A Portable and Sensitive DNA-based Electrochemical Sensor for Detecting Piconewton-scale Cellular Forces

Abstract

Background:

Cell-generated forces are a key player in cell biology, especially during cellular shape formation, migration, cancer development, and immune response. The measurement of forces exerted and experienced by cells is fundamental in understanding these mechanosensitive cellular behaviors. While cell-generated forces can now be detected based on techniques like fluorescence microscopy, atomic force microscopy, optical/magnetic tweezers, however, most of these approaches rely on complicated instruments or materials, as well as skilled operators, which could limit their potential broad applications in regular biological laboratories.

Results:

A new type of smartphone-based electrochemical sensor is developed here for cellular force measurement. In this system, a double-stranded DNA-based force probe, known as tension gauge tether, is attached to the surface of a gold screen-printed electrode, which is then incorporated into a portable smartphone-based electrochemical device. Cellular force-induced DNA detachment on the sensor surface results in multiple redox reporters to reach the surface of the electrode and generate enhanced electrochemical signals. To further improve the sensitivity, a CRISPR-Cas12a system has also been incorporated to cleave the remaining surface-attached anchor DNA strand. Using integrin-mediated tension as an example, piconewton-scale adhesion forces generated by ≤10 HeLa cells could now be reliably detected. Meanwhile, the threshold forces of these electrochemical sensors can also be modularly tuned to detect different levels of cellular forces.

Significance:

These novel DNA-based highly sensitive, portable, cost-efficient, and easy-to-use electrochemical sensors can be potentially powerful tools for detecting different cell-generated molecular forces. Functioning as complementary tools with traction force microscopy and fluorescent probes, these electrochemical sensors can be straightforwardly applied in regular biological laboratories for understanding the basic mechanical principles of cell signaling and for developing novel strategies and materials in tissue engineering, regenerative medicine, and cell therapy.

Graphical Abstract

1. Introduction

Cells adhere to the extracellular matrix and other cells during their migration, deformation, invasion, and metastasis processes [1–3]. These cellular adhesion events are often mediated by transmembrane proteins (e.g., integrins and cadherins) [4–6] that can not only bind with their target ligands but also generate piconewton (pN)-scale forces upon binding. Thanks to the technological advancements in traction force microscopy [7], micropillar arrays [8], and fluorescent tension probes [9–12], etc. during the past years, our understanding on the biological roles of cell-generated forces have been significantly improved. To detect these forces at the molecular level, various fluorescent tension probes have been engineered by integrating DNAs [9,10], polymers [13,14], or peptides [15,16] as the force-sensitive transducers. DNA-based fluorescent probes are arguably the most widely used because they can be easily synthesized and modified with different ligands and functional moieties, meanwhile, their force threshold values can also be modularly and precisely tuned.

We recently developed the first electrochemical DNA-based sensor for detecting cell-generated molecular forces [17]. Compared to fluorescent probes, these electrochemical sensors can be more portable and ready to use by a broader scientific community. However, the sensitivity of these “first-generation” electrochemical force sensors is still relatively low: to detect integrin-mediated adhesion forces, ≥10⁴ mL⁻¹ HeLa cells are needed. To improve the sensitivity of the device, herein, we designed a “second-generation” electrochemical DNA-based force sensor.

In our design, a double-stranded “tension gauge tether” (TGT) force probe is attached to the surface of a gold screen-printed electrode (Au-SPE), which is then incorporated to a portable smartphone-based Sensit Smart electrochemical device. Unlike in our previous design where the DNA probe is covalently labeled with a redox reporter [17], herein, the redox reporters are free in the solution. Each force-induced DNA detachment on the sensor surface will result in multiple redox reporters, i.e., ferrocyanide [Fe(CN)₆]⁴⁻, to reach the surface of the Au-SPE and generate enhanced electrochemical signals (Fig. 1a) [18,19]. To further improve the sensitivity, we have also incorporated a CRISPR-Cas12a system [20] to cleave the remaining surface-attached thiolated anchor TGT strand. Both “label-free” signaling and CRISPR amplification are the novel strategies in this work as compared to our previous report [17]. We expect such new powerful electrochemical sensors can be potentially used for measuring various cell-generated molecular forces.

Fig. 1.

(a) Schematic of the smartphone-based electrochemical tension gauge tether (TGT) force sensors. Cell adhesion forces rupture the double-stranded TGT structures, and as a result, in “signal-off” design, the methylene blue redox reporter exhibited a decreased electrochemical signal, while in “signal-on” system, [Fe(CN)₆]⁴⁻ can reach the surface of the electrode to generate an increased signal. (b) Cyclic voltammograms and (c) square wave voltammograms for characterizing the fabrication process of the “rigid” TGT sensors. These measurements were performed in a solution containing (v/v) 50% DMEM and 50% phosphate buffer (0.2 M, pH 7.4) and 5.0 mM [Fe(CN)₆]⁴⁻, respectively on the unmodified Au-SPE (red line), thiolated anchor DNA-modified Au-SPE (orange line), 1-hexanethiol-passivated thiolated anchor DNA-modified Au-SPE (grey line), and that after adding ligand DNA strand (green line). The cyclic voltammetry signals were recorded at the scan rate of 50 mV·s⁻¹. (d) Square wave voltammograms for characterizing the fabrication process of the “soft” TGT sensors as measured respectively on the unmodified Au-SPE (red line), 3-mercaptopropionic acid-modified Au-SPE (orange line), after attaching poly-L-lysine (grey line), after adding thiolated anchor DNA (green line), after adding ligand DNA strand (blue line), and that after the BSA blocking (purple line).

2. Materials and methods

2.1. Reagents and apparatus

Double deionized water (18.6 MΩ·cm⁻¹) was used throughout this project. The DNA oligonucleotides were custom synthesized and purified by W. M. Keck Oligonucleotide Synthesis Facility at Yale University School of Medicine, and the sequences are: (1) biotinylated ligand DNA strand: 5’-CACAGCACGGAGGCACGACAC-biotin-3’; (2) 12 pN anchor DNA strand: 5’-HS-GTGTCGTGCCTCCGTGCTGTG-3’; (3) 56 pN anchor DNA strand: 5’-GTGTCGTGCCTCCGTGCTGTG-SH-3’; (4) crRNA for Cas12a: 5’-CACAGCACGGAGGCACGAC AC-3’. The CRISPR-Cas12a (EnGen^® Lba Cas12a (Cpf1)) enzyme was brought from New England Biolabs. Hanks’ Balanced Salt Solution (HBSS), 4-(2-hydroxyethyl)-1-piperazine ethane sulfonic acid (HEPES), sodium bicarbonate, potassium chloride (KCl), sodium chloride (NaCl), dipotassium phosphate (K₂HPO₄), sulphuric acid (H₂SO₄), poly-L-lysine, ethylenediaminetetraacetic acid (EDTA), tris(2-carboxyethyl)phosphine (TCEP), ferrocyanide [Fe(CN)₆]⁴⁻, adriamycin, Dulbecco’s modified eagle medium (DMEM), streptavidin, biotinylated cyclic arginine-glycine-aspartic acid (B-cRGDfK), sulfosuccinimidyl 4-(N-maleimidomethyl)cyclohexane-1-carboxylate (sulfo-SMCC), 1-hexanethiol, and latrunculin B were obtained from Thermo Fisher Scientific and used without further purification. Gold screen-printed electrodes (Au-SPE) with working electrode made of gold, auxiliary electrode made of platinum, reference electrode, and electric contacts made of silver (dimensions: 3.4×1.0×0.05 cm³, length×width×height) were purchased from Metrohm-DropSens (Llanera, Spain). The cyclic voltammetry (CV), electrochemical impedance spectroscopy (EIS), and square wave voltammetry (SWV) studies were performed using a Sensit Smart electrochemical device from PalmSens (Houten, Netherlands).

2.2. Fabrication of the rigid electrochemical TGT sensor

The fabrication of the electrochemical TGT sensors was similar to that shown in our previous work [17]. The surface of the Au-SPE was first cleaned using 2 M H₂SO₄ solution until no change in the cyclic voltammogram was observed when scanning the potential in the range from −0.3 V to 1.2 V (Figure S1a). Then, 100 μL of TCEP-reduced 12 pN (or 56 pN) DNA anchor strand (5 μM) was dropped onto the cleaned Au-SPE surface and kept in the refrigerator for 16 h. Following incubation, the DNA-anchored Au-SPE was rinsed with copious amounts of 0.1 M phosphate buffer (pH 7.4) to wash away nonspecifically adsorbed DNA strands. 100 μL of 1-hexanethiol (100 μM) was then dropped on the surface of the electrode to block the remaining active sites on the electrode surface by incubating at 37 °C for 1 h. After that, 100 μL of the biotinylated DNA ligand strand (5.0 μM) was added to the surface of the electrode to generate double-stranded DNA probes by incubating at room temperature for 1 h. The electrode was then washed profusely with 0.1 M phosphate buffer (pH 7.4), and subsequently, 100 μL of 5.0 μM streptavidin was dropped on the surface of the electrode to interact with biotinylated DNA duplex at room temperature for 1 h. Again, after rinsing plentifully with 0.1 M phosphate buffer (pH 7.4), 100 μL of 5.0 μM biotinylated cyclic arginine-glycine-aspartic-acid-D-phenylalanine-lysine (B-cRGDfK) was dropped on the surface of the electrode to interact with streptavidin at room temperature for 1 h. Consequently, the electrode was washed with 0.1 M phosphate buffer (pH 7.4) and 100 μL of 100 μM bovine serum albumin (BSA) was casted on the surface of the electrode to block the remaining active sites. As a final step, the fabricated TGT sensor was rinsed with 0.1 M phosphate buffer (pH 7.4) and stored at 4 °C before usage.

2.3. Measurement of cell-generated forces

HeLa cells were cultured in DMEM with 10% fetal bovine serum, 100 U/mL penicillin, and 100 U/mL streptomycin in an Eppendorf Galaxy incubator at 5% (v/v) CO₂. Before measurement, the cells were first detached by adding 2 mM EDTA solution, consisting of 1×HBSS, 0.06% sodium bicarbonate, and 0.01M HEPES (pH 7.6), for 10 min. The solution was then centrifuged three times at 1,200 rpm for 7 min and re-suspended in a measuring solution containing (v/v) 50% DMEM and 50% phosphate buffer (0.2 M, pH 7.4), with a final concentration from ~100 cells/mL to 1×10⁶ cells/mL. In a typical measurement, ~1×10⁵ HeLa cells (100 μL) were added on the surface of the above-prepared TGT electrochemical sensor and allowed the cells to interact with the DNA probes for 75 min. After that, the electrode was washed with 0.1 M phosphate buffer (pH 7.4) and added with 100 μL of 5 mM [Fe(CN)₆]⁴⁻ on the surface to start recording the SWV signals of the [Fe(CN)₆]⁴⁻. SWV parameters were as follows: step potential, 20 mV; pulse amplitude, 50 mV; and frequency, 20 Hz. The impact of different experimental conditions on the response of the TGT sensors were also investigated. It is worth mentioning that the [Fe(CN)₆]⁴⁻ ions should not potentially affect cells or their surface adhesions in our system, as before adding [Fe(CN)₆]⁴⁻, cells have already been washed away from the surface of the electrode.

2.4. The treatment of cells with a force inhibition drug

We have studied how the latrunculin B treatment can influence the cellular force generation. For this purpose, ~1×10⁶ mL⁻¹ HeLa cells were first pretreated with 5–60 μM of latrunculin B for 60 min at 37 °C inside a cell culture incubator. After that, cells were separated and dispersed in cell culture media, ~1×10⁵ cells were then casted onto the surface of the TGT sensor and incubated for 75 min. After washing with 0.1 M phosphate buffer (pH 7.4) for several times, the SWV of 5 mM [Fe(CN)₆]⁴⁻ on the electrode was recorded using the above-mentioned parameters: step potential, 20 mV; pulse amplitude, 50 mV; and frequency, 20 Hz.

2.5. Cas12a-mediated signal amplification of the TGT sensors

After rupturing the ligand DNA strand of the TGT sensors by the cells, the Cas12a/crRNA conjugate was used to further cleave the remaining anchor DNA strand to additionally improve the diffusibility of [Fe(CN)₆]⁴⁻ and to increase the sensitivity of the sensor. Here, the Cas12a/crRNA conjugate was first prepared by mixing 100 nM CRISPR-Cas12a, 100 nM crRNA, and 1×NEBuffer 2.1 for 30 min. After that, 100 μL of the mixture at 30, 60, 90, or 120 nM concentration was casted for 30 min on the surface of the cell-incubated electrode or the Au-SPE conjugated with just the anchor DNA strands. The anchor DNA was cleaved during this process. After washing the electrode with 0.1 M phosphate buffer (pH 7.4) for several times, the SWVs of 5 mM [Fe(CN)₆]⁴⁻ were recorded using the following parameters: step potential, 20 mV; pulse amplitude, 50 mV; and frequency, 20 Hz.

3. Results and discussion

3.1. Design and fabrication of the TGT sensors

We first anchored a TGT force probe with T_tol of ~12 pN onto the Au-SPE surface using previously reported DNA duplex sequences [17,21]. The whole fabrication process was characterized in each step using cyclic voltammetry (CV) and square wave voltammetry (SWV) in a measuring solution containing (v/v) 50% DMEM and 50% phosphate buffer (0.2 M, pH 7.4) and 5.0 mM [Fe(CN)₆]⁴⁻ redox reporter. At a scan rate of 50 mV·s⁻¹, after initial assembly of the thiolated anchor DNA to the Au-SPE surface, the CV peak current intensities of [Fe(CN)₆]⁴⁻ decreased from I_pa ~ 85 μA and I_pc ~ −60 μA to I_pa ~ 67 μA and I_pc ~ −52 μA (Fig. 1b and S1). An increased peak-to-peak separation (ΔE) from ~0.13 V to ~0.16 V indicated an enhanced mass transfer limitation of [Fe(CN)₆]⁴⁻ at the surface of Au-SPE, due to the electrostatic repulsion between the negatively charged DNA and negatively charged [Fe(CN)₆]⁴⁻ redox reporter. Further blocking with 1-hexanethiol and then adding the complementary ligand DNA strands increased the mass transfer limitation of [Fe(CN)₆]⁴⁻, and consequently, decreased its peak intensities to I_pa ~ 49 μA, I_pc ~ −31 μA and I_pa ~ 28 μA, I_pc ~ −6 μA, respectively, with ΔE increased to ~0.19 V and ~0.41 V (Fig. 1b). The SWV results were also coherent with the CV data. After the coating of DNA and 1-hexanethiol, the peak current intensity of the [Fe(CN)₆]⁴⁻ signals (ΔI = I_peak – I_background) decreased step by step from ~480 μA to ~150 μA (Fig. 1c), indicating the successful fabrication of TGT sensors on the electrode.

As the softness of the surface can affect the force response of TGT [22,23], in addition to the above device, we also fabricated a type of “soft” TGT sensors by first casting 3-mercaptopropionic acid on the surface of the Au-SPE and then attaching poly-L-lysine, followed by the addition of DNA strands. As characterized using SWV, the intensity of the peak current first decreased from ~480 μA to ~200 μA after adding 5 μM of 3-mercaptopropionic acid (Fig. 1d), and then increased to ~275 μA upon attaching poly-L-lysine as the positively charged lysine could adsorb the negatively charged [Fe(CN)₆]⁴⁻. After further adding the anchor DNA and ligand DNA strands, the intensity of the peak current again decreased to ~160 μA and ~115 μA, due to the electrostatic repulsion between DNA and [Fe(CN)₆]⁴⁻.

3.2. TGT-based detection of cell adhesion forces

After the successful fabrication, we next applied both “rigid” and “soft” TGT sensors to detect cell-generated adhesion forces. We chose to study integrin αvβ3-mediated surface attachment of HeLa cells by modifying the corresponding ligand (B-cRGDfK) on the ligand DNA probe. After incubating ~1×10⁶ mL⁻¹ HeLa cells on the surface of either rigid or soft electrode for 90 min, the peak current intensity of the TGT sensors clearly increased in both cases (Fig. 2a, 2b, and S2a). The cellular response of rigid TGT sensor is much higher compared to the soft electrode (Fig. 2c), and thus will be used for the following studies. Meanwhile, a 75 min cellular incubation time will be applied later as the SWV signals barely changed after that. It is worth noting though that this current TGT sensor design is not quite suitable for real-time monitoring of cell adhesion forces, as a separate washing step is needed before detection.

Fig. 2.

(a) Square wave voltammetries of the “rigid” 12 pN TGT sensor before (dotted line) and after adding ~1×10⁵ HeLa cells for 15, 30, 45, 60, 75, and 90 min, respectively. (b) Square wave voltammetries of the “soft” TGT sensor before (dotted line) and after adding ~1×10⁵ HeLa cells for 15, 30, 45, 60, and 75 min, respectively. (c) Cell incubation time-dependent changes in the peak current values as measured using the rigid or soft TGT sensors. (d) Peak current changes of the rigid TGT sensor after adding from ~100 to ~1×10⁵ HeLa cells for 75 min. All these measurements were performed in a solution containing (v/v) 50% DMEM and 50% phosphate buffer (0.2 M, pH 7.4) and 5.0 mM [Fe(CN)₆]⁴⁻. Step potential, 20 mV; pulse amplitude, 50 mV; frequency, 20 Hz. Shown are the mean and standard error peak values after subtracting the background signals from four replicated tests.

We next studied the effect of HeLa cell concentrations on the TGT signals. After incubating ~100 to ~1×10⁵ HeLa cells on the rigid TGT sensors for 75 min, the SWV signals obviously increased at higher cell concentrations (Fig. 2d and S2b). By comparing these results with our previous design that use methylene blue-labeled 12 pN TGT sensors [17], the sensitivity of these new “signal-on” sensors was ~140 times higher than our first-generation “signal-off” force sensors (Fig. S3). It is worth noting that the small standard error values as measured from four replicated tests indicate the great repeatability of the sensors. To validate whether the integrin-RGD interactions are indeed responsible for the observed electrochemical signals, 1–5 μM of free cyclic arginine-glycine-aspartic acid (cRGD) was used to block the integrins on the HeLa cell membranes (Fig. 3a and S4a). As expected, after the cRGD treatment, much decreased peak current intensities were observed. Using 5 μM cRGD, the SWV signals of the TGT sensors were almost identical to that without adding the cells, suggesting that the integrin-RGD interactions are required for the activation of TGT signals.

Fig. 3.

(a) Square wave voltammetries of the 12 pN TGT sensor in the absence of HeLa cells (dotted line) or after adding ~1×10⁵ HeLa cells for 75 min. These cells have been pre-treated with 0, 1, 2.5, or 5 μM of cyclic arginine-glycine-aspartic acid (cRGD) for 75 min. (b) Square wave voltammetries of the 12 pN TGT sensor in the absence of HeLa cells (dotted line) or after adding ~1×10⁵ HeLa cells for 75 min. These cells have been pre-treated with 0, 5, 30, or 60 μM latrunculin B for 60 min. (c) Square wave voltammetries of the 56 pN TGT sensors before (dotted line) and after adding ~1×10⁵ HeLa cells for 15, 30, 45, and 60 min, respectively. (d) Peak current changes of the 56 pN TGT sensors after adding from ~100 to ~1×10⁵ HeLa cells for 60 min. All these measurements were performed in a solution containing (v/v) 50% DMEM and 50% phosphate buffer (0.2 M, pH 7.4)and 5.0 mM [Fe(CN)₆]⁴⁻. Step potential, 20 mV; pulse amplitude, 50 mV; frequency, 20 Hz. Shown are the mean and standard error peak values after subtracting the background signals from four replicated tests.

To further test if these electrochemical signals are indeed due to cell-generated forces, we treated HeLa cells with 5–60 μM of latrunculin B, a force-inhibiting drug that prevents actin polymerization and the transition of cellular G-actins into F-actins [24]. After treating with 5 μM or 30 μM of latrunculin B, much less DNA probes were ruptured from the Au-SPE surface and resulted in a significantly decreased peak current intensity from ~290 μA to ~180 μA (Fig. 3b and S4b). While interestingly, in the case of 60 μM latrunculin B treatment, the SWV signals were even lower than that in the absence of the cells. All these data suggested that the electrochemical signals have resulted from the integrin-RGD interactions by generating forces to rupture the DNA duplex.

We next wanted to study whether these electrochemical sensors can detect different levels of cell-generated forces by designing another “shear mode” TGT sensor [21] ruptured at forces ≥56 pN, termed “TGT₅₆”. TGT₅₆ shared the same sequence with the above-used 12 pN TGT, but the thiolated anchor group is located at the opposite end of the ligand strand. Our previous results suggested that integrin-generated forces during the HeLa cell attachment is not large enough to rupture these 56 pN TGT structures [17]. The TGT₅₆ results here are also consistent with our previous data. Indeed, after adding ~1×10⁵ HeLa cells on the surface of the TGT₅₆ sensor, the peak current is not increased, but instead, continuously decreased (Fig. 3c and S5a). Moreover, as the concentration of the HeLa cells increased, the SWV signals kept decreasing (Fig. 3d and S5b). Unlike 12 pN TGT sensors, TGT₅₆ exhibits a “signal-off” response, which is likely because without enough forces to rupture TGT₅₆, HeLa cells will just attach to the electrode surface through strong integrin-RGD binding. As a consequence, the mass transfer limitation of [Fe(CN)₆]⁴⁻ to the surface dramatically increased and resulted in a decreased electrochemical signal, similar to that shown in the case of 60 μM latrunculin B treatment (Fig. 3b and S4b).

3.3. CRISPR-Cas12a-incorporated TGT force sensors

To further improve the sensitivity of these TGT sensors, we have also tried to incorporate a CRISPR-Cas12a (Cpf1)/crRNA ribonucleoprotein complex to amplify the force-induced electrochemical signals. In above-designed TGT sensors, even after the removal of ligand DNA strands, the remaining anchor DNAs on the electrode surface can still partially repulse [Fe(CN)₆]⁴⁻. Inspired by some recent Cas12a/crRNA sensors [12,25,26], we want to test an idea to further reduce surface-attached anchor DNAs in the presence of cellular forces. Here, by inserting an activator sequence in the anchor DNA, the rupture of DNA duplex triggers a Cas12a/crRNA complex to recognize the activator sequence and then cleave the anchor DNA (Fig. 4a). We need to also mention that as the Cas12a/crRNA complex was added after incubating and washing away the cells, the kinetics and efficacy of cellular adhesion should not be influenced by the usage of such CRISPR-Cas system.

Fig. 4.

(a) Schematic of the Cas12a/crRNA-mediated signal amplification for the TGT force sensors. Upon experiencing cellular forces, the rupture of DNA duplex binds and activates the Cas12a/crRNA complex to cleave the surface-attached anchor DNA strands on the electrode. More [Fe(CN)₆]⁴⁻ can reach the surface of the electrode to generate electrochemical signals. (b) Peak current of the 12 pN TGT sensors after first incubating with ~1×10⁵ HeLa cells for 75 min, and then adding different concentrations of the Cas12a/crRNA complex for 30 min. (c) Peak current change of the 12 pN TGT sensor after adding from ~10 to ~1×10⁵ HeLa cells for 75 min, in the presence (blue line) or absence (orange line) of 90 nM Cas12a/crRNA complex. All these measurements were performed in a solution containing (v/v) 50% DMEM and 50% phosphate buffer (0.2 M, pH 7.4) and 5.0 mM [Fe(CN)₆]⁴⁻. Step potential, 20 mV; pulse amplitude, 50 mV; frequency, 20 Hz. Shown are the mean and standard error peak values after subtracting the background signals from four replicated tests.

Indeed, after adding increasing concentrations (from 0 to 90 nM) of Cas12a/crRNA, further enhanced SWV peak current intensities were observed upon incubating ~1×10⁵ HeLa cells on the above-mentioned 12 pN TGT sensors (Fig. 4b and S6a). In the presence of 90 nM Cas12a/crRNA, these TGT sensors are ~65% more sensitive than that without Cas12a/crRNA, i.e., ~230 times more sensitive (indicated by the slope of the detection curve) as compared to our previously developed methylene blue-labeled TGT sensors (Fig. 4c and S6b). Adhesion forces generated by ≤10 cells could now be reliably detected (Fig. 4c).

4. Conclusions

In summary, an advanced electrochemical DNA-based force sensor was fabricated in this project to detect cell-generated molecular tension forces. Using integrin-RGD interaction-mediated HeLa cell adhesion as an example, our results demonstrated that these novel electrochemical sensors could be used for highly sensitive and modular investigation of mechanicals forces and cellular adhesion events. The sensitivities of these DNA-based sensors could be further amplified by incorporating a crRNA design and CRISPR-Cas12a system. Compared to our previously demonstrated “first-generation” electrochemical DNA-based force sensors, these advanced sensors can provide much improved force detection sensitivity. Piconewton-scale forces from ≤10 cells can now be reliably detected, which was not possible using our previous sensors [17].

Moreover, with rapid response and cost-efficiency, these portable smartphone-based electrochemical devices can be straightforwardly applied in regular biological laboratories. It is worth noting that in most other methods used for detecting cellular forces, large and expensive instruments as well as skilled operators are often required. This limitation has significantly hindered their broad applications. In this case, we expect our DNA-based electrochemical sensors can function as a complementary tool to help a wide range of researchers to study cellular mechanosensing and mechanotransduction.

Highlights.

• A smartphone-based portable electrochemical sensor that measure cellular forces.

• Modular DNA-based sensor that detect adhesion forces at molecular level.

• Two mechanisms of signal amplification: label-free detection and CRISPR-Cas12a.

Data availability

Data will be made available on request.