Intestinal epithelial tight junction barrier regulation by autophagy-related protein ATG6/beclin 1

Abstract

A defective tight junction (TJ) barrier is a key pathogenic factor for inflammatory bowel disease. Previously, we have shown that autophagy, a cell survival mechanism, enhances intestinal epithelial TJ barrier function. Autophagy-related protein-6 (ATG6/beclin 1), a key protein in the autophagy pathway, also plays a role in the endocytic pathway. The constitutive role of beclin 1 in the intestinal TJ barrier is not known. In Caco-2 cells, beclin 1 was found to be coimmunoprecipitated with the TJ protein occludin and colocalized with occludin on the membrane. Treatment of Caco-2 cells with beclin 1 peptide [transactivating regulatory protein (Tat)-beclin 1] reduced TJ barrier function. Activation of beclin 1 increased occludin endocytosis and reduced total occludin protein level. In contrast, beclin 1 siRNA transfection enhanced Caco-2 TJ barrier function. In pharmacologic and genetic autophagy inhibition studies, the constitutive function of beclin 1 in the TJ barrier was found to be autophagy independent. However, de novo induction of autophagy with starvation or rapamycin prevented Tat-beclin 1-induced increase in TJ permeability and reduction in occludin level. Induction of autophagy also resulted in reduced beclin 1-occludin association. In mouse colon, beclin 1 colocalized with occludin on the epithelial membrane. Perfusion of mouse colon with beclin 1 peptide caused an increase in colonic TJ permeability that was prevented by in vivo induction of autophagy. These findings show that beclin 1 plays a constitutive, autophagy-independent role in the regulation of intestinal TJ barrier function via endocytosis of occludin. Autophagy terminates constitutive beclin 1 function in the TJ barrier and enhances the TJ barrier.

INTRODUCTION

The intestinal epithelium plays a vital role in forming a physical and interactive barrier between the intestinal mucosa and the luminal environment. The intestinal epithelium regulates transport of water, ions, and nutrients and also provides a barrier against toxins and pathogens (17). The apical intercellular tight junctions (TJs) are responsible for the paracellular barrier function and regulate transepithelial flux of ions and solutes through the paracellular space. Increased intestinal permeability caused by defects in the intestinal epithelial TJ barrier is considered to be an important pathogenic factor for the development of intestinal inflammation (1–4). The defects in the intestinal TJ barrier allow increased antigenic penetration, resulting in amplified inflammatory response in inflammatory bowel disease (IBD), celiac disease, necrotizing enterocolitis, and ischemia-reperfusion injury (1, 2). Conversely, the enhancement or retightening of the intestinal TJ barrier has been shown to benefit the resolution of intestinal inflammation in both animal models of IBD and human IBD (3–5). Our previous studies have demonstrated that autophagy plays a critical role in epithelial TJ barrier function (34).

Autophagy, a conserved cell survival mechanism, plays an important role in diverse processes such as metabolic stress, neurodegeneration, cancer, aging, immunity, and inflammatory diseases (6). Autophagy, a process of engulfment and degradation of cellular proteins including damaged organelles and long-lived and misfolded proteins, is active in the normal colonic intestinal mucosa (7). It also plays a role in intestinal cell survival during physiological stress (8). Although clinical data show a direct link between a defective intestinal TJ barrier and persistent, prolonged intestinal inflammation in patients with IBD (5, 16, 17), the role of autophagy in the regulation of the intestinal epithelial TJ barrier remains unknown. Beclin 1 (BECN1), the mammalian ortholog of autophagy-related protein-6 (ATG6), plays a central role in the autophagy cascade by interacting with over 20 cellular proteins through conformational flexibility. Biallelic loss of beclin 1 results in early embryonic death in mice, whereas monoallelic loss causes phenotypes such as increased rate of spontaneous malignancies, increased susceptibility to Alzheimer’s-like disease and respiratory syncytial virus infection, and reduced exercise endurance (14). Besides interacting with Bcl-2 and Bcl-xL in regulating autophagosome nucleation and maturation, beclin 1 has been shown to be involved in multiple vesicle trafficking and endocytosis pathways (16, 21, 43). Beclin 1 performs these functions as a core component of the class III phosphatidylinositol 3-kinase complex (PI3KC) and, also, via binding to an activating molecule in beclin 1-regulated autophagy protein-1 (Ambra1), UV radiation resistance-associated gene (UVRAG), and ATG14L (14, 27, 46). The aim of this study was to examine the role of beclin 1 in intestinal TJ barrier function and the intersection of autophagy and endocytosis in this process. Our data show that beclin 1 plays an autophagy-independent role in the constitutive regulation of intestinal TJ barrier function via endocytosis of occludin. Induction of autophagy terminates constitutive beclin 1 function in the TJ barrier and reduces TJ barrier permeability.

EXPERIMENTAL PROCEDURES

Cell culture and reagents.

Caco-2 and T84 cells obtained from American Type Culture Collection were grown on 0.4-μm-pore size, 12-mm-diameter inserts maintained at 37°C in DMEM culture medium supplemented with 10% fetal bovine serum. The transepithelial electrical resistance (TER) of the filter-grown cells was measured by an epithelial voltohmmeter (World Precision Instruments, Sarasota, FL), and Caco-2 monolayers with TER of ~450 Ω·cm² were used for all experiments. The IPEC-J2 cell line was a kind gift from Dr. Adam Moeser, Michigan State University, East Lansing, Michigan. Starvation was induced by incubation of filter-grown monolayers in Earle’s balanced salt solution (E-3024; Sigma-Aldrich). The other reagents used were autophagy inducer rapamycin (PHZ-1235; Life Technologies) and autophagy inhibitor bafilomycin A1 (sc-201550; Santa Cruz Biotechnology). Primary antibodies used in this study included occludin (33-1500; Invitrogen; and LS-C-140253-100; LifeSpan BioSciences), zona occludens protein-1 (ZO-1, 617300; Invitrogen), claudin-1 (51-9000; Invitrogen), claudin-2 (51-6100; Invitrogen), claudin-3 (34-1700; Thermo Fisher Scientific), claudin-4 (PA-5-16875; Thermo Fisher Scientific), caveolin-1 (3238; Cell Signaling Technology), early endosome antigen 1 (EEA1, E-7659; Sigma-Aldrich), Ras-related protein Rab-5 (Rab5, R-4654; Sigma-Aldrich), beclin 1 (ab-114071 and ab-207612; Abcam), microtubule-associated protein 1A/1B-light chain 3B (LC3B, L-7543; Sigma-Aldrich), ATG16 (PA5-35207; Thermo Fisher Scientific), lysosome-associated membrane protein-2 (LAMP2, NBP-2-22217; Novus Biologicals), and β-actin (MA-5-15739; Thermo Fisher Scientific). Transactivating regulatory protein (Tat)-beclin 1 peptide and Tat-inactive scrambled peptide were purchased from Novus Biologicals (NBP-2-49888 and NBP-2-49887, respectively).

Determination of paracellular permeability.

Besides the measurements of TER, paracellular permeability was determined using [³H]mannitol (molar mass = 182 g/mol) flux measurements. For determination of the apical-to-basal flux rate, a known concentration (1.5 μM) of [³H]mannitol (NET-101250-UC; PerkinElmer) was added to the apical solution, and radioactivity was measured in basal solution using a scintillation counter, as described previously (28).

Cell transfections.

Caco-2 monolayers were transiently transfected with beclin 1 siRNA (Dharmacon) using DharmaFect transfection reagent (Lafayette, CO), as described previously (35). In brief, confluent Caco-2 monolayers on 12-well Transwell plates were treated with 5 ng (0.5 nmol) of the siRNA and DharmaFect transfection reagent in Accell media (Thermo Fisher Scientific). The efficiency of silencing was confirmed by Western blot analysis after 72 h of treatment.

Western blot analysis for assessment of protein expression.

To study the protein expression, Caco-2 monolayers were rinsed with ice-cold PBS, and cells were lysed with lysis buffer (50 mM Tris·HCl, pH 7.5, 150 mM NaCl, 500 μM NaF, 2 mM EDTA, 100 μM vanadate, 100 μM phenylmethylsulfonyl fluoride, 1 μg/ml leupeptin, 1 μg/ml pepstatin A, 40 mM para-nitrophenyl phosphate, 1 μg/ml aprotinin, and 1% Triton X-100). The cell lysates were clarified (2,000 rpm, 2 min) and centrifuged (10,000 rpm, 10 min), and the supernatant was saved. Protein quantification of the extracted aliquots was performed using BCA Protein Assay Kit (Pierce, Rockford, IL). After addition of Laemmli gel loading buffer and boiling for 5 min, equal amounts of protein were loaded and separated on SDS-PAGE gel. Proteins from the gel were transferred to a nitrocellulose membrane, and the membrane was incubated for 2 h in blocking solution (5% dry milk in TBS-Tween 20 buffer). The membrane was incubated with the appropriate primary antibody in blocking solution. After being washed in TBS-0.1% Tween buffer, the membrane was incubated in appropriate secondary antibody [horseradish peroxidase (HRP)-goat anti-rabbit IgG, cat. no. 31460, or HRP-goat anti-mouse IgG, cat. no. 31430; Invitrogen] and developed using Santa Cruz Western Blotting Luminol Reagents (Santa Cruz Biotechnology) on Kodak BioMax MS film (Fisher Scientific). All the primary and secondary antibodies were diluted to the concentrations suggested by the manufacturers.

Confocal immunostaining.

Immunolocalization of various proteins was assessed by confocal immunofluorescence. Caco-2 monolayers were washed twice in cold PBS, fixed with 2% paraformaldehyde for 10 min, and permeabilized with 0.1% Triton X-100 in PBS at room temperature for 5 min. The cell monolayers were then blocked in normal serum and labeled with primary antibodies in blocking solution overnight at 4°C. After being washed with PBS, the cells were incubated in Alexa Fluor 488- (A-11029 and A-11034; Thermo Fisher Scientific; and NBP-1-72854; Novus Biologicals), Cy3- (111-165-003 and 115-165-003; Jackson ImmunoResearch Laboratories), or Cy5-conjugated (111-605-003; Jackson ImmunoResearch Laboratories) secondary antibodies. All the primary and secondary antibodies were used at the concentrations suggested by the manufacturers. ProLong Gold antifade reagent (Invitrogen), containing DAPI as a nuclear stain, was used to mount the cell filters on glass slides. The slides were examined using a confocal fluorescence microscope (Leica SP8). Images were processed with LAS X software (Leica Microsystems).

Coimmunoprecipitation analysis.

Coimmunoprecipitation studies were performed using Dynabeads Protein G, as per manufacturer’s instructions (Life Technologies). Dynabeads were incubated with primary antibody, washed, and then incubated with sample lysates. Immunoprecipitates were separated by SDS-PAGE and further analyzed by Western blot (32).

Occludin endocytosis assay.

Assay for occludin endocytosis was performed with modifications in the protocol indicated by Pierce Cell Surface Protein Isolation Kit (Thermo Fisher Scientific) and previous reports (26, 32). Briefly, cell surface proteins on cell monolayers were biotinylated with EZ-Link Sulfo-NHS-SS-Biotin (Pierce), quenched with 50 mM NH₄Cl in PBS containing 0.9 mM CaCl₂ and 0.33 mM MgCl₂ (PBS/CM) at 4°C, and incubated at 37°C for 30 min in normal media to allow endocytosis. The remaining biotin on the cell surface was stripped with 50 mM sodium 2-mercaptoethanesulfonate in 100 mM Tris·HCl (pH 8.6) containing 100 mM NaCl and 2.5 mM CaCl₂ at 4°C for 30 min and quenched with 5 mg/ml iodoacetamide in PBS/CM at 4°C for 15 min. The cells were lysed with lysis buffer (Pierce), and aliquots were taken to determine the total amount of cargo protein (occludin) expressed in the cells. Biotinylated cargo proteins were then isolated with UltraLink Immobilized NeutrAvidin Plus beads (Pierce) and analyzed by Western blot using anti-occludin antibody.

HRP and transferrin uptake assay.

Assay for HRP uptake was conducted as previously described (5). A confluent monolayer of Caco-2 cells, treated with Tat-beclin 1 or Tat-scrambled peptide, was incubated with HRP (1 mg/ml) in DMEM for 1 h at 37°C. The cells were washed twice with HBSS to remove the unbound HRP and harvested using 0.05% trypsin-EDTA. The cells were washed twice with ice-cold PBS and incubated with 1 ml buffer solution (1 mM CaCl₂, 1 mM MgCl₂, and 0.2% BSA) for 5 min on ice. The cells were then washed and sonicated in 200 µl of PBS. HRP enzyme activity was assessed by treating 5 µl of cell lysate with 95 µl of freshly made solution containing 0.5 N sodium formate, 0.75 mg/ml o-phenylenediamine, and 0.006% H₂O₂. The assay was incubated at room temperature for 5 min, and the reaction was stopped by adding 100 µl of 0.1 N H₂SO₄. The plate was then read at 490 nm, and the readings were normalized to the total protein concentration. For transferrin uptake, Caco-2 cells were serum starved for 30 min and incubated with 50 µg/ml of Alexa Fluor 488-conjugated human transferrin (T-13342; Molecular Probes). After washing, cell were fixed and stained with DAPI for nuclear staining using standard techniques.

Determination of paracellular permeability in mouse colon.

Mouse studies were approved by the Pennsylvania State University College of Medicine Institutional Animal Care and Use Committee and used 9-wk-old, male and female, C57BL/6 mice. The in vivo recycling perfusion (with Tat-beclin 1 or Tat-scrambled peptide, 7 μM) was established using a recycling colonic perfusion method, as described previously (29, 48). Briefly, the mouse colon was isolated following anesthesia with 2% isoflurane and midabdominal incision. The colon was cannulated at the proximal end and at the distal end via rectal opening with a 0.88-mm-diameter plastic tube. An external recirculating pump was used to recirculate the perfusate of Krebs-phosphate saline buffer (with Tat-beclin 1 or Tat-scrambled peptide, 7 μM) for a 2-h perfusion period at a constant flow rate (0.75 ml/min). The body temperature of the mouse was maintained at 37°C with a temperature-controlled warming blanket. Following the perfusion period, the mice were euthanized using an institutional animal care and use committee-approved method of deep anesthesia and cervical dislocation. The colonic tissue was harvested and mounted in Ussing chambers to measure paracellular permeability using [³H]mannitol, as described previously (33). For autophagy activation, rapamycin was injected intraperitoneally, 1.5 mg·kg⁻¹·day⁻¹ for 48 h.

Statistical analysis.

The values of experimental data are expressed as means ± SE and analyzed using t-test for unpaired data and ANOVA whenever required (GraphPad Prism 7 for Windows; GraphPad Software, San Diego, CA). P values of <0.05 were considered significant. Independent experiments were repeated at least three times to ensure reproducibility.

RESULTS

Association between TJ protein occludin and beclin 1 in Caco-2 cells.

To study the role of beclin 1 in epithelial TJ barrier function, we first examined localization of beclin 1 in intestinal epithelial Caco-2 cells. In confocal immunofluorescence examination, a portion of beclin 1 was found to be present on the membrane (Fig. 1A). Since occludin is an important structural and functional barrier-forming component of TJs (8, 18, 38), we also examined colocalization of beclin 1 with occludin. Beclin 1 present on the membrane showed significant colocalization with occludin. Two commercial antibodies used for localization of beclin 1 yielded similar results, and no immunofluorescence was observed when the primary antibody was omitted. Beclin 1 showed minimal colocalization with basolateral adhesion junction protein E-cadherin (Fig. 1B). These data suggest that a portion of beclin 1 is associated with the TJ protein occludin.

Fig. 1.

Localization of beclin 1 in Caco-2 monolayers. A: in confocal immunofluorescence, beclin 1 (red) present on the membrane showed significant colocalization with tight junction protein occludin (green), as can be seen in the merged (yellow) panel. The x–z-axis shows beclin 1-occludin colocalization (yellow) on the apical side of the lateral membrane; nuclei, blue, in merged panel. B: beclin 1 (red) showed minimal colocalization with basolateral adhesion junction protein E-cadherin (green). Scale bars = 10 μm. Representation of 5 areas examined in 3 independent samples.

Beclin 1 plays a role in constitutive TJ barrier function in Caco-2 cells.

To study the functional role of beclin 1 in epithelial TJ barrier function, we next examined the effect of cell-permeable Tat-beclin 1 peptide on TJ barrier function in Caco-2 cells. Tat-beclin 1 peptide activates beclin 1 by competing against its negative regulator, Golgi-associated plant pathogenesis-related protein-1 (GAPR-1; 42). Filter-grown Caco-2 cell monolayers were incubated with Tat-scrambled or Tat-beclin 1 peptide, and the TER was measured to assess TJ barrier function. Tat-beclin 1 peptide caused a time- and concentration-dependent decrease in Caco-2 TER compared with cells incubated with Tat-scrambled peptide (Fig. 2A). The paracellular permeability was also measured by apical-to-basal flux of the paracellular marker mannitol. Tat-beclin 1 peptide caused an increase in the paracellular flux of mannitol compared with Tat-scrambled peptide-treated cells (Fig. 2B). Tat-beclin 1 peptide also reduced the TER (not shown) and increased paracellular mannitol flux in two other intestinal cell lines, nontransformed IPEC-J2 and T84 cells (Fig. 2C), suggesting that the effect of Tat-beclin 1 on TJ barrier function is not restricted to Caco-2 cells. Thus, beclin 1 activation caused a decrease in intestinal epithelial TJ barrier function.

Fig. 2.

Beclin 1 plays a role in constitutive intestinal epithelial tight junction barrier function. A: Tat-beclin 1 peptide caused a time- and concentration-dependent decrease in transepithelial resistance (TER). Means ± SE from 3 independent experiments (n = 4). *P < 0.01 compared with cells treated with Tat-scrambled peptide as analyzed by one-way repeated-measures ANOVA. B: Tat-beclin 1 peptide treatment (7.5 μM, 24 h) increased paracellular mannitol flux in Caco-2 cells. *P < 0.01, two-tailed unpaired Student’s t-test; n = 3, representation of 3 independent experiments. C: Tat-beclin 1 peptide treatment (7.5 μM, 24 h) increased paracellular mannitol flux in IPEC-J2 and T84 cells. *P < 0.01 vs. scrambled peptide (Tat-scr), two-tailed unpaired Student’s t-test; n = 3 per group, representation of 3 independent experiments. D: beclin 1 siRNA transfection, but not nontarget (NT) siRNA transfection, resulted in significant reduction in beclin 1 protein expression (~52-kDa band size). β-Actin (~52-kDa band size) is shown as loading control. Blots represent 3 independent experiments. E and F: knockdown of beclin 1 with siRNA transfection significantly increased the TER (E) and reduced paracellular mannitol flux (F). *P < 0.01 vs. NT siRNA, two-tailed unpaired Student’s t-test; n = 3, representation of 3 independent experiments.

As an alternate approach to examine the constitutive role of beclin 1 in TJ barrier function, we knocked down beclin 1 by siRNA silencing in filter-grown Caco-2 monolayers. The beclin 1 siRNA transfection, but not nontarget (NT) siRNA transfection, produced a near-complete knockdown of beclin 1 protein expression (Fig. 2D). Beclin 1 siRNA, but not NT siRNA, transfection caused significant increase in Caco-2 TER (Fig. 2E). The beclin 1 siRNA transfection also resulted in reduction in mannitol flux compared with NT siRNA-transfected cells (Fig. 2F). Thus, beclin 1 plays a role in constitutive TJ barrier permeability.

Tat-beclin 1 reduces occludin level in Caco-2 cells.

The TJs consist of an array of membrane-spanning proteins (e.g., occludin and claudins) linked by cytoplasmic plaque zona occludens proteins (e.g., ZO-1) to the cytoskeleton (17). In view of the Tat-beclin 1 peptide-induced increase in TJ barrier permeability, we examined TJ protein composition by studying protein expression of occludin, ZO-1, and select claudins that are known to regulate the paracellular TJ barrier (23). Tat-beclin 1 peptide treatment caused a significant decrease in the protein level of occludin (Fig. 3, A and B). Whereas the protein level of claudin-2 showed a mild increase (Fig. 3C), claudin-1, -3, and -4 and ZO-1 levels did not change after Tat-beclin 1 peptide treatment (Fig. 3A). The protein levels of occludin, claudins, and ZO-1 did not differ between untreated control cells and Tat-scrambled peptide-treated cells.

Fig. 3.

Tat-beclin 1 reduces occludin level. A: Caco-2 cells were treated with Tat-scrambled peptide (Tat-scr) or Tat-beclin 1 peptide (Tat Becl 1) for 24 h and analyzed by Western blot for tight junction proteins. Tat-beclin 1 peptide treatment caused a significant decrease in the protein level of occludin (~65-kDa band size), but not claudin-1, -3, or -4 (~23-kDa band size) or zona occludens protein-1 (ZO-1, ~200-kDa band size). β-Actin is shown as a loading control. B and C: densitometry analysis was performed using ImageJ software to indicate relative levels of occludin (B) and claudin-2 (C) after Tat-beclin 1 treatment. *P < 0.01 vs. Tat-scr, two-tailed unpaired Student’s t-test; n = 3, representation of 3 independent experiments.

Beclin 1 is associated with endosomes in Caco-2 cells.

Beclin 1 is known to be associated with endocytic machinery and vacuolar protein sorting (14, 46). First, we examined whether beclin 1 was associated with endocytic components in Caco-2 cells. Indeed, beclin 1 was found to be colocalized with the classical endocytic markers Rab5 and EEA1 (Fig. 4A). This baseline beclin 1-Rab5 and beclin 1-EEA1 colocalization was found to be mostly in the subapical region. Moreover, Tat-beclin 1 peptide treatment resulted in alteration of Rab5 and EEA1 colocalization. In cells treated with Tat-beclin 1 peptide, Rab5 and EEA1 showed larger coalescing areas of colocalization compared with Tat-scrambled peptide-treated cells (Fig. 4B), suggesting a role of beclin 1 in endocytic function.

Fig. 4.

Association of beclin 1 with endosomes. A: beclin 1 (red) was found to be colocalized with classical endocytic markers Ras-related protein Rab-5 (Rab5, green) and early endosome antigen 1 (EEA1, green). B: Tat-beclin 1 alters localization of endosomal markers. Tat-beclin 1 peptide treatment of Caco-2 cells caused redistribution of Rab5 (green) and EEA1 (red) with their increased colocalization (yellow) compared with Tat-scrambled peptide (Tat-scr)-treated cells. Only merged signals (yellow) are retained in the colocalization panel. Scale bars = 5 μm. Representation of 10 areas examined in 4 independent experiments.

Beclin 1 regulates occludin endocytosis.

Previous studies have shown that intracellular trafficking of occludin regulates the TJ barrier (3, 10, 11, 32). In view of increased TJ permeability and reduced occludin level in Tat-beclin 1 peptide-treated cells and the association of beclin 1 with endocytic components, we postulated that beclin 1 may play a role in occludin endocytosis. To examine this possibility, localization of occludin to caveolae, a known route for occludin intracellular trafficking (18, 35, 40), was studied. In Tat-scrambled peptide-treated cells, occludin as well as beclin 1 showed baseline colocalization with caveolin-1 (Fig. 5A). In Tat-beclin 1 peptide-treated cells, occludin immunofluorescence was less intense on the membrane with an increase in cytoplasmic localization compared with Tat-scrambled peptide-treated cells (Fig. 5A). Furthermore, occludin and beclin 1 both showed increased colocalization with caveolin-1 in Tat-beclin 1 peptide-treated cells, suggesting that Tat-beclin 1 induces caveolae-mediated occludin endocytosis. In control Tat-scrambled or Tat-beclin 1-treated cells, cytoplasmic occludin did not colocalize to clathrin, another major route of endocytosis (not shown). Assessment of fluid phase process by HRP uptake (Fig. 5B) and receptor-mediated endocytosis by transferrin internalization (Fig. 5, C and D) did not show a significant difference between Tat-scrambled and Tat-beclin 1 peptide-treated cells. Thus, the Tat-beclin 1 peptide does not appear to broadly affect endocytic processes and may have a specific role in occludin endocytosis. To confirm and quantify occludin endocytosis, we used cell surface biotinylation to study the movement of occludin from the membrane to the cytosol, as described previously (32). Tat-scrambled peptide-treated and Tat-beclin 1 peptide-treated Caco-2 cells were surface biotinylated, and then endocytosis of biotin-labeled protein was allowed for various time periods. The remaining biotin on the cell surface was stripped and quenched, and the cells were scraped in lysis buffer. The biotinylated proteins were immobilized with avidin, isolated, and analyzed by Western blot. We found that in Caco-2 cells treated with Tat-beclin 1 peptide, the rate of occludin endocytosis was significantly increased compared with control cells (Fig. 6, A and B). In view of the increased endocytosis and reduced total level of occludin, we asked whether occludin is increasingly targeted to lysosomes in Tat-beclin 1-treated cells. In the coimmunofluorescence study, occludin showed increased colocalization with the lysosomal marker LAMP2 in Tat-beclin 1-treated cells (Fig. 6C). Overall, these data confirm that beclin 1 activation reduces occludin level via increased endocytosis and lysosomal degradation.

Fig. 5.

Tat-beclin 1 induces occludin endocytosis. A: Tat-beclin 1 peptide treatment of Caco-2 cells resulted in reduced intensity of occludin staining on the membrane and its increased colocalization with caveolin-1 and beclin 1 in the cytoplasm. Occludin, green; beclin 1, blue; caveolin-1, red. Images represent orthogonal view of Z-stacks. Scale bars = 5 μm. Representation of 10 areas examined in at least 3 independent experiments. B: horseradish peroxidase (HRP) uptake in Caco-2 cells was calculated as a ratio of HRP enzyme activity (using o-phenylenediamine as a substrate) and total protein concentration and is presented as arbitrary units. The amount of internalized HRP was not different between Tat-scrambled peptide (Tat-scr)- and Tat-beclin 1 peptide-treated Caco-2 cells (unpaired Student’s t-test). Results represent means ± SE for 3 independent experiments. C: fluorescence images of Tat-scrambled peptide- and Tat-beclin 1 peptide-treated Caco-2 cells subsequently assayed for the internalization of Alexa Fluor 488-conjugated transferrin (green). Nuclei are shown in blue. Representation of 10 areas examined in each of 3 independent experiments. Scale bars = 3 μm. D: quantification of studies from C showed no difference in transferrin internalization between Tat-scrambled peptide- and Tat-beclin 1 peptide-treated Caco-2 cells (unpaired Student’s t-test). Results represent means ± SE of fluorescence intensity measured in 50 cells from each of 3 independent experiments.

Fig. 6.

Tat-beclin 1 increases the rate of occludin endocytosis. A: following surface biotinylation assay, as detailed in experimental procedures, and SDS-PAGE, immunoblots were probed with anti-occludin antibody. Caco-2 cells treated with Tat-beclin 1 peptide showed an increased amount of biotinylated occludin. B: %endocytosed occludin compared with total occludin contents at indicated time points, from 3 independent experiments. *P < 0.01 vs. Tat-scrambled peptide, unpaired Student’s t-test; n = 4, representation of 3 independent experiments. C: Tat-beclin 1 induces increased lysosomal targeting of occludin. In Tat-beclin 1-treated cells, occludin showed increased colocalization with lysosomal marker lysosome-associated membrane protein-2 (LAMP2) compared with Tat-scrambled peptide (Tat-scr)-treated cells. Occludin, green; LAMP2, red. Only merged signals (yellow) are retained in the colocalization images. Scale bars = 5 μm. Representation of 20 areas examined in 3 independent experiments.

Role of autophagy in Tat-beclin 1-induced increase in TJ barrier permeability.

In previous studies, we have demonstrated that autophagy induction enhances TJ barrier function in Caco-2 cells as well as other cell lines such as T84 and MDCKII (34). Since beclin 1 is an integral component of the autophagy pathway, we examined the interplay between beclin 1 function and autophagy in the modulation of TJ barrier function. Autophagic activity as assessed by the established method of LC3 lipidation showed no significant change in the ratio of LC3-II to LC3-I after Tat-beclin 1 treatment (Fig. 7, A and B). Additionally, confocal immunofluorescence revealed no LC3 puncta formation after Tat-beclin 1 treatment of Caco-2 cells (Fig. 7C). Thus, though Tat-beclin 1 peptide is known to induce autophagy (42), the lower dose of Tat-beclin 1 (7.5 μm) did not induce autophagy in Caco-2 cells. In examination of autophagic flux, Tat-scrambled peptide- and Tat-beclin 1 peptide-treated cells showed a significant accumulation of LC3-II levels in the presence of bafilomycin A [a known inhibitor of vacuolar H⁺-ATPase that affects autolysosome formation and degradation (27, 28); Fig. 7D]. Bafilomycin A did not have a significant effect on the Tat-beclin 1-induced decrease in TER and increase in mannitol flux (Fig. 8, A and B). Furthermore, lysosomal inhibition with bafilomycin A caused an increase in occludin levels. After preincubation with bafilomycin A, the total occludin levels were greater in Tat-scrambled peptide-treated cells compared with the bafilomycin A/Tat-beclin 1 group (Fig. 8, C and D). To further examine the role of autophagy in the beclin 1-mediated Caco-2 TJ barrier, the autophagy-related protein ATG16L1, which is known to be critical for the formation of isolation membranes of autophagosomes (32, 33), was knocked down by siRNA silencing in filter-grown Caco-2 monolayers. The ATG16L1 siRNA transfection, but not NT siRNA transfection, produced a near-complete knockdown of ATG16L1 protein expression (Fig. 8E). ATG16L1 siRNA transfection had no effect on the Tat-beclin 1-induced drop in TER or increase in mannitol flux (not shown) or the Tat-beclin 1-induced decrease in occludin expression (Fig. 8, F and G). These data suggest that the constitutive role of beclin 1 in TJ barrier function is autophagy independent.

Fig. 7.

Autophagic activity in Tat-beclin 1-treated and starved Caco-2 cells. A: Caco-2 cells were incubated in normal or starvation medium for 24 h, treated with Tat-scrambled peptide (Tat-scr) or Tat-beclin 1 peptide (Tat-becln 1; 7.5 μM), and analyzed by Western blot for microtubule-associated protein 1A/1B-light chain 3B (LC3B) protein. B: the ratio of LC3-II band density to LC3-I band density was calculated using ImageJ software and is representative of >4 independent experiments. *P < 0.01, one-way ANOVA. C: confocal immunofluorescence revealed no LC3 puncta formation after Tat-beclin 1 treatment (compared with LC3 puncta formation after starvation, white arrows). LC3, red; nuclei, blue. Scale bars = 5 μm. D: Western blot for LC3 from Tat-scrambled peptide- and Tat-beclin 1 peptide-treated Caco-2 cells in the presence of bafilomycin A (BAF, 20 nM) alone or BAF and starvation. β-Actin is shown as a loading control. Representation of 3 independent experiments.

Fig. 8.

Effect of inhibition of constitutive autophagy on Tat-beclin 1-mediated changes in the tight junction barrier. A and B: autophagy inhibition by bafilomycin A (BAF; 1-h pretreatment) did not have a significant effect on the Tat-beclin 1 (Tat-becln 1)-induced decrease in transepithelial electrical resistance (TER, A) or increase in mannitol flux (B). *P < 0.01 vs. Tat-scrambled peptide (Tat-scr) and BAF/Tat-scr groups, unpaired Student’s t-test; n = 4, representation of 3 independent experiments. NSD, no significant difference. C: BAF caused an increase in occludin level in Tat-scrambled peptide-treated cells and partially attenuated Tat-beclin 1-induced reduction in occludin level. D: densitometry analysis of occludin blots in C. ^a,b,c,dP < 0.01, one-way ANOVA (Tukey’s multiple-comparison test); 3 independent experiments. E: autophagy-related protein-16-1 (ATG16L1) siRNA transfection produced a near-complete knockdown of ATG16L1 protein expression (~72-kDa band size) compared with nontarget (NT) siRNA transfection. F and G: ATG16L1 siRNA transfection had no effect on the Tat-beclin 1-induced decrease in occludin expression. Blots represent 3 independent experiments. *P < 0.01 compared with NT siRNA group, unpaired Student’s t-test.

Induction of autophagy prevents Tat-beclin 1-induced reduction in the TJ barrier.

In further experiments, de novo autophagy was induced by starvation or mammalian target of rapamycin (mTOR) inhibitor rapamycin treatment, as described by us previously (34). Induction of autophagy by starvation as well as rapamycin treatment prevented the Tat-beclin 1-induced drop in TER and increase in mannitol flux (Fig. 9, A and B). Next, we studied occludin and beclin 1 association by coimmunoprecipitation under autophagic conditions. Beclin 1 was present in occludin immunoprecipitates from control or Tat-scrambled peptide-treated cells, further supporting an association between beclin 1 and occludin (Fig. 9C). Tat-beclin 1 peptide treatment significantly increased the amount of beclin 1 present in occludin immunoprecipitates (Fig. 9, C and D). On the other hand, autophagy induction with starvation significantly reduced the amount of beclin 1 present in occludin immunoprecipitates (Fig. 9, C and D). Also, the Tat-beclin 1-induced reduction in total occludin protein level was prevented by starvation (Fig. 9C). Thus, in spite of reduction in the total occludin level, beclin 1-occludin association was increased by Tat-beclin 1 peptide treatment, and in spite of an increase in total occludin level, beclin 1-occludin association was reduced after starvation. Alternatively, autophagy induction by rapamycin also reduced beclin 1-occludin association irrespective of an increase in total occludin level (Fig. 9, E and F). Examination of beclin 1 immunoprecipitates showed similar dissociation of beclin 1-occludin after starvation or rapamycin treatment (not shown). These data indicate that the induction of autophagy changes the occludin-beclin 1 complex and alters the role of beclin 1 in constitutive occludin endocytosis and TJ barrier function.

Fig. 9.

Induction of autophagy prevents Tat-beclin 1-induced reduction in the Caco-2 cell tight junction barrier. Caco-2 cells were incubated in starvation media [Earle’s balanced salt solution (EBSS)] or treated with mammalian target of rapamycin (mTOR) inhibitor rapamycin (Rapa, 500 nM) 24 h before the Tat-beclin 1 peptide treatment to induce autophagy. A and B: starvation as well as rapamycin pretreatment prevented the Tat-beclin 1-induced drop in transepithelial electrical resistance (TER, A) and increase in mannitol flux (B). *P < 0.01 vs. all other groups, two-tailed unpaired Student’s t-test; representation of 3 independent experiments. C: in coimmunoprecipitation studies, beclin 1 was detected within occludin immunoprecipitates. Tat-beclin 1 peptide (Tat-becln 1) treatment significantly increased the amount of beclin 1 present in occludin immunoprecipitates, whereas autophagy induction with starvation significantly reduced the amount of beclin 1 present in occludin immunoprecipitates. The negative control (-ve con) lane represents nonimmune IgG control, whereas the control (con) lane represents untreated control cells. The total protein levels before immunoprecipitation showed that Tat-beclin 1-induced reduction in total occludin protein level was prevented by starvation. D: densitometry graph showing beclin 1 and occludin association, calculated as the ratio of beclin 1 to occludin within the respective immunoprecipitates. ^a,b,cP < 0.01 vs. control and each other, one-way ANOVA (Tukey’s multiple-comparison test); 3 independent experiments. E: in coimmunoprecipitation, autophagy induction by rapamycin reduced the amount of beclin 1 present in occludin immunoprecipitates. The negative control lane represents nonimmune IgG control. The total protein levels before immunoprecipitation showed that rapamycin treatment induced an increase in total occludin protein level. F: densitometry graph showing beclin 1 and occludin association, calculated as the ratio of beclin 1 to occludin within the respective immunoprecipitates. *P < 0.01 vs. control, one-way ANOVA (Dunnett’s multiple-comparison test); 3 independent experiments. IP, immunoprecipitation; Starv, starvation; Tat-scr, Tat-scrambled peptide.

In vivo induction of autophagy prevents Tat-beclin 1-induced reduction in the TJ barrier.

To examine the in vivo relevance of our cell culture results, we first studied localization of beclin 1 in mouse colonic epithelium. By confocal immunofluorescence, beclin 1 was localized prominently to the apical cell membrane in mouse colonic surface epithelial cells (Fig. 10A). Moreover, consistent with cell culture data, beclin 1 showed significant colocalization with the TJ protein occludin on the mouse colonic surface epithelial membrane. Next, we studied the effect of Tat-beclin 1 on mouse colonic TJ barrier function. To this end, full-length mouse colons were perfused in vivo with either Tat-scrambled or Tat-beclin 1 peptide (7.5 μM), as described earlier (29). Following in vivo colonic perfusion, the colonic tissues were harvested and mounted in Ussing chambers to study colonic paracellular permeability, as described earlier (31, 33). We found that the colonic paracellular mannitol flux was significantly increased in Tat-beclin 1-treated mouse colon compared with Tat-scrambled peptide-treated mouse colon (Fig. 10B). Tat-beclin 1 treatment did not cause any histological changes in mouse colon. In a subset of experiments, we used rapamycin, an mTOR inhibitor (37), to activate autophagy. Following rapamycin administration (48 h), full-length mouse colons were perfused in vivo with either Tat-scrambled or Tat-beclin 1 peptide, and the colonic tissues were harvested and mounted on the Ussing chamber to study colonic epithelial paracellular permeability. In vivo activation of autophagy by rapamycin was found to prevent the Tat-beclin 1-induced increase in mouse colonic permeability (Fig. 10B). These data indicate that Tat-beclin 1 caused an increase in mouse colonic TJ permeability that was prevented by autophagy activation.

Fig. 10.

Induction of autophagy prevents Tat-beclin 1-induced increase in mouse colonic tight junction permeability. A: in confocal immunofluorescence, beclin 1 (red) was present on the membrane of mouse colonocytes and showed significant colocalization with TJ protein occludin (green), as can be seen in the merged (yellow) panel. Nuclei, blue. Scale bars = 25 μm. B: in vivo perfusion of 9-wk-old male C57BL/6 mouse colon with experimental peptides followed by Ussing chamber studies shows that Tat-beclin 1 peptide increased paracellular mannitol flux and autophagy-inducer rapamycin (Rapa) prevented Tat-beclin 1-induced increase in mouse colonic mannitol flux. *P < 0.01 vs. all other groups, one-way ANOVA (Dunnett’s multiple-comparison test); n = 4 in each group, 2 independent experiments. Cont, control; Tat-scr, Tat-scrambled peptide.

DISCUSSION

The intestinal epithelial TJ barrier is a crucial determinant of intestinal homeostasis. A defective epithelial TJ barrier contributes to the development of intestinal inflammation by allowing access of luminal antigens and microbes to the host immune system (1–3). There is increasing evidence that autophagy is among several cellular processes that are involved in the complex pathogenesis of intestinal disorders. In particular, defective autophagy leading to impaired goblet and Paneth cell function, compromised immune responses and microbial sensing, and inadequate destruction and clearance of microbes has been proposed to be a part of the pathogenesis of Crohn’s disease (4, 6, 12, 13). We have previously shown that autophagy enhances intestinal epithelial TJ barrier function (34), which forms the first line of host defense against intestinal luminal antigens and is an important component of innate immunity. Since autophagy is a degradation pathway, it is quite conceivable that it interacts with the classical endocytic pathway. Additionally, recent studies have elucidated autophagy-independent functions of autophagy-related proteins. In particular, ATG6/beclin 1 has been shown to play a role in endocytosis (16, 46). In this study, we demonstrated the association of beclin 1 and the TJ core protein occludin on the intestinal epithelial membrane and further investigated the role of beclin 1 in intestinal epithelial TJ barrier function. Our data showed that beclin 1 plays an important role in the TJ barrier by regulating constitutive endocytosis of occludin. Our study also indicated that autophagy induction upstream of beclin 1 terminated the function of beclin 1 in constitutive occludin endocytosis.

We used Tat-beclin 1 peptide, which activates beclin 1 by competing against its negative regulator GAPR-1/glioma pathogenesis-related protein-2 (GLIPR2; 42), to investigate the role of beclin 1 in epithelial TJ barrier function. Tat-beclin 1 treatment reduced TER and increased paracellular permeability, associated with a decrease in TJ occludin protein expression. The role of occludin in TJ barrier function has been established by several in vitro and in vivo studies (1, 2, 20, 22, 47). Also, overexpression of occludin has been shown to reduce TJ permeability (20), and depletion of occludin has been shown to increase macromolecular TJ permeability (1, 22). Steady-state endocytosis and recycling of occludin (26), as well as its diffusion within the plasma membrane (41), have been shown to modulate the role of occludin in TJ barrier function. In this study, we demonstrated increased occludin endocytosis in response to beclin 1 activation, providing a mechanistic basis for reduced occludin level and increased TJ permeability.

Beclin 1 is known to be closely involved in endocytosis (16, 46). Beclin 1 is associated with endosomes in neurons, and its deficiency causes impaired endosome formation with mislocalization of phosphatidylinositol 3-phosphate, which is an integral component of endosomes (21). In other instances, beclin 1 plays a role in endocytosis in HeLa cells (44) and mediates efficient phagosome-lysosome function in Toll-like receptor signaling (39). Interestingly, these beclin 1 functions have been shown to be at least partially, if not fully, autophagy independent. In our studies, too, beclin 1 was found to be associated with endosomal markers, and its activation with Tat-beclin 1 peptide changed the localization of endosomal markers Rab5 and EEA1. Moreover, the function of beclin 1 in occludin endocytosis was autophagy independent as pharmacologic or genetic inhibition of autophagy did not affect Tat-beclin 1-mediated changes in the TJ barrier. Though bafilomycin A attenuated Tat-beclin 1-induced decrease in occludin levels to a certain extent, it did not prevent Tat-beclin 1-mediated increase in TJ permeability. This might be due to the beclin 1-mediated removal of occludin from the membrane. In line with the inhibition of the Tat-beclin 1-mediated increase in TJ permeability by autophagy, we have previously demonstrated that under autophagic conditions induced by starvation, occludin remains persistently present on the membrane (34).

Besides beclin 1, other ATG proteins have also been shown to have autophagy-independent functions; for example, Atg7 regulates p53-dependent cell cycle and cell death (13), the Atg FAK family kinase-interacting protein of 200 kDa (FIP200) protects against TNF-α-induced apoptosis (6), and Atg13 and FIP200 play a crucial role in host viral response (19). Association of Atg proteins with other protein complexes involved in membrane dynamics likely explains their diverse functions in endocytosis, cell signaling, host immune response, cell division, and cell death.

The prevention of Tat-beclin 1-induced reduction in the TJ barrier by induction of autophagy, both in vitro and in vivo, as observed in the present study is consistent with our previous report that autophagy enhances the TJ barrier (34). Autophagy has also been shown to be associated with preservation of the TJ barrier during oxidative stress, ischemia-reperfusion, and hyperglycemia and hypoxia injury (7, 15, 23, 30). Among intestinal diseases, autophagy was first linked to the pathogenesis of Crohn’s disease when genome-wide association studies identified mutations in autophagy-related genes ATG16L1 and immunity-related GTPase M (IRGM) as risk factors for Crohn’s disease (9–11). Autophagy has been shown to play a role in dendritic-epithelial cell interactions, adaptive immune response, and nucleotide-binding oligomerization domain-containing protein-2 (NOD2)-directed bacterial sensing and lysosomal destruction in IBD (12–15). Recent studies have also demonstrated that beclin 1 is expressed in normal human colon mucosa and its expression is extensively increased in colonic mucosa of patients with ulcerative colitis. Moreover, the increased levels of beclin 1 were also correlated with the severity of disease, the endoscopic classification, and the pathologic staging results (9). Other studies have also demonstrated the role of beclin 1 in intracellular trafficking of cell surface proteins (36, 43). Also, increased beclin 1 levels have been shown to be associated with disassembly of mTOR complex 2 (mTORC2) and cytodestructive autophagy (45). The mTORC2 assembly is known to regulate actin cytoskeleton and TJ localization of occludin (24, 25). Though we were able to show that beclin 1 activation induces occludin endocytosis, the mechanism of how autophagy induction alters beclin 1-occludin association dynamics needs further investigation, particularly in terms of beclin 1 complex with proteins involved in vesicular sorting including PI3KC, UVRAG, and ATG14L. Nonetheless, our studies clearly showed that beclin 1 activation increases beclin 1-occludin association, whereas autophagy induction reduces beclin 1-occludin association. Thus this study lays a foundation for further investigation of autophagy-mediated enhancement of the TJ barrier. In summary, our study showed that beclin 1 plays an autophagy-independent role in the regulation of constitutive intestinal TJ barrier function via endocytosis of occludin. Induction of autophagy terminates intrinsic beclin 1 function in the TJ barrier and reduces TJ barrier permeability.

GRANTS

This work was supported by National Institute of Diabetes and Digestive and Kidney Diseases Grants R01-DK-114024 and K01-DK-100562 and National Institute of General Medical Sciences (Grant P20-GM-121176) of the National Institutes of Health.

DISCLAIMERS

The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

P.N. conceived and designed research; M.W., A.S.G., E.S., L.L., and P.N. performed experiments; M.W., A.S.G., E.S., L.L., and P.N. analyzed data; M.W., A.S.G., T.M., and P.N. interpreted results of experiments; M.W., A.S.G., L.L., and P.N. prepared figures; M.W. and P.N. drafted manuscript; M.W., T.M., and P.N. edited and revised manuscript; P.N. approved final version of manuscript.