QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) All Versions of this Article: jpet.106.107847v1 319/2/632    most recent Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Gullberg, E. Articles by Artursson, P. Search for Related Content PubMed PubMed Citation Articles by Gullberg, E. Articles by Artursson, P. Pubmed/NCBI databases Compound via MeSH Substance via MeSH

METABOLISM, TRANSPORT, AND PHARMACOGENOMICS

Identification of Cell Adhesion Molecules in the Human Follicle-Associated Epithelium That Improve Nanoparticle Uptake into the Peyer's Patches

Department of Pharmacy, Uppsala University, Uppsala, Sweden (E.G., P.A.); Department of Biomedicine and Surgery, Linköping University, Faculty of Health Sciences, Linköping, Sweden (Å.V.K., S.Y.S., J.D.S.); and Department of Surface Biotechnology, Uppsala University, Uppsala, Sweden (M.A., K.D.C.)

Received May 27, 2006; accepted August 15, 2006.

	Abstract

Top Abstract Materials and Methods Results Discussion References

 
The aim of this study was to identify cell adhesion molecules that could serve as targets of the human follicle-associated epithelium (FAE) overlying Peyer's patches and to assess nanoparticle uptake levels across this epithelium. We first studied the expression of the mouse M-cell marker ₁-integrin and used a model of human FAE derived from intestinal epithelial Caco-2 cells and Raji B-cells to identify additional potential targets by cDNA array. The protein expression of potential targets in the model FAE and in human ileal FAE tissues was quantified by immunofluorescence. Integrin targeting was studied by investigating the transport of Arg-Gly-Asp (RGD)-coated (integrin-binding), Arg-Gly-Glu (RGE)-coated (nonintegrin-binding), and uncoated nanoparticles across ileal specimens mounted in Ussing chambers. Both ₁-integrin and the cell adhesion molecule CD9 were more abundantly expressed in the model and human FAE compared with the Caco-2 control cells or villus epithelium (VE). Uncoated nanoparticles were not taken up across either FAE or VE. General integrin targeting with RGD improved the nanoparticle transport dramatically across the FAE and to a lower extent across the VE. Compared with RGE, RGD improved transport 4-fold across the FAE. There was no difference in the transport of RGD- and RGE-coated nanoparticles across the VE. In conclusion, ₁-integrin and CD9 were identified as targets in human FAE. The difference in RGD- and RGE-mediated transport across the FAE, but not the VE, suggests that a specific integrin interaction was the dominating mechanism for improved nanoparticle uptake across the FAE., whereas charge interaction contributed substantially to the improved VE uptake.

Targeting to specific surface receptors on epithelial cells by use of receptor ligands on the particle surface has been successful in improving uptake of proteins across intestinal epithelial cells in culture (Russell-Jones, 2004). However, in vivo, the mucus layer most probably limits particle uptake across the human villus epithelium (VE). In contrast, antigen uptake across the follicle-associated epithelium (FAE), covering Peyer's patches (PP), is facilitated by the absence of mucus-producing Goblet cells and a reduced expression of hydrolases. A prominent feature of this epithelium is also the presence of specialized M-cells (Owen, 1999) that transport intact protein antigens, as well as particles, viruses, and bacteria across the epithelial cell layer (Neutra, 1999; Owen, 1999). We have recently shown an increased transport of antigens and bacteria across the human FAE compared with the VE (Keita et al., 2006).

Thus, the FAE and M-cells are prime targets when aiming at improved particle uptake and immune responses. Targeting of particles to mouse M-cells by use of various ligands, including Ulex Europaeus agglutinin 1 (Foster et al., 1998) and immunoglobulins (Smith et al., 1995), increased adherence, and transcytosis of particles across M-cells. However, the magnitude of these effects has not been assessed, and the size of the particulate carrier and accessibility of the receptor seem critical for the outcome (Mantis et al., 2000; van der Lubben et al., 2001).

In humans, little is known about surface proteins that could act as specific targets in human FAE. Lectins such as the sambucus nigra antigen seem to predominantly stain polysaccharides in the human FAE (Jepson et al., 1996), but the expression of these sugar moieties varies considerably along the intestine, as well as among individuals (Gebert et al., 1996; Sharma et al., 1996). Two putative human M-cell markers have been reported, cathepsin E (Finzi et al., 1993) and the sialyl Lewis A antigen (Giannasca et al., 1999), but unfortunately, these results could not be reproduced (Wong et al., 2003). mRNA transcripts coding for Claudin-4 and the cell adhesion protein TM4SF3 seem specific to the human FAE, but it is not known if the corresponding proteins are accessible for targeting (Lo et al., 2004).

₁-Integrin, on the other hand, marks the surface of mouse M-cells and mediates uptake of bacteria, alone or in complex with ₅-integrin (Clark et al., 1998; Tyrer et al., 2006). Both we and others have also demonstrated an increased apical expression of ₁-integrin in model FAE cell monolayers (Schulte et al., 2000; des Rieux et al., 2005). Studies in rats have suggested that targeting of particles to ₁-integrin may improve uptake across the intestinal epithelium, although these data were inconclusive regarding site of uptake (Hussain and Florence, 1998). Surprisingly, the expression of ₁-integrin has not yet been reported in human FAE, and the effect of targeting particles to ₁-integrin has not been investigated in human or mouse intestine.

The aim of this study was to identify cell adhesion molecules that could serve as targets in the human FAE. We first selected ₁-integrin and then identified additional molecules unique to the human FAE, using a specific cell adhesion/cell interaction cDNA array. Because it was impossible to isolate sufficient quantities of pure and viable epithelial cells from human FAE, we initially used a modification of an established human coculture model of the FAE, which has several phenotypic features of tissue FAE (Kernéis et al., 1997; Gullberg et al., 2000). The protein expression of ₁-integrin and the other targets identified on the array was assessed in the model and in human ileal FAE. Finally, we explored these findings for improved uptake of nanoparticles across human FAE.

	Materials and Methods

Top Abstract Materials and Methods Results Discussion References

 
Cell Culture. Caco-2 cells (ECACC, Salisbury, UK), clone 1 Caco-2 cells (Dr. Maria Rescigno, University of Milano-Bicocca, Milano, Italy), and the human Burkitt's lymphoma cell line Raji (American Type Culture Collection, Manassas, VA) were maintained as described previously (Gullberg et al., 2000). For both RNA isolation and transport experiments, Caco-2 cells were grown on polycarbonate filters (pore size, 3.0 µm; Costar, Baedvenhorp, The Netherlands), which were coated with Matrigel (BD Biosciences, San Jose, CA). To obtain the model FAE, 5 x 10⁵ Raji cells were added to the basolateral chamber of the Caco-2 cell monolayers at day 14. The cocultures were maintained for 4 to 5 days (Gullberg et al., 2000). Monocultures of Caco-2 cells grown on filters for 18 to 19 days were used as controls. The transport of fluorescence-labeled polystyrene particles (0.5 µm; Invitrogen, Carlsbad, CA) was used to verify the function of each batch of model FAE. Transepithelial electrical resistance (TEER) was used for assessment of cell monolayer integrity (see transport experiments below). The cell monolayers were prepared for immunofluorescence studies by 10-min fixation in 4% buffered formaldehyde and subsequent rinsing with Hanks' balanced salt solution (HBSS with calcium, magnesium, and sodium bicarbonate; Invitrogen). They were stored in HBSS until use.

RNA Isolation, Array Hybridization, and Gene Expression Analysis. For RNA isolation, the cell monolayers were rinsed several times with ice-cold phosphate-buffered saline (PBS; 0.01 M NaPO₄, 0.14 M NaCl, 0.003 M KCl) and scraped off the filters into RNase-inhibiting denaturing Solution (Ambion, Austin, TX). Total RNA was isolated with Ambion's Totally RNA kit according to the manufacturer's instructions. All RNA samples were DNase-treated, and control PCR was run on all RNA samples to confirm that they were free from genomic DNA. RNA samples from three pooled model FAE and Caco-2 cell batches (six filters in total) were reversibly transcribed to [³²P]cDNA and hybridized to cell-adhesion/cell-interaction arrays (BD Biosciences Clontech, Palo Alto, CA), according to the manufacturer's instructions. The primer mix used for reverse transcription only amplified transcripts from genes present on the array.

Differences in gene expression were analyzed by Atlas Image (BD Biosciences Clontech). Signals lower than the average background signal were filtered out. To compare two or more arrays, the signal intensities of all genes on the arrays were normalized according to the manufacturer's instructions.

Isolation of Peyer's Patches from Human Ileum and Preparation of Cryosections. The study was approved by the local committee of human ethics in Linköping, and all subjects gave their informed consent. Specimens were obtained from the ileum next to the ileocaecal valve during ileocolonic surgery of eight patients diagnosed with colon cancer at the University Hospital of Linköping. The median age of the patients (five men and three women) was 67 years (range, 35–87). Immediately following dissection, the specimens were placed in ice-cold oxygenated Krebs-Ringer buffer (115 mM NaCl, 25 mM NaHCO₃, 2 mM K₂HPO₄, 1.25 mM CaCl₂, 1.2 mM MgCl₂), pH 7.4, and taken to the laboratory for the experiments. A pathologist examined the specimens, which all showed macroscopically normal histology. While still immersed in ice-cold buffer, the external muscle and myenteric plexus were stripped from the surgical specimens. After identification in a dissection microscope, transilluminated from below (Keita et al., 2006), regions of VE and FAE were cut out, oriented to give cross-sections of the intestinal wall, snap-frozen in liquid nitrogen, and mounted in Cryomount. Cryosections (5 µm) were cut using a microtome (Leica, Wetzlar, Germany) and dried at room temperature on glass slides overnight, before storage at –70°C. Sections were prepared for immunofluorescence by fixing in ice-cold acetone for 10 min, followed by drying at room temperature.

Immunostaining of Fixed Cell Cultures and Cryosections. All cryosections and formaldehyde-fixed cell cultures were microwaved for 10 min in 10 mM citrate, pH 6 (₁-integrin), or 1 mM EDTA, pH 8 (CD9), as recommended by the manufacturers of the antibodies. After cooling to room temperature, the sections/cells were rinsed several times with distilled water, washed in HBSS, and permeated with 0.2% Triton X-100 for 5 min (₁-integrin) or with Cryofix (Merck, Darmstadt, Germany) for 2 min (CD9). After blocking with HBSS containing 1% fetal calf serum, the sections/cells were stained with primary antibodies to ₁-integrin (clone J10, 400 mg/ml; Biogenex, San Ramon, CA) or CD9 (clone 72F6, 1:100; Serotec, Oxford, UK). The primary antibodies were detected with secondary goat-anti-mouse IgG1 antibodies conjugated to Alexa 488 (Invitrogen). Antibodies to CD20 (clone L26, 1:100; Biomeda, Foster City, CA) were used to visualize CD20-positive B-cells in lymphoid follicles and M-cells (Yamanaka et al., 2001). CD20 was detected using Alexa568 goat-anti-mouse IgG2a (Invitrogen). Cells/sections incubated with isotype-matched antibodies were used as controls.

Immunofluorescence Analysis of Cell Cultures and Human Tissue. The immunofluorescence of ₁-integrin and CD9 in cell cultures was analyzed by optical sectioning in a confocal laser scanning microscope (Leica), using the same settings for the model FAE and Caco-2 epithelium. Fluorescence intensity profiles were made in ImageJ software (http://rsb.info.nih.gov/ij). The human tissue sections were analyzed using an Axioplan fluorescence microscope (Carl Zeiss GmbH, Jena, Germany). Pictures were captured with a CCD camera (Hamamatsu C4542–95B; Hamamatsu Corporation, Bridgewater, NJ), using the same gain and exposure times for all sections on the same slide. Relative fluorescence intensities of the epithelium were measured using ImageJ software. Two slides were prepared on different occasions for each protein and each patient. Four to eight sections, representing two to three different cell layers, were analyzed per patient. In each tissue section, all types of epithelia, FAE, VE, and villus epithelium adjacent to dome (VED) were represented, and their fluorescence intensities relative to each other was determined within the same section. Ten measurements of cell membrane fluorescence intensities were performed on equal areas of the different type of epithelia in each section. Background values obtained from control sections on each slide were subtracted from the true measured fluorescence intensities found on the same slide. The data were normalized to compare protein expression among the patients. The fluorescence intensity for each sample from a particular patient was divided by the sum of the fluorescence intensities of the total tissue sections for that patient. By two-way ANOVA, we analyzed the difference in fluorescence intensities between each type of epithelia from the same patient, and in each type of epithelia between different patients. The tissue obtained from two patients lacked villus epithelium. For these patients, only the fluorescence intensities of FAE and VED are reported.

Preparation of RGD- and RGE-Coated Polystyrene Nanoparticles. Yellow-green fluorescent polystyrene particles (average diameter, 200 nm) were purchased from Polysciences Inc. (Warrington, PA). Pluronic F108 modified with a thiol-specific pyridyl disulfide group (Cell Link) was kindly donated by Allvio Inc. (Lake Forest, CA). Two polypeptides containing the RGD motif (Cys-Gly-Arg-Gly-Asp-Ser-Tyr) and the RGE motif (Cys-Gly-Arg-Gly-Glu-Ser-Tyr) were synthesized at the Department of Medicinal Chemistry, Uppsala University. The Pluronic surfactant was adsorbed onto the particles by a hydrophobic interaction, and the peptide was subsequently attached to the hydrophilic parts of the linker using the cysteine residue. The reaction was confirmed by measuring the UV absorbance at 343 nm from released pyridyl disulfide groups. A Shimadzu UV-2102 PC UV-visible scanning spectrophotometer (Shimadzu, Kyoto, Japan), was used and the extinction coefficient at the wavelength was 8080 M^–1 cm^–1. The preparation, purification, and characterization procedures are described in detail elsewhere (Andersson et al., 2005).

Transport of Nanoparticles across the Model FAE and Caco-2 Epithelium. Transport experiments were performed in HBSS, pH 7.4. After washing and equilibration in HBSS/fetal calf serum at 37°C for 30 min, the TEER of the cell monolayers was measured. Cell monolayers with a TEER above 175 /cm² were included in the experiments. Yellow-green fluorescent polystyrene particles (2.5 x 10⁹) coated with RGD or RGE were added to the apical side. For the inhibition experiments, the cell monolayers were preincubated apically with 0.5 mM RGD peptide (H-Arg-Gly-Asp-OH; Bachem, Bubendorf, Switzerland) for 1 h at room temperature before adding the particles. Samples were collected after 1 h and analyzed in a fluorescence-activated cell scan (BD Biosciences). Analysis of transport data were performed using one-way ANOVA and Student's t tests.

Uptake and Transport of RGD-Coated Particles across Human FAE and VE. Sections of VE and FAE, identified microscopically as described above, were cut to appropriate sizes and mounted in modified Ussing chambers (Harvard Apparatus Inc., Holliston, MA). FAE segments were carefully adjusted so that the patches covered the entire exposed tissue surface area (4.9 mm²). The transport experiments were performed in Krebs Ringer buffer (see above for composition), pH 7.4, 37°C, with the addition of 10 mM mannitol on the mucosal side and 10 mM glucose on the serosal side of the segments. The segments were continuously oxygenated by circulating 95% O₂/5% CO₂ gas and equilibrated for 40 min to achieve steady-state conditions. The integrity and viability was continuously monitored by measuring the TEER, short-circuit current, and resting potential of the segments (Keita et al., 2006). Only segments that stabilized with a resting potential of less than –0.5 mV were included in the experiments. During inhibition experiments, 0.5 mM RGD peptide (see above) was added to the mucosal side of the specimens 1 h before adding the particles. Approximately 2.5 x 10⁹ particles/ml, with or without RGD and RGE coating, were added to the mucosal side of the tissue (particles without peptide carried only adsorbed Pluronic linker). After 2 h, the serosal buffer was withdrawn, and the number of particles was measured by fluorescence-activated cell scan. Analysis of transport data were performed using one-way ANOVA and Student's t tests. For qualitative assessment of particle uptake, the tissues were washed with PBS (composition, see above) to remove excess particles after 15 and 45 min and the specimens were fixed with 4% formaldehyde overnight. Subsequently, the tissues were rinsed with PBS, labeled with 22 nM phalloidin-Alexa Fluor 594 (Invitrogen) for 30 min, washed repeatedly with PBS, and the whole tissue was visualized in PBS-containing chambers under a Nikon Eclipse E600W confocal laser scanning microscope (Nikon, Melville, NY).

	Results

Top Abstract Materials and Methods Results Discussion References

 
Increased Expression of ₁-Integrin in Human FAE. Immunofluorescent staining of ₁-integrin in human Peyer's patches revealed increased expression of ₁-integrin in human FAE compared with the surrounding villi. In the FAE, ₁-integrin was strongly expressed all around the basolateral face of the epithelial cells (Fig. 1, A and B). ₁-Integrin expression could also be found on the apical surface of the FAE, typically where CD20-positive cells infiltrated the epithelium (Fig. 1, D and F) but occasionally also in epithelial cells not associated with CD20-positive lymphocytes (Fig. 1E). In contrast, ₁-integrin predominantly lined the basal part of the epithelial cells of the surrounding villi and was strongly expressed on cells in the lamina propria. This expression pattern compares well with previous studies on ₁-integrin expression in human intestinal villi (Lussier et al., 2000). Thus, it seemed that ₁-integrin had an increased accessibility in human FAE and was indeed a promising target. The next step was to identify additional targets by cDNA array using our previously established model FAE.

View larger version (31K): [in this window] [in a new window]   Fig. 1. Expression of 1-integrin in human Peyer's patches. Where appropriate, stitched lines outline the surface of the epithelium. Boxes in A and B show areas of magnification for C and D. In the FAE, 1-integrin was expressed in the basolateral membrane (A) or in the entire plasma membrane of the epithelial cells (B). In the surrounding villi (VED/VE), it was predominantly expressed on the basal aspect of the epithelial cells and in the lamina propria. C and D, magnification of the FAE showed infiltration of CD20-positive lymphocytes into the epithelium (arrows). D and F, at several sites, prominent apical expression of 1-integrin was found in these lymphocyte-associated epithelial cells (arrows). Apical expression of 1-integrin was occasionally also found in FAE cells that were not associated with CD20-positive lymphocytes (E) but most commonly (F) in FAE cells that were lymphocyte-associated (arrows).

View larger version (31K): [in this window] [in a new window]   Fig. 2. Genes with changed expression in the model FAE compared with the Caco-2 epithelium. Gene names correspond to NCBI nomenclature and have been grouped according to reported functions. Shades of gray represent the ratio of gene expression in the model FAE compared with the Caco-2 epithelium.

Assessing the protein expression of all of the genes identified by the array was beyond the scope of this study. The cell adhesion molecules CD9 and CD4 were chosen for further investigation since they exhibited the most pronounced up-regulation in expression and since both of these proteins have been reported to act as coreceptors for viruses and bacteria. CD9 is a tetraspan protein that interacts with integrins in the cell membrane (Berditchevski et al., 1999), and CD4 is a well known coreceptor for human immunodeficiency virus (Dalgleish et al., 1984). Therefore, we compared the expression of these two proteins with the expression of ₁-integrin in the model FAE.

Enhanced Protein Expression of CD9 in the Model FAE Compared with the Caco-2 Epithelium. The protein expression of CD9 was clearly enhanced in the model FAE compared with the Caco-2 epithelium (Fig. 3, A and B). In both epithelia, CD9 was expressed laterally; however, expression in the basal and apical cell membranes was most pronounced in the model FAE. We were unable to find protein expression of CD4 in any of the two epithelia (Fig. 3, C and D). Reverse transcriptase-PCR studies revealed that there were no difference in the number of CD4 transcripts between the model FAE and Caco-2 control epithelium (data not shown). Thus, the increase in CD4 gene expression was falsely detected by the cDNA array. In contrast, the increase in CD9 expression was confirmed by reverse transcriptase-PCR, with a ratio of 1.5 CD9 mRNA transcripts in the model FAE compared with the Caco-2 epithelium.

View larger version (48K): [in this window] [in a new window]   Fig. 3. Immunofluorescence of CD9, CD4 and 1-integrin in the model FAE (A, C, and E) and Caco-2 epithelium (B, D, and F). Optical xz section of cell monolayers with the apical surface of the cells oriented upwards. Each image is representative of three independent experiments. Densitometric graphs of the corresponding fluorescence intensities are shown on the right side of each image. The expression of CD9 was higher in model FAE (A) than in the Caco-2 epithelium (B). CD4 expression was not detected in the model FAE (C) or the Caco-2 epithelium (D). The expression of 1-integrin was higher in the model FAE (E) than in the Caco-2 epithelium (F) and found mainly on the apical cell surface of the model FAE. Bars, 10 µm.

The Expression of Both ₁-Integrin and CD9 Is Higher in Human FAE Than in Villus Epithelium. Qualitative assessment of CD9 expression in human Peyer's patches revealed that the FAE was indeed positive for CD9, whereas the VE was negative (Fig. 4A). In addition, CD9 was expressed by cells in the lamina propria in all specimens and could also be found in the lymphoid follicles. The expression of CD9 in VE is in agreement with a previous study of CD9 expression in the human small intestine (Sincock et al., 1997). In concordance with the results obtained in cell cultures, CD9 expression was found in both the apical and basolateral membranes of the epithelial cells in the FAE (Fig. 4B).

View larger version (33K): [in this window] [in a new window]   Fig. 4. Expression of CD9 in human Peyer's patches. Bars, 100 and 10 µm, respectively. A, CD9 was predominantly expressed in the FAE (arrows), whereas the VED showed little expression of CD9 (arrowhead). No or low expression of CD9 was found in the VE (arrowhead). However, CD9 expression was found on cells in both the lamina propria and the subepithelial dome region (arrowheads, magnification 16x). B, higher magnification (63x) of the FAE showed that CD9 was expressed in the entire plasma membrane of the epithelial cells.

Semiquantitative assessment of the expression levels of ₁-integrin and CD9 in the eight individuals studied confirmed our qualitative findings (Fig. 5). The expression levels of both proteins were 2- to 3-fold higher in the FAE than in the VE. In several of the specimens obtained, the villi adjacent to the dome (Fig. 5A, VED) also expressed CD9, whereas villi situated further away from the dome (VE) displayed no or low expression (Fig. 5B). In contrast, there was no difference in ₁-integrin expression between the VED and VE (Fig. 5C).

View larger version (25K): [in this window] [in a new window]   Fig. 5. Quantification of CD9 and 1-integrin expression in human Peyer's patches from eight individuals. A, section of ileal tissue demonstrating the three types of epithelia that were studied: FAE, VED, and VE. Four to eight sections were studied for each individual, and each data point represents the mean fluorescence intensity of CD9 (B) or 1-integrin (C) in one individual. For details, see Materials and Methods. Lines cross at the average fluorescence intensity of all patients, and bars show the 95% confidence intervals. ***, p < 0.0001; **, p < 0.01; *, p < 0.05; n.s., nonsignificant, p > 0.05

Increased Transport of Integrin-Adherent Particles across the Model and Human FAE. Our expression data suggested that our initial proposal to target ₁-integrin would indeed be a feasible strategy to improve uptake into human Peyer's patches. In addition to ₁-integrin, we identified an up-regulation of the ₅- and ₈-integrin subunits on the mRNA level in the model FAE (Fig. 2). Therefore, we chose a general integrin-targeting sequence to maximize adherence and the uptake and transport of nanoparticles. The RGD peptide motif (Arg-Gly-Asp) is known to compete with Yersinia invasin A in binding to ₁-integrin (Leong et al., 1995) but also to recognize many other integrins, including the ₅- and ₈-integrin subunits (Ruoslahti, 1996). CD9 also seemed promising, but to date, no ligands have been reported for human CD9. It was judged beyond the scope of this study to identify such molecules; thus, we decided to study the effects of targeting integrins.

Optical sectioning of whole-tissue specimens in a confocal microscope showed uptake of RGD-coated particles across the FAE as early as 15 min after application (Fig. 6A). After 45 min, the RGD-coated particles had crossed the FAE and were found deep inside the dome (Fig. 6B), whereas particles without peptide were not taken up at all (Fig. 6C). In the corresponding villus specimens, particles were mainly found associated with the luminal cell surface and inside epithelial cells at 15 min (Fig. 6D). At 45 min, we could find RGD particles also below the VE (Fig. 6E), whereas particles without peptide were not taken up (Fig. 6F). Uptake of both submicron-sized particles and aggregates of several microns were detected across both the FAE and VE. In addition, a large number of diffusely fluorescent, cell-shaped bodies were found beneath the epithelium, suggesting that particles had been taken up by macrophages or dendritic cells (Fig. 6, G and H). Thus, it was quite difficult to assess the magnitude of uptake with certainty. However, we could not at any time point detect uptake of RGD-negative particles in either FAE or villus tissue (Fig. 6, C and F), whereas at 45 min, RGD-coated particles were found inside all of the subepithelial domes of the PP and the villi studied (Fig. 6, B and E).

View larger version (25K): [in this window] [in a new window]   Fig. 6. Uptake of 200 nm RGD-coated (+RGD) and noncoated (–RGD) polystyrene particles in Peyer's patches and villus tissue. The images show optical xz sections of PP and villus tissue 15 and 45 min after addition of particles to the mucosa. Each time point is illustrated by a representative fluorescent image (magnification, 100x). F-actin staining of nonpermeabilized tissues was used to visualize the epithelium (red). The epithelial cells display green autofluorescence, whereas particles are seen in yellow, single or as aggregates. The scale bar was used to assess the position of particles within the tissue (the epithelial cell height was on average 20 µm). A, in the FAE, the RGD-coated particles where found below the epithelial cell layer already after 15 min. After 45 min, many of the particles had crossed the FAE and were found in the subepithelial layer (B), whereas noncoated particles had not been taken up at all (C). In the VE, the RGD-coated particles were mainly associated with the epithelium after 15 min (D), but after 45 min (E), particles could also be detected beneath the VE. F, noncoated particles were not taken up at all across the VE. G and H, diffusely fluorescent bodies beneath both FAE and VE probably represent cells that have taken up particles (large arrows). Bars, 10 µm.

To assess the specificity of the RGD-mediated uptake, we performed transport experiments using particles coated with either RGD or RGE (Arg-Gly-Glu). In RGE, the aspartic acid, which is most critical for integrin binding, has been substituted with glutamic acid, which confers similar charge, but abolishes binding (Leong et al., 1995).

In our model FAE, the number of transported RGD-coated particles was 20-fold higher than particles coated with RGE (Fig. 7, p < 0.01). There was no significant increase in transport of RGD-coated particles across the Caco-2 epithelium compared with RGE-coated particles. Preincubation of model FAE cell monolayers with soluble RGD (sRGD) significantly reduced the transport of RGD-coated particles.

View larger version (24K): [in this window] [in a new window]   Fig. 7. Transport of 200 nm RGD-coated (integrin-binding) and RGE-coated (nonintegrin-binding) polystyrene particles across model FAE and Caco-2 epithelium. Both model FAE and Caco-2 epithelium were preincubated with HBSS or sRGD, and the transported fluorescent particles were measured by flow cytometry. Each box shows median, interquartile range (25–75), and S.D. of 8 to 24 filters, run on three to six separate occasions. **, p < 0.01 for all groups compared with the transport of RGD-coated particles across the model FAE.

View larger version (22K): [in this window] [in a new window]   Fig. 8. Transport of 200-nm RGD- and RGE-coated polystyrene particles across human Peyer's patches and villus tissue mounted in Ussing chambers. Both PP and VE tissues were preincubated with Krebs buffer or sRGD. The number of particles found in the serosal buffer after 2 h was assessed by flow cytometry and is presented as the fraction of particles transported. This accounts for the difference in the effective number of particles added to the mucosal chamber. Each box shows median, interquartile range (25–75), and S.D. of three to five separate measurements in two individuals. **, p < 0.01 or less for all groups compared with the transport of RGD-coated particles across the FAE.

	Discussion

Top Abstract Materials and Methods Results Discussion References

It is known that various integrins mediate, and potentially regulate, uptake of particulate antigens in cell culture and animal studies (Kerr, 1999; Schulte et al., 2000). An earlier study showed that ₁-integrin expressed apically on M-cells was responsible for the adherence and uptake of Yersinia bacteria across the mouse FAE (Clark et al., 1998). Recently, Tyrer et al. (2006) confirmed and extended this finding by showing that ₁-integrin associates with ₅-integrin in mediating integrin-dependent uptake of bacteria across mouse M-cells. In our study, ₁-integrin stained the apical membranes of dome epithelial cells associated with CD20-positive lymphocytes. Thus, the transport of RGD-coated particles across the FAE was possibly mediated by M-cells. However, apical expression of ₁-integrin was occasionally also found in dome epithelial cells not associated with CD20-positive lymphocytes, and it cannot be excluded that integrins expressed on other FAE cells contributed to the increased uptake. The presence of integrins in human FAE has not been reported previously, but it is known that their expression is tightly regulated (Lussier et al., 2000).

The tetraspan proteins, such as CD9, are known to modulate both the location and the conformation of many proteins, including integrins, through binding and interactions within so-called tetraspanin webs in the cell membrane (Berditchevski and Odintsova, 1999). Specifically, earlier studies have demonstrated an association between CD9 and ₁-integrin precursors in epithelial cells (Rubinstein et al., 1997), suggesting that CD9 is involved in maturation and delivery of ₁-integrins to the cell membrane. CD9 has also been shown to play a role in protein kinase C-mediated integrin signaling (Zhang et al., 2001) and focal adhesion kinase-induced reorganization of the cortical actin cytoskeleton (Berditchevski and Odintsova, 1999). Both of these events are critical for integrin-mediated uptake of particulate antigens (Dramsi and Cossart, 1998). Interestingly, a previous study also identified an mRNA transcript of a tetraspan protein (TM4SF3) with specific expression in the human FAE (Lo et al., 2004). Thus, CD9 and possibly other tetraspan proteins may play a role in the regulation of integrin-mediated uptake.

Regardless of the tetraspan interaction, the relationship between integrin expression levels and functionality was evident in this study. The median difference and variability in RGD-mediated nanoparticle transport between the FAE and the VE (Fig. 8) compares well with the mean difference and variability in ₁-integrin expression between the FAE and the VE (Fig. 5). Our transport data also clearly demonstrate the functional involvement of integrins in the particle uptake. There was a 4-fold increase in transport between RGD- and RGE-coated particles across the FAE, which had an increased expression of ₁-integrins. However, no significant difference was found between RGD- and RGE-coated particles in VE, where we found no ₁-integrin expression. The involvement of other coexpressed RGD-binding integrins in the uptake cannot be excluded. Their expression in human FAE is not known.

Because RGD and RGE induced nanoparticle transport across the VE to the same extent, it was conceivable that a mechanism other than integrin interaction was of importance for this uptake. Because hydrophobic latex particles coated with only the hydrophilic Pluronic linker were not taken up at all (Fig. 6), we could exclude the involvement of an unspecific interaction with the latex particle surface. Furthermore, we could not find any difference in particle size or aggregation tendency between Pluronic-coated and peptide-Pluronic-coated particles during our initial characterization studies (Andersson et al., 2005), which excluded a size effect. Thus, the most likely explanation is a charge interaction between the peptide-coated particles and the VE cells or mucus. Both RGD and RGE contain amino acids that confer both positive (arginine) and negative (aspartic and glutamic acid) charge at pH 6 to 7.5, whereas the Pluronic linker is uncharged in this pH range. In addition, free RGD peptide inhibited both RGD- and RGE-mediated uptake. In the FAE, however, the transport of RGD-coated particles was much higher than the transport of RGE-coated particles. Judging from these data, the contribution of charge to the improved transport across the FAE was less than 25%, indicating that the specific interaction with integrins was of major importance (Fig. 8).

Our data certainly points to the necessity for a specific interaction when aiming at improved uptake into Peyer's patches. Despite the improvement in nanoparticle transport seen with both RGD and RGE, we did not find a significant difference in the uptake of RGE-coated particles across the FAE compared with the VE. This further supports that the FAE mediated an increased particle uptake by the specific expression of integrins and not by unspecific mechanisms, such as an increased accessibility or a higher transport capacity. The difference in CD9 expression levels between FAE and VE were even more prominent than for ₁-integrin, and it would be exciting to investigate the selectivity of a CD9 targeting sequence for the human FAE, once a ligand has been discovered.

A possible application of our findings would be to improve the uptake of subunit oral vaccine formulations. The crucial question is, therefore, if the uptake of particles, even after targeting, will be sufficient to initiate an immune response. Two hours after administration of RGD-coated particles, 1% of them had crossed the entire FAE and underlying tissue (Fig. 8). We are not aware of any previous studies of intestinal particle transport across human tissues using the Ussing chamber technique; thus, comparisons are difficult. However, our data can possibly be qualitatively compared with animal studies where particles have been instilled in ligated intestinal loops. From our confocal microscopy studies, we approximate that at least 50 RGD-coated particles of sizes up to 5 µm were found in each dome covered by FAE after 45 min. During experiments with ligated loops in the mouse intestine, using cross-linked starch particles, only one to three particles (2 µm in size) were taken up in 4 of the 10 studied domes after 45 min (Stertman, 2004). Still, a lower dose of starch particles than the one used in the loop has been shown to initiate a prominent immune response after oral administration (Strindelius et al., 2002). The RGD-mediated uptake of 50 or more particles in our study was at least 20 times higher, but whether it will be sufficient to initiate an immune response after oral administration remains to be seen.

In conclusion, by using our recently introduced technique to obtain human FAE (Keita et al., 2006), we have been able to investigate the transport of nanoparticles into human Peyer's patches and assess the relative contribution of FAE and VE to nanoparticle uptake in the human intestine. We show that uptake of nanoparticles is a rare event, also in human intestine, but that the uptake into Peyer's patches may be dramatically improved by introducing an integrin targeting moiety on the particles. Apart from integrins, the integrin-associated cell adhesion molecule CD9 has an increased expression in human FAE and also holds as a potential target, once a targeting ligand has been identified. These findings provide a platform for improving the intestinal uptake and effects of oral vaccine formulations in humans.

	Acknowledgements

 
The skillful technical assistance of Eva Sjödahl and Åsa Schoug is gratefully acknowledged. We also apologize to and acknowledge all authors whose work was not cited here due to space limitations.

	Footnotes

 
This work was supported by the Swedish National Association of Pharmaceutical Scientists, by the Swedish Research Council (Vetenskapsradet-medicin), by the Broad Medical Research Program of the Eli and Edythe L. Broad Foundation, and by AstraZeneca R&D Mölndal, Sweden

Article, publication date, and citation information can be found at http://jpet.aspetjournals.org.

doi:10.1124/jpet.106.107847.

ABBREVIATIONS: VE, villus epithelium; FAE, follicle-associated epithelium; PP, Peyer's patches; TEER, transepithelial electrical resistance; HBSS, Hanks' balanced salt solution; RGD, Arg-Gly-Asp tripeptide; RGE, Arg-Glu-Glu tripeptide; PBS, phosphate-buffered saline; PCR, polymerase chain reaction; VED, villus epithelium adjacent to dome; ANOVA, analysis of variance; sRGD, soluble RGD.

Address correspondence to: Per Artursson, Department of Pharmacy, Box 580, BMC, SE-751 23 Uppsala, Sweden. E-mail: Per.Artursson{at}farmaci.uu.se

	References

Top Abstract Materials and Methods Results Discussion References

 

Andersson M, Fromell K, Gullberg E, Artursson P, and Caldwell KD (2005) Characterization of surface-modified nanoparticles for in vivo biointeraction: a sedimentation field flow fractionation study. Anal Chem 77: 5488–5493.[Medline]

Berditchevski F and Odintsova E (1999) Characterization of integrin-tetraspanin adhesion complexes: role of tetraspanins in integrin signaling. J Cell Biol 146: 477–492.[Abstract/Free Full Text]

Brayden DJ (2001) Oral vaccination in man using antigens in particles: current status. Eur J Pharm Sci 14: 183–189.[CrossRef][Medline]

Clark MA, Hirst BH, and Jepson MA (1998) M-cell surface beta1 integrin expression and invasin-mediated targeting of Yersinia pseudotuberculosis to mouse Peyer's patch M cells. Infect Immun 66: 1237–1243.[Abstract/Free Full Text]

Dalgleish AG, Beverley PC, Clapham PR Crawford DH, Greaves MF, and Weiss RA (1984) The CD4 (T4) antigen is an essential component of the receptor for the AIDS retrovirus. Nature (Lond) 312: 763–767.[CrossRef][Medline]

des Rieux A, Ragnarsson EGE, Gullberg E, Preat V, Schneider YJ, Artursson P (2005) Transport of nanoparticles across an in vitro model of the human intestinal follicle associated epithelium. Eur J Pharm Sci 25: 455–465.[CrossRef][Medline]

Dramsi S and Cossart P (1998) Intracellular pathogens and the actin cytoskeleton. Annu Rev Cell Dev Biol 14: 137–166.[CrossRef][Medline]

Finzi G, Cornaggia M, Capella C, Fiocca R, Bosi F, Solcia E, and Samloff IM (1993) Cathepsin E in follicle associated epithelium of intestine and tonsils: localization to M cells and possible role in antigen processing. Histochemistry 99: 201–211.[CrossRef][Medline]

Foster N, Clark MA, Jepson MA, and Hirst B (1998) Ulex europaeus 1 lectin targets microspheres to mouse Peyer's patch M-cells in vivo. Vaccine 16: 536–541.[CrossRef][Medline]

Gebert A, Rothkötter H-J, and Pabst R (1996) M cells in Peyer's patches of the intestine. Int Rev Cytol 167: 91–159.[Medline]

Giannasca P, Giannasca K, Leichtner AM, and Neutra MR (1999) Human intestinal M cells display the Sialyl Lewis A antigen. Infect Immun 67: 946–953.[Abstract/Free Full Text]

Gullberg E, Leonard M, Karlsson J, Hopkins AM, Brayden D, Baird AW, and Artursson P (2000) Expression of specific markers and particle transport in a new human intestinal M-cell model. Biochem Biophys Res Commun 279: 808–813.[CrossRef][Medline]

Hamzaoui N, Kerneís S, Caliot E, and Pringault E (2004) Expression and distribution of beta 1 integrins in in vitro-induced M cells: implications for Yersinia adhesion to Peyer's patch epithelium. Cell Microbiol 6: 817–828.[CrossRef][Medline]

Hussain N and Florence AT (1998) Utilizing bacterial mechanisms of epithelial cell entry: invasin-induced oral uptake of latex nanoparticles. Pharm Res (N Y) 15: 153–156.

Jepson MA, Clark MA, Foster N, Mason CM, Bennett MK, Simmons NL, and Hirst BH (1996) Targeting to intestinal M-cells. J Anat 189: 507–516.[Medline]

Keita Å, Gullberg E, Ericson A-C, Salim SY, Wallon C, Kald A, Artursson P, and Söderholm JD (2006) Characterization of antigen and bacterial transport in the follicle-associated epithelium of human ileum. Lab Investig 86: 504–516.[CrossRef][Medline]

Kernéis S, Bogdanova A, Kraehenbuhl J-P, and Pringault E (1997) Conversion by Peyer's patch lymphocytes of human enterocytes into M cells that transport bacteria. Science (Wash DC) 277: 949–952.[Abstract/Free Full Text]

Kerr JR (1999) Cell adhesion molecules in the pathogenesis of and host defence against microbial infection. Mol Pathol 52: 220–230.[Abstract]

Leong JM, Morrissey PE, Marra A, and Isberg RR (1995) An aspartate residue of the Yersinia pseudotuberculosis invasin protein that is critical for integrin binding. EMBO (Eur Mol Biol Organ) J 14: 422–431.[Medline]

Lo D, Tynan W, Dickerson J, Scharf M, Cooper J, Byrne D, Brayden D, Higgins L, Evans C, and O'Mahony DJ (2004) Cell culture modeling of specialized tissue: identification of genes expressed specifically by follicle-associated epithelium of Peyer's patch by expression profiling of Caco-2/Raji co-cultures. Int Immunol 16: 91–99.[Abstract/Free Full Text]

Lussier C, Basora N, Bouatrouss Y, and Beaulieu JF (2000) Integrins as mediators of epithelial-cell matrix interactions in the human small intestinal mucosa. Microsc Res Tech 51: 169–178.[CrossRef][Medline]

Mantis N, Frey A, and Neutra M (2000) Accessibility of glycolipid and oligosaccharide epitopes on rabbit villus and follicle-associated epithelium. Am J Phys 278: G915–G923.[Abstract/Free Full Text]

Mutwiri G, Bowersock TL, and Babiuk LA (2005) Microparticles for oral delivery of vaccines. Exp Opin Drug Deliv 2: 791–806.[CrossRef]

Neutra MR (1999) M cells in antigen sampling in mucosal tissues. Curr Top Microbiol Immunol 236: 17–32.[Medline]

O'Hagan DT and Rappuoli R (2006) Novel approaches to pediatric vaccine delivery. Adv Drug Deliv Rev 58: 29–51.[CrossRef][Medline]

Owen RL (1999) Uptake and transport of intestinal macromolecules and microorganisms by M cells in Peyer's patches–a personal and historical perspective. Semin Immunol 11: 157–163.[CrossRef][Medline]

Rubinstein E, Poindessous-Jazat V, Le Naour F, Billard M, and Boucheix C (1997) CD9, but not other tetraspans, associates with the beta1 integrin precursor. Eur J Immunol 27: 1919–1927.[Medline]

Ruoslahti E (1996) RGD and other recognition sequences for integrins. Annu Rev Cell Dev Biol 12: 697–715.[CrossRef][Medline]

Russell-Jones GJ (2004) Use of targeting agents to increase uptake and localization of drugs to the intestinal epithelium. J Drug Target 12: 113–123.[CrossRef][Medline]

Schulte R, Kerneis S, Klinke S, Bartels H, Preger S, Kraehenbuhl JP, Pringault E, and Autenrieth IB (2000) Translocation of Yersinia entrocolitica across reconstituted intestinal epithelial monolayers is triggered by Yersinia invasin binding to beta1 integrins apically expressed on M-like cells. Cell Microbiol 2: 173–185.[CrossRef][Medline]

Sharma R, van Damme EJM, Peumans WJ, Sarsfield P, and Schumacher U (1996) Lectin binding reveals divergent carbohydrate expression in human and mouse Peyer's patches. Histochem Cell Biol 105: 459–465.[CrossRef][Medline]

Sincock PM, Mayrhofer G, and Ashman LK (1997) Localization of the transmembrane 4 superfamily (TM4SF) member PETA-3 (CD151) in normal human tissues: comparison with CD9, CD63, and alpha5beta1 integrin. J Histochem Cytochem 45: 515–525.[Abstract/Free Full Text]

Smith MW, Thomas NW, Jenkins PG, Miller NG, Cremaschi D, and Porta C (1995) Selective transport of microparticles across Peyer's patch follicle-associated M cells from mice and rats. Exp Physiol 80: 735–743.[Abstract]

Stertman L (2004) Starch Microparticles as an Oral Vaccine Adjuvant with Emphasis on the Differentiation of the Immune Response, Acta Universitatis Upsaliensis, Uppsala, Sweden.

Strindelius L, Degling-Wikingsson L, and Sjoholm I (2002) Extracellular antigens from Salmonella enteritidis induce effective immune response in mice after oral vaccination. Infect Immun 70: 1434–1442.[Abstract/Free Full Text]

Tyrer P, Foxwell AR, Cripps AW, Apicella MA, and Kyd JM (2006) Microbial pattern recognition receptors mediate M-cell uptake of a Gram-negative bacterium. Infect Immun 74: 625–631.[Abstract/Free Full Text]

van der Lubben IM, Verhoef JC, van Aelst AC, Borchard G, and Junginger HE (2001) Chitosan microparticles for oral vaccination: preparation, characterization and preliminary in vivo uptake studies in murine Peyer's patches. Biomaterials 22: 687–694.[CrossRef][Medline]

Wong NA, Herriot M, and Rae F (2003) An immunohistochemical study and review of potential markers of human intestinal M cells. Eur J Histochem 47: 143–150.[Medline]

Yamanaka T, Straumfors A, Morton H, Fausa O, Brandtzaeg P, and Farstad I (2001) M cell pockets of human Peyer's patches are specialized extensions of germinal centers. Eur J Immunol 31: 107–117.[CrossRef][Medline]

Zhang XA, Bontrager AL, and Hemler ME (2001) Transmembrane-4 superfamily proteins associate with activated protein kinase C (PKC) and link PKC to specific beta(1) integrins. J Biol Chem 276: 25005–15013.[Abstract/Free Full Text]

This Article Abstract Full Text (PDF) All Versions of this Article: jpet.106.107847v1 319/2/632    most recent Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Gullberg, E. Articles by Artursson, P. Search for Related Content PubMed PubMed Citation Articles by Gullberg, E. Articles by Artursson, P. Pubmed/NCBI databases Compound via MeSH Substance via MeSH