Rabbit Tonsil-associated M-cells Express Cytokeratin 20 and Take Up Particulate Antigen

Abstract

M-cells are believed to play a pivotal role in initiation of the immune response. These cells, located in the epithelia that overlie mucosal lymphoid follicles, are responsible for the active uptake of particulate antigens and for their translocation to the underlying lymphoid tissue. The identification of reliable markers for M-cells is therefore extremely important for the study of the initial steps that lead to the immune response. For this purpose, we studied cytokeratin 20 (CK20) expression in the epithelium of rabbit palatine tonsils by immunofluorescence, confocal microscopy, and Western blotting. CK20⁺ cells were observed in all rabbit palatine tonsils examined. By Western blotting, one CK20-immunoreactive band was identified at 46 kD on samples of proteins from the intermediate filament-enriched cytoskeletal fraction of tonsil epithelium. Double labeling of CK20⁺ cells with cell-specific markers confirmed that such cells were actually M-cells. Moreover, CK20⁺ M-cells displayed a mature phenotype (they formed pockets harboring lymphoid cells) and were functionally competent because they could take up particulate antigens from the pharyngeal lumen. We conclude that CK20 is an M-cell marker for rabbit palatine tonsils. Moreover, we can hypothesize the use of M-cells as a possible site for antigen delivery of particle-based vaccines.

Keywords

M-cells cytokeratin 20 tonsils MALT antigen transport

Mucosa-associated lymphoid tissue (MALT) represents an impressive first line of defense by the immune system that protects a huge mucosal surface area in adult human. Belonging to MALT are lymphoid follicles, both solitary and as aggregates, located along mucosal surfaces. The major follicle aggregates are anatomic structures easily recognizable to the naked eye (i.e., adenoids, tonsils, Peyer's patches, appendix). These aggregates play a key role in mucosal immunity as the sites at which the mucosal immune system meets foreign antigens (Ags) in a regulated fashion. Such encounter can lead either to the development of a mucosal immune response or to the induction of oral tolerance (Shalaby 1995; Brandtzaeg 1996; Brandtzaeg et al. 1999). In the former case, it is believed that, on antigenic stimulation, activated B-cells spread to the lamina propria of all mucosae and secrete sIgA (Shalaby 1995). Peyer's patches (PPs), as well as other aggregates, are permanently in contact with foreign Ags whose epithelial barrier crossing, if uncontrolled, could cause harm to the organism. On the other hand, an enhanced uptake of immunogens specifically engineered as oral vaccines to develop an immune response would be highly desirable. For this reason, the structure and the mechanisms by which the follicle-associated epithelium (FAE) can manage Ag sampling in a controlled way is very important to understanding of the first steps that lead to the development of mucosal immunity. Ag sampling in PPs has been largely clarified and is achieved by specialized membranous (M) cells that display very peculiar morphological and functional characteristics (Regoli et al. 1995a; Gebert et al. 1996). M-cells have deep invaginations of the basolateral membrane that form large hollows, also called pockets, where immunocompetent cells are found (Bye et al. 1984). M-cells also have remarkable transcytotic activities. If on the one hand they allow the passage of viable lymphoid cells from the pockets into the intestinal lumen (Regoli et al. 1994), on the other hand they take up Ags from the intestinal lumen and drive them into the underlying follicular dome area, thereby accomplishing their Ag-sampling function (Regoli et al. 1995b; Borghesi et al. 1996; Jepson and Clark 1998). Although several markers have been proposed to identify M-cells, their use is generally limited. In particular, whereas some markers (e.g., lectins) are site-specific due to the heterogeneity of the M-cell glycocalyx, others (alkaline phosphatase and some cytokeratins) are not “clear-cut” markers because they can also stain other cells of the FAE, albeit with different intensity (Jepson and Clark 1998). In rabbit, vimentin represents an exception (Nicoletti 2000). Vimentin is expressed by all M-cells thus far investigated, regardless of the location. On the other hand, it is not expressed by any other epithelial cell type within the FAE. M-cells have been also demonstrated in rabbit palatine tonsils because of their vimentin immunoreactivity and their ability to internalize horseradish peroxidase from the lumen (Gebert 1995,1996,1997; Gebert et al. 1996). The M-cell pattern of intermediate filament (IF) expression has been studied in detail only in the intestine. In addition to vimentin, intestinal M-cells express enterocyte-like cytokeratins (CKs) in human (Kucharzik et al. 1998), rabbit (Jepson et al. 1992), rat (Rautenberg et al. 1996), and pig (Gebert et al. 1994). The pattern of CK expression in M-cells belonging to other epithelia has not been studied yet. In particular, CK expression of M-cells in palatine tonsil appears especially intriguing because in that location M-cells are embedded in a complex epithelium that expresses a different set of CKs (Moll et al. 1982). With this in mind, we asked the question as to whether M-cells maintain the same pattern of CK expression displayed in the intestine or whether this changes according to the tissue in which they are located. We therefore investigated the expression of CK20 in M-cells of rabbit palatine tonsils. Taking advantage of vimentin immunoreactivity as a positive M-cell marker, we demonstrate that a large subset of M-cells do express CK20 and that such CK can be used as an M-cell marker. Moreover, we show that CK20⁺ M-cells are functionally competent to sample antigens from the tonsil lumen, as established by their uptake of fluorescent polystyrene microparticles. The potential role of tonsil-associated M-cells in mucosal immunity is also addressed.

Materials and Methods

Antibodies and Lectin

A goat anti-vimentin polyclonal antibody (PAb) (Hartig et al. 1997) was a kind gift from Prof. Peter Traub (Max-Planck-Institut für Zellbiologie; Rosenhof, Ladenburg, Germany). MAb anti-CK20 (clone Ks20.8) was purchased from DAKO (Glostrup, Denmark). In addition, for direct immunohistochemical (IHC) demonstration of rabbit IgG-containing cells, double-labeling grade tetramethylrhodamine isothiocyanate (TRITC)-conjugated anti-rabbit IgG was obtained from Chemicon (Temecula, CA). As secondary antibodies, fluorescein isothiocyanate (FITC)-conjugated anti-mouse IgG antibody (Fab-specific) was purchased from Sigma-Aldrich (Milan, Italy) and double-labeling grade TRITC-conjugated anti-goat and anti-mouse IgG antibodies were from Chemicon. FITC-conjugated lectin Ulex europaeus 1 agglutinin (UEA-I) specific for α-l-fucose residues was obtained from Sigma-Aldrich.

Animals and Tissues

Five male New Zealand rabbits (1.5–2 kg bw) (Charles River Italia Laboratories; Rome, Italy) were used for this study. Animals were treated according to current Italian law. Palatine tonsils were sampled from animals previously anesthetized with an IV injection of pentobarbital (30 mg/kg) and sacrificed with an overdose of the same anesthetic. Some animals, anesthetized as above, were tracheotomized and, to expose palatine tonsils, the anterior wall of the pharynx was opened along the median plane. Forty μl of a PBS solution containing 2.5 × 10⁹/ml FITC-conjugated Fluoresbrite polystyrene microparticles (Polysciences; Eppelheim, Germany) were placed with a micropipette into the tonsil crypts for 15 min. At this point, tonsils were quickly removed, briefly rinsed in PBS to wash away the excess microparticles, and processed as all other samples. Specimens were mounted on Tissue-Tek OCT (Sakura Finetek Europe; Zoeterwoude, The Netherlands) and immediately frozen in isopentane prechilled in liquid nitrogen. Samples were stored at −80C or directly transferred to a cryostat (Frigocut N 2800; Reichert-Jung, Wetzlar, Germany) in which consecutive sections (7–10 μm thick) were cut at −20C and mounted on SuperFrost Plus slides (Menzel-Gläser; Mannheim, Germany). Sections were air-dried, fixed with cold acetone for 10 min at −20C, and stored at −80C. Sections from tonsils incubated with Fluoresbrite microparticles were directly stored without the fixation step with acetone.

Immunohistochemistry, Lectin Histochemistry, and Confocal Microscopy

Sections for single-labeling experiments were placed in PBS for 10 min and then blocked with 1% BSA for 10 min at RT. Then the slides were incubated overnight at 4C with anti-CK20 at a dilution of 1:100 and for 90 min with 1:100 FITC-conjugated anti-mouse IgG antibody. Almost the same protocol was followed with unfixed sections from tonsils incubated with 0.5-μm-diameter fluorescein Fluoresbrite microparticles. The only change was that the secondary antibody employed was a TRITC-conjugated anti-mouse IgG antibody.

Sections for vimentin/CK20 double-labeling experiments were incubated overnight at 4C with goat anti-vimentin PAb at a 1:50 dilution. After washing, sections were incubated for 3 hr at RT with TRITC-conjugated anti-goat IgG at a dilution of 1:100. Slides were then incubated for 18 hr at 4C with MAb anti-CK20 diluted 1:100 and, after washing, sections were finally incubated for 3 hr at RT with FITC-conjugated anti-mouse IgG. Controls for all IHC experiments were performed by omission of the primary antibodies or by their replacement with the appropriate non-immune serum.

Sections for CK20/rabbit IgG double-labeling experiments were incubated for 18 hr at 4C with MAb anti-CK20 diluted 1:100. After washing, sections were incubated for 3 hr at RT with FITC-conjugated anti-mouse IgG. Slides were rinsed and incubated with TRITC-conjugated anti-rabbit IgG.

Some sections were double labeled using UEA-I agglutinin and the MAb anti-CK20. For this purpose, slides were processed as follows. After washing in PBS, sections were blocked with 10% goat serum for 10 min at RT and then incubated overnight at 4C with MAb anti-CK20 (diluted 1:100). After rinsing in PBS, sections were blocked as above and incubated for 4 hr at RT with TRITC-conjugated antimouse IgG (diluted 1:50). Slides were thoroughly rinsed and then fixed with a solution of 1% glutaraldehyde and 2% paraformaldehyde in PBS for 10 min at RT. After washing in PBS, sections were dehydrated in increasing concentrations of ethanol and left in 100% ethanol for 10 min at RT. Slides were then rehydrated in a series of decreasing concentrations of ethanol and washed in PBS. Then the sections were blocked with 2.5% BSA for 20 min at RT and incubated overnight at 4C with UEA-I at a concentration of 5 μg/ml in 1% PBS-BSA. Controls were performed by preincubation of UEA-I with fucose or mannose separately.

Unfixed sections (10–25-μm thick) of tonsils incubated with fluorescent polystyrene microparticles were washed with PBS, blocked with 10% goat serum for 10 min at RT, and then incubated overnight at 4C with MAb anti-CK20 (diluted 1:100). After rinsing in PBS, sections were incubated for 3 hr at RT with TRITC-conjugated anti-mouse IgG (diluted 1:100) and thoroughly washed.

Fluorescent staining was observed with a Zeiss Axioplan microscope or a TCS 4D Leica laser scanning confocal microscope.

SDS-PAGE and Western Blotting of IF-enriched Cytoskeletal Fractions

To confirm the IHC results, Western blotting analysis with anti-CK20 antibody was carried out on proteins of the IF-enriched cytoskeletal fraction of palatine tonsil epithelium separated by SDS-PAGE. As a positive control, proteins of the IF-enriched cytoskeletal fraction of the intestinal epithelium were separated and blotted in parallel with tonsil samples. IF-enriched cytoskeletal fractions were prepared as previously reported (Achtstaetter et al. 1986), with minor modifications. Briefly, tonsils were removed and their epithelia were isolated by microdissecting away the submucosa, including lymphoid follicles. Then the epithelia were minced and placed in 1 ml of homogenization buffer (96 mM NaCl, 8 mM KH₂PO₄, 5.6 mM Na₂PO₄2H₂O, 1.5 mM KCl, 10 mM EDTA, 0.1 mM dithiothretol (DTT), 2.5 mg/ml aprotinin, 100 mM PMSF, pH 6.8). Intestinal epithelium, isolated by gentle scraping of the intestinal surface with a scalpel, was placed in 1 ml of the above homogenization buffer. Both samples (tonsil and intestinal epithelia) were homogenized with Dounce homogenizers and then filtered through four layers of gauze. Three ml of very high salt buffer (2 M KCl, 200 mM NaCl, 10 mM Tris-HCl, 0.1 mM DTT, 2.5 mg/ml aprotinin, 100 mM PMSF, pH 7.4) was added to each homogenate and the obtained suspensions were stirred for 30 min on ice. Samples were next homogenized once more in Dounce homogenizers and centrifuged at 10,000 × g for 20 min at 4C. After removal of the supernatants, each pellet was resuspended in 3 ml of high salt buffer (10 mM Tris-HCl, 140 mM NaCl, 1.5 M KCl, 5 mM EDTA, 0.5% w/v Triton X-100, 2.5 mg/ml A, 100 mM PMSF, pH 7.6) with Dounce homogenizers. Samples were gently agitated for 30 min on ice, centrifuged at 10,000 × g for 20 min at 4C, and resuspended in 3 ml of high salt buffer. These steps were repeated once more but the obtained pellets, instead of being resuspended in high salt buffer, were washed in PBS, pH 7.4, with 0.1 mM DTT. After a centrifugation at 10,000 × g for 20 min at 4C, the supernatants were discarded and the final pellets were stored at −20C until use.

IF-enriched cytoskeletal fractions were separated by electrophoresis through a 10% polyacrylamide gel according to Laemmli (1970) and transferred to nitrocellulose in a Bio-Rad Transblot apparatus (Bio-Rad; Hercules, CA). A 5% (w/v) solution of skimmed milk in TBS was used to quench nonspecific protein adherence of the antibodies. Membranes were incubated overnight at RT with 1:500 anti-CK20 antibody in 5% (w/v) solution of skimmed milk in TBS. Specific bands were detected using an electrochemiluminescence kit (Roche Diagnostics; Milan, Italy).

Results

Palatine Tonsil Epithelium Contains CK20-immunoreactive Cells

Immunocytochemical experiments carried out with anti-CK20 MAb on sections of rabbit palatine tonsils displayed strong staining on a subset of epithelial cells that were irregularly distributed along the crypt epithelium. CK20⁺ cells did not appear to be a homogeneous population in terms of distribution, morphology, and intensity of the immunofluorescent staining. Some cells appeared to be stretched through several layers of the epithelium, whereas others displayed very long processes that outlined great hollows, and still others appeared to be cuboidal cells (Figure 1a). In contrast, the adjacent pharyngeal epithelium did not display CK20⁺ cells, confirming that such cells were specifically located in the crypt epithelium (Figure 1b). Even though the antibody employed in this study has been well characterized (Moll et al. 1990,1992), to verify whether it recognized actual CK20 we carried out Western blotting analysis of the IF-enriched cytoskeletal fraction of tonsil epithelium. The same protocol was followed with intestinal epithelium as a positive CK20-immunoreactive control. Western blotting analysis with anti-CK20 MAb of IF-enriched cytoskeletal fractions revealed a single band in tonsil samples co-migrating as expected at 46 kD with immunoreactive CK20 from intestinal epithelium (Figure 2). Interestingly, samples of the intestinal epithelium exhibited three CK20⁺ bands of very similar molecular weight, probably representing different post-translational modifications of CK20 monomers (Figure 2).

CK20-immunoreactive Cells Are M-cells

Overall, the morphology of some CK20⁺ cells that outlined great hollows suggested the possibility that these cells were M-cells. To verify this, sections of rabbit palatine tonsils were double-labeled for the simultaneous detection of CK20 and vimentin, which is a well-known rabbit M-cell marker. A previously characterized goat anti-vimentin PAb (Hartig et al. 1997) was used for double-labeling experiments. Double immunofluorescent staining revealed that all CK20⁺ cells were also vimentin-positive (Figure 3). Confocal microscopic analysis at high magnification showed that vimentin and CK20 were arranged in the form of two distinct networks, with vimentin always being located deeper than CK20 within the cell (Figure 4). UEA-I agglutinin has been reported to be a further marker of rabbit tonsil M-cells (Gebert 1996). In our hands, however, lectin UEA-I did not display a high level of specificity for M-cells. Indeed, other cells (i.e., some squamous cells) labeled with UEA-I. Nevertheless, double-stained sections with anti-CK20 MAb and UEA-I enabled us to establish that CK20 labeling closely resembled the UEA-I pattern of staining (Figure 5). Preincubation of UEA-I with fucose drastically reduced labeling. Preincubation of UEA-I with mannose did not influence the labeling pattern of UEA-I.

Figure 1

(a) Section of rabbit palatine tonsil immunostained with anti-CK20 antibody. CK20-immunoreactive cells span through all the layers of the crypt epithelium. Many of them, mainly located on the superficial layers of the epithelium, display large hollows outlined by moderately stained cytoplasmic processes (arrows). (b,c) A portion of epithelium located at the edge of the tonsil crypt double stained with anti-CK20 (b) and anti-vimentin antibodies (c). CK20⁺ cells and vimentin⁺ cells can be spotted within the epithelium on the right half of the figures that correspond to the crypt epithelium (C). L, pharyngeal lumen; P, pharyngeal epithelium. Bar = 25 μm.

A well-known feature of M-cells is their ability to harbor lymphoid cells in large pockets created by invagination of basolateral membranes. For this reason, to investigate whether the hollows outlined by some tonsil CK20⁺ cells actually contained immunocompetent cells, we double labeled sections with anti-CK20 MAb and anti-rabbit IgG PAb. Some CK20⁺ M-cell pockets contained IgG-expressing cells (Figure 6) confirming that CK20-immunoreactive M-cells retained their ability to harbor lymphoid cells.

CK20-immunoreactive M-cells Efficiently Take Up Luminal Particles

To verify whether tonsil CK20⁺ M-cells were functionally capable of taking up particles from the lumen, we incubated rabbit tonsils with fluorescent latex beads for 15 min. Scanning sections of tonsil epithelium along single focal (XY) planes by confocal microscopy showed that even after such a short time some fluorescent particles were taken up from the lumen and could be found in some CK20⁺ cells located mainly, but not exclusively, in the most superficial layers of crypt epithelium (Figure 7a). These results were further supported by scanning sections along specific XZ planes passing through fluorescent beads. These projections demonstrated that fluorescent beads were actually located deep in the thickness of tissue sections, either in or outside of M-cells (Figures 7b and 7e). To confirm the integrity of CK20⁺ cells containing fluorescent beads, whole cells were scanned through 21 different focal planes. These planes, placed one on top of the other, allowed a 3D reconstruction of the cells that could be virtually tilted and observed under different views. This procedure confirmed the presence of fluorescent beads inside CK20⁺ M cells (Figures 7f–7l).

Figure 2

Western blotting with anti-CK20 antibody. Lane 1, IF-enriched cytoskeletal fraction of rabbit palatine tonsil epithelium. Lane 2, molecular weight standards (250 kD, 150 kD, 100 kD, 75 kD, 50 kD, 37 kD, 25 kD). Lane 3, IF-enriched cytoskeletal fraction of rabbit intestinal epithelium. One CK20-immunoreactive band migrating at 46 kD can be detected in Lane 1. Three CK20-immunoreactive bands migrating at 46, 47, and 48 kD can be detected in Lane 3.

Figure 3

Double immunofluorescent staining on sections of rabbit palatine tonsil. (a,c, vimentin); (b,d, CK20). (a,b) Many cells located in the superficial layer of the epithelium co-express vimentin and CK20. Arrows outline the inner border of the epithelium. (c,d) Higher magnification of a small portion of the epithelium. A double-stained cell (arrowhead) is provided with a large pocket. L, pharyngeal lumen. Bars = 50 μm.

Figure 4

Confocal microscopy of double-stained sections of rabbit palatine tonsil epithelium. M-cells can be identified by the co-expression of vimentin (red) and CK20 (green). Vimentin and CK20 form two clearly independent networks of IFs. CK20-immunoreactive IFs are localized mainly at the periphery of the cells, whereas vimentin is prevalently perinuclear. (a) M-cells of the intermediate layers of the epithelium. (b,c) To show the different arrangements of the IF networks, one M-cell of the superficial layer of the epithelium has been scanned along different focal planes passing through the nucleus (b) or through the periphery of the cell (c). Bar = 5 μm.

Discussion

The present study demonstrates that M-cells of rabbit palatine tonsils express CK20 and take up particulate antigens from the pharyngeal lumen. This report is based on the simultaneous detection in the same cells of CK20 and other well-known rabbit M-cell markers, such as high levels of fucose residues and vimentin (Jepson and Clark 1998).

Figure 5

Double fluorescent staining on section of rabbit palatine tonsil. (a) UEA-I; (b) CK20. The pattern of labeling of the two fluorochromes is almost the same because both stainings are located mainly on the middle and superficial layers of the epithelium. Moreover, the certain identification of double-labeled cells is possible in some cases (arrows). L, pharyngeal lumen. Bar = 50 μm.

Specificity of CK20 staining is confirmed by the following elements: control experiments (see Materials and Methods) result in labeling suppression; confocal microscopy demonstrates that CK20 immunoreactivity is localized, as expected, to a complex of cytoskeletal elements that differ from the vimentin network; and Western blotting of the IF-enriched cytoskeletal fraction displays only a single band of the appropriate molecular weight co-migrating with intestinal CK20-immunoreactive proteins.

Vimentin/CK20 double-labeling experiments show that CK20-immunoreactive cells also express vimentin. The converse, however, does not apply. Clearly, many vimentin⁺/CK20^– cells should be classified as cells that massively infiltrate the epithelium (i.e., macrophages and other immunocompetent cells) but, for the time being, we cannot rule out the possibility that some of these cells may correspond to a minor subset of tonsil M-cells. However, because most of these cells are located in the lower layers of the epithelium, an alternative possibility is that they may represent immature M-cells. Unfortunately, the uncertainty of assigning vimentin⁺/CK20^– cells to the population of M-cells or to the infiltrating immunocompetent cells prevents us from producing any quantitative or semiquantitative data that can reliably determine the ratio between vimentin⁺/CK20^– M-cells and the entire M-cell population.

Figure 6

Confocal microscopy of double-stained sections of rabbit palatine tonsil epithelium. M-cells, stained with anti-CK20 MAb (green), display pockets that contain immunocompetent cells. A subset of lymphoid cells has been stained with anti-rabbit IgG antibody (red). One M-cell (arrows) outlines a large hollow that harbors three lymphoid cells. Bar = 10 μm.

Even though we have identified CK20⁺ cells provided with fucose residues that label with specific lectins, the use of UEA-I in our hands did not prove useful to confirm that CK20⁺ cells corresponded to the entire population of M-cells. Indeed, probably due to the peculiar protocol that we had to follow to obtain an acceptable double-staining of both CK20 and fucose residues, the labeling with UEA-I appeared largely unspecific, involving a population of epithelial cells much broader than was expected and that was previously reported (Gebert 1997). Clearly, CKs cannot be a useful tool for M-cell identification when M-cells are embedded in an epithelium with the same profile of CK expression (Kucharzik et al. 1998). However, when M-cells are located in epithelia with a different CK profile (i.e., composed epithelia), the presence of M-cell-specific CKs can be used as a valuable marker. This is the case for M-cells from palatine tonsils, in which we have demonstrated specific CK20 staining. Even though with the limitations due to the possible existence of a subpopulation of vimentin⁺/CK20^– M cells, in palatine tonsils CK20 can be added to the very short list of documented M-cell markers.

Figure 7

Confocal microscopy of sections from rabbit palatine tonsil incubated with fluorescent microparticles (green) and stained with anti-CK20 MAb (red). One section (a) was scanned through a single focal plane. Many fluorescent beads are located in the pharyngeal lumen and on the surface of the epithelium, but some of them (arrowheads) are already situated within the epithelium (E). L, lumen. (b,c) One 10-μm-thick section scanned through a single focal plane along which a bead can be found (b). The level of the focal plane shown in b is demonstrated by the visualization of the orthogonal XZ plane (c) passing through the bead itself. The bead appears located almost at the border between the tissue and the slide on the edge of a CK20⁺ cell. (d–i) One 25-μm section scanned through 21 XY focal planes and one XZ plane. One fluorescent bead (arrowhead) appears located within a CK20⁺ cell, as shown by two orthogonal planes passing through the bead itself (d, Y plane; e, XZ plane). The CK20⁺ IF network of the cell looks continuous through all XY planes scannerized. The same cell (arrows), rebuilt placing XY planes one on the top of the other (f) were virtually tilted by different degrees (g–i). (j) Cell tilted by 90°. (i) Cell tilted almost by 180°. Bars = 5 μm.

From a functional point of view, we demonstrated that CK20⁺ M-cells have the ability to form wide pockets harboring lymphoid cells and to take up fluorescent polystyrene microparticles from the pharyngeal lumen. These features are considered distinctive of functionally competent M-cells (Bye et al. 1984; Kernéis et al. 1997; Gebert et al. 1999). Indeed, not all tonsil CK20⁺ M-cells display the characteristic pockets, this aspect being mainly confined to the M-cells located close to the pharyngeal lumen. This observation could be in accordance with a progressive maturation of M-cells moving from the basal layers of the tonsil epithelium towards the surface, somewhat resembling what occurs elsewhere. In the intestine, for example, M-cells originate from dome-surrounding intestinal crypts and, during their migration along the flanks of the dome, progressively acquire morphological and functional maturity (Bye et al. 1984; Gebert et al. 1999).

We have reported the presence of IgG⁺ cells in M-cell pockets. This observation confirms that the cells harbored within CK20⁺ M-cell hollows are actually immunocompetent cells, corroborating the acquisition by M-cells of a functionally mature phenotype. However, several cells contained in M-cell pockets do not stain with anti-IgG antibodies. This was expected because of the heterogeneous population of lymphoid cells infiltrating M-cells. Indeed, in this population, which accounts mostly for T and B memory cells (Farstad et al. 1994), IgG⁺ B-cells correspond only to a subset of cells, even though they are well represented at least in human tonsils (Liu et al. 1995).

A very important function of M-cells, and perhaps the most important, is the ability to take up a large set of microorganisms and even inert particles that are steered into the pockets where they meet immunocompetent cells (Jepson and Clark 1998; Meynell et al. 1999; Nicoletti 2000). Exactly what happens in the pockets is not yet clear, but it is believed that important interactions may occur between antigens and memory B-cells that might substantially contribute to the initiation of the immune response (Liu et al. 1995; Yamanaka et al. 2001). Therefore, the demonstration that tonsil CK20⁺ M-cells are able to take up fluorescent microparticles is particularly important to demonstrate their full operational capabilities. The kinetics of microparticle uptake appear to be extremely rapid because even after 15 min some of the particles can be found in M-cells situated even in middle layers of the epithelium. However, probably owing to the short time of exposure to the fluorescent beads, the number of beads that could be spotted in the middle layers of epithelium was low. Therefore, to evaluate the actual impact of this phenomenon on the generation of the immune response, longer incubation time points must be investigated. The ability of M-cells to sample material from the pharyngeal lumen was first observed by Gebert (1995). In that case, horseradish peroxidase was used as tracer. To our knowledge, this is the first time that active transport of particulate antigen has been documented. The ability of tonsil-associated M-cells to carry out antigen-sampling activity might have important implications for the development of new strategies for the delivery of certain types of vaccine. It has been recently shown that significant levels of specific IgA antibody to Streptococcus mutans were detected in the saliva of human volunteers after tonsillar administration of a glucosyltransferase-enriched preparation (Childers et al. 2002). In addition, it is noteworthy that high levels of a specific salivary antibody to other pathogens, such as S. sobrinus, can be induced via tonsillar immunization and that the immunological function of palatine tonsils can be modulated by the use of regulatory cytokines, such as IL-1 (Kokuryo et al. 2002). These findings point to tonsils as a possible site for antigen delivery to obtain a local immune response against oral pathogens. In this context, our observation that tonsil-associated M-cells can take up particles opens the way to the development of a particle-based vaccine that can be delivered in a simple and effective way.

Footnotes

Acknowledgements

Supported by intramural funds from the University of Siena (PAR 2003 quota servizi) to E. Bertelli, “fondi di liberalità” from L. Fonzi, and by a grant from the Biotechnology and Biological Sciences Research Council (UK) to C. Nicoletti.

We are indebted to Prof. F. Rosati (Dept of Evolutionary Biology, University of Siena, Siena) for supplying us with UEA-I lectin.

References

Achtstaetter

Hatzfeld

Quinlan

Parmelee

Franke

(1986) Separation of cytokeratin polypeptides by gel electrophoretic and chromatographic techniques and their identification by immunoblotting. Methods Enzymol 134:355–371.

Borghesi

Regoli

Bertelli

Nicoletti

(1996) Modifications of the follicle-associated epithelium by short-term exposure to a non-intestinal bacterium. J Pathol 180:326–332.

Brandtzaeg

(1996) History of oral tolerance and mucosal immunity. Ann NY Acad Sci. 778:1–27.

Brandtzaeg

Baekkevold

Farstad

Jahnsen

Johansen

Nilsen

Yamanaka

(1999) Regional specialization in the mucosal immune system: what happens in the microcompartments?. Immunol Today 20:141–151.

Bye

Allan

Trier

(1984) Structure, distribution, and origin of M cells in Peyer's patches of mouse ileum. Gastroenterology 86:789–801.

Childers

Tong

Desanayaka

Kirk

Michalek

(2002) Humans immunized with Streptococcus mutans antigens by mucosal routes. J Dent Res 81:48–52.

Farstad

Halstensen

Fausa

Brandtzaeg

(1994) Heterogeneity of M cell-associated B and T cells in human Peyer's patches. Immunology 83:457–464.

Gebert

(1995) Identification of M-cells in the rabbit tonsil by vimentin immunohistochemistry and in vivo protein transport. Histochem Cell Biol 104:211–220.

Gebert

(1996) M-cells in the rabbit tonsil exhibit distinctive glycoconjugates in their apical membranes. J Histochem Cytochem 44:1033–1042.

10.

Gebert

(1997) M cells in the rabbit palatine tonsil: the distribution, spatial arrangement and membrane subdomains as defined by confocal lectin histochemistry. Anat Embryol 195:353–358.

11.

Gebert

Fassbender

Werner

Weissferdt

(1999) The development of M cells in Peyer's patches is restricted to specialized dome-associated crypts. Am J Pathol. 154:1573–1582.

12.

Gebert

Rothkotter

Pabst

(1994) Cytokeratin 18 is an M-cell marker in porcine Peyer's patches. Cell Tissue Res 276:213–221.

13.

Gebert

Rothkotter

Pabst

(1996) M cells in Peyer's patches of the intestine. Int Rev Cytol 167:91–159.

14.

Hartig

Huang

Janetzko

Shoeman

Grub

Traub

(1997) Binding of fluorescence- and gold-labeled oligodeoxyribonucleotides to cytoplasmic intermediate filaments in epithelial and fibroblast cells. Exp Cell Res 233:169–197.

15.

Jepson

Clark

(1998) Studying M cells and their role in infection. Trends Microbiol 6:359–365.

16.

Jepson

Mason

Bennett

Simmons

Hirst

(1992) Co-expression of vimentin and cytokeratins in M cells of rabbit intestinal lymphoid follicle-associated epithelium. Histochem J 24:33–39.

17.

Kernéis

Bogdanova

Kraehenbuhl

J-P

Pringault

(1997) Conversion by Peyer's patch lymphocytes of human enterocytes into M cells that transport bacteria. Science. 277:949–951.

18.

Kokuryo

Inoue

Fukuizumi

Tsujisawa

Tominaga

Fukuda

(2002) Evaluation of IL-1 as a mucosal adjuvant in immunization with S. sobrinus cells by tonsilar application in rabbits. Oral Microbiol Immunol 17:163–171.

19.

Kucharzik

Lugering

Schmid

Schmidt

Stoll

Domschke

(1998) Human intestinal M cells exhibit enterocyte-like intermediate filaments. Gut 42:54–62.

20.

Laemmli

(1970) Cleavage of the structural proteins during the assembly of the head of bacteriophage T₄ . Nature. 227:680–685.

21.

Liu

Y-J

Barthélémy

de Bouteiller

Arpin

Durand

Banchereau

(1995) Memory B cells from human tonsils colonize mucosal epithelium and directly present antigen to T cells by rapid up-regulation of B7—1 and B7—2. Immunity 2:239–248.

22.

Meynell

Thomas

James

Holland

Taussig

Nicoletti

(1999) Up-regulation of microsphere transport across the follicle-associated epithelium of Peyer's patch by exposure to Streptococcus pneumoniae R36a. FASEB J 13:611–619.

23.

Moll

Franke

Schiller

Geiger

Krepler

(1982) The catalog of human cytokeratins: patterns of expression in normal epithelia, tumors and cultured cells. Cell 31:11–24.

24.

Moll

Löwe

Laufer

(1992) Cytokeratin 20 in human carcinomas. A new histodiagnostic marker detected by monocolonal antibodies. Am J Pathol. 140:427–447.

25.

Moll

Schiller

Franke

(1990) Identification of protein IT of the intestinal cytoskeleton as a novel type I cytokeratin with unusual properties and expression patterns. J Cell Biol 111:567–580.

26.

Nicoletti

(2000) Unsolved mysteries of intestinal M cells. Gut 47:735–739.

27.

Rautenberg

Cichon

Heyer

Demel

Schmidt

(1996) Immunocytochemical characterization of the follicle-associated epithelium of Peyer's patches: anti-cytokeratin 8 antibody (clone 4.1.18) as a molecular marker for rat M cells. Eur J Cell Biol 71:363–370.

28.

Regoli

Bertelli

Borghesi

Nicoletti

(1995a) Three-dimensional (3D-) reconstruction of M cells in rabbit Peyer's patches: definition of the intraepithelial compartment of the follicle-associated epithelium. Anat Rec 243:19–26.

29.

Regoli

Borghesi

Bertelli

Nicoletti

(1994) A morphological study of the lymphocyte traffic in Peyer's patches after an in vivo antigenic stimulation. Anat Rec 239:47–54.

30.

Regoli

Borghesi

Bertelli

Nicoletti

(1995b) Uptake of a gram-positive bacterium (Streptococcus pneumoniae R36a) by the M cells of rabbit Peyer's patches. Anat Anz 177:119–124.

31.

Shalaby

(1995) Development of oral vaccines to stimulate mucosal and systemic immunity: barriers and novel strategies. Clin Immunol Immunopathol 74:127–134.

32.

Yamanaka

Straumfors

Morton

Fausa

Brandtzaeg

Farstad

(2001) M cells pockets of human Peyer's patches are specialized extension of germinal centers. Eur J Immunol 31:107–117.