Linking E-cadherin mechanotransduction to cell metabolism through force-mediated activation of AMPK

Abstract

The response of cells to mechanical force is a major determinant of cell behaviour and is an energetically costly event. How cells derive energy to resist mechanical force is unknown. Here, we show that application of force to E-cadherin stimulates liver kinase B1 (LKB1) to activate AMP-activated protein kinase (AMPK), a master regulator of energy homeostasis. LKB1 recruits AMPK to the E-cadherin mechanotransduction complex, thereby stimulating actomyosin contractility, glucose uptake and ATP production. The increase in ATP provides energy to reinforce the adhesion complex and actin cytoskeleton so that the cell can resist physiological forces. Together, these findings reveal a paradigm for how mechanotransduction and metabolism are linked and provide a framework for understanding how diseases involving contractile and metabolic disturbances arise.

Figure 1. AMPK is activated in response to force applied to E-cadherin

a and b, MCF10A cells were incubated with magnetic beads coated with IgG or E-cadherin extracellular domains (E-cad). The cells were left resting(-) or a magnet was used to generate tensional forces (+). a, AMPK immunoprecipitates were subjected to in vitro kinase assay with its substrate, SAMS peptide. SAMA=control peptide. Cmpd. C indicates cells pretreated with the AMPK inhibitor, Compound C. b, total cell lysates were immunoblotted with antibodies that recognize AMPK or AMPK phosphorylated in its activation loop (pAMPK). shControl indicates cells treated with scrambled shRNAs. shAMPK1and shAMPK2 indicate cells infected with two separate shRNAs targeting AMPK. c, shear stress was applied to MDCK cells, and AMPK and pAMPK were monitored by immunoblotting. d, tensional forces (+) were applied to MCF10A cells pretreated with blebbistatin (Blebbi) or E-cadherin function blocking antibodies (HECD-1). Total cell lysates were probed with antibodies against pAMPK, AMPK, phospho-myosin light chain (pMLC), or MLC. e, MCF10A cells were left resting (-) or treated (+) with Calyculin A (to stimulate myosin II-dependent increased contractility). Total cell lysates were immunoblotted as described in d. f and g, Tensional forces were applied to MCF10A cells as described in a. The beads were recovered and co-precipitation of AMPK (f) and pAMPK (g) with E-cadherin were examined by immunoblotting. The graphs beneath the image show the average ± SEM for 3 independent experiments. *, #, and ## indicate p-values of <0.01, 0.05 and 0.005, respectively. Unprocessed scans of blots are shown in Supplementary Figure 5.

Figure 2. LKB1 is recruited to the cadherin adhesion complex in response to force and activates and recruits AMPK

MCF10A cells (a,b,d,g) or MDCK cells (e) were incubated with beads coated with IgG or E-cadherin extracellular domains (E-cad) and left resting (-) or stimulated (+) with tensional force using a permanent magnet. In other experiments, MDCK cells (c and f) were left resting (-) or exposed to shear stress (+). a, the cells were lysed and co-precipitation of LKB1 with the E-cadherin-coated magnetic beads was examined. b, Co-immunoprecipitation of E-cadherin (E-cad) with LKB1 was monitored using immunoblotting. c, The cells were fixed, permeabilized and stained with antibodies against E-cadherin or LKB1. The co-localization of LKB1 with E-cadherin was examined using confocal microscopy. Scale bar = 20μm. d-f, The cells were lysed, and whole cell lysates were immunoblotted with the indicated antibodies. shLKB1 denotes cells with LKB1 silenced. shLuc indicates cells expressing a vector control cDNA, and cl.11 and cl.14 indicate two clonal cell lines lacking LKB1. g, the cells were lysed, and pAMPK and AMPK co-purification with the E-cadherin-coated magnetic beads was examined by immunoblotting. The graphs beneath each image show the average ± SEM for 3 independent experiments.* and # indicate p-values of <0.01, and <0.05, respectively. Unprocessed scans of blots are shown in Supplementary Figure 5.

Figure 3. LKB1 and AMPK are upstream of Abl-mediated phosphorylation of Y822 vinculin and Rho-mediated contractility

a, schematic of the signal transduction cascade from E-cadherin to Rho-mediated contractility. b-h, MCF10a cells were incubated with beads coated with IgG or E-cadherin extracellular domains (E-cad) and left resting (-) or stimulated (+) with tensional force using a permanent magnet. b,c,d,e, whole cell lysates were probed by immunoblotting with antibodies that recognize phosphorylation of CrkL at the Abl-specific site (b and c, pCrkL) or phosphorylation of vinculin Y822 (d and e, pY822). f, AMPK was immunoprecipitated and vinculin and Abl recruitment were examined by immunoblotting. g, Active Rho (Rho–GTP) was isolated with GST–RBD and analyzed by western blotting. h, total cell lysates were immunoblotted with antibodies against myosin light chain (MLC) or MLC phosphorylated at Serine 19 (pMLC). shLKB1 denotes cells expressing shRNAs against LKB1. shControl indicates cells treated with scrambled shRNAs. shAMPK1and shAMPK2 indicate cells infected with two separate shRNAs targeting AMPK. The graphs beneath each image show the average ± SEM for 3 independent experiments. **, * and # indicate p-values of <0.001, <0.01, and 0.05, respectively. Unprocessed scans of blots are shown in Supplementary Figure 5.

Figure 4. Force-induced AMPK stimulates glucose uptake and increases intracellular ATP levels

a and b, MCF10A (a) or MDCK (b) cells were incubated paramagnetic beads coated with IgG or E-cadherin extracellular domains (E-cad). Tensile forces were applied to the beads using a magnet, the cells were lysed, and the amount of a fluorescently-labelled 2-deoxyglucose analog taken up into the cells was monitored using a fluorimeter. c, MDCK cells were left resting (no shear) or exposed to shear stress (shear), and the amount of glucose taken up into the cells was monitored as described in a. d and e, Total ATP levels in cells treated as described in a and b were monitored as described in the experimental procedures. Cmpd C indicates cells treated with the AMPK inhibitor Compound C, Oligo indicates cells treated with the ATP synthase inhibitor, Oligomycin A, and 2-DG indicates cells incubated in the presence of 2-deoxyglucose. The graphs represent average glucose uptake or intracellular ATP for at least three representative experiments ± SEM. # and ## indicate p-values of <0.05 and < 0.005, respectively. n.s. indicates that there is no statistical differences between groups.

Figure 5. Force-induced increases in ATP reinforce the actin cytoskeleton and the E-cadherin adhesion complex to modulate barrier formation

a-c, MDCKII cells (n=80) or two clonal MDCKII cells lines (cl.11 and cl.14, n=63 and 52 respectively) lacking LKB1 were left untreated, treated with inhibitors of AMPK (Compound C=Cmpd. C, n=62) or ATP synthesis (Oligo A,n=44 or Carbonyl cyanide-4-(trifluoromethoxy)phenylhydrazone=FCCP, n=26, or incubated in low glucose containing media (Low Gluc, n=25). The cells were then left resting (no shear) or exposed to physiological shear stress. The cells were fixed, stained with antibodies against E-cadherin or Texas-Red phalloidin, and examined by confocal microscopy. The graphs in b and c represent the average corrected fluorescence intensity of E-cadherin (b, E-cad) or F-actin (c) in junctions.The data are represented as a box and whisker plot with median, 10^th, 25^th, 75^th, and 90^th percentiles shown. Scale bars=20 μm. d, MDCKII cells were grown to confluence and then incubated overnight in low calcium containing media. The formation of cell-cell junctions was then stimulated by adding growth media to the cells. The trans-epithelial resistance across the epithelial monolayer was monitored using a voltmeter at the indicated times (hours). **,*, and # indicate p-values of <0.001, <0.01 and <0.05, respectively.