A modular spring-loaded actuator for mechanical activation of membrane proteins

Abstract

How cells respond to mechanical forces by converting them into biological signals underlie crucial cellular processes. Our understanding of mechanotransduction has been hindered by technical barriers, including limitations in our ability to effectively apply low range piconewton forces to specific mechanoreceptors on cell membranes without laborious and repetitive trials. To overcome these challenges we introduce the Nano-winch, a robust, easily assembled, programmable DNA origami-based molecular actuator. The Nano-winch is designed to manipulate multiple mechanoreceptors in parallel by exerting fine-tuned, low- piconewton forces in autonomous and remotely activated modes via adjustable single- and double-stranded DNA linkages, respectively. Nano-winches in autonomous mode can land and operate on the cell surface. Targeting the device to integrin stimulated detectable downstream phosphorylation of focal adhesion kinase, an indication that Nano-winches can be applied to study cellular mechanical processes. Remote activation mode allowed finer extension control and greater force exertion. We united remotely activated Nano-winches with single-channel bilayer experiments to directly observe the opening of a channel by mechanical force in the force responsive gated channel protein, BtuB. This customizable origami provides an instrument-free approach that can be applied to control and explore a diversity of mechanotransduction circuits on living cells.

Fig. 1. Design and assembly of a DNA-based Nano-winch.

A Schematic illustration of the assembly strategy of the DNA-based Nano-winch. Double-helical DNA domains are represented by cylinders and are packed on a honeycomb lattice. The Nano-winch comprises two origami in a 1:2 trimer, a central Piston-cylinder and two Landing Legs. The central piston-cylinder core origami has eight strands to anneal to the inner face of the Landing Leg origami (i). To prevent toppling, two ~30 nm six-helix bundles project at 90° from the origami landing legs and 45° away from each other to lay parallel to a surface and maximize area coverage to retain an upright position. They are reinforced with a dsDNA strut. Single-stranded DNA connectors link the top and bottom of the cylinder to the backstop and the piston tip, respectively (ii). These single-stranded DNA connectors loops act as entropic springs with stiffness kDNA and exert defined force which is mechanically translated through the rigid origami to the tip of the piston coupled to a molecular target with stiffness kprotein. The length of these connectors can be adjusted by storing the excess scaffold in reservoir loops on the backstop (iii). The tip of the piston positions up to three ligand moieties targeting specific cell surface receptors (iv). Extension of the Piston results in equivalent and opposing compressive force through the Landing Legs. B 1.0% agarose gel with 11 mM MgCl₂ on which the following samples were electrophoresed: M, 1-kb ladder, Piston-cylinder monomer, Landing Leg monomer, and Piston-cylinder incubated with two-fold molar excess of Landing Legs with a concentration gradient from 11 mM to 35 mM MgCl₂. Schematic representations and reference-free class averages from single-particle TEM micrographs of individual components (particle sets available in Supplementary Figs. 5, 8, and 9). Fully assembled Nano-winch particles were visualized both laterally and from above. C Each landing leg can be modified with eight cholesterol moieties, for a total of 32-modifications. Example TEM images of Nano-winches functionalized with cholesterol moieties adhering to small unilamellar vesicles through the landing legs. Distortions of liposomes is an effect of the process of sample preparation for negative stain TEM. White bars represent 50 nm. All TEM analyses were conducted at least three times for each sample. Source data are provided as a Source Data file.

Fig. 2. Autonomous single-stranded DNA Nano-winch characterization.

A Fully assembled trimer with single-stranded 97 nt connectors exists in an equilibrium state between ~5–25 nm with corresponding reference-free class average calculated from single-particle TEM micrographs. Measured distance r, correlates directly to the distance d between the surface and the tip of the piston. Nano-winches tuned with (B) 97 nt, (C) 60 nt, and (D) 30 nt connector strands and the corresponding distance d distributions and reference-free class averages (particle sets available in Supplementary Figs. 20, 22). The length of the connectors are adjusted by storing the excess scaffold in reservoir loops on the backstop, in red. E The forces exerted by a single (n = 1) top (orange) and bottom (blue) 97 nt ssDNA molecule are plotted as a function of extension x defined as the distance between the bottom of the piston and the bottom of the cylinder. The force of the linear spring is plotted for the values of kprotein (kpr) = (0.1, 0.5, 1) pN/nm. F Total force as a function of x for the three values used for kpr. The corresponding values of the equilibrium state of x, x_eq, are shown as vertical gray dashed lines in both a) and b) (Supplementary Note 1). G The total force applied on the linear spring as a function of kpr, when n = 6 molecules of 97 nt are present at the top and at the bottom of the structure.

Fig. 3. Autonomous DNA Nano-winch activation of integrin signalling.

A The transmembrane receptor integrin (blue) exists as a compact αβ heterodimer. Integrins transmit applied mechanical stresses, between 1 and 15 pN, and recruits additional proteins to assemble focal adhesions including Focal Adhesion Kinase (FAK), which becomes phosphorylated at residue Y397 after mechanical stimulation of integrin. Addition of two antibodies with donor, D, and acceptor, A, labels allows detection of phosphorylated FAK in a LRET assay. Both antibodies bind to phosphorylated FAK (Y397-P) eliciting a detectable high LRET signal, whereas only a single antibody binds in the absence of phosphorylation yielding a low LRET signal. B MCF-7 cells in suspension were 1, left untreated control, 2, incubated with RGD conjugated oligonucleotide, 3, incubated with cRGD functionalized Piston-cylinder origami, 4, incubated with non-functionalized Nano-winches, 5, incubated with cRGD functionalized Nano-winch. Cells were then lysed and FAK phosphorylation. The background signal, R₀, of antibodies alone was subtracted from the signal of lysed cells in experimental and control conditions calculated from ratios of acceptor and donor fluorescence intensities, R_AD. Results are the average of at least three independent experiments. Error bars represent the standard deviation, statistical significance was determined by one-way analysis of variance with comparison to the untreated control (***P < 0.001). Source data are provided as a Source Data file.

Fig. 4. Remotely activated double-stranded DNA Nano-winch characterization and activity.

A Addition of extension oligonucleotides complementary to the connector strands between the backstop and cylinder ratchets the piston from the membrane surface. Annealing of the oligonucleotides to these connectors transitions them from single-stranded, with a persistence length of ~1 nm, to double-stranded, which is fifty times stiffer, at 50 nm. B Nano-winches tuned with 30-bp, 60-bp, and 97-bp connectors between the backstop and cylinder with corresponding distance d measurements and reference-free class averages (particle sets available in Supplementary Figs. 33, 35, and 37). C The mean distance extension of 30-bp, 60-bp, and 97-bp connectors (red dots) overlaid on the WLC model of Piston-cylinder extension (black line) enables the user to rationally define the connectors length for a required distance d. Error bars represent standard deviation. D Nano-winch was folded with a DNA hairpin tethering the backstop to the cylinder (red strands) requiring F1/2 of approximately 20pN to unzip. This hairpin serves as a benchmark to cross-validate the force applied by annealing of the extension oligonucleotides estimated by coarse-grained molecular dynamics simulation. E The distance d of tethered Piston-cylinders with the hairpin were measured before (blue) and after (red) incubation with extension oligonucleotides (particle sets available in Supplementary Figs. 39–40). F BtuB is a mechanically gated β-barrel channel occluded by a globular plug domain. The N-terminal 49 residues (red) are dislocated to form a channel and a N- terminal linker was engineered with a cysteine residue into the protein to serve as a linker for conjugation with a thiolated oligonucleotide. Addition of extension oligonucleotides retracts the piston and unfolds part of the plug domain to open a channel. G Selected traces and corresponding trace count histogram of BtuB-oligo reconstituted into planar lipid bilayers with Nano-winches after addition of 400 nM extension oligonucleotides. Closed channels (blue) transitioned to an open state (red) over two seconds after addition of 400 nM extension oligonucleotides. H Reducing the connection between BtuB and the Nano-winch allowed the plug domain to refold and transition from an open channel back to a closed state. Breaks in the traces represent preparations for thoroughly mixing the cis compartment. Arrows on the graphs indicate where the extension oligonucleotides and the DTT reducing agent solutions are added. Source data are provided as a Source Data file.