The Mechanically-induced Catalytic Amplification Reaction for Readout of Receptor-Mediated Cellular Forces

Abstract

Mechanics play a fundamental role in cell biology, but detecting piconewton forces is challenging due to the lack of accessible and high throughput assays. Herein we report the mechanically-induced catalytic amplification reaction (MCR) for readout of receptor-mediated forces in cells. Mechanically labile DNA duplexes presenting ligands are surface immobilized such that specific receptor forces denature the duplex and thus expose a blocked primer. Amplification of primers is achieved using an isothermal polymerization reaction and quantified by fluorescence readout. As a proof-of-concept, the assay was used to test the activity of a mechano-modulatory drug and integrin adhesion receptor antibodies. To the best of our knowledge, this is the first example of a catalytic reaction triggered in response to molecular piconewton forces. The MCR may transform the field of mechanobiology by providing a new facile tool to detect receptor-specific mechanics with the convenience of the PCR.

Graphical Abstract

Herein we report the mechanically-induced catalytic amplification reaction (MCR) for readout of receptor-mediated forces in cells. To the best of our knowledge, this is the first example of a catalytic reaction triggered in response to molecular piconewton forces.

Coupling between mechanical forces and chemistry at interfaces plays a profound role in biological processes ranging from biofilm formation to stem cell differentiation, and wound healing.^[1] To understand these types of chemo-mechanical coupling processes, it is necessary to develop methods to quantify cellular forces. This is challenging because molecular forces in biochemical processes are transient and tend to range from ~1–100 pN, which is sufficient to drive conformational changes in proteins but insufficient to dissociate covalent bonds.^[2] Therefore, forces in biochemical systems are difficult to detect and map.

We previously developed molecular tension-based fluorescence microscopy (MTFM) to image forces transmitted by cell surface receptors in living cells.^[3] The initial tension probes were comprised of an extendable polyethylene glycol (PEG) spring, flanked by a fluorophore and a spectroscopically-matched quencher.^[4] pN forces extend the mean end-to-end distance of the polymer, which reduces energy transfer through an R⁻⁶ distance-dependent relationship. Next generation probes utilized oligonucleotides,^[5] elastin like polypeptides,^[6] and engineered proteins^[7], and also employed gold nanoparticle quenchers to extend energy transfer distances and enhanced sensitivity.^[7–8]

Nonetheless, the sensitivity of MTFM is limited due to the energy transfer-based readout and the transient nature of cellular forces. For example, current probes require high-end microscopy systems with single-photon counting EMCCDs coupled with high-numerical aperture (NA) objectives to detect changes in energy transfer efficiency. Accordingly, high-throughput screening of drugs that target mechanical processes is prohibitive. Likewise, screening the mechanical phenotype of a library of cells is currently a challenge. Therefore, there is a pressing need to develop strategies that transduce pN forces into an easily quantifiable, and amplified chemical signal. As a corollary, catalytic amplification strategies, such as PCR and ELISA, provide the foundation of modern molecular and cellular biology, and equivalent assays for mechanics may transform the field of mechanobiology.

Enzymes that respond to specific pN mechanical inputs are widespread in Nature.^[9] In contrast, synthetic systems that transduce defined pN forces into a catalytic output are rare (Scheme 1A). To the best of our knowledge, the only examples of synthetic catalytic reactions that are mechanically triggered are based on polymer structures that initiate mixing of an enzyme and its substrate.^[10] These systems are sensitive to bulk forces, measured in units of kPa, rather than molecular pN events.

Scheme 1.

The mechanically-induced catalytic amplification reaction (MCR).

Herein we present the mechanically-induced catalytic amplification reaction (MCR) to amplify the signal associated with pN forces applied by cell surface receptors. The strategy depends on a blocked initiator of an enzymatic reaction that is exposed through the action of mechanical forces. Given the fidelity and sensitivity of PCR, we aimed to leverage DNA amplification as a proof-of-concept readout for MCR. The mechanically responsive element was a DNA duplex inspired by Wang and Ha’s tension gauge tether (TGT) assay.^[11] In their assay, an immobilized DNA duplex denatures due to cellular forces exceeding the tension tolerance, T_tol (defined as the minimum force needed to denature DNA when applied for 2 s). TGTs are a powerful tool in defining the mechanical forces needed for receptor activation.

As illustrated in Scheme 1B, a DNA duplex modified with a ligand is immobilized. When cells are plated on the surface, adhesion receptors engage their ligands and apply mechanical forces (F_receptor). Receptor-mediated tension exceeding the T_tol exposes the blocked primer for amplification. We demonstrate MCR using isothermal amplification (rather than PCR) to minimize background due to thermal denaturation of the blocked primer. For amplification, an 81-mer linear DNA template is hybridized and circularized by T4 ligase (Table S1 for DNA sequences). Next, the primer strand is replicated with isothermal rolling circle amplification (RCA).^[12] Under optimal conditions, the RCA reaction replicates a circular template thousands of times generating a long tandem repeat of DNA (Figure S1). The repetitive amplified product is then visualized by fluorescence in situ hybridization (FISH), an established technique for sensitive nucleic acid detection with high specificity.^[13] Quantification of the product can be achieved by direct surface imaging or by release of fluorescent oligonucleotides followed by high-throughput plate reader measurements (Scheme 1B). In principle, each mechanical rupturing event is transduced and amplified into hundreds of fluorescent oligonucleotides.

Because immobilization imposes a steric constraint to polymerases, we first quantified the efficiency and selectivity of RCA on a surface. 5′ thiol modified primers with a T₁₀ spacer were immobilized onto gold films,^[14] and amplified as described in Scheme 1B. Surface imaging of hybridized FISH probes in the amplified samples revealed a fluorescent monolayer with a 15.7±4.9% coefficient of variation (CV) in intensity (Figure 1A), which is likely due to heterogeneous efficiency of polymerization on the surface. In contrast, the non-amplified samples showed a 4.9±0.3% CV, demonstrating that the hybridization of complement to the primer strand is relatively more homogeneous. Importantly, the fluorescence signal in the amplified primer samples showed 102±4 fold increase compared to non-amplified samples (Figure 1B). Solution amplification shows ~1000 fold replication (Figure S1),^[12a] thus surface confinement inhibits polymerase activity and reduces the overall amplification efficiency. The ~ 100 fold enhancement in signal represents the maximum amplification of a mechanically triggered dehybridization event into a chemical output.

Figure 1.

Selectivity and efficiency of surface initiated isothermal amplification. Schematic and representative epi-fluorescence images (A) and plot of mean signal (B) from positive (primer), and negative controls (non-amplified sample, and scrambled primer). Error bars represent S.E.M. obtained from three independent experiments (total 30 images). Scale bars = 10 μm.

We next used MCR to detect forces mediated by integrins, which are a family of heterodimeric cell surface receptors that mediate cell adhesion and migration.²⁸ Integrins physically bridge the cellular cytoskeleton with the extracellular matrix, and accordingly experience pN forces.^[15] Assays that allow for screening of compounds that modulate integrin tension are of potential significance.

First, we quantified integrin-mediated denaturation of immobilized DNA duplexes. 5′-Cy3B, 3′-biotin labeled complement was hybridized to the primer (Figure 2A). Biotin-streptavidin binding was used to present the cyclic Arg-Gly-Asp-D-Phe-Lys (cRGDfK) peptide, a high-affinity ligand for integrin receptors. In this geometry, mechanical forces denature the duplex in an unzipping mode with a predicted T_tol = 12 pN. An identical primer sequence hybridized to a complement with 3′-Cy3B, 5′-biotin leads to denaturation in a shearing mode with a predicted T_tol = 56 pN. Note that the surface presents chemically identical probes with differing mechanical tolerance. After plating NIH/3T3 fibroblast cells on these surfaces for 1 h, we observed a loss in fluorescence that colocalized with the cell footprint as indicated by reflection interference contrast microscopy (RICM) (Figure 2B). Minimal loss in fluorescence was observed when the cRGDfK adhesion peptide was withheld (not shown). We quantified the decrease in Cy3B fluorescence under individual cells (Figure 2C) and found a greater fraction of the T_tol = 12 pN duplexes were denatured (13.7±0.9% decrease in fluorescence) compared to that of the T_tol = 56 pN duplex (6.5±0.45% decrease in fluorescence). The data shows differential mechanical denaturation of DNA duplexes, with a two-fold difference in DNA loss when comparing the 12 pN to 56 pN duplexes.

Figure 2.

(A) Schematic of mechanically labile duplexes used to study integrin-mediated forces. (B) Representative RICM and fluorescence images of cells cultured on T_tol = 12 pN and T_tol = 56 pN surfaces. The negative signal observed in the Cy3B channel colocalized with cell adhesion patterns in RICM. Yellow line shows the cell edge as determined from RICM. Scale bars = 10 μm. (C) Plot quantifying loss of Cy3B fluorescence, which indicates mechanical DNA denaturation (n = 10 cells).

To catalytically amplify exposed primers, we plated 100,000 cells to the 12 and 56 pN surfaces (surface area = 68.58 mm²) and allowed them to spread for 1 h. This cell density corresponds to a full monolayer (680 μm² available per cell assuming each cell can spread ~900 μm²). We then performed MCR and imaged the fluorescently tagged probes by epi-fluorescence microscopy. As shown in Figure 3A and B, a significant fluorescence signal was observed on the surface. Therefore, primer amplification can readily be used to detect integrin-driven denaturation of blocked primers. Note that MCR was performed in standard conditions (20 mM Tris-OAc, 50 mM KOAc, 10 mM Mg(OAc)₂, 100 μg/ml BSA, pH = 7.9) as media compatible with cells (such as DMEM, PBS, and HEPES) inhibit polymerase activity which is needed for MCR (Figure S2). Therefore, cells are absent during readout likely due to the multiple washing and incubation steps in MCR buffer.

Figure 3.

MCR to report integrin forces. Schematic and representative fluorescence images of (A) duplexes with T_tol = 12 pN, (B) T_tol = 56 pN, (C) duplexes lacking the cRGDfK peptide, and (D) scrambled duplexes. (E) Bar graph showing the average MCR signal from 30 different images from three independent samples. Error bars represent the S.E.M. of the results. (F) Bar graph showing the intensity of the eluent following release of FISH probes from the surfaces and detected by a microplate reader. Error bars represent the S.E.M. of the results from three independent experiments. Scale bars = 10 μm.

Controls using duplexes lacking cRGDfK (Figure 3C) and scrambled duplexes non-complementary to the template (Figure 3D) confirmed the specificity of MCR. In these controls, we doped the DNA surface with 10% (by incubation concentration) single stranded DNA labelled with cRGDfK to mediate cell adhesion. The cell density was nearly identical on all the tested surfaces in Figure 3, indicating that the density of cRGDfK ligands was sufficient to trigger cell adhesion prior to MCR readout (Figure S3). All controls showed low signal, ~150 fold lower than that generated by the 12 pN surface (Figure 3C and D). The background signal observed in Figure 3C is likely due to amplification of single stranded primers exposed due to spontaneous dissociation of DNA duplexes. Confirming this result, we found a ~3% loss of fluorescently labelled DNA duplexes from the surface when incubated in cell imaging media for 3 hrs at 37°C (Figure S4).

The bar graphs in Figures 3E and F show the results of quantifying the MCR signal using imaging-based, and plate reader-based readouts, respectively. For plate reader-based readout, the bound FISH probes were released by de-hybridization with nanopure water, and then transferred to a 96-well plate where fluorescence was quantified. Importantly, the fluorescence intensity was normalized to the maximum MCR signal obtained from a monolayer of primer (~3.5 x 10⁴ primers per μm²⁻ (Supplementary Note 1), which is consistent with the literature precedent^[16]) prepared in the same batch. The differences between image-based and plate-reader based readouts are likely due to differential levels of background and sensitivity; with the image-based readout likely more sensitive. The 12 pN duplex showed ~5-fold and 2.7-fold greater signal than that of the 56 pN duplex in Figures 3E and 3F, respectively, consistent with the mechanically-induced dehybridization data in Figure 2.

We next demonstrated the ability to conduct drug screens where we measured the effect of a drug on integrin mechanics rather than cell viability. We investigated the non-muscle myosin II inhibitor blebbistatin, which diminishes myosin contractility and thus reduces forces transmitted by focal adhesions. We pre-treated NIH/3T3 cells with a range of blebbistatin concentrations (10 nM – 10 μM) for 15 min and then incubated the cells onto the surface with the 12 pN duplexes for 1 h, which was then followed by MCR readout. Brightfield imaging indicated that cells become more rounded with increasing drug dose (Figure S5). This is confirmed by F-actin staining which showed more disorganized and shorter actin filaments at the highest blebbistatin doses (Figure 4A). Correspondingly, the MCR signal displayed a dose-dependent relationship where the highest blebbistatin concentrations generated the lowest MCR signal (Figure 4B and C). The MCR signal is a direct readout of integrin tension, measuring the dose-dependent dissipation of actomyosin contractile forces.

Figure 4.

Representative fluorescence images showing (A) F-actin staining and (B) MCR signal for NIH/3T3 cells treated with increasing concentrations of blebbistatin (0.01 – 10 μM). (C) Bar chart showing the MCR signal in response to increasing concentration of blebbistatin. (D) Bar chart quantifying MCR signal in the presence of anti-α₅β₁, anti-α_vβ₃ or both antibodies relative to the sample without antibodies. Error bars represent the S.E.M. from n=30 images from 3 independent samples in surface imaging readout, and n = 3 independent samples for plate reader based readout. Scale bars = 10 μm.

To further highlight the utility of MCR, we measured the MCR signal on the 12 pN duplex surfaces in response to inhibiting different integrin subtypes. The two major integrin subtypes mediating adhesion of NIH/3T3 fibroblasts are α_vβ₃ and α₅β₁ that display divergent cellular functions.^[17] In surface-based imaging, anti-α_vβ₃ antibody treatment reduced the MCR signal by 59.6±4.1%, while anti-α₅β₁ antibody treatment reduced the MCR signal by 14.6±5.6% (Figure 4D, blue bars and Figure S6). Incubation with both antibodies led to the greatest reduction in MCR signal (64.4±4.3%). Plate reader measurement showed a similar trend (Figure 4D, red bars). The differential MCR signal following antibody blocking is likely due to a number of factors. First, α_vβ₃ integrins play a more important role in mediating the adhesion of fibroblasts.^[18] Second, the experiment is performed after 1 h of cell incubation, and α_vβ₃ integrins are thought to initiate cell adhesion.^[19] Finally, α_vβ₃ integrins show higher affinity toward the cRGDfK ligand (K_d ~ nM) compared to that of α₅β₁.^[20] Taken together, we demonstrate the first example of screening drugs that target cellular mechanics using a catalytic amplification assay.