Lymphocytes accelerate epithelial tight junction assembly: role of AMP-activated protein kinase (AMPK)

Abstract

The tight junctions (TJs), characteristically located at the apicolateral borders of adjacent epithelial cells, are required for the proper formation of epithelial cell polarity as well as for sustaining the mucosal barrier to the external environment. The observation that lymphocytes are recruited by epithelial cells to the sites of infection [1] suggests that they may play a role in the modulation of epithelial barrier function and thus contribute to host defense. To test the ability of lymphocytes to modulate tight junction assembly in epithelial cells, we set up a lymphocyte-epithelial cell co-culture system, in which Madin-Darby canine kidney (MDCK) cells, a well-established model cell line for studying epithelial TJ assembly [2], were co-cultured with mouse lymphocytes to mimic an infection state. In a typical calcium switch experiment, the TJ assembly in co-culture was found to be accelerated compared to that in MDCK cells alone. This accelaration was found to be mediated by AMP-activated protein kinase (AMPK). AMPK activation was independent of changes in cellular ATP levels but it was found to be activated by the pro-inflammatory cytokine TNF-alpha. Forced suppression of AMPK, either with a chemical inhibitor or by knockdown, abrogated the accelerating effect of lymphocytes on TJ formation. Similar results were also observed in a co-culture with lymphocytes and Calu-3 human airway epithelial cells, suggesting that the activation of AMPK may be a general mechanism underlying lymphocyte-accelerated TJ assembly in different epithelia. These results suggest that signals from lymphocytes, such as cytokines, facilitate TJ assembly in epithelial cells via the activation of AMPK.

Figure 1. Lymphocyte co-culture facilitates MDCK tight junction assembly.

(a) Representative pictures of ZO-1 relocation to cell–cell junctions in MDCK cells cultured alone and co-cultured with lymphocytes at various time points (0h, 0.5h, 1h, 1.5h, 2h, 2.5h, 3h) after calcium switch. (Scale bar: 30 µm). (b) Quantification of ZO-1 relocation to cell–cell junctions (length of ZO-1/cell (µm)) in MDCK cells cultured alone and co-cultured with lymphocytes at various time points (0h, 0.5h, 1h, 1.5h, 2h, 2.5h, 3h) after calcium switch. The asterisks denote significant differences detected in the presence vs. the absence of lymphocytes by Student's t test (*, p<0.05; **, p<0.01).

Figure 2. AMPK is activated in the epithelium-lymphocyte co-culture and by TNF-α.

(a) Western blot results showing increased level of AMPK phosphorylation in MDCK-lymphocyte co-culture and Calu-3-lymphocyte co-culture as compared to MDCK and Calu-3 cultured alone, respectively. Total AMPK remained unchanged, indicating that AMPK is activated in these co-cultures, with beta-actin used as loading control. The corresponding statistical analysis is on the right panel (*, p<0.05; ***, p<0.001). (b) ATP assay showing no significant changes of cellular ATP levels in MDCK-lymphocyte co-culture vs. MDCK cells alone (‘ns’ stands for no significance). Cell lysates from MDCK cells with and without lymphocytes were subjected to an ATP assay. The ATP contents were normalized to the total protein amount. Data were from three experiments. (c) Lysates prepared from MDCK cells maintained in α-MEM in the absence or presence of TNF-α were blotted with the indicated antibodies. Treatment with AICAR served as the positive control for AMPK activation. (d) The activated AMPK was represented as a relative ratio to total AMPK. The phosphorylation of AMPK is significantly increased by exposure for 2 hours to TNF-α treated, as assessed by Student's test (**, p<0.01).

Figure 3. AMPK inhibitor, Compound C, abolishes the lymphocyte-facilitation of tight junction assembly.

(a) Representative pictures of ZO-1 relocation to cell–cell junctions in MDCK cells cultured alone and co-cultured with lymphocytes, with or without the AMPK inhibitor, Compound C, (A. MDCK; B. Co-culture; C. Co-culture+Compound C; D. Co-culture+DMSO) at various time points (0h, 0.5h, 1h, 2h) after calcium switch. (Scale bar: 30 µm). (b) Quantification of ZO-1 relocation to cell–cell junctions (length of ZO-1/cell (µm)) in MDCK cells cultured alone and co-cultured with lymphocytes, with or without AMPK inhibitor, Compound C, at various time points (0h, 0.5h, 1h, 2h) after calcium switch. The asterisks denote significant differences detected in (C) Co-culture+Compound C vs. (D) Co-culture+DMSO by Student's t test (*, p<0.05; **, p<0.01).

Figure 4. shRNA-mediated AMPK knockdown inhibits the lymphocyte-facilitation of tight junction assembly.

(a) Representative pictures of ZO-1 relocation to cell–cell junctions in (A) Vector control cells; (B) Co-cultured vector control cells; (C) shRNA AMPK cells and (D) Co-cultured shRNA AMPK cells at various time points (0h, 1h, 2h, 3h) after calcium switch. (Scale bar: 30 µm). (b) Quantification of ZO-1 relocation to cell–cell junctions (length of ZO-1/cell (µm)) in culture conditions as indicated in (a) at various time points (0h, 1h, 2h, 3h) after calcium switch. The asterisks denote significant differences detected in (B) co-cultured vector control cells vs. (D) co-cultured shRNA AMPK cells by Student's t test (*, p<0.05; **, p<0.01).