Toll-like receptor 2-mediated peptidoglycan uptake by immature intestinal epithelial cells from apical side and exosome-associated transcellular transcytosis

Abstract

Peptidoglycan (PGN) is a potent immune adjuvant derived from bacterial cell walls. Previous investigations suggest that intestinal epithelium may absorb PGN from the lumen. Nonetheless, how PGN is taken up and crosses intestinal epithelium remains largely unclear. Here, we first characterized PGN transport in vitro using IEC-18 and HT29-CL19A cells, which represent less mature epithelial cells in intestinal crypts. With fluorescent microscopy, we visualized internalization of dual-labeled PGN by enterocytes. Engulfed PGN was found to form a complex with PGN recognition protein-3, which may facilitate delivering PGN in vivo. Utilizing electronic microscopy, we revealed that uptake of apical PGN across intestinal epithelial monolayers was involved in phagocytosis, multivesicular body formation, and exosome secretion. We also studied transport of PGN using the transwell system. Our data indicated that apically loaded PGN was exocytosed to the basolateral compartment with exosomes by HT29-CL19A cells. The PGN-contained basolateral exosome extracts induced macrophage activation. Through gavaging mice with labeled PGN, we found that luminal PGN was taken up by columnar epithelial cells in crypts of the small intestine. Furthermore, we showed that pre-confluent immature but not post-confluent mature C2BBe1 cells engulfed PGN via a toll-like receptor 2-dependent manner. Together, our findings suggest that (1) crypt-based immature intestinal epithelial cells play an important role in transport of luminal PGN over the intestinal epithelium; and (2) luminal PGN is transcytosed across intestinal epithelia via a toll-like receptor 2-mediated phagocytosis-multivesicular body-exosome pathway. The absorbed PGN and its derivatives may facilitate maintenance of intestinal immune homeostasis.

Figure 1. Internalization of PGN by IEC-18 cells.

IEC-18 cells were treated with PGN that was dual labeled with BacLight-Green Stain and biotin. After 2 h, the cells were washed, fixed, and stained with strepavidin-Alex633 as described in the Methods. Under a fluorescent microscope, internalized PGN particles (marked with circles) were visualized with BacLight-Green not Alex633 optics, whereas extracellular PGN particles (marked with arrows) were visualized with both BacLight-Green and Alex633 optics. The phase contrast appearance and merged image of the green and red signal is shown in the merged panel. Control panel shows the merged image of cells which were treated with unlabeled PGN. Original magnification, X20. Data are representative of three separate experiments.

Figure 2. PGLYRP-3 is expressed in intestinal epithelial cells.

(A) PGLYRP-3 is constitutively expressed in human intestinal epithelial cells. Total cellular proteins were isolated from HT29-CL19A and undifferentiated C2BBe1 cells. Thirty micrograms of protein were subjected to SDS-PAGE, transferred to membranes, and analyzed using Western blot analysis with anti-PGLYRP-3 mAb. Lane 1, positive control provided by Imgenex Corp; Lane 2, protein extracted from HT29-CL19A cells; Lane 3, protein extracted from undifferentiated C2BBe1 cells. (B) Cellular localization of PGLYRP-3 protein in undifferentiated C2BBe1 cells. The cells were stained with immunofluorescence using mAb against human PGLYRP-3 followed by counterstaining the nuclei with DAPI as described in the Methods. Their staining profiles were merged. Original magnification, X63. Data in each panel are representative of two separate experiments.

Figure 3. Internalized PGN is colocalized with PGLYRP-3 in intestinal epithelial cells.

Pre-confluent C2BBe1 cells were exposed to BacLight-Green Stain labeled-PGN. After 4 h, the cells were washed, fixed, and stained with immunofluorescence using murine mAb against human PGLYRP-3 as a primary antibody and Alex633-labeled anti-mouse IgG as the secondary antibody. Nuclei were counterstained with DAPI. The cells were examined under a fluorescent microscope. (A) Internalized PGN appeared under BacLight-Green optic. (B) PGYLRP3 was visualized under Alex633 optic. (C) Nucleus was visualized using DAPI filter set. (D) The images of panels A – C were merged using Photoshop software. Original magnification, X63. Data in each panel are representative of three separate experiments.

Figure 4. Internalization and cellular fate of PGN within intestinal epithelial cells.

HT29-CL19A cell monolayers were exposed to PGN for 2 (Panels A – B), 6 (Panel C) and 24 h (Panels D – E) from apical sides. At the end of the treatments, the cells were washed, fixed, and processed for EM. Original magnification, X12,000 (A – B), X20,000 (C – D), and X25,000 (E). P, PGN. White arrows indicate phagosomes. Black arrows indicate phagolysosomes. Asterisks indicate MVB. Data in each panel are representative of three separate experiments.

Figure 5. Absorption of PGN occurs in crypts of the small intestines.

Mice (n=3) were gavaged with BacLight-Green stain labeled PGN (Panel A) or the vehicle (Panel B). After 90 min, they were sacrificed. The small intestines were processed from cryosections, counterstaining with DAPI (blue), and examination under a fluorescent microscope. Yellow arrows indicate labeled PGN (green) in the crypt area and white arrows indicate labeled PGN (green) in the lamina propria. X10. Merged images are shown. Data in each panel are representative of two separate experiments.

Figure 6. PGN is engulfed by immature but not mature intestinal epithelial cells.

Cells were cultured in transwell systems. PGN was added to apical compartments of pre-confluent C2BBe1 cells (Panel A), post-confluent C2BBe1 cells (Panel B), and post-confluent HT29-CL19A cells (Panel C). After 2 h, the cells were washed, fixed, and processed for EM. Original magnification, X7,000 (A) and X12,000 (B, C). P, PGN. The arrow indicates engulfed-PGN underwent degradation in a phagolysosome. Data in each panel are representative of three separate experiments.

Figure 7. TLR2 facilitates phagocytosis of PGN by immature intestinal epithelial cells.

Pre-confluent C2BBe1 cells were pretreated with medium alone (Panel A), anti-TLR2 antibody (20 μg/ml, Panel B), or isotype control IgG (20 μg/ml, Panel C) for 30 min. Then, cells were co-incubated with PGN for 2 h. At the end of treatments, cells were extensively washed with PBS and processed for EM examination as described in Methods. Original magnification, X12,000. P, PGN. Data in each panel are representative of two separate experiments.

Figure 8. PGN is transcellular transported by intestinal epithelial cells via an exosome associated mechanism.

HT29-CL19A cells were cultured on transwell filters for three weeks. The cell monolayers with stable TER were treated from apical side with PGN. The TER of monolayers was measured prior and after PGN treatment (Panel A). Twenty-four hours after PGN exposure, exosomes were isolated from basolateral and then processed for quantifying PGN with SLP-HS assay as described in the Methods (Panel B). n=3. ND, Not detected. Data in each panel are representative of two separate experiments.

Figure 9. Macrophages are activated by exosomes isolated from basolateral medium of PGN-treated HT29-CL19A monolayer.

(A) PGN-contained exosomes induced pseudopod formation in macrophages. Peritoneal macrophages from C57BL/6J mice were treated for 6 h with culture medium alone, boiled-exosome extracts isolated from basolateral medium of controls HT29-CL19A monolayer cultures (Control-Exosome), or boiled-exosome extracts isolated from basolateral medium of PGN-treated HT29-CL19A monolayer cultures (PGN-Exosome). The cells were examined under an inverted microscope. X40. Data in each panel are representative of three separate experiments. (B) PGN-contained exosomes induced IL-6 production in macrophages. Raw 264.7 cells were subjected to treatment with medium alone, exosome isolated from control HT29-CL19A cells (Control-Exosome), or PGN-contained exosomes (PGN-Exosome). After 24 h, the culture supernatants were processed for measuring the IL-6 level as described in Materials and Methods. n = 4. ND, Not detected. Results are the means ± SEM. **, P < 0.01 compared with the control-exosome group.