The JI Acurri Cytometers
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     
 

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Mantis, N. J.
Right arrow Articles by Mehta, S.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Mantis, N. J.
Right arrow Articles by Mehta, S.
The Journal of Immunology, 2004, 172: 6838-6845.
Copyright © 2004 by The American Association of Immunologists

Oligosaccharide Side Chains on Human Secretory IgA Serve as Receptors for Ricin1

Nicholas J. Mantis2, Stephanie A. Farrant and Simren Mehta3

Gastrointestinal Cell Biology Laboratory, Children’s Hospital Boston, and Department of Pediatrics, Harvard Medical School, Boston, MA 02115


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
Secretory IgA (sIgA) Abs are polymeric Igs comprised of two or more IgA monomers joined together at their C termini and covalently associated with a 70-kDa glycoprotein called secretory component. As the predominant Ig type in gastrointestinal sections, sIgA Abs are centrally important in adaptive immunity to enteropathogenic bacteria, viruses, and toxins. In this study, we demonstrate that sIgA Abs may also function in innate defense against ricin, a naturally occurring, galactose-specific plant lectin with extremely potent shiga toxin-like enzymatic activity. In lectin blot overlay assays, we found that ricin bound to secretory component and the H chain of human IgA, and this binding was inhibited by the addition of excess galactose. The toxin also recognized IgM (albeit with less affinity than to IgA), but not IgG. Ricin bound to both human IgA1 and IgA2, primarily via N-linked oligosaccharide side chains. At 100-fold molar excess concentration, sIgA (but not IgG) Abs inhibited ricin attachment to the apical surfaces of polarized intestinal epithelial cells grown in culture. sIgA Abs also visibly reduced toxin binding to the luminal surfaces of human duodenum in tissue section overlay assays. We conclude that sIgA Abs in mucosal secretions may serve as receptor analogues for ricin, thereby reducing the effective dose of toxin capable of gaining access to glycolipid and glycoprotein receptors on epithelial cell surfaces.


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
The gastrointestinal tract represents the largest mucosal surface in the human body. As part of the normal diet, the gastrointestinal tract is continually exposed to bacterial- and plant-derived Ags and toxins. Accessibility of microbial pathogens and enterotoxins to glycoprotein and glycolipid receptors on the apical surfaces of intestinal enterocytes is limited by several physical barriers, including mucus (1) and the filamentous brush border glycocalyx (FBBG)4 (2, 3). In addition, gastrointestinal secretions contain secretory IgA (sIgA) Abs at concentrations that are estimated to exceed hundreds of micrograms per milliliter (4, 5, 6). sIgA Abs are polymeric Igs comprised of two or more IgA monomers joined together at their C termini and covalently associated with a 70-kDa glycoprotein derived from the polymeric Ig, called secretory component (SC) (7, 8). sIgA Abs protect mucosal surfaces by several mechanisms, including ‘immune exclusion,’ steric hindrance, and by directly binding to epitopes on pathogens and toxins involved in receptor recognition (9).

In addition to their well-appreciated role in adaptive immunity, there is emerging evidence that sIgA Abs may also be important in innate defense (10). Both the IgA H chain and SC are heavily glycosylated molecules with N- and/or O-linked oligosaccharide side chains (see Fig. 1), which have been proposed to function as ligands for bacterial adhesins (11, 12, 13). For example, carbohydrates on SC bound to Clostridium difficile toxin A and effectively inhibited toxin binding to epithelial brush border membrane receptors (14). Considering the diversity of oligosaccharide side chains present on sIgA, it is likely that these oligosaccharide side chains may also function as receptors for plant-derived toxins that are constituents of the normal human diet or that may be consumed accidentally. One such class of toxins is the so called ribosome inhibiting protein (RIP), which is produced by >50 species of plants, including wheat, maize, and barley (15, 16).



View larger version (26K):
[in this window]
[in a new window]
  		FIGURE 1. Schematic of human IgA subclasses and SC with respective oligosaccharide side chain modifications. The two classes of human IgA are shown as dimers joined together by J chain and SC. SC is comprised of five (I–V) Ig domains (yellow ovals). N-linked oligosaccharide side chains on IgA subclasses and J chain are depicted as Y-shaped open squares, whereas N-linked oligosaccharide side chains on SC are shaded in blue. O-linked oligosaccharide side chains on the IgA1 hinge region are depicted as solid black ball-and-stick-like figures.

 
Like shiga toxin, RIPs are N-glycosidases that irreversibly inactivate the 28S rRNA of the mammalian 60S ribosomal subunit, thereby arresting host cell protein synthesis (17). One of the most potent members of the RIP family is ricin, which is produced naturally by castor beans (Ricinus communis) (18). Although castor beans are not a part of the normal human diet, there in heightened interest in understanding innate immunity (especially on mucosal surfaces) to this toxin because it has been classified by the National Institutes of Health as a category B select agent that could potentially be spread via aerosol or food/water supplies (19). A member of the A-B family of toxins, ricin is comprised of a single enzymatic ‘A’ subunit and a single lectin-like ‘B’ subunit (20). The ricin B subunit binds with micromolar affinity to {beta}(1, 4)-linked galactose residues and with slightly lower affinity to {beta}(1, 3)-linked galactose residues (21, 22, 23, 24), thereby mediating toxin attachment to both glycolipids and glycoproteins on epithelial cell surfaces. Once bound to the cell surface the toxin is then rapidly internalized into host cells via multiple endocytic pathways (20).

In this paper we tested the hypothesis that sIgA Abs via their carbohydrate side chains may serve as receptor analogues for ricin. Accordingly, we demonstrate that ricin binds primarily to N-linked galactose-containing side chains on SC and human IgA1 and IgA2, but not on human IgG. Moreover, human sIgA Abs independent of their variable domains reduced toxin attachment to the apical surfaces of intestinal epithelial cells grown in culture and to the luminal surfaces of human duodenal biopsies. We conclude that galactose-containing oligosaccharide side chains on human sIgA present in mucosal secretions can serve as receptors for ricin, thereby reducing the effective dose of toxin capable of gaining access to glycolipid and glycoprotein receptors on epithelial cell surfaces. These data further underscore the important role sIgA Abs may play in innate immunity on mucosal surfaces.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
Chemical reagents

Unlabeled and biotinylated derivatives of R. communis agglutinin (RCA)-I and ricin (RCA-II) were purchased from Vector Laboratories (Burlingame, CA). HBSS and HEPES were purchased from Sigma-Aldrich (St. Louis, MO). Tween 20 was obtained from Bio-Rad (Torrance, CA) and paraformaldehyde (16%) was purchased from Electron Microscopy Sciences (Fort Washington, PA). DispoDialysers (m.w. cutoff 10,000) for ricin dialysis were purchased from Spectrum Laboratories (Gardena, CA).

Abs and protein reagents

Purified, human colostral IgA and BSA were purchased from Sigma-Aldrich. Purified human plasma Igs (IgA, IgM, and IgG) and myeloma proteins (IgA1 and IgA2) were obtained from Calbiochem (San Diego, CA). N-linked oligosaccharide side chains were removed from IgA1 using peptide-N-glycosidase F (PNGase F; Calbiochem) under denaturing conditions. Purified, recombinant human IgA Abs lacking N- or O-linked oligosaccharide side chains were kindly provided by Dr. K. Chintalacharuvu (Department of Microbiology and Immunology, University of California, Los Angeles, CA). The IgA1 Ab lacking N-linked side chains was originally described by Chuang and Morrison (25), and the IgA1 Ab lacking O-linked side chains that carry a deletion of the entire hinge region are described by Rifai et al. (26).

ELISA and lectin binding assays

Nunc Maxisorb F96 microtiter plates (Fisher, Pittsburgh, PA) were coated with Igs (0.1 µg/well) in PBS (pH 7.4) overnight at 4°C. Microtiter plates were washed with PBS in Tween 20 (0.05% v/v) and blocked for 1 h with Ig-free BSA (2% w/v in PBS with Tween 20), and then treated with biotinylated ricin or RCA-I for 1 h at room temperature. After washing, plates were developed with avidin-HRP (0.4 µg/ml), and one component tetramethylbenzidine colorimetric substrate (Kirkegaard & Perry Laboratories, Gaithersburg, MD). Plates were read at 450 nm using a SpectriMax 250 microtiter plate reader (Molecular Devices, Sunnyvale, CA) and accompanying Softmax software. Average values and SE between duplicate samples were calculated using Microsoft Excel 2002 (Redmond, WA).

Lectin blots

Human sIgA or plasma IgA Abs were solubilized in Laemmeli sample buffer containing 2-ME (5% v/v), boiled for 10 min, then size-fractionated on a precast SDS-polyacrylamide gradient (4–20%) gel (Bio-Rad). Igs were transferred to nitrocellulose membrane (0.45 µm pore size; Bio-Rad) via electroelution. The nitrocellulose membranes were washed with PBS with Tween 20, blocked for 1 h with BSA (2% w/v in PBS with Tween 20), then incubated with biotinylated ricin (2 µg/ml) or biotinylated RCA-I for 1 h at room temperature. To determine whether lectin binding was dependent on oligosaccharide side chains, biotinylated ricin was preincubated with galactose before addition to the membranes. To visualize bound lectins, membranes were probed with avidin-HRP (0.4 µg/ml) and developed using ECL kit (Amersham Pharmacia Biotech, Piscataway, NJ). Proteins were visualized on polyacrylamide gels using Gel Code blue (Pierce, Rockford, IL).

Ricin binding to T84 cell monolayers

The human adenocarcinoma cell line T84 was kindly provided by W. Lencer (Children’s Hospital, Boston, MA). T84 cells were seeded on rat tail collagen-coated 0.33 cm2 Transwell inserts (3.0 µm pore size; Costar, Cambridge, MA) and maintained at 37°C (5% CO2) for ~7–10 days to allow them to establish confluent, polarized monolayers, as previously described (27). Monolayers were used for ricin binding assays when they achieved a transepithelial resistance that exceeded 1000 {Omega}/cm2. Before being was used for binding studies, ricin was dialyzed for 12 h to remove residual sodium azide added by the manufacturer as a preservative.

For biotin-ricin binding assays, T84 cell monolayers were washed three times with HBSS containing HEPES (20 mM, pH 7.4) then cooled to 4°C in a cold room. Biotinylated ricin diluted in HBSS to a final volume of 200 µl was applied to the upper (apical) chamber of Transwell inserts and incubated at 4°C for 40 min. For competitive inhibition assays, biotinylated ricin was incubated with appropriate concentrations of Abs or BSA at room temperature for 1 h, and then chilled on ice for 20 min before application to T84 cell monolayers. After ricin addition, monolayers were washed with cold HBSS, fixed with paraformaldehyde (4% v/v in PBS) and labeled with streptavidin-FITC (2 µg/ml; Pierce). Transwell filters were removed from the inserts using a razor blade and mounted right-side up on glass microscope slides (Fisher), mounted with a coverslip, and visualized using a Zeiss Axiophot microscope equipped with epifluorescence. Images were captured using a SPOT camera and accompanying software (Diagnostic Instruments, McHenry, IL), and then cropped and framed using Adobe Photoshop (Adobe Systems, Mountain View, CA).

For quantitative binding assays, ricin (2 mg/ml) was labeled with 125I (New England Nuclear, Boston, MA) using IodoBeads (Pierce). Free, unbound 125I was removed from the ricin mixture using disposable 10-ml bed volume polyacrylamide desalting columns (Pierce), followed by dialysis for 12 h against 4 L of PBS. The concentration of 125I-labeled ricin was determined using the bicinchoninic acid assay kit (Pierce). For binding assays, 125I-labeled ricin (50 nM) was incubated at room temperature with molar excess (see figures for exact concentrations) human colostral sIgA, plasma IgG, or Ig-free BSA for 1 h, cooled on ice for 20 min, then applied (200 µl per well) to the apical surfaces of T84 cell monolayers. When calculating molar excess concentrations, we assumed the m.w. of sIgA to be 400,000 and IgG to be 160,000. After 40 min, the monolayers were washed three times with cold HBSS to remove unbound ricin, fixed briefly with paraformaldehyde (4% v/v in PBS), and then subjected to a gamma counter. All binding assays were performed in duplicate, and average values and SE between samples were calculated using StatView 5.0.1 Software (SAS Institute, Cary, NC).

Ricin binding to human duodenal tissue sections

Paraffin sections of human duodenum biopsies were obtained from the Department of Pathology at Children’s Hospital, Boston, MA, with prior approval from the Internal Review Board. Biotinylated ricin (10 µg/ml) or biotinylated RCA-I (5 µg/ml) were incubated with human sIgA or plasma IgG for 1 h at room temperature before being applied to deparaffinized tissue sections (5 µm thick) of human duodenum. After 1 h, sections were washed with PBS and labeled with streptavidin-FITC (50 µg/ml; Pierce). Sections were fixed with 4% paraformaldehyde for 10 min, mounted with a glass coverslip using Mowiol (28), and visualized using a Zeiss Axiophot microscope, as previously described.


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
Ricin binds to human IgA H chain and SC

We used a lectin blot overlay assay to determine whether ricin binds to one or more of the polypeptides of human plasma IgA or sIgA. Plasma IgA from serum and sIgA from colostrum were size fractionated by gradient SDS-PAGE under reducing conditions, then transferred to nitrocellulose membrane by electroelution. Nitrocellulose membranes were probed with biotinylated ricin or RCA-I and developed with avidin-HRP. Plasma IgA is a heterotetrameric protein comprised two identical {alpha} H chain and two identical L chains (Fig. 1) that by reducing SDS-PAGE analysis appeared as two distinct bands of ~55 kDa and 25 kDa, respectively (Fig. 2). Lanes containing sIgA had an additional band of ~70 kDa that corresponded to SC (Fig. 2). The 15-kDa joining (or J chain) protein was not visible with Coomassie stained gels, although it was readily detected by Western blot analysis with anti-human J chain Abs (data not shown). By lectin blot overlay, ricin (and RCA-I) labeled H chain and SC, but not L chain or J chain (Fig. 2). Ricin binding to both H chain and SC could be reduced to undetectable levels by preincubating ricin with a millionfold molar excess galactose, demonstrating that ricin recognizes human IgA H chain and SC via their galactose oligosaccharide side chains. Although there was no detectable difference in ricin affinity for SC vs H chain in the lectin blot assay, we observed by indirect ELISA that ricin bound to SC with approximately twice the affinity that it bound to plasma IgA (data not shown).



View larger version (69K):
[in this window]
[in a new window]
  		FIGURE 2. Ricin binding to human IgA H chain and SC by lectin blot. Purified human plasma IgA or sIgA from colostrum were size fractionated by reducing, SDS-PAGE and stained with Coomassie (A) or transferred to nitrocellulose membranes (B) and probed with biotinylated ricin. A, By SDS-PAGE, secretory IgA fractionated into a 25 kDa, 55 kDa, and 70 kDa protein species corresponding to L chain (LC), IgA {alpha} H chain (HC), and SC, respectively. Human plasma IgA fractionated into two bands corresponding to H chain and L chain. B, In lectin overlay assays, biotinylated ricin bound to SC and H chain, but not L chain. In the lectin blot overlays, SC appeared as a doublet; the minor, low m.w. band represents an alternative proteolytic cleavage product of SC. Both the high and low m.w. forms of SC reacted with anti-human SC Abs (data not shown).

 
Ricin binds IgA and IgM, but not IgG

Although sIgA Abs constitute the predominant Ig type in human intestinal secretions, IgG and IgM are also present, albeit at significantly lower levels (5). To test whether ricin recognizes these other Ig classes, microtiter plates were coated with equal amounts of human plasma IgA, IgM, or IgG, and then probed with biotinylated ricin (Fig. 3A). Ricin bound to IgM, but had little or no affinity for IgG. For comparative purposes, we also probed Ig-coated ELISA plates with RCA-I, a nontoxic, galactose-specific lectin from castor beans closely related to ricin, but with slightly different carbohydrate recognition specificity (21). In contrast to ricin, RCA-I bound all three Ig classes to varying degrees, with highest apparent affinity for IgM (Fig. 3B). The binding of ricin and RCA-I to IgA was inhibited by the addition of exogenous galactose (data not shown).



View larger version (16K):
[in this window]
[in a new window]
  		FIGURE 3. Ricin and RCA-I binding to human IgA, IgG, and IgM. Microtiter plates were coated with human plasma IgA, IgM, or IgG and probed with biotinylated ricin (A) or biotinylated RCA-I (B), then developed using avidin-HRP (see Materials and Methods). IgA ({blacktriangleup}), IgG (•), and IgM ({blacksquare}) are represented. Data are means of a single representative experiment performed in duplicate.

 
Ricin binds human IgA1 and IgA2 via N-linked oligosaccharide side chains

Human IgA is comprised of two subclasses, IgA1 and IgA2. Both subclasses are present in intestinal secretions in roughly equal amounts (29), and both are glycosylated (30) (Fig. 1). Human IgA1 has two N-glycosylation sites, and nine potential O-glycosylation sites within the serine-threonine rich hinge between C{alpha}1 and C{alpha}2 (13, 30). Human IgA2 has three (and four in some isoforms) N-linked side chains, but no O-linked side chains. To test whether ricin binds to both human IgA subclasses, microtiter plates were coated with purified myeloma IgA1 or IgA2, and then probed with biotinylated ricin. Ricin bound to both IgA subclasses, although its apparent affinity was approximately eight times greater for IgA1 than IgA2 (Fig. 4A). Similar results were obtained when we used two other, noncommercial IgA1 and IgA2 myeloma preparations (data not shown). In contrast, RCA-I bound equally well to IgA1 and IgA2 (Fig. 4B). From these data, we conclude that both human IgA1 and IgA2 can serve as ligands for ricin.



View larger version (14K):
[in this window]
[in a new window]
  		FIGURE 4. Ricin and RCA-I recognition of human IgA1 and IgA2. Microtiter plates were coated with purified human IgA1 or human IgA2 then probed with biotinylated ricin (A) or biotinylated RCA-I (B), and then developed using avidin-HRP (see Materials and Methods). IgA1 (•) and IgA2 ({blacksquare}) are represented. Data are means ± SE of a single representative experiment performed in triplicate.

 
To determine whether ricin recognizes N- and/or O-linked oligosaccharide side chains on human IgA1, we performed lectin blot assays using myeloma IgA1 in which we had selectively removed N-linked glycans with PNGase F. Under reducing and denaturing conditions, the H chain of IgA1 treated with PNGase F migrated with greater mobility than mock-treated IgA1, demonstrating that the N-linked oligosaccharide side chains were successfully removed by enzymatic treatment (Fig. 5A). In the corresponding lectin blot assays, ricin strongly labeled the H chain of untreated IgA1 and mock-treated IgA1, but only weakly labeled the H chain of IgA1 treated with PNGase F (Fig. 5B).



View larger version (65K):
[in this window]
[in a new window]
  		FIGURE 5. Ricin binding to human IgA1 lacking N-linked oligosaccharide side chains by lectin blot assay. N-linked oligosaccharide side chains were removed from human myeloma IgA1 by treatment with PNGase F. A, IgA1 (1.5 µg/lane, lane 1), IgA1 mock-treated with PNGaseF (2 µg/lane, lane 2), or IgA1 treated with PNGaseF (2 µg/lane, lane 3) were size-fractionated on a 5–20% SDS-polyacrylamide gradient gel and then stained with Coomasie blue. The H chain from IgA1 treated with PNGaseF (lane 3; solid arrowhead with asterisk) migrated with a faster mobility than the H chain from untreated or mock-treated IgA1 (lanes 1 and 2; filled arrowhead), indicating that the N-linked glycans were successfully removed by enzymatic treatment. In contrast, the L chains from all three IgA1 preparations migrated with identical mobility (unfilled arrowhead). B, IgA1 preparations were sized-fractionated by SDS-PAGE as in A, transferred to nitrocellulose, and the blotted with biotinylated ricin, as described in Fig. 2.

 
To verify these results by an alternative method, we tested whether ricin bound to recombinant human IgA1 Abs lacking N- or O-linked side chains (25, 26). Microtiter plates were coated with purified, recombinant IgA1 Abs and then probed with biotinylated ricin. Ricin bound to fully glycosylated recombinant IgA1 and IgA1 lacking O-linked oligosaccharide side chains, but was reduced in binding to IgA1 lacking N-linked oligosaccharide side chains (Fig. 6A). Similar results were obtained when the plates were probed with RCA-I, although RCA-I showed some residual reactivity toward IgA1 lacking N-linked side chains (Fig. 6B). Collectively, these data indicate that ricin binds primarily to N-linked rather than O-linked oligosaccharide side chains on human IgA1.



View larger version (15K):
[in this window]
[in a new window]
  		FIGURE 6. Ricin and RCA-I binding to recombinant IgA1 lacking N- or O-linked oligosaccharide side chains. Microtiter plates coated with a recombinant human IgA1 Ab lacking N- or O-linked oligosaccharide side chains were probed with biotinylated ricin (A) and biotinylated RCA-I (B), and then developed using standard ELISA reagents. IgA1 ({blacksquare}), IgA1 lacking O-linked side chains (•), IgA1 lacking N-linked side chains ({blacktriangleup}). Data are means of a single representative experiment performed in duplicate.

 
Ricin binding to the apical surfaces of polarized intestinal epithelial cell monolayers in vitro is inhibited by sIgA

The fact that oligosaccharide side chains on sIgA can serve as ligands for ricin raised the possibility that sIgA in mucosal secretions could competitively inhibit toxin binding to epithelial cell surfaces. To test this hypothesis in vitro, biotinylated ricin was incubated with increasing concentrations of human colostral IgA, and then applied to the apical side of T84 intestinal epithelial cell monolayers. Toxin binding to epithelial cell surfaces was visualized by fluorescence microscopy after the monolayers were labeled with streptavidin-FITC. In the absence of sIgA, ricin labeled all cells within the T84 cell monolayers, although there was variation in staining intensity among individual cells or clusters of cells (Fig. 7). A similar pattern of staining was observed previously when RCA-I was applied to the apical surfaces of polarized human Caco-2 intestinal epithelial cell monolayers, and most likely reflects inherent cell-to-cell variation in apical glycosylation levels (3, 31). Preincubation of ricin with 1- to 100-fold molar excess concentrations of sIgA for 1 h at room temperature reduced toxin binding to epithelial surfaces to undetectable levels in a dose-dependent manner (Fig. 7). However, preincubation of ricin with 100-fold molar excess human plasma IgG had no effect on toxin attachment to T84 cell monolayers. The ability of sIgA Abs to reduce toxin binding to epithelial cell surfaces was specific to ricin, as sIgA Abs in 100-fold molar excess had no affect on the attachment of cholera toxin to its ganglioside receptor GM1, as determined by fluorescence microscopy (data not shown).



View larger version (120K):
[in this window]
[in a new window]
  		FIGURE 7. sIgA reduces ricin binding to the apical surfaces of T84 cell monolayers. Biotinylated ricin (5 µg/ml) was incubated for 1 h with PBS (a) or indicated molar excess concentrations of sIgA (b–d) or plasma IgG (e), and then applied to the apical surfaces of filter grown T84 cell monolayers at 4°C. The monolayers were washed after 1 h, labeled with streptavidin-FITC, and then visualized en face by fluorescence microscopy. f, Control monolayers were not treated with biotinylated ricin. Bar equals 20 µm.

 
The inhibitory effect of sIgA on toxin binding to epithelial cell surfaces was confirmed quantitatively using 125I-labeled ricin (Fig. 8). Binding of 125I-labeled ricin to the apical surfaces of T84 cell monolayers was reduced to background levels by preincubation of the toxin with between 10- and 100-fold molar excess sIgA. In contrast, preincubation of 125I-labeled ricin with 100-fold molar excess human plasma IgG or BSA had no effect on toxin attachment to T84 cells (Fig. 8).



View larger version (12K):
[in this window]
[in a new window]
  		FIGURE 8. sIgA reduces 125I-labeled ricin binding to the apical surfaces of T84 cell monolayers. 125I-labeled ricin (50 nM) was incubated for 1 h with PBS (bar on far left) or indicated molar excess concentrations of colostral sIgA, plasma IgG, or BSA, and then applied to apical surfaces of filter grown T84 cell monolayers at 4°C. The monolayers were washed after 1 h, and the filters were removed and subjected to gamma irradiation counting. Data are means of a single representative experiment performed in duplicate. The horizontal dashed line indicates the background values obtained from filters not exposed to 125I-labeled ricin.

 
sIgA reduced ricin binding to the luminal surfaces of human duodenum

Intestinal epithelial cells grown in culture do not completely recapitulate enterocytes in vivo. In particular, the apical surfaces of enterocytes in vivo are coated with a 500-nm thick FBBG that is absent from epithelial cells grown in culture (2, 32). Rich in glycoproteins, the FBBG represents a potential high affinity and high avidity binding site for ricin. To determine whether sIgA can prevent toxin attachment to the luminal surfaces of human intestinal epithelium, biotinylated ricin (10 µg/ml) was preincubated with human sIgA Abs and overlaid onto tissue sections of human duodenum. In the absence of sIgA, ricin labeled the luminal surfaces of intestinal villi strongly and uniformly (Fig. 9a). Preincubation of ricin with sIgA at either 20 µg/ml or 200 µg/ml (corresponding to 0.3- or 3-fold molar excess sIgA, respectively) reduced toxin binding to intestinal villi in a dose-dependent manner (Fig. 9, b and c). For technical reasons, we were unable to test sIgA at concentrations exceeding 200 µg/ml. Nonetheless, these data demonstrate that sIgA Abs in 3-fold molar excess can reduce ricin attachment to the luminal aspects of intestinal enterocytes in vivo.



View larger version (58K):
[in this window]
[in a new window]
  		FIGURE 9. sIgA reduces ricin binding to the luminal surfaces of intestinal villi. Biotinylated ricin (10 µg/ml) was incubated for 1 h with PBS (a) or indicated concentrations of sIgA (b and c), and then applied to deparaffinized tissue sections of human duodenum. Sections were then labeled with streptavidin-FITC and visualized by fluorescence microscopy. a, Ricin labeled the luminal aspects of intestinal villi strongly and uniformly. Ricin also weakly labeled cells within the lamina propria. b and c, Preincubation of ricin with indicated concentrations of sIgA Abs reduced toxin attachment to luminal aspects of the intestinal villi. Arrowheads highlight areas of visible toxin labeling.

 

   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
As the predominant Ig class in gastrointestinal secretions, sIgA Abs are of central importance in mucosal immunity to enteropathogenic bacteria, viruses, and toxins (6, 9). In addition, there is growing evidence that sIgA Abs may also be important in innate immunity (10, 14, 33, 34, 35). In support of this concept, we demonstrated in this study that ricin recognizes galactose-containing oligosaccharide side chains on human IgA and SC, and that sIgA Abs at physiologically relevant concentrations can reduce ricin attachment to the apical surfaces of intestinal epithelial cells. Based on these data we propose that sIgA Abs in mucosal secretions can serve as ricin receptor analogues, thereby reducing the effective dose of toxin capable of attaching to epithelial cell surfaces. These data underscore the possibility that sIgA Abs via their diverse oligosaccharide side chains constitute an important line of innate defense that aids in protecting the gastrointestinal epithelium from exposure to plant and bacterial-derived toxins.

Human sIgA has several structural features that make it particularly well suited to serve as a receptor for toxins like ricin. Oligosaccharide side chains account for >10% of the molecular mass of the IgA H chain (30) and 15–20% of SC (11, 36). In contrast, glycans comprise <3% of the mass of human IgG (37). The N- and O-linked oligosaccharide side chains on human SC, IgA1, and IgA2 collectively are extremely diverse and complex (11, 12, 13), which allows them to be recognized by a wide variety of lectins and toxins. Although some bacterial adhesins have been shown to be specific for only one subclass of IgA (14, 33, 34, 35), we found that ricin recognized both IgA1 and IgA2 (although the toxin bound IgA1 with significantly higher affinity than IgA2). It is important to note that although we have not calculated the exact affinity of ricin for sIgA in this study, the specific association constants for ricin binding to an array of oligosaccharide side chains, including one similar to those found on SC and IgA, has been reported by Baenziger and Fiete (21), and range from ~1 x 106 to 45 x 106 M–1. Finally, molecular modeling vividly demonstrates the extent by which N- and O-linked glycan side chains project from Fc{alpha} and SC, making them highly accessible to lectins and viral or bacterial adhesins (12, 13). This is in contrast to IgG in which the glycans are proposed to face inward, and be sequestered between the H chains (12). Ricin did not detectably bind to human J chain, despite the fact that J chain contains a single N-glycan site that constitutes ~8% of the molecule m.w. (13, 38). This may be because >75% of the N-glycan side chains are sialylated (13) and sialic acid substituents markedly reduced ricin binding to complex oligosaccharide side chains (21).

We postulate that sIgA Abs can prevent ricin attachment to the intestinal epithelium in vivo by at least two different mechanisms. As the in vitro binding assays used in this study demonstrate, sIgA can serve as a receptor analog and thereby competitively inhibit in a dose-dependent manner ricin binding to glycolipid and/or glycoprotein ligands on the apical surfaces of intestinal epithelial cells. Indeed, this is the proposed mechanism by which sIgA in maternal milk prevents attachment of C. difficile toxin A to colonocyte brush border membranes of suckling infants (14). Second, the ricin B subunit has two galactose recognition domains (39) that could potentially promote cross-linking of sIgA in the intestinal lumen. Aggregated IgA-Ab complexes diffuse poorly through the mucus layer overlying the epithelium (40), and are eventually eliminated from the intestinal lumen by peristalsis; a phenomenon referred to as immune exclusion (9). In vitro, sIgA2 via its mannose-rich oligosaccharide side chains agglutinate strains of Escherichia coli expressing mannose-specific type 1 fimbriae (35), suggesting that immune exclusion may be a primary mechanism by which sIgA Abs function in innate defense against enteric pathogens and toxins with multivalent lectin binding activities. Our data also raise the possibility that SC itself (not complexed with IgA) may have a role in innate defense against ricin, as free SC is found in mucosal secretions and human milk (10). Indeed, Dallas and Rolfe (14) determined that free SC is sufficient to prevent C. difficile toxin A from binding to intestinal brush border membranes. The exact mechanisms by which sIgA and free SC may prevent ricin binding to intestinal mucosal surfaces needs to be addressed in a relevant animal model.

Ricin is a potential biowarfare agent for which there is currently no antidote or immunotherapy available. Several strategies have been proposed to prevent or treat mucosal ricin exposure, including passive neutralizing mAb therapy (41, 42, 43). Whereas the current emphasis has been on anti-ricin IgG Abs, we propose that neutralizing anti-ricin sIgA Abs may be more effective at protecting mucosal surfaces. For example, from this study we would predict that anti-ricin IgA Abs could both neutralize toxin via the Fv region and agglutinate the toxin via carbohydrate side chains on Fc{alpha} and SC, thereby promoting toxin clearance from mucosal surfaces. Furthermore, sIgA Abs have the additional benefit over IgG in that they are multivalent, protease resistant, and when applied passively can localize within the mucus layer overlying epithelial surfaces (36, 44). Indeed, recent technological advances now make it possible to produce recombinant human sIgA Abs of desired Ag specificity on a commercial scale (45).

Whereas we assume that the formation of sIgA-ricin complexes in vivo would be beneficial to the host, we have recently shown that intestinal M cells can mediate the selective uptake and transcytosis of sIgA (and potentially sIgA-Ag complexes) across the intestinal epithelium (46). This raises the possibility that a small amount of toxin could gain access to the mucosa by ‘piggy-backing’ on sIgA. Whereas M cells have been shown to transcytose ricin more rapidly than enterocytes (47), it has not been determined whether this was because M cells simply have a naturally higher level of endocytosis as compared with enterocytes, or whether uptake was promoted by the association of the toxin with sIgA. After M cell transcytosis, ricin-sIgA complexes would be expected to be sampled by underlying dendritic cells (48, 49), although neither the biological nor immunologic consequences of this interaction are known. We are using a murine model of gastrointestinal ricin poisoning (N. J. Mantis and S. A. Farrant, unpublished observations) to define in vivo the contribution of sIgA in innate defense.


   Acknowledgments
 
We acknowledge Wayne Lencer, Children’s Hospital (Boston, MA) for providing us with T84 cell monolayers, Bonny Dickinson (Children’s Hospital) for technical assistance in ricin iodination, Jessica Wagner for assistance with microscope imaging analysis, and Natalie Anosova and Marian Neutra (Children’s Hospital) for critically reading this manuscript. We also extend special thanks to Koteswara Chintalacharuvu (University of California, Los Angeles) for providing us with human recombinant IgA Abs and Pamela Kozlowski (Children’s Hospital) for purified human myelomas.


   Footnotes
 
1 This work is supported in part by the National Institutes of Health Grant K01 DK59295 (to N.J.M.). Core services used in this study were supported by the Harvard Digestive Diseases Center, National Institutes of Health Grant DK034854. Back

2 Address correspondence and reprint requests to Dr. Nicholas J. Mantis at the current address: Division of Infectious Disease, Wadsworth Center, P.O. Box 22002, Albany, NY 12201-2002. E-mail address: nmantis{at}wadsworth.org Back

3 Current address: Program in Biology and Biomedical Sciences, Washington University, St. Louis, MO. Back

4 Abbreviations used in this paper: FBBG, filamentous brush border glycocalyx; RIP, ribosome inactivating protein; RCA, Ricinus communis agglutinin; sIgA, secretory IgA; SC, secretory component; H chain, heavy chain; L chain, light chain; PNGase F, peptide-N-glycosidase F. Back

Received for publication December 15, 2003. Accepted for publication March 18, 2004.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Deplancke, B., H. R. Gaskins. 2001. Microbial modulation of innate defense: goblet cells and the intestinal mucus layer. Am. J. Clin. Nutr. 73:1131S.[Abstract/Free Full Text]
  2. Frey, A., K. T. Giannasca, R. Weltsin, P. J. Giannasca, H. Reggio, W. I. Lencer, M. R. Neutra. 1996. Role of the glycocalyx in regulating access of microparticles to apical plasma membranes of intestinal epithelial cells: implications for microbial attachment and vaccine targeting. J. Exp. Med. 184:1045.[Abstract/Free Full Text]
  3. Mantis, N. J., A. Frey, M. R. Neutra. 2000. Accessibility of glycolipid and oligosaccharide epitopes on rabbit villus and follicle-associated epithelium. Am. J. Phys. 278:G915.[Abstract/Free Full Text]
  4. Conley, M. E., D. L. Delacroix. 1987. Intravascular and mucosal immunoglobulin A: two separate but related systems of immune defense?. Ann. Intern. Med. 106:892.
  5. Kozlowski, P. A., R. M. Lynch, R. R. Patterson, S. Cu-Uvin, T. P. Flanigan, M. R. Neutra. 2000. Modified wick method using Weck-Cel sponges for collection of human rectal secretions and analysis of mucosal HIV antibody. J. Acquired Immune Defic. Syndr. 24:297.
  6. Mestecky, J., M. W. Russell, C. O. Elson. 1999. Intestinal IgA: novel views on its function in the defense of the largest mucosal surface. Gut 44:2.[Free Full Text]
  7. Mestecky, J., I. Moro, B. J. Underdown. 1999. Mucosal Immunoglobulins. P. Ogra, and J. Mestecky, and M. E. Lamm, and W. Strober, and J. Bienenstock, and J. R. McGhee, eds. Mucosal Immunology 133. Academic Press, San Diego.
  8. Mostov, K.. 1994. Transepithelial transport of immunoglobulins. Annu. Rev. Immunol. 12:63.[Medline]
  9. Lamm, M. E.. 1997. Interactions of antigens and antibodies at mucosal surfaces. Annu. Rev. Microbiol. 51:311.[Medline]
  10. Phalipon, A., B. Corthesy. 2003. Novel functions of the polymeric Ig receptor: well beyond transport of immunoglobulins. Trends Immunol. 24:55.[Medline]
  11. Hughes, G. J., A. J. Reason, L. Savoy, J. Jaton, S. Frutiger-Hughes. 1999. Carbohydrate moieties in human secretory component. Biochim. Biophys. Acta 1434:86.[Medline]
  12. Mattu, T. S., R. J. Pleass, A. C. Willis, M. Kilian, M. R. Wormald, A. C. Lellouch, P. M. Rudd, J. M. Woof, R. A. Dwek. 1998. The glycosylation and structure of human serum IgA1, Fab, and Fc regions and the role of N-glycosylation on Fc{alpha} receptor interactions. J. Biol. Chem. 273:2260.[Abstract/Free Full Text]
  13. Royle, L., A. Roos, D. J. Harvey, M. R. Wormald, D. van Gijlswijk-Janssen, R. M. el Redwan, I. A. Wilson, M. R. Daha, R. A. Dwek, P. M. Rudd. 2003. Secretory IgA N- and O-glycans provide a link between the innate and adaptive immune systems. J. Biol. Chem. 278:20140.[Abstract/Free Full Text]
  14. Dallas, S. D., R. D. Rolfe. 1998. Binding of Clostridium difficile toxin A to human milk secretory component. J. Med. Microbiol. 47:879.[Abstract]
  15. Barbieri, L., M. G. Battelli, F. Stirpe. 1993. Ribosome-inactivating proteins from plants. Biochim. Biophys. Acta 1154:237.[Medline]
  16. Nielsen, K., R. S. Boston. 2001. Ribosome-inactivating proteins: a plant perspective. Annu. Rev. Plant Physiol. Plant Mol. Biol. 52:785.[Medline]
  17. Endo, Y., K. Mitsui, M. Motizuki, K. Tsurugi. 1987. The mechanism of action of ricin and related toxins on eukaryotic ribosomes. J. Biol. Chem. 262:5908.[Abstract/Free Full Text]
  18. Pappenheimer, A. M., Jr, S. Olsnes, A. A. Harper. 1974. Lectins from Abrus precatorius and Ricinus communis. I. Immunochemical relationships between toxins and agglutinins. J. Immunol. 113:835.[Abstract/Free Full Text]
  19. The NIAID Biodefense Research Agenda for Category B and C Priority Pathogens 2003 National Institutes of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, MD.
  20. Sandvig, K., B. van Deurs. 2000. Entry of ricin and Shiga toxin into cells: molecular mechanisms and medical perspectives. EMBO J. 19:5943.[Medline]
  21. Baenziger, J. U., D. Fiete. 1979. Structural determinants of Ricinus communis agglutinin and toxin specificity for oligosaccharides. J. Biol. Chem. 254:9795.[Free Full Text]
  22. Debray, H., D. Decout, G. Strecker, G. Spik, J. Montreuil. 1981. Specificity of twelve lectins towards oligosaccharides and glycopeptides related to N-glycosylproteins. Eur. J. Biochem. 117:41.[Medline]
  23. Green, E. D., R. M. Brodbeck, J. U. Baenziger. 1987. Lectin affinity high-performance liquid chromatography: interactions of N-glycanase-released oligosaccharides with Ricinus communis agglutinin I and Ricinus communis agglutinin II. J. Biol. Chem. 262:12030.[Abstract/Free Full Text]
  24. Nicolson, G. L., J. Blaustein. 1972. The interaction of Ricinus communis agglutinin with normal and tumor cell surfaces. Biochim. Biophys. Acta 266:543.[Medline]
  25. Chuang, P. D., S. L. Morrison. 1997. Elimination of N-linked glycosylation sites from the human IgA1 constant region: effects on structure and function. J. Immunol. 158:724.[Abstract]
  26. Rifai, A., K. Fadden, S. L. Morrison, K. R. Chintalacharuvu. 2000. The N-glycans determine the differential blood clearance and hepatic uptake of human immunoglobulin (Ig)A1 and IgA2 isotypes. J. Exp. Med. 191:2171.[Abstract/Free Full Text]
  27. Apter, F. M., W. I. Lencer, R. A. Finkelstein, J. J. Mekalanos, M. R. Neutra. 1993. Monoclonal immunoglobulin A antibodies directed against cholera toxin prevent the toxin-induced chloride secretory response and block toxin binding to intestinal epithelial cells in vitro. Infect. Immun. 61:5271.[Abstract/Free Full Text]
  28. Harlow, E., D. Lane. 1999. Using Antibodies Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.
  29. Crago, S. S., W. H. Kutteh, I. Moro, M. R. Allansmith, J. Radl, J. J. Haaijman, J. Mestecky. 1984. Distribution of IgA1-, IgA2-, and J chain-containing cells in human tissues. J. Immunol. 132:16.[Medline]
  30. Tomana, M., W. Niedermeier, J. Mestecky, F. Skvaril. 1976. The differences in carbohydrate composition between the subclasses of IgA immunoglobulins. Immunochemistry 13:325.[Medline]
  31. Giannasca, K. T., P. J. Giannasca, M. R. Neutra. 1996. Adherence of Salmonella typhimurium to Caco-2 cells: identification of a glycoconjugate receptor. Infect. Immun. 64:135.[Abstract]
  32. Maury, J., C. Nicoletti, L. Guzzo-Chambraud, S. Maroux. 1995. The filamentous brush border glycocalyx, a mucin-like marker of enterocyte hyper-polarization. Eur. J. Biochem. 228:323.[Medline]
  33. de Oliveira, I. R., A. N. de Araujo, S. N. Bao, L. G. Giugliano. 2001. Binding of lactoferrin and free secretory component to enterotoxigenic Escherichia coli. FEMS Microb. Lett. 203:29.[Medline]
  34. Schroten, H., C. Stapper, R. Plogmann, H. Kohler, J. Hacker, F. G. Hanisch. 1998. Fab-independent antiadhesion effects of secretory immunoglobulin A on S-fimbriated Escherichia coli are mediated by sialyloligosaccharides. Infect. Immun. 66:3971.[Abstract/Free Full Text]
  35. Wold, A. E., J. Mestecky, M. Tomona, A. Kobata, H. Ohbayashi, T. Endo, C. S. Eden. 1990. Secretory immunoglobulin A carries oligosaccharide receptors for Escherichia coli type 1 fimbrial lectin. Infect. Immun. 58:3073.[Abstract/Free Full Text]
  36. Phalipon, A., A. Cardona, J. P. Kraehenbuhl, L. Edelman, P. J. Sansonetti, B. Corthesy. 2002. Secretory component: a new role in secretory IgA-mediated immune exclusion in vivo. Immunity 17:107.[Medline]
  37. Jefferis, R., J. Lund. 1997. Glycosylation of antibody molecules: structural and functional significance. Chem. Immunol. 65:111.[Medline]
  38. Baenziger, J. U.. 1979. Structure of the oligosaccharide of human J chain. J. Biol. Chem. 254:4063.[Abstract/Free Full Text]
  39. Rutenber, E., B. J. Katzin, S. Ernst, E. J. Collins, D. Mlsna, M. P. Ready, J. D. Robertus. 1991. Crystallographic refinement of ricin to 2.5 A. Proteins 10:240.[Medline]
  40. Saltzman, W. M., M. L. Radomsky, K. J. Whaley, R. A. Cone. 1994. Antibody diffusion in human cervical mucus. Biophys. J. 66:508.[Medline]
  41. Casadevall, A.. 2002. Passive antibody administration (immediate immunity) as a specific defense against biological weapons. Emerging Infect. Dis. 8:833.[Medline]
  42. Houston, L. L.. 1982. Protection of mice from ricin poisoning by treatment with antibodies directed against ricin. J. Toxicol. Clin. Toxicol. 19:385.[Medline]
  43. Poli, M. A., V. R. Rivera, M. L. Pitt, P. Vogel. 1996. Aerosolized specific antibody protects mice from lung injury associated with aerosolized ricin exposure. Toxicon 34:1037.[Medline]
  44. Weltzin, R., T. P. Monath. 1999. Intranasal antibody prophylaxis for protection against viral disease. Clin. Microbiol. Rev. 12:383.[Abstract/Free Full Text]
  45. Corthesy, B.. 2002. Recombinant immunoglobulin A: powerful tools for fundamental and applied research. Trends Biotechnol. 20:65.[Medline]
  46. Mantis, N. J., M. C. Cheung, K. R. Chintalacharuvu, J. Rey, B. Corthesy, M. R. Neutra. 2002. Selective adherence of IgA to murine Peyer’s patch M cells: evidence for a novel IgA receptor. J. Immunol. 169:1844.[Abstract/Free Full Text]
  47. Neutra, M., T. Phillips, E. Mayer, D. Fishkind. 1987. Transport of membrane-bound macromolecules by M cells in follicle-associated epithelium of rabbit Peyer’s patch. Cell Tissue Res. 247:537.[Medline]
  48. Geissmann, F., P. Launay, B. Pasquier, Y. Lepelletier, M. Leborgne, A. Lehuen, N. Brousse, R. C. Monteiro. 2001. A subset of human dendritic cells expresses IgA Fc receptor (CD89), which mediates internalization and activation upon cross-linking by IgA complexes. J. Immunol. 166:346.[Abstract/Free Full Text]
  49. Iwasaki, A., B. Kelsall. 2001. Unique functions of CD11b+, CD8a+, and double negative Peyer’s patch dendritic cells. J. Immunol. 166:4884.[Abstract/Free Full Text]



This article has been cited by other articles:


Home page
 Infect. Immun.Home page
J. M. Yoder, R. U. Aslam, and N. J. Mantis
Evidence for Widespread Epithelial Damage and Coincident Production of Monocyte Chemotactic Protein 1 in a Murine Model of Intestinal Ricin Intoxication
Infect. Immun., April 1, 2007; 75(4): 1745 - 1750.
[Abstract] [Full Text] [PDF]

Home page
 J. Immunol.Home page
B. Corthesy
Roundtrip Ticket for Secretory IgA: Role in Mucosal Homeostasis?
J. Immunol., January 1, 2007; 178(1): 27 - 32.
[Abstract] [Full Text] [PDF]

Home page
 Infect. Immun.Home page
N. J. Mantis, C. R. McGuinness, O. Sonuyi, G. Edwards, and S. A. Farrant
Immunoglobulin A Antibodies against Ricin A and B Subunits Protect Epithelial Cells from Ricin Intoxication.
Infect. Immun., June 1, 2006; 74(6): 3455 - 3462.
[Abstract] [Full Text] [PDF]

Home page
 J. Biol. Chem.Home page
C. Perrier, N. Sprenger, and B. Corthesy
Glycans on Secretory Component Participate in Innate Protection against Mucosal Pathogens
J. Biol. Chem., May 19, 2006; 281(20): 14280 - 14287.
[Abstract] [Full Text] [PDF]

Home page
 J. Immunol.Home page
M. J. Lewis, R. J. Pleass, M. R. Batten, J. D. Atkin, and J. M. Woof
Structural Requirements for the Interaction of Human IgA with the Human Polymeric Ig Receptor
J. Immunol., November 15, 2005; 175(10): 6694 - 6701.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Mantis, N. J.
Right arrow Articles by Mehta, S.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Mantis, N. J.
Right arrow Articles by Mehta, S.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS