg of CPE for 20 min at 37 °C to form large complex (9,
10). Control CaCo-2 cells were prepared similarly, except that they
were incubated without CPE. Both the CPE-treated and control CaCo-2
cell samples were then extracted for 10 min at room temperature with
2.5 ml of 2× SDS sample buffer (without -mercaptoethanol). The
resultant extracts were separately electrophoresed overnight at 8 mA on
preparative scale (1.5 mm thick), 4% acrylamide gels containing SDS.
After electrophoresis, an ~3-cm-wide vertical strip was cut from both
the large complex-containing and control gels. To identify where the
two major CPE-containing large complex species (see "Results") were
localized in each gel strip, proteins present in the strips were
electrotransferred onto nitrocellulose membranes, and those
membranes were immunoblotted with anti-CPE antibody. Horizontal gel slices corresponding to the location (if present) of
each large complex species were then carefully excised from both the
large complex-containing and the control preparative gels. To reduce
contamination of the gel slice containing the ~155-kDa complex by the
~135-kDa complex or smaller proteins, only the upper half of the band
containing the ~155-kDa large complex was excised from preparative gels.
Proteins in the excised gel slices were then electroeluted, as
described previously (31). Briefly, each gel slice was inserted into a
sack of Spectra Pore 3 dialysis membrane (Spectrum; molecular mass
cut-off, 3,500 Da) containing 25 mM Tris, 0.192 M glycine, 0.1% (w/v) SDS, pH 8.3. These sacks were then
sealed and electrophoresed at 50 mA for 3 h at 4 °C. Eluted
proteins were removed from the dialysis sacks and concentrated to a
final volume of 200 µl, using Centricon 3 microconcentrators
(Amicon), before storage at 
20 °C.
Prior to electrophoresis, a 100-µl aliquot of each eluted protein
sample, along with 100 µl of freshly prepared SDS lysate of CaCo-2
cells, was treated with 6 M urea and 2× SDS sample buffer (with -mercaptoethanol) at 90 °C for 20 min. Those samples were then boiled for 5 min to further denature proteins. The denatured samples were then electrophoresed, along with prestained markers (Bio-Rad), on 10% acrylamide gels containing SDS. The separated proteins on those gels were transferred onto nitrocellulose membranes, which were blocked with Blotto and immunoblotted with either anti-CPE antibody or anti-occludin antibody, as described earlier.
Immunoprecipitation Analyses
Immunoprecipitation of Small Complex--
To form small complex
in the absence of appreciable levels of large complex (4), scraped
CaCo-2 cells (~2.6 × 107 cells) were suspended in
cold PBS containing Ca2+ and Mg2+. A 200-µl
aliquot of that cell suspension was treated with 0.5 µg of
125I-CPE, in the presence of a 100-fold excess of unlabeled
CPE, for 5 min at 4 °C. The remainder of the cell suspension was
treated with only 0.5 µg of 125I-CPE for 5 min at
4 °C. As a control, CaCo-2 cells (2.4 × 106) were
similarly incubated for 5 min at 4 °C in PBS containing Ca2+ and Mg2+ but lacking either
125I-CPE or native CPE. After this initial incubation
period, the control and CPE-treated cells were both extracted, in the
presence of protease inhibitors (leupeptin and aprotinin, at 10 µg/ml
each, along with 1 mM phenylmethylsulfonyl fluoride), with
200 µl of 1% Triton X-100 (Sigma) in PBS for 20 min at 4 °C (4).
The extracted material was then aliquoted (200 µl/sample), and each sample was then preincubated with 15 µl of suspended acrylic beads containing protein A (Sigma) for 1 h at 4 °C to reduce
nonspecific binding to the beads during our subsequent
immunoprecipitation procedure. All but one of these samples were then
treated at 4 °C overnight, in the presence of protease inhibitors
(as described above), with 20 µg of either affinity-purified rabbit
IgG raised against CPE, human occludin antibody, or NRS IgG. The
remaining aliquot was incubated similarly, except for the omission of
any antibody. To precipitate antigen-antibody complexes, 40 µl of the
suspended protein A-coupled acrylic beads were then added to each
sample for 2 h at 4 °C. After microcentrifugation, supernatants were removed, mixed with 40 µl of 5× native PAGE sample buffer (4),
and then electrophoresed (without sample boiling) on 6% acrylamide
gels containing 0.1% Triton X-100. After electrophoresis, those gels
were autoradiographed on x-ray film at 80 °C.
The pelleted acrylic beads with bound antibody-antigen complexes were
washed four times with PBS containing Ca2+,
Mg2, and 0.5% Triton X-100, and a fifth time with PBS
containing only Ca2+ and Mg2+. Material bound
to the washed beads was eluted by boiling for 5 min in 100 µl of SDS
sample buffer (with -mercaptoethanol). The boiled bead mixture was
then centrifuged, and the resultant supernatant was analyzed using 10%
acrylamide gels containing SDS. Those gels were then subjected to
Western immunoblotting using antibodies raised against the C terminus
of human occludin, as described earlier. The blots were then stripped
of occludin antibody using antibody elution buffer (Pierce) and
reprobed by Western blotting using anti-CPE IgG, as described earlier.
Immunoprecipitation of Large Complex--
Scraped CaCo-2 cells
(2.4 × 107) were treated with CPE (6 µg) for 20 min
at 37 °C to allow formation of large complex. These CPE-treated
cells, along with similarly prepared control (non-CPE-treated) Caco-2
cells, were then extracted with 1% Triton X-100 in PBS containing
Ca2+ and Mg2+ for 20 min at room temperature in
the presence of protease inhibitors (leupeptin and aprotinin, both at
10 µg/ml, as well as phenylmethylsulfonyl fluoride, at 1 mM). After microcentrifugation of these mixtures, the
supernatant was aliquoted into nine tubes, which were each preincubated
with 15 µl of acrylic beads coupled with protein A for 1 h at
4 °C. After the beads were removed by centrifugation, the
supernatant was aliquoted; one supernatant sample was then kept at
4 °C without the addition of any antibody. The remaining supernatant
samples were incubated overnight at 4 °C, either without any
antibody or with 20 µg of affinity-purified polyclonal rabbit IgG
antibodies raised against purified CPE, rabbit polyclonal antibodies
raised against the C-terminal region of human occludin, or NRS IgG.
After this overnight incubation, 40 µl of a suspension containing
acrylic beads coupled with protein A were then added to each sample,
and the samples were further incubated for 2 h at 4 °C to
precipitate antibody-antigen complexes. These samples were then
microcentrifuged, and the pelleted acrylic beads were washed four times
with PBS containing Ca2+, Mg2+, and 0.5%
Triton X-100 and one time with PBS containing Ca2+ and
Mg2+, but no Triton X-100. The washed beads were then
boiled for 5 min in 2× SDS sample buffer, and the eluted proteins in
the boiled samples were electrophoresed on 10% acrylamide gels
containing SDS. Western blotting with occludin antibodies was then
performed on these gels, as described above.
Analysis of the Large Complex Forming Ability of rCPE and
Attenuated rCPE Mutants
Recombinant Escherichia coli transformants expressing
rCPE or a G49D, S59L, or R116S rCPE point mutant were prepared in a previous study (20). These four rCPE species, all of which contain a
six-histidine tag, were employed in the current study on the basis of
previous characterization studies (20) showing that (i) rCPE retains
all toxic properties of native CPE, (ii) the R116S mutant is strongly
attenuated for toxicity because of decreased large complex forming
ability, and (iii) the G49D and S59L point mutants are completely
nontoxic, because of their inability to form large complex.
Recombinant E. coli expressing the desired rCPE species were
cultured in 5 liters of SOB medium containing ampicillin (100 µg/ml).
After centrifugation and sonication, the resultant lysates were
chromatographed over a Talon resin affinity column. The amount of rCPE
species present in each affinity-enriched preparation was then
determined by Western blotting, as described previously (20). To
quantitate the amount of rCPE species present in each affinity-enriched
preparation, densitometric scans were performed on each Western blot
with a Scan Jet Plus (Hewlett Packard), using the Deskscan II 2.3 program. Peak area integrations were determined using one-dimensional
Process and Report Program (Zeineh Biomedical Instruments). These
quantitative Western immunoblot analysis also indicated that rCPE
species represented at least 60-70% of the total protein present in
each affinity-enriched sample.
Scraped CaCo-2 cells (2 × 106) were treated for 20 min at 37 °C with 0.5 µg of either rCPE, an attenuated rCPE
mutant, or native CPE. These CaCo-2 cells were then extracted with SDS
sample buffer, and extracts were run (without sample boiling) on 4%
acrylamide gels containing SDS. The gels were then Western blotted with
CPE antibody, as described above.
Analysis of 125I-CPE Binding to CaCo-2, Rat-1/R12,
and FL-22 Occludin Transfectants
FL-22 occludin transfectants, Rat-1/R12 parental cells, or
CaCo-2 cells were inoculated into 60-mm tissue culture dishes (Falcon) at a density of 106/dish and grown to confluency (~4-6
days). These confluent cultures were washed twice with warm HBSS and
then treated for 15 min at either 37 °C or room temperature with 3 ml of warm HBSS containing 125I-CPE (0.5 µg), with or
without a 100-fold excess of unlabeled CPE. After this CPE treatment,
each culture was washed twice with HBSS, and adherent cells were
scraped from dishes in 2 ml of HBSS and combined with nonadherent cells
(if any) collected during washing. The total number of cells present in
each sample was then counted using a hemocytometer, and the radiation
associated with these samples was determined using a 
counter
(Packard). As described previously (5), 125I-CPE specific
binding was then calculated by subtracting radioactive counts present
in cell samples that had been co-treated with a 100-fold excess of
unlabeled CPE (nonspecific binding) from radioactive counts present in
similar cell samples that had received only 125I-CPE
treatment (total binding). Specific binding always represented >60%
of total binding for CaCo-2 cells.
86Rb Release Assay to Determine the CPE Sensitivity
of FL-22 Occludin Transfectants
24-Well plates (Corning) containing confluent monolayers of
FL-22, parental Rat-1/R12 liver fibroblast cells, or CaCo-2 cells were
labeled with 4 µCi/well of 86RbCl (NEN Life Science
Products) for 2 h at 37 °C. After this radiolabeling, the
cultures were washed twice with warm HBSS and then incubated at
37 °C for 15 min with 2 ml of warm HBSS containing increasing
amounts (0.5-16 µg) of native CPE. After that CPE treatment, the
culture supernatant was removed, and supernatant radioactivity was
counted in a Beckman counter.
As described previously (5), the percentage of maximal 86Rb
release induced by CPE treatment was calculated as 100 × (86Rb release in CPE-treated wells spontaneous
86Rb release)/(maximal 86Rb release
spontaneous 86Rb release). Maximal release corresponds to
the total cytoplasmic radioactivity present at the start of each
experiment and was always ~5 × 104 cpm/well.
Spontaneous release represents the radioactivity released from
monolayers in the absence of CPE treatment and was always ~1 × 104 cpm/well.
| |
RESULTS |
CPE Western Immunoblot Analysis of Large Complex Formation by
CaCo-2 Cells Treated with CPE at 37 °C--
Previous studies (4,
32) have shown that CaCo-2 cells are sensitive to CPE, suggesting that
human intestinal cell line represents an especially germane in
vitro model for studying CPE action. However, before CaCo-2 cells
become widely employed for in vitro studies of CPE action,
it was important to demonstrate that those cells exhibit similar
responses to CPE treatment as other in vitro models of CPE
action. Prior to our current study, it had already been established
that CaCo-2 cells can specifically bind CPE and form CPE-containing
small complex (4).
However, it had not yet been determined whether CaCo-2 cells treated
with CPE at 37 °C form CPE-containing large complex. Therefore, the
initial experiment of our current study utilized a well established CPE
Western immunoblot assay (19, 20) to evaluate large complex formation
by CaCo-2 cells treated with CPE at 37 °C. Results from that CPE
Western immunoblot analysis indicated that SDS extracts from CaCo-2
cells treated with CPE at 37 °C contain significant levels of
material that reacts with CPE antibodies and migrates more slowly than
free CPE (Fig. 1A, compare
lanes 1-4 with CPE lane). Formation of that high
Mr, immunoreactive material required the
presence of CaCo-2 cells, because similar high
Mr material was not detected when 0.5 µg (Fig.
1A, CPE lane) or more (data not shown) of free
CPE was subjected to CPE Western blotting, i.e. the high
Mr, immunoreactive material shown in lanes 1-4 of Fig. 1A does not represent CPE aggregates.
Furthermore, no high Mr material that reacts
with CPE antibodies was present in similarly prepared SDS extracts from
control (i.e. non-CPE-treated) CaCo-2 cells (Fig.
1A, cells lane), strongly suggesting that the high Mr, immunoreactive material detected in
CPE-treated cells does not result from nonspecific cross-reactivity
between CPE antibodies and a CaCo-2 cell protein(s).
View larger version (38K):
[in this window]
[in a new window]
|
Fig. 1.
Western immunoblot analysis (6% acrylamide,
SDS-containing gels, without sample boiling) of SDS extracts from
CPE-treated CaCo-2 cells. CaCo-2 cells were treated with CPE (0.5 µg) for 20 min at 37 °C, extracted with SDS buffer,
electrophoresed on 6% acrylamide gels containing SDS (no sample
boiling), and then Western immunoblotted with either CPE antibodies
(A) or occludin antibodies (B). Lanes
1-4 of A and B contain 5, 10, 50, and 100 µl, respectively, of SDS extracts from CPE-treated CaCo-2 cells. The
Cells lane of both panels contains 50 µl of SDS extracts
from control (non-CPE-treated) CaCo-2 cells. The arrow in
B indicates the location of high Mr,
occludin antibody-reactive material induced by CPE treatment.
|
|
On the 6% acrylamide gels containing SDS (no sample boiling) that were
used in our Fig. 1 experiment, the high Mr, CPE
antibody-reactive material present in CaCo-2 cells treated with CPE at
37 °C appears to migrate as an ~160-kDa species, matching previous
size estimates of the CPE-containing large complex formed in both
rabbit intestinal brush border membranes and Vero cells (9). Formation
of similar large complex material was observed when CaCo-2 cells were
treated with 125I-CPE at 37 °C (data not shown);
formation of that high Mr, radiolabeled species
was inhibited when CaCo-2 cells were treated simultaneously with
125I-CPE and a 100-fold excess of unlabeled CPE (data not
shown), i.e. formation of CPE-containing large complex in
CaCo-2 cells, like large complex formed in Vero cells and intestinal
brush border membranes, involves specifically bound CPE.
Besides demonstrating that CaCo-2 cells form large complex, the initial
CPE Western immunoblot results shown in Fig. 1A also revealed that small volumes of SDS extracts of CaCo-2 cells treated with CPE at 37 °C (Fig. 1A, lane 1) run as a
relatively sharp, discrete band, whereas larger volumes of similarly
prepared SDS extracts run as a more diffuse smear (Fig. 1A,
lane 4). However, similar smearing of large complex material
could be observed (data not shown) if the chemiluminescence exposure
time was increased for lane 1 sample of Fig. 1A.
Occludin Western Immunoblot Analysis to Evaluate whether Occludin
Is Present in Large Complex Formed when CaCo-2 Cells Are Treated with
CPE at 37 °C--
For reasons described in the Introduction, we
hypothesized that occludin, an ~65-kDa tight junction protein, might
be present in CPE-containing large complex. Our Fig. 1A
results demonstrating that CaCo-2 cells treated with CPE at 37 °C
form large complex permitted us to test our hypothesis using CaCo-2
cells, which was fortuitous because (i) CaCo-2 cells have a human
origin and antibodies against human occludin are available commercially
and (ii) CaCo-2 cells are known to express occludin (33).
To begin evaluating whether the CPE-containing large complex material
present in CaCo-2 cells contains occludin, SDS extracts prepared from
CaCo-2 cells that had been treated with CPE at 37 °C were
electrophoresed on 6% acrylamide gels containing SDS (no sample
boiling) and immunoblotted using occludin antibodies (Fig. 1B). Those immunoblots clearly demonstrated the presence of
a high Mr species reactive with occludin
antibodies (Fig. 1B, lanes 1-4). Formation of
that high Mr material reactive with occludin antibody was dependent on CPE treatment, because a similar species was
absent from SDS extracts that had been similarly prepared from control
(non-CPE-treated) CaCo-2 cells (Fig. 1B, cells
lane). Furthermore, the high Mr species
that reacts with occludin antibodies in CPE-treated CaCo-2 cells did
not result from nonspecific cross-reactivity between the occludin
antibody and CPE, because that antibody did not react with free CPE
(Fig. 1B, CPE lane).
Immunoblotting of 4% Acrylamide Gels Containing SDS to Determine
whether CPE-containing Large Complex of CaCo-2 Cells Contains Multiple
Species--
Our Fig. 1A results indicating that
CPE-containing large complex material from CaCo-2 cells can run as a
smear could suggest that large complex material is a heterogeneous mix
of several CPE-containing species. That hypothesis received additional,
although still indirect, support from comparing the Fig. 1
(A and B) immunoblot results, i.e. on
6% acrylamide gels containing SDS (no sample boiling), the high
Mr species reactive with occludin antibodies (Fig. 1B, lane 2) appears to only partially
co-migrate with the smear of high Mr material
that reacts with CPE antibodies (Fig. 1A, lane
4).
To evaluate the possibility of large complex heterogeniety more
definitively, SDS extracts prepared from CaCo-2 cells treated with CPE
at 37 °C were electrophoresed (without sample boiling) on SDS gels
containing less (4%) acrylamide, which should improve resolution of
high Mr, protein species. When those 4%
acrylamide gels containing SDS were immunoblotted with CPE antibodies,
multiple high Mr species reactive with CPE
antibodies were detected (Fig. 2,
left panel). Those high Mr species
were absent (Fig. 2, left panel) from either (i) similarly
processed, control CaCo-2 cells (Cells lane), which strongly
suggests that the high Mr species present in
CPE-treated CaCo-2 cells contain CPE, or (ii) a sample of free CPE that
had been similarly electrophoresed and Western blotted (CPE
lane), strongly suggesting that the high Mr
species observed in CPE-treated CaCo-2 cells do not correspond to
aggregates of free CPE.
View larger version (31K):
[in this window]
[in a new window]
|
Fig. 2.
Kinetics of large complex formation in
CPE-treated CaCo-2 cells at 37 °C. After incubation of CaCo-2
cells with CPE for desired time periods, CPE-treated cells were
extracted with SDS, electrophoresed on 4% acrylamide gels containing
SDS, and then Western immunoblotted with either CPE antibodies or
occludin antibodies, as indicated. The time of CPE treatment is shown
at the top of the gel, whereas the migration of myosin (212 kDa) and -galactosidase (122 kDa) markers are indicated in the
center space between the two blots. The double,
open, and closed arrows indicate the location of
the ~200-, ~155-, and ~135-kDa large complexes,
respectively.
|
|
Two high Mr, large complex species are
particularly abundant in the CPE immunoblot shown in the left
panel of Fig. 2. On the 4% acrylamide SDS-containing gels used in
the Fig. 2 experiment, these two major large complex species migrated
with apparent sizes of ~155 and ~200 kDa. In addition to those
~155- and ~200-kDa large complex species, which were reproducibly
detected in all five repetitions of the Fig. 2 CPE immunoblot
experiment, a minor high Mr species with an
apparent size of ~135 kDa was also discernible in most repetitions of
that experiment.
The results shown in the left panel of Fig. 2 indicate that
the major ~200- and ~155-kDa large complex species can both be detected within the first minute of CPE treatment of CaCo-2 cells at
37 °C. When discernible, the ~135-kDa large complex species became
clearly visible within 10 min of CPE treatment using our standard
chemiluminescent detection protocol (Fig. 2, left panel). However, if longer chemiluminescent detection periods were used, the
~135-kDa large complex species could often be detected within 1 min
of continuous CPE treatment at 37 °C (data not shown).
To further evaluate the kinetics of formation of the three
CPE-containing large complex species, densitometric analysis was performed on immunoblots from three independent repetitions of the Fig.
2 (left panel) experiment. Those densitometric analyses revealed that the amounts of the ~135-, ~155-, and ~200-kDa
CPE-containing large complexes present in CaCo-2 cells all
progressively increased during the first 30 min of continuous CPE
treatment at 37 °C. However, between 30 and 90 min of continuous CPE
treatment, the amount of the ~200- and ~155-kDa species declined by
about 33 and 25%, respectively, whereas levels of the ~135-kDa
species increased by ~20%. Consequently, whereas the ~200-,
~155-, and ~135-kDa large complex species had represented ~25,
~65, and ~10%, respectively, of large complex material present in
CaCo-2 cells after 30 min of continuous CPE treatment at 37 °C, the
ratio of these species changed to ~20, ~60, and ~15%,
respectively, after 90 min of continuous CPE treatment.
Kinetic relationships between the large complex species were further
investigated in an experiment where CaCo-2 cultures were treated with
CPE for 10 min at 37 °C, unbound toxin was removed, and the cultures
were further incubated for up to another 80 min. When those cells were
subjected to CPE Western immunoblot analysis (Fig.
3A), all three CPE-containing
large complexes remained detectable throughout the duration of the
experiment. However, levels of the ~200- and ~155-kDa large
complexes declined by about 50% over the 80-min time span of this
experiment, whereas the amount of the ~135-kDa large complexes
increased by >30% during the experimental period.
View larger version (19K):
[in this window]
[in a new window]
|
Fig. 3.
Formation and stability of large complex
species in CaCo-2 cells. CaCo-2 cells were treated with CPE for 10 min at 37 °C. Unbound CPE was then removed (as indicated by
arrow at the top of gel), and cells were further
incubated in CPE-free medium for up to an additional 80 min
(i.e. total CPE treatment time, 90 min). At the indicated
times, aliquots of the CPE-treated cells were extracted with SDS,
electrophoresed on 4% acrylamide gels containing SDS (no sample
boiling), and then Western immunoblotted with either CPE antibodies
(A) or occludin antibodies (B). The total time of CPE
treatment is indicated on the x axis, whereas the migration
of myosin (212 kDa) and -galactosidase (122 kDa) markers are
indicated in the center space between the two blots. The
double, open, and closed arrows
indicate, respectively, the location of the ~200-, ~155-, and
~135-kDa large complexes.
|
|
Occludin Immunoblot Analysis to Identify Which Large Complex
Species Contain Occludin--
Because our Fig. 1B results
suggested that occludin is associated with large complex material, it
was important to determine which of the large complex species
identified in the left panels of Figs. 2 and 3A
might contain occludin. Therefore, SDS extracts from CaCo-2 cells
treated with CPE at 37 °C were electrophoresed on 4% acrylamide
gels containing SDS (no sample boiling) and subjected to Western
immunoblotting using occludin antibodies. Five repetitions of the
occludin Western immunoblotting experiment detected only a single, high
Mr band of ~200 kDa (Fig. 2, right
panel). That occludin antibody-reactive band always precisely
co-migrated with the ~200-kDa complex species detected using CPE
antibodies (Fig. 2, left panel). Co-migration of the
occludin-antibody and CPE-antibody complexes of ~200 kDa was
confirmed by stripping the occludin antibodies from these blots and
reprobing with CPE antibodies (data not shown).
With respect to kinetics, the ~200-kDa large complex species that
reacts with occludin antibodies became detectable within 1 min of
continuous CPE treatment of CaCo-2 cells at 37 °C (Fig. 2,
right panel). Levels of that ~200-kDa species apparently
containing occludin then steadily increased through the first 30 min of
continuous CPE treatment before declining. When CaCo-2 cells were
exposed to CPE for 10 min, washed free of unbound CPE, and then
reincubated in CPE-free medium at 37 °C, an ~40% decline in the
amount of the ~200-kDa large complex was noted between 20 and 80 min
post-removal of unbound CPE (Fig. 3B).
Denaturing SDS-PAGE Analysis of the ~155- and ~200-kDa Large
Complexes Isolated by Preparative Electrophoresis--
The results
shown in the left panel of Fig. 2 strongly suggested that
treatment of CaCo-2 cells with CPE at 37 °C induced formation of two
major large complex species, including an ~155-kDa large complex
species that apparently contains CPE but not occludin and an ~200-kDa
species that apparently contains both CPE and occludin. However, it
remained possible that occludin might be present in the ~155-kDa
species but had not been detected by Western blotting because it was
inaccessible to occludin antibodies. It also remained possible that the
reactivity of the commercial occludin antibody preparation shown in
Figs. 2 (right panel) and 3B with the ~200-kDa
large complex species might result from nonspecific cross-reactivity
betweeen that antiserum and some CaCo-2 cell protein other than occludin.
To evaluate those possibilities and confirm that CPE is actually
present in both the ~155- and ~200-kDa large complex species, gel
slices containing the ~155- and ~200-kDa large complexes were excised from preparative SDS-PAGE gels that had been run under the same
conditions used for Figs. 2 and 3 experiments (note that similar
preparative electrophoresis isolation of the ~135-kDa large complex
was not attempted because of the limited abundance of that complex and
its close migration to both the ~155-kDa large complex and to the dye
front, which would cause significant cross-contamination of the
isolated ~135-kDa complex with other proteins). When proteins eluted
from these preparative gel slices were re-electrophoresed on denaturing
SDS-PAGE gels containing urea and then subjected to Western immunoblot
analysis with CPE antibodies, the results (Fig.
4A) clearly demonstrated that
both the ~155- and ~200-kDa complexes contain a 35-kDa protein that
co-migrates with purified CPE and reacts with CPE antibodies,
i.e. CPE is present in both the ~155- and ~200-kDa large
complexes.
View larger version (31K):
[in this window]
[in a new window]
|
Fig. 4.
Western immunoblot analysis of denaturing
SDS-PAGE gels run with proteins electroeluted from preparative gels
containing SDS extracts prepared from either CPE-treated or control
CaCo-2 cells. CaCo-2 cells were incubated with (Large
Complex) or without (Control) CPE at 37 °C,
extracted with SDS sample buffer (without boiling), and electrophoresed
on separate 4% preparative gels. Gel slices corresponding to regions
containing the ~155- and ~200-kDa large complexes were excised from
the gels containing both CPE-treated and control CaCo-2 SDS extracts.
Proteins in the gel slices were electroeluted at 50 mA for 3 h,
the electroeluents were treated with 6 M urea at 90 °C
and then boiled for 5 min in 2× SDS sample buffer (with
-mercaptoethanol). The boiled samples were then run on 10%
acrylamide gels containing SDS, and those gels were subjected to
Western immunoblotting with either anti-CPE IgG or occludin antibodies,
as indicated. The migration of molecular mass markers is shown in the
center space between the two immunoblots; the migration of
CPE and occludin are noted by the closed and open
arrows, respectively. Note that the material at the top
of some lanes of these CPE Western immunoblots probably represents
incompletely dissociated large complex material (data not shown).
|
|
When similar denaturing SDS-PAGE Western immunoblot analyses were
performed with the same commercial occludin antibody preparation used
in our Fig. 2 (right panel) and 3B experiments, a
single immunoreactive band of ~65 kDa, which matches the expected
size of occludin, was detected in SDS extracts freshly prepared from control CaCo-2 cells. The Fig. 4B occludin Western
immunoblot analysis also demonstrated that a substantial amount of that
~65-kDa protein reactive with occludin antibodies was also present in the sample eluted from gel slices containing the ~200-kDa large complex of CPE-treated CaCo-2 cells, i.e. preparative gel
slices with the ~200-kDa large complex contain occludin, as well as
CPE. The absence of immunoreactivity in corresponding gel slices from preparative gels run with SDS extracts prepared from control cells indicates that the presence of occludin in the sample eluted from gel
slices containing the ~200-kDa large complex can be specifically attributed to CPE treatment.
When proteins eluted from preparative gel slices containing the
~155-kDa band of CPE-treated CaCo-2 cells were similarly analyzed by
occludin Western immunoblots, a trace amount of occludin was detectable
(Fig. 4B). However, those trace levels of occludin do not
appear to be associated with CPE treatment, because similar trace
levels of occludin were also detected in samples eluted from gel slices
of corresponding regions of preparative gels run with SDS extracts
prepared from control CaCo-2 cells. The trace levels of occludin that
appear to be contaminating gel slices with the ~155-kDa large complex
(and corresponding regions of gel slices from control gels) probably
resulted from smearing of free occludin during preparative
electrophoresis, as can be observed in the Fig. 3B
immunoblot, which used similar 4% acrylamide gels containing SDS.
Immunoprecipitation Analysis of CaCo-2 Cells Treated with CPE at
37 °C--
The results shown in Figs. 2-4 strongly suggested that
CPE and occludin co-localize in a common (~200 kDa) complex. To
formally prove that CPE is physically associated with occludin in
CaCo-2 cells, an immunoprecipitation experiment was performed using SDS extracts of CaCo-2 cells treated with CPE at 37 °C. Results from that experiment (Fig. 5) show that, as
would be expected, occludin antibodies specifically immunoprecipitated
a protein of ~65 kDa, which matches the expected size of occludin and
reacts with occludin antibodies, from both CaCo-2 cells treated with
CPE at 37 °C and control (non-CPE-treated) CaCo-2 cells. In
contrast, CPE antibodies only immunoprecipitated an ~65-kDa protein,
which reacts with occludin antibodies, from CaCo-2 cells that had been
treated with CPE at 37 °C. The ability of CPE antibodies to
immunoprecipitate occludin from CPE-treated, but not control, CaCo-2
cells confirms that CPE and occludin can physically associate in CaCo-2
cells treated with CPE at 37 °C.
View larger version (24K):
[in this window]
[in a new window]
|
Fig. 5.
Occludin Western immunoblot analysis of
denaturing SDS-PAGE gels run with material immunoprecipitated from
control or large complex-containing CaCo-2 cells. CaCo-2 cells
incubated with (Large Complex) or without
(Control) CPE at 37 °C were extracted with 1% Triton
X-100 at 4 °C. Aliquots of those extracts were then incubated
without or with 100 µg/ml of anti-CPE IgG, occludin antibodies, or
NRS IgG. The extracts were then further incubated with protein A beads
to immunoprecipitate antigen-antibody complexes. Those mixtures were
microcentrifuged, and the pelleted beads were washed and extracted with
SDS sample buffer (with -mercaptoethanol) and boiled for 5 min. The
boiled samples were electrophoresed on 10% acrylamide gels containing
SDS and Western immunoblotted with occludin antibody. The samples shown
for both the control and large complex samples include
immunoprecipitate obtained using anti-CPE IgG (lane 1),
immunoprecipitate obtained using occludin antibodies (lane
2), immunoprecipitate obtained using NRS-IgG (lane 3),
and materials bound to beads in the absence of any antibody (lane
4). The CPE lane contains 0.5 µg of free CPE. The open
arrow on the right side indicates the location of the
~65-kDa band that appears to correspond to occludin.
|
|
Immunoprecipitation Analysis of CaCo-2 Cells Treated with CPE at
4 °C--
Results from our Fig. 1-5 experiments indicated that CPE
and occludin can co-localize in an ~200-kDa large complex. However, because there are known to be several steps in CPE action preceding large complex formation, it was possible that CPE and occludin might
interact before large complex formation. Therefore, an experiment was
performed to evaluate whether occludin is present in the CPE-containing small complex that forms upon or soon after CPE binding and precedes large complex formation.
To evaluate whether occludin is present in CPE-containing small
complex, CaCo-2 cells were treated with 125I-CPE at
4 °C, a temperature where small complex formation occurs but large
complex formation is effectively inhibited (4). Those 125I-CPE-treated CaCo-2 cells were subjected to
immunoprecipitation analysis (Fig. 6),
which was necessary because small complex cannot be reliably detected
by immunoblotting (20).
View larger version (44K):
[in this window]
[in a new window]
|
Fig. 6.
Immunoprecipitation experiments with CaCo-2
cells containing small complex. a, autoradiographs of
material present in supernatants (left) or
immunoprecipitates (right) from immunoprecipitation
experiments performed using control CaCo-2 cells (lanes 2-5
and 11) or CaCo-2 cells treated with 125I-CPE at
4 °C (lanes 1 and 6-9). The cells were then
extracted with 1% Triton X-100; the resultant Triton X-100 extracts
were incubated without antibody (lanes 2 and 6)
or with anti-CPE IgG (lanes 3 and 7), occludin
antibodies (lanes 4 and 8), or NRS-IgG
(lanes 5 and 9). After overnight incubation at
4 °C, the mixtures were incubated with protein A beads. After
microcentrifugation, supernatants were extracted with native sample
buffer containing Triton X-100 (without boiling), electrophoresed on
6% Triton X-100 native gels, and subjected to autoradiography.
Pelleted beads were washed and boiled for 5 min with SDS sample buffer
(containing -mercaptoethanol), and the boiled samples were run on
10% acrylamide gels containing SDS. Those gels were dried and
autoradiographed. For comparison, freshly prepared Triton X-100
lysates, i.e. lysates that were not incubated overnight,
from either CaCo-2 cells treated with 125I-CPE at 4 °C
(lane 1) or control CaCo-2 cells (lane 11) are
also shown. Migration of molecular mass markers is indicated to the
left of the gel containing the immunoprecipitates. The
open and closed arrows show the migration of
small complex and free CPE, respectively. Note that no radioactive
material was present in lanes containing Triton X-100 extracts of
CaCo-2 cells treated with 125I-CPE in the presence of
100-fold excess native CPE (data not shown), i.e. all
rardioactive material shown in lanes 1 and 6-9
is attributable to specifically bound 125I-CPE.
b, Western immunoblot analysis of material
immunoprecipitated from CaCo-2 cells containing small complex. The
immunoprecipitated material prepared during the experiments in
a was also subjected to denaturing SDS-PAGE (with sample
boiling) on 10% acrylamide gels and then immunoblotted with anti-CPE
IgG (left) or occludin antibody (right). Lanes
shown include: material present before immunoprecipitation of SDS
extracts from small complex-containing (lane 1) or control
(lane 11) CaCo-2 cells; material from SDS extracts of
control (lane 2) or small complex-containing (lane
6) CaCo-2 cells that bound to protein A beads in the absence of
any antibody; material immunoprecipitated by anti-CPE IgG from SDS
extracts of control (lane 3) or small complex-containing
(lane 7) CaCo-2 cells; material immunoprecipitated by
occludin antibody from SDS extracts of control (lane 4) or
small complex-containing (lane 8) CaCo-2 cells; and material
immunoprecipitated by NRS-IgG from SDS extracts of control (lane
5) or small complex-containing CaCo-2 cells (lane 9).
Lane 10 contains free 125I-CPE. The closed
arrow on left side depicts the migration of ~65-kDa
occludin, whereas the open arrow indicates the migration of
125I-CPE.
|
|
Results shown in lane 6 of the left panel of Fig.
6a confirm that CaCo-2 cells treated with CPE at 4 °C
form small complex, consistent with previous studies (4). Results shown
in lanes 6-9 of the left and right
panels of Fig. 6a also (i) indicate that CPE
antibodies, but not occludin antibodies (or NRS IgG), can
immunoprecipitate small complex and (ii) confirm that intact 125I-CPE is present in the small complex that was
immunoprecipitated by CPE antibodies.
The inability of occludin antibodies to immunoprecipitate small complex
in our Fig. 6a experiments was consistent with the absence
of occludin from small complex. However, it was conceivable that
occludin might be present in small complex but had not been detected by
immunoprecipitation techniques because it was inaccessible to occludin
antibodies. To test that possibility, CaCo-2 cells treated with CPE at
4 °C (a condition where small complex forms; see Fig. 6a)
were subjected to immunoprecipitation using either CPE antibodies,
occludin antibodies, or NRS IgG, and the resultant immunoprecipitates
were subjected to denaturing Western immunoblot analysis, first using
occludin antibodies (Fig. 6b, right panel) and
then, after stripping, CPE antibodies (Fig. 6b, left
panel). Consistent with our Fig. 6a results, the CPE
Western immunoblot results shown in Fig. 6b indicate that
CPE antibodies specifically immunoprecipitated a 35-kDa protein
(lane 7), which co-migrates with purified CPE (lane
10), from CaCo-2 cells containing small complex. Neither occludin
antibodies nor NRS IgG were able to immunoprecipitate that 35-kDa
protein, confirming the identity of that protein as CPE. The failure of
occludin antibodies to immunoprecipitate CPE in lane 8 of
Fig. 6b is consistent with our Fig. 6a results
showing that occludin antibodies did not immunoprecipitate either small
complex or free CPE.
The occludin Western immunoblot results shown in the right
panel of Fig. 6b confirm (as expected) that occludin
antibodies can immunoprecipitate a protein of ~65 kDa, which matches
the expected size of occludin and reacts with occludin antibodies, from
both control CaCo-2 cells and CaCo-2 cells containing small complex.
However, CPE antibodies (which can immunoprecipitate small complex; see
Fig. 6a) failed to immunoprecipitate a similar ~65-kDa
protein from either control or small complex-containing CaCo-2 cells,
conclusively demonstrating that occludin is not present in small complex.
Use of Attenuated rCPE Point Mutants to Investigate the Biologic
Importance of the ~155- and ~200-kDa Large Complexes--
A
previous study (20) from our laboratory demonstrated that (i) S59L and
G49D rCPE point mutants are completely nontoxic because they fail to
form large complex material on 6% acrylamide gels containing SDS and
(ii) an R116S rCPE point mutant with strongly attenuated cytotoxic
activity forms significantly reduced amounts of large complex material
on 6% acrylamide gels containing SDS. To begin dissecting the biologic
importance of the ~155- and ~200-kDa large complex species, the
ability of these three rCPE mutants and rCPE to form one or both of the
newly identified large complexes was determined using CPE Western blots
of 4% acrylamide gels containing SDS.
Results from that immunoblot experiment (Fig.
7) indicate that, like native CPE (Figs.
2 and 3), rCPE forms two predominate large complex species. Reminiscent
of our earlier observations with native CPE, the faster-migrating of
those two rCPE-containing large complex species was also more abundant.
These results strongly suggest that rCPE, which possesses biologic
activity similar to that of native CPE (20), also exhibits similar
large complex forming ability as the native enterotoxin. Fig. 7 results
also indicates that the two major large complex species formed by rCPE migrate more slowly than the ~155- and ~200-kDa large complexes made by native CPE, probably because of the ~5 kDa of vector-encoded amino acids present on rCPE.
View larger version (63K):
[in this window]
[in a new window]
|
Fig. 7.
CPE Western immunoblot analysis of large
complex formation in CaCo-2 cells treated with rCPE, native CPE, or
rCPE point mutants. CaCo-2 cells were treated with native CPE,
rCPE, G49D, S59L, or a R116S rCPE point mutant at 37 °C to form
large complex material. Those treated cells were extracted with SDS
sample buffer (without boiling) and electrophoresed on 4% acrylamide
gels containing SDS. The gels were then subjected to Western
immunoblotting using anti-CPE antibodies. The location of free rCPE is
shown in A; free G49D, S59L, and R116S mutants also migrated
at the dye front in these blots (data not shown). B shows
the migration of SDS extracts of CaCo-2 cells treated with native CPE
or one rCPE species. No immunoreactivity was present in SDS extracts of
control CaCo-2 cells (data not shown). C shows an
overexposure of the SDS extract from CaCo-2 cells treated with the
R116S mutant. The double, open, and closed
arrows show, respectively, migration of the ~200-, ~155-, and
~135-kDa large complex species.
|
|
When similar Western blot analyses were performed with CaCo-2 cells
treated at 37 °C with one of our rCPE point mutants, it was observed
that (i) neither the nontoxic S59L and G49D mutants induced formation
of any large complex species and (ii) the attenuated R116S mutant made
reduced amounts of both major large complex species. These rCPE point
mutants results are consistent with both major large complex species of
rCPE (and, by extension, the corresponding large complex species made
by native CPE) participating in CPE-induced cytotoxic activity.
Analysis of the CPE Sensitivity of Occludin-transfected Rat
Fibroblasts--
Because our Fig. 7 results were consistent with the
occludin-containing, ~200-kDa large complex playing a role in
CPE-induced cytotoxicity, it was considered important to address
whether the presence of occludin is sufficient to confer CPE
sensitivity to a naturally CPE-insensitive host cell. That question was
assessed by stably transfecting Rat-1/R12 fibroblasts, which are not
naturally CPE-sensitive (Fig.
8c) because of an inability to
bind CPE (Fig. 8d), to express occludin.
View larger version (64K):
[in this window]
[in a new window]
|
Fig. 8.
Studies using the FL-22 occludin
transfectants. a, expression of occludin protein in
Rat-1/R12, FL-22, and CaCo-2 cells. Confluent cultures of Rat-1/R12 and
FL-22 cells grown in the presence (+ Doxy) or absence
( Doxy) of doxycycline and CaCo-2 cells were harvested by
gentle scraping. The harvested cells were extracted with SDS sample
buffer (containing -mercaptoethanol) and boiled for 5 min. The
extracts were electrophoresed on 10% acrylamide gels containing SDS,
and these gels were Western blotted using occludin antibody. The
immunoreactive band shown for FL-22 cells and CaCo-2 cells is ~65 kDa
in size (markers not shown). b, immunofluorescence
localization of occludin and ZO-1 in FL-22 cells. Panel A
shows immunofluorescence with ZO-1 rabbit polyclonal antibody, which
(as expected) is localized at regions of cell-cell contact. Panel
B shows immunofluoresence using anti-VSV-G monoclonal antibody,
which is also localized at cell-cell contacts (note that the occludin
expressed by FL-22 cells is tagged with a C-terminal VSV-G tag).
c, CPE sensitivity of Rat-1/R12, FL-22, and CaCo-2 cells.
Confluent cultures of Rat-1/R12, FL-22, and CaCo-2 cells were labeled
with 86RbCl and then treated with increasing amounts of CPE
for 15 min. Radioactivity in culture supernatants was then determined
using a counter. Data are expressed as percentage of maximal
release, after correction for background spontaneous 86Rb
release from corresponding control (non-CPE-treated) cells. The results
shown include: Rat-1/R12 cells ( ), FL-22 cells ( ), and CaCo-2
cells ( ). These results represent the means of three separate
experiments, with each point determined in duplicate. Error
bars represent the standard error (S.E.) of the mean; points
without error bars had values too small to depict. d, CPE
Western immunoblot analysis of large complex formation by Rat-1/R12,
FL-22, and CaCo-2 cells. CaCo-2 cells, Rat-1/R12, and FL-22 cells were
treated with CPE at 37 °C to form large complex material and then
extracted with 2× SDS sample buffer (without boiling). Those extracts
were electrophoresed on 4% acrylamide gels containing SDS (no sample
boiling), and the gels were subjected to Western blotting using
anti-CPE IgG. Lanes shown include CPE-treated CaCo-2 cells and
Rat-1/R12 and FL-22 cells grown in either the presence (+ Doxy) or absence ( Doxy) of doxycycline. The
CPE lane shows the migration of free CPE on this SDS gel.
e, binding of 125I-CPE to Rat1-/R12, FL-22, and
CaCo-2 cells. Confluent cultures of Rat-1/R12, FL-22, or CaCo-2 cells
were incubated with 125I-CPE in the presence or absence of
100-fold excess native CPE at 37 °C for 15 min. After washing, the
cultures were harvested by scraping, and radioactivity was determined
by counting. An aliquot of the harvested cells was also counted
using a hemocytometer. Specific binding was calculated by subtracting
nonspecific binding (i.e. 125I-CPE binding in
the presence of excess native CPE) from total binding (i.e,
125I-CPE binding in the absence of native CPE). Data shown
are the mean ng of 125I-CPE specifically bound (± standard
error) during three independent experiments, with duplicate data points
in each experiment.
|
|
Occludin expression by the resultant transfectants, named FL-22 cells,
was confirmed by Western immunoblotting experiments (Fig.
8a). Those Western immunoblots clearly demonstrate that FL-22 cells, but not parental Rat-1/R12 cells, express occludin. Densitometric analysis of several repetitions of that Fig.
8a immunoblot experiment indicated that FL-22 cells grown in
the absence of doxycycline produce about half as much occludin/cell as
do CaCo-2 cells, whereas FL-22 cells grown in the presence of
doxycycline (which should decrease occludin expression) still produce
~35-40% as much occludin/cell as do CaCo-2 cells.
When immunofluorescence analysis (Fig. 8b) was performed to
evaluate the distribution of occludin in FL-22 cells, it was observed that occludin primarily localizes to areas of cell-cell contact in
FL-22 cells. The presence of occludin in areas of cell-cell contact is
confirmed by its co-localization with ZO-1, which is also known to be
present in regions of cell-cell contact between nearly all mammalian
cells (including fibroblasts) (28).
The availability of FL-22 cells, which express occludin at regions of
cell-cell contact allowed us to test whether the presence of occludin
in a naturally CPE-insensitive cell is sufficient to confer CPE
sensitivity. When the CPE sensitivity of FL-22 cells was evaluated
using our standard 86Rb release assay for CPE-induced
membrane permeability changes (5), little CPE-induced 86Rb
release was detected. For example, Fig. 8c shows that ~1
µg of CPE was sufficient to induce 50% of maximal 86Rb
release from CPE-sensitive CaCo-2 cells, but even 16 µg of CPE
induced only ~10% of maximal 86Rb release from FL-22
cells, even when those cells were grown in the absence of doxycycline
(where more occludin is expressed; Fig. 8a).
When our standard Western immunoblot large complex detection assays
were employed to investigate whether the presence of occludin in FL-22
cells permits formation of any large complex species, FL-22 cells
treated with CPE at 37 °C did not appear to form detectable levels
of any large complex species (Fig. 8d). However, large complexes of ~155 and ~200 kDa were readily detected in CaCo-2 cells treated with CPE at 37 °C (Fig. 8d), confirming
that the electrophoresis and immunoblotting procedures used in this
particular experiment had worked correctly.
Finally, the absence of significant levels of CPE antibody-reactive
material in CPE-treated FL-22 cells suggested that those transfectants
cannot specifically bind CPE. The inability of FL-22 cells to
specifically bind CPE was then confirmed using 125I-CPE
binding experiments (Fig. 8e), which failed to detect
specific binding of 125I-CPE either to FL-22 cells or to
their parental Rat-1/R12 cells, at either 37 °C (Fig. 8e)
or room temperature (data not shown). The 125I-CPE binding
assay used in this experiment was able to reliably detect
125I-CPE specific binding, because 125I-CPE
specific binding was demonstrated to CaCo-2 cells at both 37 °C and
room temperature (Fig. 8e and data not shown).
| |
DISCUSSION |
As mentioned in the Introduction, identifying the eucaryotic
protein constituents of CPE-containing large complex material is
important for elucidating CPE action. Prior to the current study, only
one eucaryotic protein component of large complex material had been
identified, i.e. certain members of the 22-kDa claudin
family of TJ proteins have been linked to large complex material (8).
However, ample evidence suggested the presence of other eucaryotic
proteins in large complex material. For example, early affinity
chromatography studies (9, 21, 22) using immobilized CPE reported that
~45-50- and ~65-70-kDa eucaryotic proteins are associated with
large complex material. The presence of ~45-50- and ~65-70-kDa
eucaryotic proteins in large complex material received additional
support when eucaryotic proteins matching those two sizes were also
detected in large complex samples (9) that had been partially purified
by either gel filtration or SDS-PAGE (without sample boiling). Our
current study now confirms that a eucaryotic protein of ~65-70 kDa
is present in large complex material and, more importantly, identifies
that protein as occludin.
Before discussing the pathophysiologic relevance of establishing the
presence of occludin in large complex material, it deserves brief
mention that our Fig. 7 immunoblot results may provide some additional
information regarding the composition of large complex material. Those
immunoblots show that the large complex species (discussed below)
formed by rCPE exhibit substantially slower electrophoretic migration
than the corresponding large complex species formed by native CPE.
Those migration differences, which reflect the presence of ~5 kDa of
vector-encoded sequences on rCPE, appear to be greater than would be
expected from an ~5-kDa increase in size for each rCPE large complex
species. Therefore, unless the presence of vector-encoded sequences
induces anomolous electrophoretic mobility characteristics to
rCPE-containing large complex species, the significantly slower
migration of rCPE large complex species versus native CPE
large complex species may suggest that more than one molecule of CPE is
present per large complex species.
Returning to the possible significance of establishing that occludin is
present in large complex material, our FL-22 transfectant results
suggest that the presence of occludin in regions of cell-cell contact
is not sufficient to confer either sensitivity to CPE-induced cytotoxicity or binding ability to a mammalian cell. The failure of
FL-22 cells to bind or respond to CPE does not appear to be attributable to the presence of a VSV-G tag on the occludin expressed by those cells, because (i) that tag is located on the cytoplasmic C
terminus of occludin (28), where it should not contact CPE because the
enterotoxin remains on the surface of plasma membranes (31, 34, 35) and
(ii) recombinant occludin tagged with the same C-terminal VSV-G
sequences has been shown (28) to confer adhesiveness to fibroblasts,
i.e. VSV-G tagged occludin is biologically active.
Demonstrating the presence of occludin in large complex material
provides new evidence for the ability of CPE to interact with TJ
proteins. Specifically, coupling our results establishing that occludin
is present in large complex material formed by CaCo-2 cells with data
from other recent papers linking claudins to large complex material, it
now becomes evident that CPE directly or indirectly interacts with both
of the major structural proteins (claudins and occludin) comprising TJs.
It also appears that different regions of the CPE protein are involved
in interactions with claudin versus occludin. Specifically, previous studies (8, 23, 25) have shown that C-terminal CPE sequences
are sufficient to bind claudins, which is consistent with studies
demonstrating that (i) the C terminus of CPE contains a receptor
binding domain (19, 36-38) and (ii) some claudins can serve as a
functional CPE receptor (8, 25). In contrast, our data indicate that
CPE interacts with occludin after binding. Furthermore, our results
with rCPE point mutants, which apparently retain a native conformation
(20), indicate that sequences in the N-terminal half of the native CPE
protein are necessary for occludin to become localized in the large
complex material. The apparent involvement of N-terminal CPE sequences
in the formation of occludin-containing large complex material argues
that formation of that large complex material does not simply result
from the previously reported claudin-occludin interactions (26) but
instead requires active participation of the CPE protein. It remains to be determined whether N-terminal CPE sequences directly contact occludin during formation of the occludin-containing large complex material or, instead, CPE indirectly induces localization of occludin in large complex material by affecting some other eucaryotic protein, which then interacts with occludin.
Establishing that N-terminal CPE sequences are necessary for the
interactions of toxin with occludin is especially interesting given
recent reports (23, 39) indicating that both CPE and C-terminal CPE
fragments can induce structural alterations in TJs and that these
structural alterations may have pathophysiologic consequences,
e.g. they may increase epithelial paracellular permeability. However, it should be noted that although both native CPE and C-terminal CPE fragments have been observed to induce structural changes in TJs, results from one recent study (39) suggest that TJ
structural changes are induced much more quickly by native CPE. It is
possible that this apparently greater ability of native CPE
(versus C-terminal CPE fragments) to induce TJ structural alterations results from the ability of native CPE, which has the
N-terminal sequences important for localizing occludin in the
~200-kDa complex (see below), to interact with both claudin and
occludin. If the ability of CPE (but not C-terminal CPE fragments) to
interact with occludin explains why native CPE induces TJ
rearrangements faster than C-terminal CPE fragments, it would support
previous data (28) suggesting that occludin as well as claudins (23) play important roles in maintaining the normal structure of TJs.
Perhaps the single most significant result of the current study for
understanding CPE action is the discovery of multiple large complex
species. The discovery of multiple large complex species of ~135,
~155, and ~200 kDa in size mandates changes in our thinking about
CPE action. For example, the presence of multiple large complex species
in polarized epithelium could help explain CPE's diverse biologic
effects, which are now known to include both cytotoxic effects
resulting from toxin-induced membrane permeability alterations in host
cells (5-7, 11), as well as effects on tight junctions (23,
,
TOP
ABSTRACT
INTRODUCTION
