Gastroenterology
Volume 131, Issue 1 , Pages 47-58, July 2006

Transcytosis of IgE–Antigen Complexes by CD23a in Human Intestinal Epithelial Cells and Its Role in Food Allergy

  • Hongxing Li

      Affiliations

    • Department of Medicine/Division of Clinical Immunology, Mount Sinai School of Medicine, New York, New York, USA
    • Immunobiology Center, Mount Sinai School of Medicine, New York, New York, USA
    • ,
  • Anna Nowak–Wegrzyn

      Affiliations

    • Department of Pediatrics, Mount Sinai School of Medicine, New York, New York
    • ,
  • Zachary Charlop–Powers

      Affiliations

    • Department of Pediatrics, Mount Sinai School of Medicine, New York, New York
    • ,
  • Wayne Shreffler

      Affiliations

    • Department of Pediatrics, Mount Sinai School of Medicine, New York, New York
    • ,
  • Mirna Chehade

      Affiliations

    • Department of Pediatrics, Mount Sinai School of Medicine, New York, New York
    • ,
  • Sunil Thomas

      Affiliations

    • Department of Medicine/Division of Clinical Immunology, Mount Sinai School of Medicine, New York, New York, USA
    • ,
  • Giulia Roda

      Affiliations

    • Department of Medicine/Division of Clinical Immunology, Mount Sinai School of Medicine, New York, New York, USA
    • ,
  • Stephanie Dahan

      Affiliations

    • Department of Medicine/Division of Clinical Immunology, Mount Sinai School of Medicine, New York, New York, USA
    • ,
  • Kirk Sperber

      Affiliations

    • Department of Medicine/Division of Clinical Immunology, Mount Sinai School of Medicine, New York, New York, USA
    • Immunobiology Center, Mount Sinai School of Medicine, New York, New York, USA
    • ,
  • M. Cecilia Berin

      Affiliations

    • Department of Pediatrics, Mount Sinai School of Medicine, New York, New York
    • Corresponding Author InformationAddress requests for reprints to: M. Cecilia Berin, PhD, Pediatric Allergy and Immunology, Box 1198, Mount Sinai School of Medicine, One Gustave L. Levy Place, New York, New York 10029.

Received 19 September 2005; accepted 23 March 2006. published online 30 March 2006.

Article Outline

Background & Aims: Secreted immunoglobulins play an integral role in host defense at mucosal surfaces, and recent evidence shows that IgG can participate in antigen sampling from the intestinal lumen. We examined whether IgE also could facilitate transepithelial antigen sampling. Methods: Stool samples from food-allergic patients undergoing oral food challenge were analyzed for CD23 and food-specific IgE. CD23 isoform expression on primary human intestinal epithelial cells (IEC) was analyzed by polymerase chain reaction. The role of CD23 isoforms in transcytosis of antigen and IgE–antigen complexes was assessed using polarized human T84 cells retrovirally transfected with CD23a or CD23b. Results: CD23 was expressed constitutively on IECs, and food-allergic patients had increased levels of soluble CD23 and food-specific IgE in the stool after challenge. CD23a, but not CD23b, was expressed by primary human IECs. We show in transcytosis assays that CD23a, but not CD23b, acts as a bidirectional transporter of IgE. In addition, specific IgE facilitated the uptake of antigen from the apical surface of an epithelial monolayer by diverting antigen from delivery to lysosomes. Finally, delivery of antigen–IgE complexes across the epithelial barrier could induce the degranulation of rat basophil leukemia cells transfected with the human high-affinity IgE receptor. Conclusions: These studies show that CD23a is expressed normally on human IECs, and in the presence of IgE can function as an antigen-sampling mechanism capable of activating subepithelial mast cells. IgE may serve as a secretory immunoglobulin that in concert with CD23 participates in food-induced pathophysiology of the gastrointestinal tract.

Abbreviations used in this paper:  BSA, bovine serum albumin , ELISA, enzyme-linked immunosorbent assay , FITC, fluorescein isothiocyanate , IEC, intestinal epithelial cell , IL, interleukin , NIP, 4-hydroxy-3-nitro-phenacetyl , OFC, oral food challenge , RBL, rat basophil leukemia , RT-PCR, reverse-transcription polymerase chain reaction , sCD23, cleaved CD23 , TNF, tumor necrosis factor

 

IgE is present in human intestinal secretions of individuals with food allergies,1 but it is unclear whether IgE could have a function in the intestinal lumen as shown for other immunoglobulin isotypes. Studies in rats infected with Trichinella spiralis have shown that IgE is a major secreted immunoglobulin,2 with greater output into the intestinal lumen than is observed in draining lymph or plasma. This IgE secretion mechanism was shown to be interleukin (IL)-4 dependent, and specific for the IgE isotype.3 Studies in rodent models of food allergy have suggested a potential function for luminal IgE, which is the uptake of lumenal antigen facilitated by CD23 expressed on the intestinal epithelial cells (IECs).4, 5 Facilitated uptake of antigen was shown to be a necessary step in the induction of rapid hypersensitivity reactions in the intestine,4, 5 but it is not known if a similar mechanism exists in humans.

See CME Quiz on page 276.

CD23 is a type II integral membrane glycoprotein with a carboxy terminal C-type lectin head that binds its ligand, IgE, in a calcium-dependent manner.6 Trimerization of membrane-bound CD23 results in higher-affinity binding of IgE, and lower-affinity binding is observed with CD23 monomers or soluble CD23 (sCD23).7, 8 CD23 exists in 2 forms in humans, CD23a and CD23b.9, 10 CD23a is expressed constitutively on mature activated B cells whereas CD23b is induced by IL-4 or IL-13. CD23b is a splice variant of CD23a, and differs from CD23a because it lacks a tyrosine motif in the N-terminal cytoplasmic region that is thought to be important for endocytic signaling. CD23a and CD23b are generated by the use of different transcriptional initiation sites.9, 10 CD23a and CD23b have been shown to be different functionally, with only CD23a reported to be capable of facilitating endocytic uptake of IgE in humans,11 although CD23b splice variants have been shown to endocytose IgE in a murine intestinal epithelial system.12, 13 In humans, CD23b is expressed on monocytes, eosinophils, and Langerhans cells and plays a role in antibody-dependent cell-mediated cytotoxicity against IgE-opsonized parasites. The use of CD23-deficient and transgenic mice have shown that membrane CD23 can function as a negative regulator of IgE production in mice,14, 15, 16, 17 although the use of anti-CD23 antibodies in fact decreases serum IgE levels in mice and human beings.18 The administration of antigen in the presence of antigen-specific IgE can enhance T-cell and B-cell responses to that antigen in a CD23- and B-cell–dependent manner, presumably through IgE-mediated enhanced uptake and presentation of antigen by B cells bearing CD23.19

It has been shown previously by immunohistochemical staining that CD23 is expressed constitutively on human IECs, and that CD23 expression is enhanced in patients with inflammatory bowel disease or milk protein–induced enteropathy.20 Although CD23 has been shown to play a role in phagocytosis and antigen-capture and presentation in a variety of cell types, the potential function of CD23 on IECs remains unclear. Murine models of food allergy have shown that CD23 is involved in IgE-mediated uptake of antigen across the intestinal epithelium,4, 5 suggesting that CD23 functions as a transepithelial immunoglobulin transporter. In this study, we found that CD23a, but not CD23b, was expressed constitutively in human intestinal epithelium, and that CD23 was detectable in stool of patients undergoing a specific food challenge. In this study we show that CD23a functions as a bidirectional transporter of IgE and can capture IgE–antigen complexes and deliver them in an immunologically intact form across the intestinal epithelial barrier.

Back to Article Outline

Materials and Methods 

Reagents 

Recombinant human sCD23 protein was obtained from Spring Bioscience (Fremont, CA), recombinant human IL-4 and IL-13 from R&D Systems (Minneapolis, MN), bafilomycin from Calbiochem (San Diego, CA), nocodazole from US Biological (Swampscott, MA), and wortmannin from A.G. Scientific (San Diego, CA).

Epithelial Cell Isolation 

IECs were isolated as previously described.21 In brief, colon specimens were washed extensively in phosphate-buffered saline (PBS), followed by stripping of the mucosa from the underlying submucosa. The mucosa was cut into small pieces and treated with 1 mmol/L dithiothreitol (Sigma-Aldrich, St. Louis, MO) for 15 minutes to remove mucus. The tissue then was washed in PBS and incubated twice for 30 minutes in medium containing 3 mg/mL Dispase II (Roche Diagnostics, Alameda, CA) at 37°C, 5% CO2. The supernatant (released IECs) was collected and washed in medium twice. The viability of the isolated IECs was greater than 95%. Primary cells were used either immediately for RNA isolation and CD23 detection by reverse-transcription polymerase chain reaction (RT-PCR), or were stimulated with IL-4 and IL-13 for 6 hours before RNA isolation and CD23 detection by RT-PCR. The purity of epithelial cell preparations were checked routinely by flow cytometry using the epithelial cell–specific monoclonal antibody B9, with a greater than 95% purity for B9-staining cells. The viability of epithelial cells was greater than 60% at 6 hours after stimulation, and was not different between stimulated and unstimulated cells.

Patient Characteristics 

Stool samples were collected from patients after undergoing graded oral food challenges (OFCs) at the General Clinical Research Center at Mount Sinai School of Medicine according to a previously published protocol.22 Informed consent was obtained from patients and the study was approved by the Institutional Review Board of Mount Sinai School of Medicine. A total of 9 patients (age range, 3–17 y) underwent challenge with egg (n = 6) or milk (n = 3). Patients had a history of clinical reactivity to egg or milk, and were rechallenged owing to decreasing food-specific IgE levels. All patients had either detectable food-specific IgE by Immuno-CAP IgE detection (Pharmacia Diagnostics, Uppsala, Sweden) or a positive skin prick test, and 8 of 9 had multiple food allergies. All patients had stool collected after reacting to food challenge. Six of 9 patients had cutaneous symptoms, 4 of 9 had respiratory symptoms, and 2 of 9 had gastrointestinal symptoms, with 3 of 9 having symptoms affecting more than 1 organ system. Symptoms occurred less than 2 hours after OFC. Reactivity to OFC is the gold standard for food-allergy diagnosis, and the combination of positive food-specific IgE, skin-prick test, and the immediate onset of symptoms after challenge characterizes these patients as having IgE-mediated food allergy. Stool samples were obtained from nonallergic children (n = 5; age range, 2.5–9 years) as controls. Table 1 summarizes the patient characteristics.

Table 1. Patient Characteristics
PatientAge, yOFC allergenFood IgE in serum, kIU/LSkin prick test wheal diameter, mmSymptoms after OFC
14.3Egg.57NDC
211Milk>10012R
38Egg1.586C
47Milk.976C
53Egg1.415C and R
63.5Egg<.354C
74Egg1.15NDGI and R
84Egg2.52GI and R
917Milk.410C

ND, not done; C, cutaneous; R, respiratory; GI, gastrointestinal.

Stool Extracts 

Stool samples were placed immediately in PBS containing complete protease inhibitor cocktail (Roche Diagnostics, Indianapolis, IN). Samples were frozen at −80°C before preparation of extracts. To prepare protein extracts, stool samples were vortexed for 1 minute in PBS/complete protease inhibitor, with the addition of glass homogenization beads. Samples then were placed on a rocker and incubated overnight at 4°C. Samples were centrifuged, and supernatants were concentrated using a Centricon 10-kilodalton column (Millipore, Billerica, MA). Protein concentration was assayed by Coomassie protein assay reagent from Pierce (Rockford, IL).

Epithelial Cell Lines and Transfection 

Cell culture 

The human epithelial cell lines T84 and Caco-2 were used for the study. Cells were cultured in RPMI 1640 medium supplemented with 10% fetal bovine serum and penicillin/streptomycin. T84 cells were polarized by growing on polyethylene terephthalate track-etched membranes (.4 μm) cell-culture inserts (BD Falcon, Bedford, MA) in 6-well plates. The cells were cultured in RPMI medium containing 5% fetal calf serum, with 2 mL medium in the insert and 2.5 mL in the well. The medium was replaced every day until day 10. The integrity of tight junctions was monitored by transepithelial resistance to passive ion flow as measured by voltmeter (EVOM; World Precision Instruments, Sarasota, FL) and immunostaining for zonula occludens (ZO-1). Day 10 monolayers with transepithelial resistance of 1000 to 1500 Ω · cm2 were used in this study.

Establishment of Stable CD23a and CD23b Transfected Cell Lines 

Human CD23a cDNA was available on a pDualGC expression vector (Stratagene, La Jolla, CA), CD23b first-strand cDNA amplification was achieved by using forward primer 5′-GGAATCCATGAATCCTCCAAGCCAGGAGATCGAGGAGCTT-3′, reverse primer: 5′-GGGGCGGCCGCTCAAGAGTGGAGAGGGGCAGAGGG-3′. Both open reading frames of CD23a and CD23b were subcloned into a retroviral expression vector that also coded a puromycin-resistant gene. Human 293 EbnaT cells were transfected by using calcium phosphate with 2.5 μg of plasmid pMD.G encoding vesicular stomatitis virus G protein and 7.5 μg of plasmid pMD.OGP encoding gag-pol (both kindly provided by Adrian Ting, Mount Sinai School of Medicine), together with 10 μg of the retroviral expression construct encoding either the human CD23a or CD23b. At 48 hours posttransfection, the viral supernatant was collected, centrifuged at 800 × g, and used to infect T84 cells. A quantity of 2 × 106 T84 cells was resuspended in 10 mL of viral supernatant in the presence of 4 μg/mL of Polybrene. Infected T84 cells were cultured for 48 hours and then selected by using 2 μg/mL of puromycin for 10–14 days. The drug-resistant cells were pooled and analyzed for the expression of CD23 by immunostaining with anti-CD23 antibody.

Real-Time RT-PCR for CD23 mRNA 

The total RNA of the cell was extracted by NucleoSpin RNA purification Kit (BD Biosciences, San Jose, CA). Real-Time RT-PCR was performed on a Lightcycler (Roche Diagnostics) according to the manufacturer’s instructions. We designed primers based on the human CD23a and CD23b sequences reported by Yokota et al:11 CD23a: (forward): 5′-ATGGAGGAAGGTCAATATTC-3′, (reverse): 5′-TCCAGCTGTTTTAGACTCTG-3, and CD23b, (forward): 5′-ATGAATCCTCCAAGCCAG-3′, (reverse): 5′-CACAGGAGAAGCAGAGTCAG-3′. β-actin mRNA expression was used as reference gene for the relative quantification. Actin primers were as follows: (forward) 5′-CATCGTGCACGGCATCGTCA-3′, (reverse) 5′-TAGCACAGCCTGGATAGCAAC-3′ PCR was performed with an annealing temperature of 55°C for 30 cycles (conventional PCR) or 40 cycles (real-time PCR). The optimized 1-step RT and PCR amplication was performed in a 20 μL final volume containing 10 μL QuantiTect SYBR Green mix (Qiagen, Valencia, CA), which gave a final concentration of 2.5 mmol/L MgCl2, 1 μmol/L of each primer and .2 μL QuantiTect RT mix, which contains 2 enzymes that allow highly efficient and sensitive RT. Finally, 2 μL of RNA (100 ng) was added to 18 μL of RT-PCR mixture in each capillary tube and amplified as instructed. To confirm amplification specificity, the PCR products were subjected to a melting curve analysis. Conventional RT-PCR was performed using the same primers with Qiagen OneStep RT-PCR kit and PCR products were analyzed on 2% agarose gels and visualized by ethidium bromide staining.

Immunoblotting for CD23 

Cell lysates from isolated human IECs were loaded on a 7.5% sodium dodecyl sulfate–polyacrylamide gel, underwent electrophoresis, and transferred to a nitrocellulose membrane. The membrane was probed with a polyclonal rabbit anti-human CD23 antibody (AnaSpec, San Diego, CA), and detected with a horseradish-peroxidase–linked anti-rabbit antibody and developed using Pierce Super-Signal West Dura Extended Duration Substrate.

Transcytosis Assays 

To determine transcytosis of IgE by CD23 isoforms, polarized untransfected, CD23a-transfected, or CD23b-transfected T84 cells were polarized on filter supports. Purified human IgE (10 μg/mL), 4-hydroxy-3-nitro-phenacetyl (NIP)-labeled bovine serum albumin (BSA) (10 μg/mL), or BSA NIP preincubated with anti-NIP IgE (10 μg/mL) was added to the apical or basal well of the culture inserts. Cells were incubated at 37°C for 30, 60, or 90 minutes before removal of media from the opposite side of the culture insert and immunoblotting for IgE. Immunoblotting for IgE or BSA NIP was performed as outlined earlier for CD23, using anti-human IgE or anti-NIP antibodies.

sCD23 Detection 

sCD23 protein concentration in epithelial cell lysates was assayed by enzyme-linked immunosorbent assay (ELISA; BD Bioscience OptEIA Set). In stool extracts, CD23 protein concentration was expressed per mg of extract protein concentration.

Immunostaining and Confocal Microscopy 

Cells were cultured in transwell inserts as described previously, and washed 3 times in .5% Triton/PBS and then fixed in ice-cold 4% paraformaldehyde for 5 minutes. Then cells were dipped in an ice-cold acetone for 30 seconds and quickly rehydrated in PBS. Inserts were blocked in 5% BSA/.5% Triton/PBS and then stained with primary antibody diluted in blocking buffer for 2 hours at room temperature, cells were washed 2 times in .5% Triton/PBS and incubated with second antibody for 1 hour in the dark at room temperature. The image was scanned with a Leica TCS-SP confocal laser scanning microscope (Wetzlar, Germany) and analyzed with Volocity image analysis software. Confocal laser scanning microscopy was performed at the MSSM-Microscopy Shared Resource Facility, supported with funding from NIH-NCI shared resources grant (5R24 CA095823-04), NSF Major Research Instrumentation grant (DBI-9724504), and NIH shared instrumentation grant (1 S10 RR0 9145-01).

Rat Basophil Leukemia Degranulation Assay 

Rat basophil leukemia (RBL) cells transfected with the α-chain of the human FcϵRI IgE receptor (RBL-25/30) kindly were provided by Stefan Vieths (Paul-Ehrlich Institute, Langen, Germany). Cells (1 × 105) were plated in 96-well flat-bottom plates and allowed to adhere overnight. Supernatant from transcytosis experiments containing IgE–antigen complexes was added to the RBL cells for 60 minutes. Released β-hexosaminidase was measured in supernatants using p-nitrophenyl-N-acetyl-β-d-glucosaminide (Sigma) dissolved in .1 mol/L phosphate buffer (pH 4.5) as substrate.

Statistics 

The statistical difference between groups (patients vs nonatopic controls) was tested by Student t test. Pearson correlation was used to test the relationship between CD23 and food-specific IgE in the stool samples. A P value of less than .05 was considered significant. Data are expressed as mean ± SD.

Back to Article Outline

Results 

CD23 Protein Is Expressed Constitutively by Human Gastrointestinal Epithelial Cells 

It has been reported previously that CD23 is expressed constitutively by human colonic epithelial cells and that expression is increased in inflammatory bowel disease or cow milk–induced enteropathy.20 We confirmed constitutive expression on human duodenal biopsy specimens (data not shown) by immunostaining for CD23. We used immunoblotting to confirm CD23 expression in human IECs isolated from colonic resections. As shown in Figure 1A, human IECs (freshly isolated, and T84 and Caco-2 cells) constitutively expressed a 45-kilodalton band corresponding to CD23.

  • View full-size image.
  • Figure 1. 

    CD23 expression in normal and food-allergic individuals. (A) Immunoblotting for CD23 in lysates obtained from primary human large IECs, T84 and Caco-2 human colonic epithelial cell lines, and B cells (positive control) and 293T cells (negative control). (B) Stool protein extracts from food-allergic patients (●) or nonatopic controls (□) were assayed for CD23, IL-4, IL-13, and TNF-α by ELISA. Units of measurement are indicated on the x-axis, and were normalized to the protein content in each stool extract. *P < .05 compared with nonatopic control. (C) Stool levels of food-specific IgE as measured by Immuno-CAP in stool extracts was plotted against stool CD23 content. r2 = .76; P = .0011.

CD23 Is Present in the Stool After OFC in Food-Allergic Patients 

IgE-mediated immediate hypersensitivity reactions to food are diagnosed based on the detection of food-specific IgE and positive reaction to OFC.23 Therefore, biopsy specimens are not available from patients who can be diagnosed in this manner. To assess the presence of CD23 in the intestinal tract of food-allergic patients, we obtained stool samples from patients undergoing OFC to milk or egg. Patient characteristics are shown in Table 1. Extracts were prepared from stool samples of patients with food-specific serum IgE who reacted to their OFC with convincing symptoms. Extracts were assayed for sCD23 by ELISA. As shown in Figure 1B, sCD23 was detectable in stool extracts from food-allergic patients, but not nonatopic controls, suggesting either an increased level of CD23 expression on epithelial cells or an allergen-induced shedding of CD23 in food-allergic patients. Stool cytokine levels also were assayed to determine if cytokines known to be involved in the regulation of CD23 also were increased. We observed detectable levels of IL-4 and tumor necrosis factor (TNF)-α but not IL-13 in stool extracts after OFC (Figure 1B).

Food-Specific IgE Is Present in the Stool of Food-Allergic Patients 

For CD23 to have a function at the apical surface of IECs, IgE also would need to be present within the intestinal lumen. In the stool extracts of patients reacting to the OFC we measured food-specific IgE by Immuno-CAP. We detected food-specific stool IgE in the majority of patients (Figure 1C), showing the availability of IgE antibodies to interact with CD23 on the apical surface of the gastrointestinal epithelial cell layer. There was a strong correlation between the level of CD23 and food-specific IgE in the stool (r2 = .76, P = .0011).

The CD23a Isoform Is Expressed by Human IECs 

We used isoform-specific primers to determine which CD23 isoform was expressed by intestinal epithelial cells (Figure 2A). Human CD23b lacks the tyrosine signaling motif that is thought to mediate endocytic uptake. Primary IECs were isolated from normal tissue (normal area of tissue from cancer resections [n = 3]), or from tissue obtained from patients with inflammatory bowel disease (either ulcerative colitis [n = 3] or Crohn’s disease [n = 3]). In all cases, CD23a, but not CD23b, was detected by RT-PCR. This also was found to be true for 3 IEC lines: T84, Caco-2, and HT-29. In contrast, B cells expressed both CD23a and CD23b mRNA. The finding that CD23b is expressed strongly in B cells, but is not expressed in the freshly isolated IECs, argues against any possibility that our finding of CD23a expression was caused by B-cell contamination of the epithelial cell preparations.

  • View full-size image.
  • Figure 2. 

    Regulation of CD23 expression in human IECs. (A) RT-PCR for CD23a or CD23b isoforms was performed. B cells were used as positive control. T84, Caco-2, and HT-29 human IEC lines were used, as well as primary isolated human large IECs from normal tissue (resections for cancer), or patients with ulcerative colitis (uc) or Crohn’s disease (cd). (B) Real-time PCR of CD23b or CD23a expression in B cells, human IEC lines, or primary human IECs treated with IL-4 or IL-13 for 6 hours. Data are expressed as percentage of expression in unstimulated controls. □, IL-4; ■, IL-13. (C) CD23 protein expression as measured by ELISA in cell lysates. Cells were stimulated with IL-4 or IL-13 overnight (18 hours) before preparation of cell lysates. Bars indicate mean ± SD. □, None; ■, IL-4; ▩, IL-13.

It has been shown that allergen-specific lymphocytes isolated from the gastrointestinal tract of patients with food allergy secrete Th2 cytokines,24 and we observed increased levels of IL-4 and TNF-α in stool extracts from patients undergoing OFC. Because IL-4 and IL-13 have been shown to regulate the expression of CD23b in B cells,25 we assessed the regulation of CD23a and CD23b in primary IECs and epithelial cell lines by IL-4 and IL-13. CD23a and CD23b mRNA levels were assessed by real-time PCR. As previously shown, stimulation with IL-4 or IL-13 significantly enhanced CD23b but not CD23a expression in B cells. Stimulation of primary epithelial cells with IL-4 or IL-13 modestly enhanced CD23a expression, but did not induce the expression of CD23b (Figure 2B). T84, Caco-2, and HT-29 cells also modestly up-regulated CD23a, but did not express CD23b in response to IL-4 or IL-13 stimulation. CD23 protein expression was assayed in cell lysates by ELISA, and we observed a significant increase in CD23 protein expression by primary IEC and cell lines after IL-4 or IL-13 stimulation (Figure 2C). We also tested the regulation of CD23 expression by TNF-α, which also was increased in stool samples of patients with food allergy. TNF-α also up-regulated CD23 protein expression (data not shown), but there was no synergy observed between TNF-α and IL-4 stimulation.

CD23a Facilitates Bidirectional Transcytosis of IgE and Mast Cell Degranulation 

To determine the function of CD23 isoforms in IECs, we made stable retroviral CD23a and CD23b transfectants using the T84 cell line. T84 cells form well-differentiated monolayers with tight junctions and can be used as a model system for transcytosis studies. Untransfected or transfected T84 cells were grown on transwell inserts and polarity was established with an electrical resistance of more than 1000 Ohms · cm2. Western blotting showed a strong up-regulation (>40-fold) of CD23 protein expression in both CD23a- and CD23b-transfected cells, with comparable levels of CD23 protein expression in both transfectants (data not shown). Staining with anti-CD23 antibodies and confocal microscopy confirmed increased expression of CD23 in the transfected T84 cells (Figure 3), and localization of membrane CD23 using ZO-1 costaining showed apical and basolateral expression of CD23a, whereas CD23b was expressed predominantly on the basolateral aspect of the epithelial cell monolayers (Figure 3). Although untransfected cells (expressing endogenous CD23a) had some CD23 expression at the apical surface, CD23b-transfected cells showed an absence of apical staining, suggesting that apical expression of endogenous CD23a perhaps was inhibited by the presence of CD23b. The confocal images show a substantial amount of CD23 protein in the cytoplasm, as has been shown for other immunoglobulin receptors such as FcRn.

  • View full-size image.
  • Figure 3. 

    Confocal microscopy detection of CD23. T84 cells, either untransfected or transfected with CD23a or CD23b, were polarized and stained with ZO-1 (red) to show the level of the tight junction, DAPI to stain nuclei (blue), and CD23 was detected by immunostaining and detection with FITC (green). The XY sections were taken at the level of the ZO-1 staining, and the XZ reconstructions are shown at the bottom. The red staining indicates the tight junctions at the apical pole of the epithelial cells.

To determine the role of CD23 isoforms in transcytosis of IgE, we added purified human IgE to the apical or basolateral well of the transwell, and assessed the appearance of IgE in the opposite well by immunoblotting at 30, 60, or 90 minutes after IgE addition. Untransfected T84 cells constitutively express CD23a, and there was detectable transcytosis of IgE in both apical-to-basal and basal-to-apical directions. Overexpression of CD23a resulted in enhanced transcytosis of IgE in a bidirectional manner (Figure 4). In contrast, transfection with CD23b did not enhance transcytosis of IgE, and consistently inhibited IgE transcytosis compared with untransfected T84 cells. Transcytosis of chicken IgY or BSA was unaffected by transfection with CD23a or CD23b (data not shown).

  • View full-size image.
  • Figure 4. 

    Transcytosis of IgE by T84 monolayers. T84 cells constitutively expressing CD23a were untransfected or transfected with CD23a or CD23b. Cells were grown on filter supports in transwells to polarize the cells. IgE was added to the apical (A→B) or basal (B→A) wells of the transwell, and supernatant from the opposite transwell was immunoblotted for IgE over time (30–90 minutes). The lower graph shows the ratio of optical density units of transfected/untransfected monolayers from the 90-minute timepoint. Mean ± SD of 3 experiments is shown. □, T84; ■, +CD23a; ▩, +CD23b.

The ability of transcytosed IgE to deliver antigen across T84 monolayers was tested by performing transcytosis assays of antigen (NIP-labeled BSA) in the presence or absence of anti-NIP BSA (Figure 5A). BSA NIP was transcytosed across untransfected epithelial cells, but this was enhanced markedly by addition of anti-NIP IgE. Antigen–IgE complexes were transcytosed from apical-to-basal, but not basal-to-apical directions. Overexpression of CD23a, but not introduction of CD23b, in T84 cells enhanced apical-to-basal transcytosis of antigen–IgE complexes (Figure 5B).

  • View full-size image.
  • Figure 5. 

    IgE-facilitated antigen transcytosis. (A) BSA NIP was added to the apical chamber of polarized (untransfected) T84 cells in the presence (Ag-IgE) or absence (Ag) of anti-NIP IgE. Basolateral supernatant was immunoblotted for BSA NIP over time. (B) Detection of BSA-NIP appearance in basolateral supernatants over time after addition of BSA NIP plus anti-NIP IgE to untransfected T84 cells, or T84 cells transfected with CD23a or CD23b. The lower graph shows the ratio of optical density units of transfected/untransfected monolayers from the 90-minute timepoint. Mean ± SD of 3 experiments is shown. (C) Degranulation of human FcϵR1-transfected RBL cells as measured by β-hexosaminidase release. Basolateral supernatant from transwells incubated on the apical side with BSA NIP (antigen), anti–BSA-NIP-IgE (IgE), or both (Complex) were added to RBL cells. As positive control, cells were stimulated with phorbol myristate acetate (PMA) and ionomycin. Shown is a representative experiment of 2, with error bars indicating triplicate variation. □, 30 minutes; ▩, 60 minutes; ■, 120 minutes.

We tested the ability of transcytosed antigen–IgE complexes to activate mast cells. We cultured RBL cells expressing human FcϵRI with basolateral well supernatant obtained from transcytosis assays at either 30, 60, or 120 minutes after addition of IgE to the apical well, and assessed degranulation of RBL cells by β-hexosaminidase assay. We found that basolateral supernatant collected at 120 minutes from CD23a-transfected T84 cells induced degranulation of RBL cells (Figure 5C), showing the potential for CD23a-mediated transcytosis of antigen–IgE complexes to result in activation of allergic effector cells below the intestinal epithelium. These results also show that IgE is released from the epithelial cells together with the antigen in an intact form capable of activating effector cells.

Regulation of CD23a-Mediated Transcytosis by IL-4 and sCD23 

IL-4 up-regulated the expression of CD23 on IECs. We tested the impact of IL-4 stimulation on transcytosis of IgE and IgE–antigen complexes. We observed that treatment with IL-4 increased the transcytosis of IgE across T84 cells (Figure 6A). This also was observed for IgE-facilitated antigen uptake (Figure 6B). We next examined the influence of sCD23 on transcytosis of IgE (Figure 6A) or antigen in the presence of IgE (Figure 6B). Incubation with 5 μg/mL of sCD23 inhibited, but did not completely block, the transcytosis of both bidirectional transport of IgE and IgE-facilitated antigen transport.

  • View full-size image.
  • Figure 6. 

    Effect of IL-4 and sCD23 on transcytosis of IgE and antigen–IgE complexes. (A) Polarized T84 monolayers were treated with IL-4 overnight before addition of IgE to the apical (A→B) or basal (B→A) wells of the transwell. Alternatively, IgE was preincubated with sCD23 before addition to the transwell. Supernatant from the opposite transwell was sampled over time and blotted for IgE. The graph on the right shows the ratio of optical density units of treated/untreated monolayers from the 90-minute timepoint. Mean ± SD of 3 experiments is shown. □, None; ■, IL-4; ▩, sCD23. (B) As described previously, but in place of IgE, BSA NIP plus anti-NIP IgE was added to the apical transwell, and blotting for BSA NIP was performed. The graph on the right shows the ratio of optical density units of treated/untreated monolayers from the 90-minute timepoint. Mean ± SD of 3 experiments is shown.

IgE Diverts Antigen From Lysosomal Compartments 

To determine if IgE influences the trafficking of antigen within the IECs, we tracked the uptake of fluorescein isothiocyanate (FITC)–BSA–NIP in the presence or absence of anti-NIP IgE. By using lysotracker to label lysosomes, we observed that antigen taken up in the absence of IgE prominently colocalized with lysotracker, indicating entry into the lysosomes (Figure 7). In contrast, antigen applied to cells in the presence of IgE did not track to lysosomes, as shown by the lack of colocalization of lysotracker and FITC-labeled antigen. Therefore, enhanced transcytosis of antigen in the presence of IgE likely results from prevention of normal degradation of proteins within lysosomes.

  • View full-size image.
  • Figure 7. 

    Intracellular trafficking of antigen. T84 cells were grown on coverslips and incubated with lysotracker (red) before addition of FITC–BSA–NIP (Ag) or FITC–BSA–NIP plus anti-NIP IgE (Ag + IgE). Lysosomes are stained red (top row), antigen is shown in green (second row), nuclei were stained with DAPI (blue, third row), and a merged image is shown in the bottom row with colocalization of antigen and lysosomes shown as yellow. Magnification, 1000×.

Transport Pathway 

We next examined the role of endosomal trafficking in CD23a-mediated IgE transcytosis. T84 cells were incubated in the presence or absence of bafilomycin A1 (.1 μmol/L), which inhibits endosome acidification but does not interfere with membrane trafficking, as evidenced by studies using transcytosis of cholera toxin and botulinum toxin in T84 cells. As shown in Figure 8, bafilomycin slightly inhibited apical-to-basal and basal-to-apical IgE transport, but transport of antigen in the presence of IgE was attenuated significantly by bafilomycin. We tested other compounds that inhibit vesicular transport including nocodazole and wortmannin. Nocodazole causes microtubules to depolymerize and has been shown to interfere with transcytosis of IgG by FcRn. Nocodazole had no effect on transcytosis of IgE or antigen–IgE complexes (data not shown). Wortmannin selectively inhibits PI3-kinase activity and causes lysosomes to be secreted rather than sorted through the late endosomes to lysosomes from the trans Golgi network. It also decreases the rate of endocytosis. Wortmannin blocks the activity of the Rab GTP-binding proteins and prevents endosome fusion and binding of guanosine triphosphate to Rab5. In our system, wortmannin had no effect on the transport of IgE or the IgE immune complexes.

  • View full-size image.
  • Figure 8. 

    Effect of bafilomycin on transcytosis of IgE. Polarized T84 cells were pretreated with bafilomycin before addition of IgE (apical or basal) or IgE–antigen complexes (apical only). IgE or antigen were detected in the opposite transwell. The graph on the right shows optical density units from the 90-minute timepoint. Mean ± SD of 3 experiments is shown. □, Control; ■, bafilomycin.

Back to Article Outline

Discussion 

Food allergies are a growing clinical problem, and recent estimates suggest that 3.5%–4% of Americans are affected by IgE-mediated food allergies.23 Food allergens cross-link antigen-specific IgE bound to the surface of mast cells and trigger the release of multiple mediators, including histamine, which can induce symptoms in the gastrointestinal tract, airways, and skin, and in more severe cases can induce generalized anaphylactic shock. Before allergens can trigger reactions in the body, they first must escape degradation by digestive enzymes and pass through the single layer of columnar epithelial cells that form the lining of the gastrointestinal tract. It has been shown that IgE is present in intestinal secretions of patients with IgE-mediated food allergies,1, 26 however, in the absence of lumenal effector cells such as mast cells, the functional consequence of lumenal IgE remained unclear. Studies in rodent systems have shown that expression of the low-affinity IgE receptor CD23 on IECs facilitate the uptake of allergen into epithelial cells in an IgE-dependent manner, and that blocking CD23 can interfere with allergen-induced symptoms in the gastrointestinal tract.4, 5 In this study, we examined the interaction of IgE and CD23 in food-allergic patients and in human IEC lines and show that the CD23a isoform acts as a bidirectional IgE transporter that can facilitate apical-to-basal transcytosis of functional IgE–antigen complexes capable of activating subepithelial effector cells.

We first examined the expression of CD23 by human IECs and the presence of IgE in stool samples from food-allergic patients to determine if IgE and CD23 interactions potentially could be relevant to human disease. We assessed CD23 expression in primary colonic epithelial cells by immunoblotting and found that in the majority of samples a band corresponding to CD23 was detectable showing constitutive expression. We were not able to obtain intestinal biopsy samples from patients with IgE-mediated food allergies and therefore looked for sCD23 in stool samples as a measure of intestinal CD23 expression. We found that food-allergic patients undergoing OFC had significantly increased levels of CD23 in stool, whereas CD23 in stool of normal pediatric controls was undetectable. In the absence of biopsy specimens it cannot be concluded if this is caused by increased expression of CD23 in the intestine or caused by increased shedding of CD23 from the surface of the cells. However, Kaiserlian et al20 have shown previously by immunohistochemical staining that CD23 expression is increased in cow milk–induced enteropathy, which supports our hypothesis that CD23 expression by IECs is up-regulated in IgE-mediated food allergy. Stool samples from patients undergoing OFC, but not controls, also had detectable levels of TNF-α and IL-4, suggesting that local inflammation may drive the increased expression of CD23. For CD23 to have a potential function in facilitating uptake of antigen via IgE, food-specific IgE would need to be present in the intestinal lumen. We confirmed the presence of milk, egg, or soy-specific IgE in the stool extracts of patients challenged and reactive to those foods. This is consistent with a previous report showing IgE was present in intestinal secretions of patients with food allergy.1 Although CD23 was found in stool from patients undergoing OFC, there did not appear to be any correlation between type of symptoms (affecting the gastrointestinal tract, skin, or respiratory tract), however, the number of patients in the study (9) is not sufficient for a formal assessment of the correlation of CD23 with type of symptoms or symptom severity.

We next examined the isoform of CD23 expressed by IECs by RT-PCR. CD23a and CD23b have been shown to have unique functions, with CD23a involved in endocytic uptake of IgE and CD23b involved in phagocytosis and antibody-dependent cell-mediated cytotoxicity.11 Recent studies have provided conflicting data on the expression of CD23 isoforms in human IECs. Montagnac et al27 recently showed constitutive expression of CD23a in IEC lines, with CD23b being induced on polarization. In contrast, Tu et al28 recently reported that CD23b was expressed constitutively in human IEC lines but CD23a was undetectable. We observed that CD23a was expressed constitutively in 3 colonic epithelial cell lines (T84, Caco-2, and HT-29), and that CD23b could not be detected even after polarization. In addition, we report that CD23a was expressed constitutively in primary human IECs from normal tissue or tissue from patients with inflammatory bowel disease. The discrepancy between our findings and those of Tu et al28 are not clear, with both studies reporting co-expression of CD23a and CD23b in B cells, and the study of Tu et al28 confirming selective CD23b expression in a monocyte cell line whereas we showed CD23b up-regulation in response to IL-4 in B cells. We performed RT-PCR using the primers reported by Tu et al,28 and confirmed our findings of CD23a but not CD23b expression in human IEC lines and primary isolates. Therefore, there may be some variation between the culture conditions or cell lines used between laboratories; however, our finding of selective CD23a expression in primary isolates of human IECs suggests that CD23a is the biologically relevant isoform.

We examined the regulation of CD23a or CD23b by Th2 cytokines and TNF-α, and observed that IL-4 and IL-13 up-regulated the expression of CD23a mRNA in IECs whereas CD23b was up-regulated in B cells but not induced in IECs. CD23 protein expression was enhanced by IL-4, IL-13, and TNF-α, and there was activation of STAT-6 (data not shown). This is consistent with the finding in mice that the CD23 promoter contains STAT-6 and NF-κB binding sites.29 We hypothesize that local cytokine release by lamina propria cells may regulate the expression of CD23 on IECs. Antigen-specific Th2-cytokine–producing T lymphocytes24 have been shown to occur in the lamina propria of children with food-allergic disorders, and antigen-specific TNF-α release from peripheral blood mononuclear cells has been shown in children with allergy to cow’s milk.30 Therefore, each exposure to allergen potentially could enhance the expression of CD23 on the epithelial cell layer by activating mucosal immune cells.

Studies in mouse models of food allergy have shown that CD23 is involved in IgE-mediated uptake of allergen across the epithelial barrier, and that this function was mediated by the CD23b isoform.12, 13 We assessed the role of CD23a and CD23b isoforms in bidirectional transcytosis of IgE across human IECs by overexpressing CD23a or CD23b in T84 cells. We found that overexpression of CD23a, but not introduction of CD23b, resulted in enhanced transport of IgE in both apical-to-basal and basal-to-apical directions. In addition, transport of antigen–IgE complexes also was enhanced by overexpression of CD23a, and these complexes were capable of mast cell degranulation. Montagnac et al27 showed recently that human CD23a, but not CD23b, was involved in endocytic uptake of IgE in clathrin-coated pits, however, they did not address transcytosis in their study. In contrast to our findings, they found that CD23a was localized predominantly to the basolateral membrane of transfected MDCK cells. It is possible that differences between human T84 and canine MDCK cells may explain some of this inconsistency, but in the absence of transcytosis studies it is difficult to assess whether CD23a was polarized functionally.27 We did not address the impact of ligand binding on CD23a or CD23b localization, but our transcytosis studies together with our confocal analysis indicates that CD23a is present functionally on both apical and basolateral surfaces. Despite the functional expression of CD23a on both apical and basolateral surfaces, IgE–antigen complexes were not transported in a bidirectional manner. This suggests that additional signals may be provided by antigen–IgE complexes beyond those of IgE binding alone.

Our findings and those of Montagnac et al27 highlight differences between the function of human and mouse isoforms of CD23. Specifically, Yu et al12 found that CD23b and CD23b splice variants, but not CD23a, were expressed in mouse epithelial cells in response to allergic sensitization. The splice variants of CD23b, but not full-length CD23b, mediated transcytosis of free IgE, and CD23b was capable of inducing transcytosis of antigen–IgE complexes, indicating that different splice variants performed different functions of IgE transport.12, 13 In contrast, we found that CD23a was the only isoform capable of transcytosis and was capable of facilitating transport of both IgE and IgE–antigen complexes. Therefore, in the human intestinal epithelial system there does not appear to be segregation of function between different isoforms. However, taking these studies together it is clear that CD23, whether it is a splice variant of the CD23b isoform in mice or the classic CD23a isoform in human beings, can facilitate bidirectional IgE transport across IEC barriers.

We observed that antigen was diverted from lysosomes during IgE-/CD23-mediated transcytosis, therefore bypassing an additional barrier function of IECs. The pathway also was shown to be sensitive to bafilomycin, which collapses pH gradients, but not nocodazole, which disrupts microtubules, or wortmannin, which inhibits PI3K. Both IgA transcytosis by the polymeric immunoglobulin receptor and IgG transport through the FcRn also have been shown to be sensitive to agents that interfere with pH gradients. Nocodazole has been shown to inhibit IgA transport, but there are differing reports of the role of microtubules in FcRn-mediated IgG transport, depending on the cell line used in experiments.31 Similar to FcRn and pIgR, CD23 also is internalized via clathrin-coated pits.27 The intracellular trafficking mechanism of CD23 has just begun to be dissected; however, the similar ability of CD23 and FcRn to mediate bidirectional transcytosis suggests possible similarities in their transport mechanism and function as important pathways of antigen sampling in the intestinal mucosa.

Back to Article Outline

 

The authors thank Lloyd Mayer, Hugh Sampson, and Alexander Grishin for helpful discussions.

Back to Article Outline

References 

  1. Belut D , Moneret-Vautrin DA , Nicolas JP , Grilliat JP . IgE levels in intestinal juice . Dig Dis Sci . 1980;25:323–332
  2. Negrao-Correa D , Adams LS , Bell RG . Intestinal transport and catabolism of IgE (a major blood-independent pathway of IgE dissemination during a Trichinella spiralis infection of rats) . J Immunol . 1996;157:4037–4044
  3. Ramaswamy K , Hakimi J , Bell RG . Evidence for an interleukin 4-inducible immunoglobulin E uptake and transport mechanism in the intestine . J Exp Med . 1994;180:1793–1803
  4. Yang PC , Berin MC , Yu LC , Conrad DH , Perdue MH . Enhanced intestinal transepithelial antigen transport in allergic rats is mediated by IgE and CD23 (FcepsilonRII) . J Clin Invest . 2000;106:879–886
  5. Yu LC, Yang PC, Berin MC, Di Leo V, Conrad DH, McKay DM, et al. Enhanced transepithelial antigen transport in intestine of allergic mice is mediated by IgE/CD23 and regulated by interleukin-4 . Gastroenterology . 2001;121:370–381
  6. Beavil AJ , Edmeades RL , Gould HJ , Sutton BJ . Alpha-helical coiled-coil stalks in the low-affinity receptor for IgE (Fc epsilon RII/CD23) and related C-type lectins . Proc Natl Acad Sci U S A . 1992;89:753–757
  7. Dierks SE , Bartlett WC , Edmeades RL , Gould HJ , Rao M , Conrad DH . The oligomeric nature of the murine Fc epsilon RII/CD23. Implications for function . J Immunol . 1993;150:2372–2382
  8. Gould H , Sutton B , Edmeades R , Beavil A . CD23/Fc epsilon RII (C-type lectin membrane protein with a split personality?) . Monogr Allergy . 1991;29:28–49
  9. Yokota A, Kikutani H, Tanaka T, Sato R, Barsumian EL, Suemura M, et al. Two species of human Fc epsilon receptor II (Fc epsilon RII/CD23) (tissue-specific and IL-4-specific regulation of gene expression) . Cell . 1988;55:611–618
  10. Kikutani H, Inui S, Sato R, Barsumian EL, Owaki H, Yamasaki K, et al  Molecular structure of human lymphocyte receptor for immunoglobulin E . Cell . 1986;47:657–665
  11. Yokota A, Yukawa K, Yamamoto A, Sugiyama K, Suemura M, Tashiro Y, et al. Two forms of the low-affinity Fc receptor for IgE differentially mediate endocytosis and phagocytosis (identification of the critical cytoplasmic domains) . Proc Natl Acad Sci U S A . 1992;89:5030–5034
  12. Yu LC , Montagnac G , Yang PC , Conrad DH , Benmerah A , Perdue MH . Intestinal epithelial CD23 mediates enhanced antigen transport in allergy (evidence for novel splice forms) . Am J Physiol . 2003;285:G223–G234
  13. Montagnac G, Yu LC, Bevilacqua C, Heyman M, Conrad DH, Perdue MH, et al. Differential role for CD23 splice forms in apical to basolateral transcytosis of IgE/allergen complexes . Traffic . 2005;6:230–242
  14. Payet M , Conrad DH . IgE regulation in CD23 knockout and transgenic mice . Allergy . 1999;54:1125–1129
  15. Payet ME , Woodward EC , Conrad DH . Humoral response suppression observed with CD23 transgenics . J Immunol . 1999;163:217–223
  16. Texido G , Eibel H , Le Gros G , van der Putten H . Transgene CD23 expression on lymphoid cells modulates IgE and IgG1 responses . J Immunol . 1994;153:3028–3042
  17. Yu P , Kosco-Vilbois M , Richards M , Kohler G , Lamers MC . Negative feedback regulation of IgE synthesis by murine CD23 . Nature . 1994;369:753–756
  18. Flores-Romo L, Shields J, Humbert Y, Graber P, Aubry JP, Gauchat JF, et al  Inhibition of an in vivo antigen-specific IgE response by antibodies to CD23 . Science . 1993;261:1038–1041
  19. Getahun A , Hjelm F , Heyman B . IgE enhances antibody and T cell responses in vivo via CD23+ B cells . J Immunol . 2005;175:1473–1482
  20. Kaiserlian D , Lachaux A , Grosjean I , Graber P , Bonnefoy JY . Intestinal epithelial cells express the CD23/Fc epsilon RII molecule (enhanced expression in enteropathies) . Immunology . 1993;80:90–95
  21. Mayer L , Shlien R . Evidence for function of Ia molecules on gut epithelial cells in man . J Exp Med . 1987;166:1471–1483
  22. Sicherer SH . Food allergy (when and how to perform oral food challenges) . Pediatr Allergy Immunol . 1999;10:226–234
  23. Sampson HA . Update on food allergy . J Allergy Clin Immunol . 2004;113:805–820
  24. Beyer K , Castro R , Birnbaum A , Benkov K , Pittman N , Sampson HA . Human milk-specific mucosal lymphocytes of the gastrointestinal tract display a TH2 cytokine profile . J Allergy Clin Immunol . 2002;109:707–713
  25. Punnonen J , Aversa G , Cocks BG , de Vries JE . Role of interleukin-4 and interleukin-13 in synthesis of IgE and expression of CD23 by human B cells . Allergy . 1994;49:576–586
  26. Brown WR , Borthistle BK , Chen ST . Immunoglobulin E (IgE) and IgE-containing cells in human gastrointestinal fluids and tissues . Clin Exp Immunol . 1975;20:227–237
  27. Montagnac G, Molla-Herman A, Bouchet J, Yu LC, Conrad DH, Perdue MH, et al. Intracellular trafficking of CD23 (differential regulation in humans and mice by both extracellular and intracellular exons) . J Immunol . 2005;174:5562–5572
  28. Tu Y, Salim S, Bourgeois J, Di Leo V, Irvine EJ, Marshall JK, et al. CD23-mediated IgE transport across human intestinal epithelium (inhibition by blocking sites of translation or binding) . Gastroenterology . 2005;129:928–940
  29. Tinnell SB , Jacobs-Helber SM , Sterneck E , Sawyer ST , Conrad DH . STAT6, NF-κB and C/EBP in CD23 expression and IgE production . Int Immunol . 1998;10:1529–1538
  30. Heyman M, Darmon N, Dupont C, Dugas B, Hirribaren A, Blaton MA, et al. Mononuclear cells from infants allergic to cow’s milk secrete tumor necrosis factor alpha, altering intestinal function . Gastroenterology . 1994;106:1514–1523
  31. Rojas R , Apodaca G . Immunoglobulin transport across polarized epithelial cells . Nat Rev Mol Cell Biol . 2002;3:944–955

 Supported by National Institutes of Health grants AI-44236 and AI-45343 (K.S.) and DK-071576 (M.C.B.).

PII: S0016-5085(06)00711-6

doi:10.1053/j.gastro.2006.03.044

Gastroenterology
Volume 131, Issue 1 , Pages 47-58, July 2006

Article Tools

* Image Usage & Resolution