Bile acids modulate tight junction structure and barrier function of Caco-2 monolayers via EGFR...

1

Bile acids modulate tight junction structure and barrier function of Caco-2

monolayers via EGFR activation

Francesco Raimondi, Pasquale Santoro, 2Maria Vittoria Barone, Serena Pappacoda,

Maria Luisa Barretta, 2Merlin Nanayakkara, Carmela Apicella, Letizia Capasso and

Roberto Paludetto.

Division of Neonatology, Department of Pediatrics, University “Federico II” - Naples, Italy;

2: Department of Pediatrics and European Laboratory for the investigation of Food-

Induced Diseases (ELFID), “Federico II” University, Naples, Italy

of 40Articles in PresS. Am J Physiol Gastrointest Liver Physiol (January 31, 2008). doi:10.1152/ajpgi.00043.2007

Copyright © 2008 by the American Physiological Society.

2

Running title: Bile acids modulate intestinal permeability via EGFR activation

Corresponding author: Pasquale Santoro,. Università degli Studi di Napoli “Federico II”

Dipartimento di Pediatria – Divisione di Neonatologia – Via S. Pansini, 5 - Napoli, Italy

Email address: [email protected]

of 40

3

Abstract

Intestinal and systemic illnesses have been linked to increased gut permeability. Bile

acids, whose luminal profile can be altered in human disease, modulate intestinal

paracellular permeability. We have investigated the mechanism by which selected bile

acids increase gut permeability using a validated in vitro model. Human intestinal Caco-2

cells were grown in monolayers and challenged with a bile acids panel. Transepithelial

electrical resistance (TEER) and luminal to basolateral fluxes of 10 kDa Cascade blue

conjugated-dextran were used to monitor paracellular permeability. Immunoprecipitation

and immunoblot analysis were employed to investigate the intracellular pathway.

Redistribution of tight junction proteins was studied by confocal laser microscopy.

Micromolar concentrations of cholic, deoxycholic (DCA), chenodeoxycholic acid (CDCA)

but not ursodeoxycholic (UDCA) decreased TEER and increased dextran flux in a

reversible fashion. Co-incubation of 50 µM CDCA or DCA with EGF, anti-EGF monoclonal

antibody or specific src inhibitor PP-2, abolished the effect. 50 µM of either CDCA or DCA

also induced EGFR phosphorylation, occludin dephosphorylation and occludin

redistribution at the tight junction level in the same timeframe and reversible fashion. We

conclude that selected bile acids modulate intestinal permeability via EGFR auto-

phosphorylation, occludin dephosphorylation and rearrangement at the tight junction level.

The effect is mediated by the src family kinases and is abolished by EGF treatment.

These data also provide support to the role of bile acids in the genesis of necrotizing

enterocolitis and the protective effect of EGF treatment.

of 40

4

Keywords: Bile acids; Caco-2; EGFR; EGF; intestinal permeability; tight junctions

of 40

5

Introduction

Tight junctions (TJs) are the main structures limiting molecular diffusion across the

intestinal epithelium. Failure of barrier function may result from increased paracellular

permeability and lead to passage of antigens or microbes; this has been linked to the

pathogenesis of different illnesses, such as sepsis or food allergy. In particular, necrotizing

enterocolitis (NEC), the most frequent and feared gastrointestinal emergency in premature

newborns, although multifactorial in origin, has been related to both intestinal barrier

dysfunction and to bile acids (BAs) metabolism (14, 15, 9). A consistent body of evidence

suggests that BAs are able to interact with the cells plasma membrane leading to ligand-

independent activation of epithelial growth factor receptor (EGFR) and intracellular

signalling (22, 30, 3). Previous papers also suggested a role for BAs in modulation of

transepithelial permeability (1). As EGFR may play a role in TJ rearrangement (4), we

have hypothesized that BAs, EGFR, TJs rearrangement and increased paracellular

permeability may indeed be connected.

In the present paper, using the in vitro model of Caco-2 cells grown in monolayers, we

provide evidence of a role for BAs in modulating intestinal paracellular permeability by the

rearrangement of occludin, a structural protein of TJs (16). We demonstrate that the effect

is EGFR-dependent and is prevented by EGF treatment.

of 40

6

Materials and Methods

Reagents: cell culture chemicals were obtained from GIBCO-Life Technologies (Milano,

Italy). 10 kDa Cascade blue®-Dextran was purchased from Invitrogen (Carlsbad, CA;

USA). Unconjugated bile acids, atropine and TEXAS RED-phalloidin were purchased from

Sigma (St. Louis, MO; USA). Trypan blue, as well as all other reagents of analytical grade

were also purchased from Sigma.

Proteins were determined by the protein assay from Bio-Rad Laboratories GmbH (Munich,

Germany). ECL-PLUS was from Amersham Biosciences, UK limited, England.

Protein inhibitor cocktail tablets were from Roche Diagnostics (Mannheim, Germany).

Antibodies: anti-Occludin (H279 polyclonal, rabbit); anti-EGFR 1005, polyclonal, rabbit:

(employed for western blots), and clone 528, mouse, monoclonal (employed for both

inhibition experiments and immunoprecipitation); anti-phosphotyrosine (PY-99, mouse,

monoclonal); anti ERK1-2 (mouse, monoclonal); and protein A/G PLUS-Agarose. All the

above antibodies were from Santa Cruz Biotechnology (Santa Cruz, CA; USA).

Mouse anti-Occludin-FITC, was from Zymed Laboratories, (San Francisco, CA; USA).

Secondary antibodies (both anti-rabbit and anti-mouse) were from Amersham Biosciences,

UK limited, England. Anti-src (monoclonal, mouse), anti-phospho-src (rabbit, monoclonal),

anti ErbB-2 (rabbit, polyclonal), anti phospho ErbB-2 (rabbit, polyclonal) were from Cells

Signaling Technology (Danvers, MA; USA). The phosphatidylinositol 3-kinase inhibitor

LY294002; the src family kinases inhibitor PP-2 and the MAP kinase-kinase (MEk) inhibitor

PD98059 were from Alexis Biochemicals Corporation (Vinci, Italy).

Cell culture: human Caco-2 intestinal cells were purchased from the “Istituto

Zooprofilattico della Lombardia e dell’Emilia” (Brescia, Italy). Cells were grown as

previously described (20, 21) in Dulbecco’s modified Eagle’s medium (DMEM) containing

of 40

7

25 mmol/L glucose and supplemented with 10% foetal bovine serum, 1% non-essential

amino acids, 2 mmol/L L-glutamine, 1% penicillin-streptomycin, and 1% sodium pyruvate.

Cells were maintained in a humidified atmosphere (95%) of 5% CO2 in air at 37 °C. Single-

cell suspensions were obtained from 70-80% confluent cultures by incubation with 0.05%

trypsin and then seeded at 105 cells/cm2 onto 13- or 25-mm glass coverslips, detachable

polycarbonate microporous cell culture inserts (Snapwells, 12-mm diameter, 0.4 mm pore

size; Costar; Cambridge, MA, USA) and 24-well plates (Multiwell 24 well; Becton &

Dickinson, Franklin Lakes, NJ, USA) according to experimental needs. Since vectorial

electrolyte transport requires cells to grow in a polarized fashion with structured

intercellular tight junctions, this requires Caco-2 cells to be cultured for at least 21 days

before experiments (approximately 11-13 days after confluence). Before experiments, cells

were starved overnight in a low-serum (0,5%) media. Conventional medium was replaced

by phenol red- and serum-free medium to avoid interferences with fluorescent tracer’s

measurements. The apical side of human intestinal cell line Caco-2 monolayers was

challenged with chenodeoxycolic acid (CDCA), ursodeoxycholic acid (UDCA), cholic acid

(CA), and deoxycholic acid (DCA).

Cell viability assays: BAs cytotoxicity was investigated using two well-validated assays:

the LDH release and the Trypan blue exclusion tests were performed as previously

described (19).

Transepithelial electrical resistance experiments: transepithelial electrical resistance

(TEER) over a 24-hour period was measured as previously described (19). Briefly, after

removing the Snapwells from the incubator and allowing a 10 minutes acclimatation under

a sterile hood, the entire Snapwell was placed into a resistance chamber (Snap-Endhom;

World Precision Instruments, FL, USA) and connected to a voltmeter (Millipore, Billerica,

of 40

8

MA, USA). A 1,0 cm2 planar electrode was used for all measurements. The screw cap

allowed the planar electrode to remain at the same distance from the monolayers in

repeated measurements. Monolayers with a basal reading between 350 and 450 ohm/cm2

were used for the study.

Paracellular permeability assay: the apical side of Snapwell plates were pre-incubated

with 50 µM of a selected BA for 10’ to allow TJ rearrangement. Then 2 mg/ml of Cascade

blue®-Dextran (m.w.=10 kDa) were added. After 30’ basolateral compartment medium

was replaced with fresh medium, collected and the amount of diffused marker measured

with a Perkin Elmer (Wellesley, MA, USA) 2000 fluorescence spectrophotometer (λex= 365

nm; λem= 440nm). After 120’ the basolateral compartment medium was collected again

and analysed as described above.

Assays with inhibitors: the addition of 50 µM BAs was preceded by a 20’ pre-incubation

with either: a) the anti-EGFR blocking antibody (2 µg/ml; clone 528); b) the selective

inhibitor of src family kinases PP-2 (10 µM; prepared as 10mM stock solution and stored at

–20 °C); c): the MAPk inhibitor PD 098059 (50 µM; prepared as 25 mM stock solution in

DMSO and stored at –20 °C); e): the Pi-3k inhibitor LY294002 (25 µM, prepared as a

1000x stock solution); f): atropine, a muscarinic receptor inverse agonist (0.1 µM). EGF

(1,5; 15 and 80 nM) was added together with the appropriate BA.

Immunoprecipitation and immunoblot analysis: after 10’, 20’ and 120’ of incubation

with the appropriate BA, monolayers were washed twice and scraped with ice cold PBS

without calcium and magnesium. All subsequent steps were performed at 4°C. Lysis was

performed on a rotating shaker for 30’ in either TRIS/HCl TRITON-X 100 buffer for what

concerns EGFR (50 mM TRIS/HCl, pH= 7.4; 1 mM EDTA; 1% TRITON X-100; 50 mM

NaCl; 5 mM MgCl2; 1 mM EGTA; 1 mM Na3VO4; with protease inhibitors cocktail) or in a

of 40

9

NP-40 extraction buffer for experiments with occludin and src (25 mM HEPES; 150 mM

NaCl; 4 mM EDTA; 25 mM NaF; 10 mM sodium pyrophosphate; 1 mM Na3VO4, 1% NP-

40; with protease inhibitors cocktail). Suspensions were centrifuged again for 30’ and the

pellet was discarded. The supernatants were then incubated overnight on a rotating

shaker with either anti-EGFR antibody (Santa Cruz, clone 528; 1.5 µg/mg) or anti-occludin

(3 µg/mg of proteins) or anti-src (mouse; 1:100). After the addition of protein A/G plus-

Agarose (120’ on a rotating shaker), samples were centrifuged (1’; 14.000 r.p.m.) and the

pellet was washed three times with the appropriate ice-cold lysis buffer, resuspended in

sample buffer and boiled 5’. After discarding protein A/G plus-agarose by centrifugation,

samples were run on a 10% (or 7,5% for what concerns EGFR and ErbB-2) SDS-PAGE

and proteins transferred onto a nitrocellulose membrane. Blocking was performed for 2

hours with 5% non-fat dried milk and membranes were incubated in 3% BSA-TBST with

anti-phosphotyrosine (1:1000) or anti-phopsho-src (rabbit, 1:1000). Membranes were

washed in TBST and incubated with horseradish peroxidase-conjugated secondary

antibody (45’; 1:6000 or 45’; 1:2000 for what concerns, respectively, anti-phosphotyrosine

and phospho-src) in 3% non-fat dried milk/TBST and detected with ECL-PLUS according

to the manufacturer’s instructions. Blots were stripped (100 mM 2-mercaptoethanol, 2%

SDS, and 62,5 mM TRIS/HCl pH 7,6) for 30 minutes at 60 °C and then washed in TBST

before reprobing with anti-EGFR (1:1000; rabbit, Clone 1500) or anti-occludin (1:500) or

anti-src (rabbit, 1:1000) overnight in 3% non-fat dried milk on an orbital shaker.

Membranes were washed again in TBST and incubated with horseradish peroxidase-

conjugated secondary antibody (45’; 1:5000 or 45’; 1:2000 for, respectively, anti-occludin

and anti-src) in 3% non-fat dried milk/TBST and detected with ECL-PLUS.

of 40

10

ERK1-2 activation: differentiated Caco-2 monolayers were treated with 50 µM CDCA.

After 10’, 20’ and 120’ cells were washed twice and scraped with ice cold PBS without

calcium and magnesium. All subsequent steps were performed at 4°C. Lysis was

performed for 30’ on a rotating shaker in RIPA buffer (50 mM TRIS pH=7,4; 150 mM NaCl;

1 mM EDTA; 0,25% sodium deoxycholate; 1% NP-40, with the protease inhibitors

cocktail). Suspensions were centrifuged again for 30’ and the pellet was discarded.

Sample were resuspended in sample buffer, boiled 5’ and then loaded onto an 10% SDS-

PAGE gel. Proteins were transferred onto a nitrocellulose membrane. Blocking was

performed for 2 hours with 5% BSA/TBST and membranes were incubated with the

primary antibodies (1:1000) for 2 hours followed by incubation with the appropriate

secondary antibody for 45’ at room temperature. ECL-PLUS detected the immunoreactive

bands according to the manufacturer’s instructions.

Confocal microscopy: Caco-2 cells were grown on glass coverslips, challenged with

50µM CDCA for 20 or 120 minutes and then rinsed three times with 0.01 M pH 7.4 PBS.

Cells were fixed in 3% paraformaldehyde for 15 minutes, permeabilized with 0.2% Triton

X–100 for 5 minutes, blocked in 1% FBS/PBS for 30 minutes at RT, and then rinsed again

in PBS. Staining with monoclonal fluorescent FITC anti-human occludin antibody

(Invitrogen, Carlsbad, CA; USA) 1:50 was performed overnight, at 4°C in a dark, humid

chamber. Staining with TEXAS-RED phalloidin (1:1000) was performed for 45 minutes at

RT. Monolayers were washed three times 10’ in 0.01M pH 7.4 PBS, mounted and then

examined using a confocal fluorescence microscope (Zeiss, LSM 510, Leipzig, Germany).

Statistical analysis. Results are presented as mean±SD. The data were analyzed by

using a commercial software package (GraphPad PRISM; GraphPad Software

Incorporated, San Diego, CA; USA). All data sets passed the normality test and the one-

of 40

11

way analysis of variance, together with the Tukey post hoc test, was used to compare

different groups of data. A p<0.05 was considered statistically significant.

of 40

12

RESULTS

Selected BAs induce TEER decrement and increase paracellular permeability: as

previously reported (23), none of the BAs tested is toxic at a concentration of 50 –100 µM.

A non-significant increase of LDH release and the number of trypan blue positive cells is

observed at 200 µM for all studied BAs, except UDCA. Concentrations above 400 µM

resulted toxic for the cells (not shown). As shown in figure 1A, 50 µM CDCA, DCA, CA, but

not UDCA, induced a transient and statistically significant (p<0.05, n=5) TEER decrement

after 20’ that reverted within 120’. No further changes were observed up to 24 hours.

As TEER decrement is only an indirect method to study TJ efficiency, we tested CDCA,

DCA and CA for their ability to induce a paracellular flux of macromolecules, using a 10

kDa fluorescent dextran. Figure 1B shows that all tested BAs, with the exclusion of UDCA,

were able to promote the apical to basolateral passage of dextran.

Similarly to the TEER decrement, the increase of paracellular permeability occurs within

30 minutes of stimulation and returns to control levels within 120 minutes.

BAs decrease TEER and increase paracellular permeability via EGFR activation:

CDCA and DCA, the most effective BAs tested in the previous set of experiments, were

used in further studies. Both the anti-EGFR blocking antibody (Anti-EGFR 528) and 15 nM

EGF inhibited BA-induced TEER decrement (Figure 2A) and the passage of Cascade

blue® -Dextran from the apical to basolateral side of the chamber (Figure 2B). EGF 1,5

nM was not effective. 80 nM EGF was as effective as 15 nM EGF (not shown).

BAs induce EGFR phosphorylation and occludin de-phosphorylation in a similar

fashion: as shown (figure 3, panel A), the addition of CDCA increased EGFR

phosphorylation, with a peak at 20’. After 120’, phosphorylation level returned similar to

of 40

13

control level. While DCA behaved as CDCA, UDCA induced negligible EGFR

phosphorylation (Figure 3, panel B).

Occludin, which is required in a phosphorylated state at the tight junction level (6), became

de-phosphorylated after monolayers treatment with CDCA (Figure 4 panel A). The effect is

transient, and occurred with a similar timeframe to the TEER decrement, the increase of

paracellular permeability and the EGFR phosphorylation. Again, DCA stimulation of Caco-

2 monolayers gave results similar to CDCA, whereas UDCA stimulation was not effective

(Figure 4 panel B).

Immunoblot with anti-phosphotyrosine antibodies (p-Tyr) or anti occludin (Tot.) of Caco-2

cells lysates, immunoprecipitated with anti-occludin monoclonal antibody, demonstrated

that both anti-EGFR blocking antibody and EGF inhibited the dephosphorylation of

occludin induced by 50 µM CDCA (figure 4, panel C). The stimulation of Caco-2

monolayers with EGF alone or anti-EGFR did not induce occludin dephosphorylation.

These results are in agreement with both the TEER decrement and the increased

paracellular permeability shown in figure 2. The time course of EGFR activation in

response to both EGF and EGF + CDCA is shown in figure 5. EGF induced (Figure 5 A) a

sustained and continuous increase of phosphorylation up to 120 minutes. Co-stimulation

with both CDCA and EGF caused a mixed phosphorylation pattern, with a peak at 20

minutes and a still sustained phosphorylation after 120 minutes. Figure 5 B shows the

densitometric analysis of three separate experiments and compares the effect of CDCA

and EGF alone and in combination on EGFR phosphorylation.

BAs-induced activation of EGFR decreases TEER via the activation of src family

kinases: rearrangement of tight junctions can occur via several biochemical pathways,

including MAPk (4), src (5) and PI-3k (25). Using selective inhibitors, we show that while

of 40

14

pre-treatment with PP-2, a src family kinases inhibitor, was able to abolish BAs-induced

TEER decrement, both the inhibitors of ERK 1-2 (PD98059) and Pi-3k (LY294002) had no

effect (Figure 6, panel A). Unrelated to a permeability effect, we report that CDCA is able

to induce a transient phosphorylation of ERK1-2 after 20’ on Caco-2 monolayers (not

shown).

Blots for phosphorylated src of cell lysates stimulated with 50 µM CDCA and

immunoprecipitated for occludin, revealed an increase of p-src recovered with occludin

(Figure 6 panel B). Blots for total src gave similar results (not shown), indicating that, upon

CDCA stimulation, occludin binds a greater amount of activated src.

In separate experiments we investigated the possible role of muscarinic receptor in BAs

induced TEER decrement with the use of 0,1µM atropine, a muscarinic receptor inverse

agonist. Results, shown in Figure 7, exclude a possible involvement of such receptor in

BAs activation of EGFR.

BAs induce a rearrangement of occludin at the tight junction level: confocal

microscopy allows scanning thin planes of a sample along the Z optical axis by controlled

micrometer increments. Figure 8 shows the results of a treaty Caco-2 monolayers with 50

µM CDCA. As shown in panel A, the scanning from top to bottom of an untreated Caco-2

monolayer demonstrates a distinct, continuous line of fluorescence spanning through a

limited number of sections (3 - 4) describing an homogenous occludin ring along the entire

perimeter of the cells. Panel B shows a monolayer treated for 20 minutes with 50 µM

CDCA. Most planes have a patchy organization of intracellular occludin and its staining is

not limited to some sections, but it can be observed in almost all the optical sections

suggesting that a significant occludin redistribution had occurred. Figure 8, panel C, shows

the redistribution of occludin along the z-axis occurring upon 50 µM CDCA stimulation.

of 40

15

Results of a double staining approach with anti-FITC-occludin antibody and TEXAS RED-

phalloidin, are shown in figure 9. In this figure we report the images of Caco-2 monolayers

after 50 µM CDCA treatment and co-incubation with EGF, anti-EGFR and PP-2. As shown

and in agreement with results reported in figures 2, 4 (panel D) and 6 (panel A), 50 µM

CDCA-induced occludin and actin rearrangement is inhibited by EGF, anti-EGFR and PP-

2.

of 40

16

Discussion

BAs play an important role in normal digestive processes and they also have been used as

therapeutic agents for the last decades (2,15). The biological effects depend on the

chemical structure of the individual BA and its concentration. In the millimolar range,

CDCA and DCA may induce cell damage due to membrane solubilization and promote the

loss of cholesterol and phospholipides (13). At lower concentrations, BAs affect host

defence (29), cause hydroperoxide generation and impair the function of the electron

transport chain in isolated rat liver mitochondria (27, 26). They may also function as

immunosuppressive agents (17) and are able to induce apoptosis and carcinogenesis.

DCA, in particular, stimulates the proliferation of colonic epithelium (11). BAs may also

activate EGFR in human cholangiocyte cell line (30) and induce intracellular signalling in

HCT116 cells (22). Remarkably, CDCA epimer, UDCA, does not exhibit any of these

effects.

The standard concentration of 50 µM used in our experiments is relevant to most clinical

situations. In fact, although normal serum concentration of each specific unconjugated bile

acid ranges between 0.2 and 1 µM, it may increase up to 150 µM in chronic cholestatic

disease (24). Moreover, tissue concentrations of about 250 µM are common in cholestasis

(17, 12).

The mechanisms underlying the multifaceted effects of bile acids are under intensive

investigations. Recently, using an in vivo rat model, BAs have been linked to the intestinal

epithelial barrier dysfunction and pathogenesis of NEC. It has also been reported a

protective effect of EGF treatment in the genesis of NEC (15,9).

In this paper we provide evidence that treatment with DCA and CDCA, but not UDCA, can

induce EGFR phosphorylation in Caco-2 monolayers resulting in an increased paracellular

of 40

17

permeability via occludin de-phosphorylation and cytoskeletal rearrangement at the TJ

level.

EGFR is an intriguing molecule involved in many fundamental biological processes such

as cell proliferation, differentiation, tissue healing, cell cycle progression, and cancer.

EGFR is a member of the ErbB family of receptor tyrosine kinase, which consist of an

extra-cellular ligand-binding domain, a transmembrane lipophilic domain and an

intracellular tyrosine-kinase domain. Phosphorylation of the tyrosine kinase portion leads

to receptor auto-phosphorylation and consequent homodimerization or heterodimerization

with receptors of the same family and, finally, to protein activation and signalling. EGFR

downstream effectors include ERK1/2, the phosphatidylinositol 3-kinase pathway, the src

family of kinases, and the JAKs and STATs pathways.

The link between BAs, EGFR and increased paracellular permeability is provided by the

blocking effect of anti EGFR antibodies. It is not entirely clear whether this inhibition is

ligand dependent since it is not known, at the best of our knowledge, if the binding of the

528 blocking antibody to the receptor is also able to inhibit the receptor auto-

phosphorylation or if it might have some effects on receptor mobility in the membrane,

influencing the receptor dimerization and/or auto-phosphorylation, all critical events to

distinguish between a ligand dependent or independent EGFR activation.

We also show that EGF and BAs trigger EGFR in a different fashion: BAs-induced

activation is rapid, peaks at 20 minutes and returns to baseline within 120 minutes (Figure

3), whereas EGF-induced phosphorylation shows a sustained phosphorylation of the

receptor up to 120 minutes (Figure 5). When the receptor was stimulated with both EGF

and CDCA, densitometry showed a mixed pattern of phosphorylation with a peak at 20

minutes and a still sustained phosphorylation at 120 minutes (Figure 5). These results

of 40

18

demonstrate that EGFR activation can be modulated in intensity and duration of activation

depending on the nature of the stimulus. We speculate that this signal modulation may

influence the cellular response. In other terms, EGFR may trigger different pathways

depending on the modality of the activation, namely either the BA’s induced transient

activation involved in permeability effects, or the sustained EGF mediated activation

initiating multiple cascades. The addition of EGF, the natural ligand, may stabilize the

formation of dimers, inducing a prolonged signalling and, in turn, a different response.

However, the mechanism by which EGF prevents the effects of BAs requires further

experiments; e.g. we cannot exclude that EGF may have additional modes of action, such

as influencing receptor internalization.

We tried to address the issue of how BAs could be linked to EGFR phosphorylation. Since

EGFR may form heterodimers with other members of the ErbB family and that ErbB-2 is

the preferred heterodimerization partner (8), we investigated the possibility that BAs could

induce EGFR/ErbB2 heterodimers. This hypothesis was abandoned when no band

compatible with the ErbB-2 receptor was found in our EGFR co-immunoprecipitation

experiments (not shown). We then also considered the possibility that BAs induced

signalling via muscarinic receptor and trans-activation of EGFR, as described in others cell

types (7, 28). We provide direct negative evidence against this hypothesis. In fact,

atropine, a specific muscarinic receptor inverse agonist was not able to abolish TEER

decrement (Figure 7). The precise mode of interaction of BAs, EGF, EGFR and its

phosphorylation requires further studies and may well need physics and chemistry

expertises that are currently beyond our reach.

We, though, have shed some light into the biochemical events coupling BAs induced-

EGFR activation to paracellular permeability. MAPk, src and Pi-3k are potential targets of

of 40

19

EGFR, and they have been described to modulate TJ in a variety of cells types. In

particular, MAPk may be transiently activated by BAs (4). The involvement of MAPk and

PI-3k pathways was excluded by inhibition experiments (Figure 6A). We found that only

the specific inhibitor of the src family kinases, PP-2, was able to inhibit BAs-induced TEER

decrement. Previous papers described a role of non-receptor tyrosine kinases in occludin

phosphorylation/dephosphorylation and disassembly/reassembly at the TJ level but

conflicting data have been generated (6; 25; 10). Our data show that activated src co-

immunoprecipitates with occludin, and that the amount found in the immunoprecipitate

varies in a time-dependent fashion with a peak at 20 minutes after CDCA stimulation;

suggesting that activated src plays a main role in TJ rearrangement (Figure 6 B).

The effect of BAs on TJ rearrangement was finally studied by confocal microscopy. It has

been described that TJ are connected to the actin cytoskeleton by zonula occludens

proteins and cingulin. Other authors reported morphological changes of both actin and

occludin during TJ rearrangement (16; 18). Our data clearly show that CDCA induces a

marked redistribution of intracellular occludin. Untreated Caco-2 cells show a linear

distribution of occludin along the cell perimeter. In monolayers incubated with 50 µmol/L

CDCA, no organized structure was detected and fluorescence was distributed in a

scattered and clusterized fashion along all sections (Figure 8). Accordingly, double

staining with occludin and actin shows a regular immunofluorescence pattern in control,

while in the sample treated with CDCA the regular fluorescence pattern along the cell

membrane is missing for both actin and occludin, and actin stress-fibers can also be

observed (Figure 9). Treatment with EGF, EGFR and PP2 reversed the

immunofluorescence pattern to the control level (Figure 9).

of 40

20

In conclusion, EGFR is a crucial intermediate of BAs modulation of Caco-2 tight junctions.

BAs-induced increase of paracellular permeability can be abolished both by preventing

EGFR activation with a specific inhibitor or with the engagement of the receptor by the

proper ligand, indicating that either EGFR full activation or its complete inhibition blocks

BAs effect on intestinal epithelial permeability.

Our in vitro data provide a useful paradigm to the recent in vivo observations suggesting a

role for BAs in the genesis of NEC; at the same time, they offer a biological mechanism on

the potential of EGF in the prevention and therapy of the most frequent gastrointestinal

emergency in neonates (15; 9).

of 40

21

References

1 Araki Y, Katoh T, Ogawa A, Bamba S, Andoh A, Koyama S, Fujiyama Y, Bamba T. Bile

acids modulates transepithelial permeability via the generation of reactive oxygen

species in the Caco-2 cell line. Free radical biology and medicine. 39: 769-780, 2005

2 Balistrieri FW. Bile acid therapy in pediatric hepatobiliary disease: the role of

ursodeoxycholic acid. J Pediatr Gastroenterol Nutr. 24:573-589, 1997

3 Baroni GS, Ridolfi F, Hannivoort R, Saccomanno S, Homan M, De Minicis S, Jansen

PLM, Candelaresi C, Benedetti A, and Moshage H. Bile acids induce hepatic stellate

cell proliferation via activation of the epidermal growth factor receptor.

Gastroenterology 128:1042–1055, 2005

4 Basuroy S, Sheth A, Elias B, Naren AP, and Rao R. MAPk interacts with occludin and

mediates EGF-induced prevention of tight junction disruption by hydrogen peroxide.

Biochem J 393:69-77, 2006

5 Basuroy S, Sheth P, Kuppuswamy D, Balasubramanian S, Ray RM, and Rao RK.

Expression of Kinase-inactive c-src delays oxidative stress-induced disassembly and

accelerates calcium-mediated reassembly of tight junctions in the Caco-2 cell

monolayers. J Biol Chem 278:11916-11924, 2003

6 Chen YH, Lu Q, Goodenough DA and Jeansonne B. Non receptor Tyrosine kinase c-Yes

interacts with occludin during tight junction formation in canine kidney epithelial cells.

Mol Biol Cell 13:1227-1237, 2002

7 Cheng K, Raufman JP. Bile acids-induced proliferation of human colon cancer cell line is

mediated by transactivation of epidermal growth factor receptors. Biochem Pharmacol

70:1035-47, 2005

of 40

22

8 Citri A, Skaria KB, and Yarden Y. The deaf and the dumb: the biology of ErbB-2 and

ErbB-3. Exp cell res 284:54-65, 2003

9 Clark JA, Doelle SM, Halpern MD, Saunders TA, Holubec H ; Dvorak K, Boitano SA and

Dvorak B. Intestinal barrier failure during experimental necrotizing enterocolitis :

protective effect of EGF treatment. Am. J. Physiol Gastrointest Liver Physiol 291:G938-

G949, 2006

10 Clump DA, Qazi IH, Sudol M, Flynn DC. c-Yes response to growth factor activation.

Growth factors 23:263-272, 2005

11 Debruine PR, Bruineel EA, Li X, Zimber A, Gespach C, Mareel MM. The role of bile

acids in carcinogenesis. Mutat Res. 2001; 480-481:359-369

12 Greim H, Trulzsch D. Czygan P., Rudick J., Hutter F., Schaffner F., Popper H.

Mechanism of cholestasis 6: bile acids in human liver with or without biliary obstruction.

Gastroenterology 65:846-850, 1972

13 Guldutuna S, Zimmer G, Imhof M, Batti S, You T, and Leuschner U. Molecular aspect

of membrane stabilization by ursodeoxycholate. Gastroenterology 104:1736-1744,

1993

14 Hackam DJ, Upperman JS, Grishin A, Ford HR. Disordered enterocyte signaling and

intestinal barrier dysfunction in the pathogenesis of necrotizing enterocolitis. Semin

Pediatr Surg. 2005 Feb;14(1):49-57

15 Halpern MD,Holubec H; Saunders TA; Dvorak K; Clark JA; Doelle SM; Ballatori N; and

Dvorak B. Bile Acids Induce Ileal Damage During Experimental Necrotizing

Enterocolitis. Gastroenterology; 130:359-372, 2006

of 40

23

16 Kale G, Naren AP, Shet P, and Rao RK. Tyrosine phosphorylation of occludin

attenuates its interaction with ZO-1, ZO-2 and ZO-3. Biochem Biophys Res Commun

302:324-329, 2003

17 Lacaille F and Khazal P. The Immunosuppressive Effect of Ursodeoxycholic Acid: a

Comparative In Vitro Study on Human Peripheral Blood Mononuclear Cells.

Hepatology 18:165-172, 1993

18 Lee HS, Namkoong K, Kim KJ, Cheong YH, Kim SS, Lee WB, Kim KY. Hydrogen

peroxide-induced alterations of tight junction proteins in bovine brain microvascular

endothelial cells. Microvasc Res. 2004; 68:231-8

19 Raimondi F, Crivaro V, Capasso L, Maiuri L, Santoro P, Tucci M, Barone MV,

Pappacoda S, Paludetto R. Unconjugated bilirubin modulates the intestinal epithelial

barrier function in a human-derived in vitro model. Pediatr Res 60:30-3, 2006

20 Raimondi F, Kao JP, Fiorentini C, Fabbri A, Donelli G, Gasparini N. Enterotoxicity and

cytotoxicity of Vibrio parahaemolyticus thermostable direct hemolysin in in vitro

systems. Infect Immun 68:3180-85, 2000

21 Raimondi F, Santoro P, Maiuri L, Londei M, Annunziata S, Ciccimarra F. Reactive

nitrogen species modulate the effects of rhein, an active component of senna laxatives,

on human epithelium in vitro. J Pediatr Gastroenterol Nutr 34: 529-34, 2002

22 Samira JL, Akare S, Ali MA, Mash EA Jr, Meuillet E, Martinez JD. Deoxycholic acid

induces intracellular signalling through membrane perturbation. J Biol Chem

281:14948-60, 2006

23 Santoro P, Raimondi F, Annunziata S, Paludetto R, Annella T, and Ciccimarra F.

Unconjugated bile acids modulates adult and neonatal neutrophil chemotaxis induced

ion vitro by N-formyl-Met-Leu-Phe-peptide. Pediatr Res 51:392-396, 2002

of 40

24

24 Setchell KDR, Lawson AM, Blackstock EY, Murphy GM. 1982 Diurnal changes in

serum unconjugated bile acids in normal man. Gut 23:637-642, 1982

25 Sheth P, Kuppuswarmy S, Li C, Naren AP, and Rao RK. Role of phosphatidylinositol 3-

kinase in oxidative stress-induced disruption of tight junction. J Biol Chem 278; 49239-

49245, 2003

26 Sokol RJ, McKIm IM Jr., Goff MC, Ruyle SZ, Devereaux MW, Han D, Packer L, and

Everson G. Vitamin E Reduces Oxidant Injury to Miochondria and the Hepatotoxicity of

Taurochenodeoxicholic Acid in the Rat. Gastroenterology 114:164-174, 1998

27 Sokol RJ, Winklhofer-Roob BM, Devereaux MW, and McKim Jr. JM. Generation of

hydroperoxides is isolated rat hepatocytes and hepatic mitochondria exposed to

hydrophobic bile acids. Gastroenterology 109:1249-1256, 1995

28 Stenson WF, Eason RA, Riehl TE and Turk J. Regulation of paracellular permeability in

Caco-2 cell monolayers by protein kinase C. Am. J. Physiol 265:G955-62, 1993

29 Swain MG, Tjandra K, Kanwar S, and Kubes P. Neutrophil Adhesion is Impaired in a

Rat Model of Cholestasis. Gastroenterology 109:923-932, 1995

30 Wernerburg NW, Yoon JH, Higuchi H, and Gores GJ. Bile acids activate EGF receptor

via a TG-α-dependent mechanism in human cholangiocyte cell line. Am J Physiol

285:G31-G36, 2003

of 40

Figure 1. Selected bile acids induce TEER decrement and increase paracellular permeability. Panel A: Time course of 50 µM luminally applied BAs. UDCA was not

effective; CDCA, DCA and CA induced a transient TEER decrement, which peaked at 20 minutes. Within 120 minutes TEER returned to baseline. n=10; (*): p< 0.05 versus

control; data are presented as percent decrease of TEER. Panel B: Spectrophotometric analysis of dextran fluxes (2mg/ml; applied as described in materials and methods). CDCA, DCA, and CA induced a paracellular passage of 10 kDa cascade bleu conjugated dextran. The effect occurred within 30 minutes of stimulation and returned to baseline

within 120 minutes. n=10; (*): p< 0.05 versus control; data are presented as percentual decrease of TEER.

173x116mm (600 x 600 DPI)

of 40

119x98mm (600 x 600 DPI)

of 40

Figure 2. BAs decrease TEER and increase paracellular permeability via EGFR activation. Panel A: Both anti-EGFR blocking antibody (clone 528; 2µg/ml) and 15 nM EGF blocked

CDCA and DCA (50 µM)-induced TEER decrement (time: 20 minutes). Panel B. Paracellular fluxes of dextran (2 mg/ml), measured as described in paracellular permeability

assays , were abolished in the presence of both anti-EGFR (clone 528; 2µg/ml) and 15 nM EGF. n=5; (*): p<0.05 versus control; results were normalized to the TEER value of

the control. 111x87mm (600 x 600 DPI)

of 40

145x150mm (600 x 600 DPI)

of 40

Figure 3. BAs induce EGFR phosphorylation. Immunoblot with anti-phosphotyrosine antibodies (p-Tyr) or anti EGFR (Tot.) of Caco-2 cells lysates immunoprecipitated with

anti-EGFR monoclonal antibody (clone 528). Panel A: Time course of CDCA (50 µM) induced EGFR activation. Total EGFR (m.w.: 170 kDa) was revealed with a polyclonal

antibody after stripping the blots; and activated EGFR was revealed with anti-phosphotyrosine monoclonal antibody. EGFR phosphorylation peaked at 20 minutes and

returned to baseline within 120 minutes. Panel B: figure shows total EGFR and EGFR activation at 20 minutes induced by DCA, CDCA and UDCA (50 µM). Only CDCA and DCA elicited an increment of EGFR phosphorylation, while UDCA had no effect. Figures are

representative of three separate experiments, showing similar results. Numbers at the bottom of the blots report densitometric data and represent the increase of tyrosine phosphorylation versus controls, and are normalized for the total protein amount.

254x190mm (96 x 96 DPI)

of 40

Figure 4. BAs induce occludin de-phosphorylation. Immunoblot with anti-phosphotyrosine antibodies (p-Tyr) or anti occludin (Tot.) of Caco-2 cells lysates immunoprecipitated with

anti-occludin monoclonal antibody. Panel A: Time course of occludin phosphorylation induced by 50 µM CDCA treatment. Total occludin was revealed after stripping the membranes, and phosphorylated occludin was revealed with anti-phosphotyrosine

monoclonal antibody. CDCA elicited a transient de-phosphorylation of occludin at 20 minutes. After 120 minutes occludin returned similar to control. Panel B: at 20 minutes DCA and CDCA (50 µM) induced a similar decrement of occludin phosphorylation, while

UDCA (50 µM) was similar to control. Figures are representative of three separate experiments, showing similar results. Panel C: Both anti EGFR and EGF inhibited the

dephosphorylation of occludin induced by CDCA. The stimulation of Caco-2 monolayers with EGF alone or anti-EGFR had no effect on occludin phosphorylation. Numbers at the

bottom of the blots report densitometric data and represent the increase of tyrosine phosphorylation versus the control, and are normalized for the total protein amount.

Figure is representative of three separate experiments showing similar results. Panel D: Densitometric analysis of three similar experiments in the form of a bar chart comparing

the effects of CDCA alone, EGF alone, and their combination on EGFR phosphorylation. 254x190mm (96 x 96 DPI)

of 40

254x190mm (96 x 96 DPI)

of 40

124x126mm (600 x 600 DPI)

of 40

Figure 5. BAs and EGF trigger EGFR in a different fashion. Panel A: Time course of EGFR activation in response to EGF (15 nM) and EGF + CDCA (50 µM). Immunoblot of Caco-2

lysates immunoprecipitated with anti-EGFR monoclonal antibody. Activated EGFR (P-Tyr.)was revealed with anti-phosphotyrosine antibodies; EGFR (Tot.) was revealed with anti-EGFR polyclonal antibody. Numbers at the bottom of the blots report densitometric data

and represent the increase of tyrosine phosphorylation versus the control, and are normalized for the total amount of the protein. Panel B: Densitometric analysis of three separate experiments comparing the effect of CDCA, EGF and their combination on EGFR phosphorylation. Phosphorylation is expressed as fold-increase over the corresponding

untreated sample. Results are expressed as mean ± standard deviation. 254x190mm (96 x 96 DPI)

of 40

121x92mm (600 x 600 DPI)

of 40

Figure 6. BAs decrease TEER and induce occludin de-phosphorylation via the activation of Src family kinases. Panel A: Pre-treatment of monolayers with the selective inhibitors of MAPK (PD 098059; 50 µM) and Pi-3k (LY294002; 25 µM) did not block TEER decrement. The selective inhibitor of src family kinases (PP-2; 10 µM) prevented TEER decrement.

n=5; (*): p<0.05. Panel B: Blots for phosphorylated Src (p-Src) of cell lysates stimulated with 50 µM CDCA and immunoprecipitated for occludin revealed an increase of p-Src

recovered with occludin. Numbers at the bottom of the blots report densitometric data and represent the increase of tyrosine phosphorylation versus the control, and are

normalized for the total protein amount. 175x124mm (600 x 600 DPI)

of 40

254x190mm (96 x 96 DPI)

of 40

Figure 7. The muscarinic receptor is not involved in BAs-induced activation of EGFR. Pre-treatment of monolayers with 0,1 µM atropine did not affect BAs induced TEER decrement

(n=3; p<0.05). 91x57mm (600 x 600 DPI)

of 40

Figure 8. BAs induce a rearrangement of occludin at the tight junction level. Optical sections along the Z axis (1micron): A) Untreated Caco-2 cells show a linear distribution

of occludin along the cell perimeter restricted to sections 3 - 4. B) In monolayers incubated for 20 minutes with 50µM CDCA, no organized structure is detected and

fluorescence is scattered along all sections (40x objective). DCA gave similar results (not shown). Figure is representative of five separate experiments showing similar results. C)

Confocal fluorescence localization of occludin along the z-axis. 254x190mm (96 x 96 DPI)

of 40

Figure 9. Effect of EGF, Ab-EGFR and PP-2 on occludin rearrangement at the TJ level. Treated Caco-2 monolayers were stained with both anti-FITC-occludin antibody and

TEXAS RED-phalloidin. The rearrangement of occludin and actin induced by the incubation of monolayer with 50 µM CDCA is inhibited by EGF, anti-EGFR and PP-2.

261x175mm (150 x 150 DPI)

of 40

Bile acids modulate tight junction structure and barrier function of Caco-2 monolayers via EGFR...

Documents

Transcript of Bile acids modulate tight junction structure and barrier function of Caco-2 monolayers via EGFR...