Cell-cell contact formation governs Ca2+ signaling by TRPC4 in the vascular endothelium: evidence for a regulatory TRPC4-beta-catenin interaction

Abstract

TRPC4 is well recognized as a prominent cation channel in the vascular endothelium, but its contribution to agonist-induced endothelial Ca(2+) entry is still a matter of controversy. Here we report that the cellular targeting and Ca(2+) signaling function of TRPC4 is determined by the state of cell-cell adhesions during endothelial phenotype transitions. TRPC4 surface expression in human microvascular endothelial cells (HMEC-1) increased with the formation of cell-cell contacts. Epidermal growth factor recruited TRPC4 into the plasma membrane of proliferating cells but initiated retrieval of TRPC4 from the plasma membrane in quiescent, barrier-forming cells. Epidermal growth factor-induced Ca(2+) entry was strongly promoted by the formation of cell-cell contacts, and both siRNA and dominant negative knockdown experiments revealed that TRPC4 mediates stimulated Ca(2+) entry exclusively in proliferating clusters that form immature cell-cell contacts. TRPC4 co-precipitated with the junctional proteins beta-catenin and VE-cadherin. Analysis of cellular localization of fluorescent fusion proteins provided further evidence for recruitment of TRPC4 into junctional complexes. Analysis of TRPC4 function in the HEK293 expression system identified beta-catenin as a signaling molecule that enables cell-cell contact-dependent promotion of TRPC4 function. Our results place TRPC4 as a Ca(2+) entry channel that is regulated by cell-cell contact formation and interaction with beta-catenin. TRPC4 is suggested to serve stimulated Ca(2+) entry in a specific endothelial state during the transition from a proliferating to a quiescent phenotype. Thus, TRPC4 may adopt divergent, as yet unappreciated functions in endothelial Ca(2+) homeostasis and emerges as a potential key player in endothelial phenotype switching and tuning of cellular growth factor signaling.

FIGURE 1.

Membrane presentation of TRPC4 is enhanced by formation of cell-cell contacts and divergently affected by EGF in subconfluent and confluent HMEC-1. A, top, proteins from total cell lysates (lanes 1 and 4), biotinylated fractions (lanes 2, 3, 5, and 6), and non-biotinylated fractions (lane 7) at the indicated conditions were subjected to SDS-PAGE and immunoblotted with TRPC4 antibody. Results are representative of four individual experiments. Bottom, mean values ± S.E. of TRPC4 immunoreactivity detected in biotinylated fractions before and after stimulation with EGF (100 ng/ml, 20 min). *, statistically significant differences between stimulated and unstimulated conditions (p < 0.05); #, statistically significant difference between the unstimulated subconfluent state and the unstimulated confluent state (p < 0.05). B, time course of TER increase in cultured HMEC-1 cells representing the transition from proliferating, subconfluent to barrier forming, confluent populations. TER levels (arrows) were taken as a basis to distinguish proliferating, subconfluent cells from mature barriers characterized by a constant, maximal TER level. The trace represents mean TER ± S.E. (n = 6). d, days.

FIGURE 2.

EGF-induced Ca²⁺ entry is altered by endothelial phenotype transitions. A, representative records of fura-2 fluorescence ratio in single (migrating state; open circles), subconfluent contact-forming (proliferating clusters; closed circles), and confluent (quiescent, barrier-forming; squares) cells during a Ca²⁺ readdition after stimulation with EGF (100 ng/ml; arrow). Ca²⁺ readdition (elevation of extracellular Ca²⁺ from nominally free to 2 mm) is indicated. Basal Ca²⁺ entry into non-stimulated cells is shown for proliferating clusters (unstimulated). The inset shows inhibition of Ca²⁺ entry into EGF-stimulated subconfluent cells by TRPC4 siRNA silencing. Responses of a TRPC4-silenced cell (filled circles) and a control (scrambled; open circles) are shown. B, mean peak increases in fura-2 emission ratios (Δ values ± S.E., n ≥ 45) obtained during Ca²⁺ readdition. C, mean rates of Ca²⁺-sensitive fluorescence changes during the initial phase of reentry. *, statistically significant difference (p < 0.05) versus the single, migrating phenotype. D, TRPC4 immunofluorescence images of single (migrating; left), subconfluent contact-forming (proliferating; center), and confluent (quiescent; right) HMEC-1 cells (nuclei stained by 4′,6-diamidino-2-phenylindole). The arrows indicate positions of enhanced TRPC4 immunofluorescence. Scale bars, 10 μm.

FIGURE 3.

Formation of immature cell-cell contacts in proliferating HMEC-1 enables TRPC4-dependent Ca²⁺ entry. A, Ca²⁺ entry into thrombin or EGF-stimulated single HMEC-1 cells. Left, representative recordings of fluo-4 fluorescence intensity in single cells transfected with pcDNA3-vector (control; open circles), TRPC4 (closed circles), or a functional dominant negative TRPC4 fragment (DNTRPC; squares). Time courses during acute stimulation with 0.5 unit/ml thrombin (arrow) and Ca²⁺ readdition (2 mm) are shown. Center, mean values of peak fluo-4 intensity (n > 6; open columns, control; hatched columns, TRPC4; filled columns, DNTRPC). Right, mean increases in fura-2 fluorescence ratios (n ≥ 6) obtained during Ca²⁺ readdition after stimulation with EGF (100 ng/ml). B, left, representative recordings of fura-2 fluorescence ratios in the subconfluent, contact-forming state is shown for controls (sham-transfected; open circles) and DNTRPC-transfected (closed circles) cells, HMEC-1 cells were challenged with EGF (100 ng/ml; arrow), as indicated. Right, increases in fura-2 fluorescence ratios (mean ± S.E.; n ≥ 8). C, left, representative recordings of fura-2 fluorescence ratios in the confluent, barrier-forming state is shown for controls (sham-transfected; open circles) and DNTRPC-transfected (closed circles) cells, and HMEC-1 cells were challenged with EGF (100 ng/ml; arrow), as indicated. Right, increases in fura-2 fluorescence ratios (mean ± S.E.; n ≥ 8).

FIGURE 4.

Association of VE-cadherin and β-catenin with TRPC4 depends on cell-cell contact formation. A, top, TRPC4 was detected in total cell lysates (from the left, lanes 1 and 4) immunocomplexes precipitated with anti-VE-cadherin (lanes 2, 3, 5, and 6) and lysates precipitated only with beads (lane 7). Results are a representative of four experiments. Bottom, mean values ± S.E. of the TRPC4 immunoreactivity detected in precipitates obtained from cells before and after stimulation with EGF (100 ng/ml, 20 min). B, top, TRPC4 was detected in total cell lysates (from the left, lanes 1 and 4), immunocomplexes precipitated with anti-β-catenin (lanes 2, 3, 5, and 6), and lysates precipitated only with beads (lane 7). Results are a representative of four experiments. Bottom, mean values ± S.E. of the TRPC4 immunoreactivity detected in precipitates obtained from cells before and after stimulation with EGF (100 ng/ml, 20 min). * and #, statistically significant differences (p < 0.05) between stimulated versus unstimulated cells and subconfluent state versus confluent state, respectively. IP, immunoprecipitation.

FIGURE 5.

Cell-cell contact formation and β-catenin expression determines the cellular targeting of TRPC4. Fluorescence images of single (top images) or contact-forming (bottom images) HMEC-1 (A and B) and HEK293 (C and D) cells transfected with either CFP-TRPC4 (red) or GFP-β-catenin (green) alone (A and C) or double transfected with both constructs (B and D). The arrows indicate enhanced fluorescence of CFP-TRPC4 or GFP-β-catenin, respectively. A, fluorescence images of single transfected (single TF) HMEC-1 cells. B, fluorescence images of double transfected (double TF) HMEC-1 cells. C, fluorescence images of single transfected HEK293-cells. D, fluorescence images of double transfected HEK293 cells. Images from cells expressing GFP and CFP have been normalized for channel bleed-through by linear spectral unmixing. Scale bars, 10 μm. For B and D, relative values of fluorescence overlap as a measure for co-localization are shown for defined cellular regions (total cell, nuclear region, and cell-cell contact area). The columns represent the overlap of the two fluorophores (percentage of pixels) within selected cellular areas (mean ± S.E.).

FIGURE 6.

Interaction between TRPC4 and VE-cadherin within cell-cell contacts in the HEK293 expression system. This figure shows donor (CFP), acceptor (YFP), and raw FRET images as well as a color-coded map of FRET intensity (FRET index) after background subtraction. A, FRET analysis of CFP-TRPC4 and GFP-β-catenin transiently expressed in HEK293 cells. B, FRET analysis of CFP-TRPC4 and YFP-VE-cadherin transiently expressed in HEK293 cells. The arrows indicate enhanced FRET signal at cell-cell contact positions and cell-cell junctions. Images are representative of 14–19 individual FRET measurements.

FIGURE 7.

β-catenin enables cell-cell contact-dependent promotion of Ca²⁺ entry into TRPC4-expressing HEK293 cells. Ca²⁺ entry into thrombin-stimulated HEK293 cells during a typical Ca²⁺ readdition protocol. A, representative recordings of fura-2 fluorescence ratios from single or contact-forming HEK293 cells stably expressing TRPC4 (T4-60), transfected with either pcDNA3 (control) or β-catenin. Ca²⁺ readdition (from nominally free to 2 mm) was performed as indicated after a 5-min stimulation with thrombin (0.5 unit/ml; pretreatment). Shown are mean peak increases in fura-2 fluorescence ratios (Δ values ± S.E.) obtained during Ca²⁺ readdition to single or contact-forming T4-60 (B) as well as HEK293 cells (C). *, statistical significance (p < 0.05) versus contact-forming controls as well as single cells. WT, wild type.

FIGURE 8.

Proposed model of phenotype-dependent TRPC4 function in growth factor-stimulated endothelial cells via interaction of the channel with β-catenin and targeting into cell-cell contacts. TRPC4 is for a large part sequestered in intracellular compartments and unavailable for Ca²⁺ signaling in single cells (contact deficient; upper left). By contrast, formation of immature cell adhesions promotes surface targeting of β-catenin-TRPC4 complexes and enables further recruitment of channels into the plasma membrane and Ca²⁺ entry function (immature contact; upper right). Once mature barriers are formed (mature contact; lower panel), TRPC4 resides for a large part in junctional complexes that are rapidly retrieved from the cell surface during growth factor stimulation and are barely available for contribution to global Ca²⁺ signaling.