Vinculin transmits high-level integrin tensions that are dispensable for focal adhesion formation

Abstract

Focal adhesions (FAs) transmit force and mediate mechanotransduction between cells and the matrix. Previous studies revealed that integrin-transmitted force is critical to regulate FA formation. As vinculin is a prominent FA protein implicated in integrin tension transmission, this work studies the relation among integrin tensions (force), vinculin (protein), and FA formation (structure) by integrin tension manipulation, force visualization and vinculin knockout (KO). Two DNA-based integrin tension tools are adopted: tension gauge tether (TGT) and integrative tension sensor (ITS), with TGT restricting integrin tensions under a designed T_tol (tension tolerance) value and ITS visualizing integrin tensions above the T_tol value by fluorescence. Results show that large FAs (area >1 μm²) were formed on the TGT surface with T_tol of 54 pN but not on those with lower T_tol values. Time-series analysis of FA formation shows that focal complexes (area <0.5 μm²) appeared on all TGT surfaces 20 min after cell plating, but only matured to large FAs on TGT with T_tol of 54 pN. Next, we tested FA formation in vinculin KO cells on TGT surfaces. Surprisingly, the T_tol value of TGT required for large FA formation is drastically decreased to 23 pN. To explore the cause, we visualized integrin tensions in both wild-type and vinculin KO cells using ITS. The results showed that integrin tensions in FAs of wild-type cells frequently activate ITS with T_tol of 54 pN. With vinculin KO, however, integrin tensions in FAs became lower and unable to activate 54 pN ITS. Force signal intensities of integrin tensions reported by 33 and 43 pN ITS were also significantly reduced with vinculin KO, suggesting that vinculin is essential to transmit high-level integrin tensions and involved in transmitting intermediate-level integrin tensions in FAs. However, the high-level integrin tensions transmitted by vinculin are not required by FA formation.

Figure 1

Calibrate integrin ligand strength required for FA formation with PLL-PEG-TGT surfaces. (A) A PLL-PEG-TGT platform is developed to knock down integrin tensions without disrupting cell adhesion. TGT is immobilized on glass surfaces coated with composite polymer PLL-PEG, which supports cell adhesion by electrostatic adsorption and suppresses non-specific integrin-substrate interaction. (B) Molecular structure of PLL-PEG. PLL is a long peptide consisting of ∼100 lysines, and 20% of lysines are conjugated with biotinylated PEG polymers. (C) T_tol values are selectable in the range of 12–54 pN by changing the biotin location on the TGT. (D) HeLa cells were incubated on PLL-PEG-TGT surfaces for 90 min, with vinculin and talin imaged to mark adhesion clusters. Talin and vinculin are highly colocalized in the clusters. (E) Cells adhered on all surfaces, including the TGT-null PLL-PEG surface, with comparable cell spreading areas. Ten to 20 cells were analyzed for each condition. (F) Vinculin cluster areas were analyzed by Python-based code. (G) Vinculin cluster area monotonically increases along with T_tol values of TGT. The cluster area increases most significantly from T_tol of 43 to 54 pN. Fifty to 70 clusters were analyzed for each T_tol condition. The error bar defines the data range (excluding outliers) and the inner box defines the standard deviation and the median value. This applies to all other error bars in this paper. To see this figure in color, go online.

Figure 2

Time series of integrin and zyxin clusters formed on PLL-PEG-TGT surfaces. (A) HeLa cells were plated on TGT surfaces with T_tol of 12 and 54 pN, respectively, and then fixed at various time points to observe the development of integrin and zyxin clusters. Integrin β₃ marks initial cluster formation, and zyxin supposedly marks mature focal adhesion. (B) Integrin β₃ cluster area versus cell incubation time. The left side shows cluster areas on the 12 pN TGT surface from 10 to 60 min, and the right side shows cluster areas on the 54 pN TGT surface from 10 to 60 min. (C) Zyxin cluster area versus cell incubation time. The results demonstrate that both integrin β₃ and zyxin clusters were able to grow larger by time on the 54 pN TGT surfaces but not on the 12 pN TGT surfaces. To see this figure in color, go online.

Figure 3

HeLa cells with vinculin knocked out and wild-type HeLa cells on PLL-PEG-TGT surfaces. (A) Both vinculin KO cells and wild-type cells express talin-GFP which reports FA formation. Cells were incubated on the surfaces for 90 min. (B) Vinculin immunostaining confirms that vinculin is indeed knocked out in the vinculin KO cells. (C) Talin cluster areas versus T_tol values of TGT. The cluster area increases significantly with T_tol values changing from 12 to 23 pN. (D) Talin cluster areas versus T_tol values of TGT. The cluster area increases significantly with T_tol value changing from 43 to 54 pN. To see this figure in color, go online.

Figure 4

Visualize integrin tensions in vinculin KO cells using integrative tension sensor (ITS). (A) Schematics of ITS application. ITS shares similar DNA constructs to TGT, but serves to image integrin tensions, not to modulate integrin tensions. ITS is conjugated with a quencher-dye pair to visualize integrin tensions. The surface is co-coated with fibronectin (FN) which provides stable integrin ligands for cell adhesion. (B) Vinculin KO HeLa and wild-type HeLa cells incubated on FN-coated surfaces for 90 min. FAs were marked by talin-GFP and F-actin was stained with phalloidin. Stress fibers connecting FAs are visible in the wild-type cells but few in vinculin KO cells. (C) Vinculin KO HeLa and wild-type asasaHeLa cells incubated on ITS surfaces with T_tol of 54 pN for 90 min. FAs were marked by talin-GFP and integrin tension signals were reported by 54 pN ITS. The third row is an overlay of talin (green) and ITS (magenta). (D) Relative fluorescence intensities of ITS signal (T_tol = 54 pN) between wild-type and vinculin KO HeLa cells. Integrin tension signal reported by 54 pN ITS in vinculin KO cells is nearly zero. To see this figure in color, go online.

Figure 5

Integrin tension signals in vinculin KO and wild-type HeLa cells on a series of ITS surfaces. (A) Vinculin KO cells incubated on 12–54 pN ITS surfaces for 90 min, with talin marking FAs. (B) Wild-type cells incubated on 12–54 pN ITS surfaces for 90 min. (C) Relative fluorescence intensities of ITS signals in vinculin KO cells. The 54 pN ITS showed nearly zero signals, but ITS surfaces with other T_tol values showed observable force signals. (D) Relative fluorescence intensities of ITS signals in wild-type cells. To see this figure in color, go online.