Mechanotransduction of Magnetically Applied Forces Leads to Mechanical Stiffening of Platelet‐Rich Plugs

ABSTRACT

The mechanobiology of platelets governs early hemostasis, where their contractile forces compact and stiffen the nascent plug. However, the dynamic changes in the stiffness of a nascent platelet‐rich plug have been difficult to measure. Here, we introduce a microfluidic assay integrating microscale blocks with magnetic microposts to assess platelet forces and plug stiffness. Rotation of a multipole ring magnet beneath the microfluidic channel generates oscillatory motion in the magnetic microposts. As blood enters the channel and platelets aggregate around the blocks and microposts, the oscillatory motion is progressively dampened, providing a direct measure of plug stiffness. Additionally, the deflections of the microposts are used to quantify platelet‐generated forces during intermittent periods without magnetic actuation. With this approach, we find that the rate of stiffening and overall plug stiffness correlate not only with the magnitude of platelet forces, but also with the magnitude of the magnetically applied load. In contrast, pharmacological inhibition with acetylsalicylic acid or blebbistatin markedly reduced the process of plug stiffening, indicating that platelet activation and myosin‐based contractility are required. Together, these findings establish a novel microfluidic platform for probing platelet biomechanics under pathophysiological shear and demonstrate that platelet mechanotransduction governs the mechanical maturation of early hemostatic plugs.

A microfluidic device with magnetic microposts was used to assess platelet mechanobiology. The oscillatory actuation of posts with embedded iron particles was dampened by the dynamic stiffening of the platelet‐rich plugs. This stiffening process increased with applied load, but inhibiting platelet activation (ASA) or myosin contractility (blebbistatin) abolished the effect, which emphasizes the role of platelet mechanotransduction in primary hemostasis.

1. Introduction

Platelets play a central role in hemostasis by rapidly forming a hemostatic plug at sites of vascular injury [1, 2]. This plug initially consists of a loose aggregate of platelets that adhere to each other through molecular bridges formed by surface receptors and adhesive ligands [3]. Subsequent activation of the coagulation cascade drives the formation of a polymerized fibrin mesh that encases platelets and red blood cells together, transforming the nascent plug into a mature clot [4]. Beyond the well‐established biochemical pathways that drive the formation of a hemostatic plug, there is a growing need to understand the mechanobiology of platelets as they are able to sense and respond to physical cues within the developing plug and surrounding hemodynamic environment [5, 6, 7, 8, 9].

Platelet‐generated forces are pivotal for a hemostatic plug as they provide biomechanical changes that strengthen the structure. Upon activation, platelets reorganize their actin cytoskeletal filaments and activate myosin motor proteins to generate contractile forces that act upon the vessel wall or neighboring platelets in a nascent plug [5, 10, 11, 12, 13, 14, 15, 16, 17, 18]. The ensuing compaction of the plug not only reduces its drag forces but also increases the adhesion between aggregated platelets and confines soluble agonists within the intercellular spaces [19, 20, 21, 22, 23]. Disruption of myosin‐generated forces, whether through pharmacological inhibition or genetic defects, results in unstable plugs and impaired hemostasis, underscoring the essential contribution of platelet forces to thrombus integrity [24, 25].

The stiffness of the extracellular environment profoundly influences the force generation of platelets, ultimately shaping thrombus architecture and stability. Experimental evidence has found that platelets exert greater contractile forces and exhibit enhanced spreading on stiffer substrates, promoting integrin clustering and cytoskeletal remodeling through RhoA/ROCK and myosin [26, 27, 28, 29, 30]. Similarly, the evolving elasticity of a mature, fibrin‐based clot provides feedback that regulates platelet contractility, creating a dynamic interplay between platelet forces and fibrin network mechanics [31, 32, 33].

Platelets, long viewed primarily as circulating mediators of hemostasis, are now understood to be sensitive to the mechanical forces they encounter within the bloodstream [5, 6, 7, 8]. Shear gradients, rapid changes in flow, and tensile forces generated during thrombus formation all elicit biophysical responses that shape platelet adhesion, activation, and cytoskeletal organization [34, 35, 36, 37, 38, 39, 40]. Platelets respond to force through their adhesion receptors, which initiate intracellular signaling and aggregation [38, 39, 40, 41, 42, 43]. This mechanotransduction response reflects a broader principle seen across a diverse array of cell types that similarly rely on force‐dependent signaling through integrins, membrane receptors, and their actin network to interpret and react to their mechanical environment [44].

The mechanotransduction pathways that connect platelet forces, applied forces, and extracellular elasticity ensure that hemostatic plugs maintain sufficient structural integrity to resist shear forces. In general, soft clots fail to staunch the bleeding, while stiff ones are associated with thrombosis, stroke, or myocardial infarction [45, 46]. The elastic modulus of clots made from whole blood can range from 0.1 to 2 kPa, while those made from platelet‐rich plasma range from 0.6 to 11 kPa [46, 47]. While the stiffness of a clot can depend upon the concentration of red blood cells and fibrinogen, there is a strong contribution by platelets to clot stiffness [48, 49, 50]. Platelets have a cellular elasticity of approximately 10 kPa, and their contractile forces can increase the stiffness of a clot by stretching its fibrin strands to induce strain hardening [12, 18, 29, 31, 50, 51, 52, 53]. However, to the best of our knowledge, the stiffness of a nascent platelet‐rich plug is unknown.

Here, we present the development of an approach to measure the stiffness of a plug and the contractile force of platelets during shear‐induced aggregation. The approach builds upon a microfluidic assay that uses microscale blocks and microposts to measure the forces within a platelet‐rich plug that forms under shear flow [13]. Since magnetic material has been added to polydimethylsiloxane (PDMS) microstructures previously [54, 55, 56, 57, 58, 59], we embedded iron particles into the microposts and used them as magnetic actuators that apply external loads to aggregated platelets. With our assay, we find that the stiffness of a plug not only depends on the strength of platelet forces, but also the stiffening process is mediated by platelet activation and myosin‐based forces in a manner that is dependent on the strength of the magnetically applied load.

2. Results

2.1. Overview of Microfluidic Approach

The microfluidic device contained microchannels with an array of microscale blocks and posts made from polydimethylsiloxane (PDMS) with embedded iron particles (Figure 1a). Magnetic actuation of the posts was accomplished by rotating a multipole ring magnet that was positioned underneath the microfluidic channel. The microfluidic devices were made by a soft lithography process in which iron particles were concentrated in the upper portion of the microscale blocks and posts using an applied magnetic field (Figure 1b). The microchannels were 50 µm in height, 200 µm in width, and 3 cm in length. The blocks were 25 µm in length, 25 µm in height, and 15 µm in width. The posts were 25 µm in height, 6 µm in diameter, and positioned 9 µm behind each block. Due to the lack of space between the stage and objective of the microscope, a stepper motor and timing belt system was utilized to control the rotation of a ring magnet (Figure 1c). In this manner, the magnetic field acting on the embedded iron particles varied with the rotation θ of the ring magnet, thereby causing the post to deflect due to the change in the magnetic force acting on it (Figure 1d).

FIGURE 1.

Microfluidic assay to quantify mechanotransduction response of platelets during plug formation. a) Each microchannel contains an array of magnetic block and post sensors. A ring multipole ring magnetic is situated underneath the microchannels to magnetically actuate the microposts by its rotation. b) The manufacturing process involves multiple steps where 1) a negative mold is cast in PDMS from a SU‐8 master made through photolithography, 2) after silane treatment, a 50:50 mixture by weight of 800‐nm iron particles and PDMS is poured onto the mold, 3) the mold is centrifuged at 1000 × G for 3 min to drive the particles to the bottom of the holes and then a 10 mT magnet is used to consolidate the particles into larger groups, 4) excess Fe‐PDMS is removed with a glass slide, 5) the uncured PDMS is cast onto a plasma‐treated glass slide and by baking at 110°C for more than 15 h, and then 6) the negative mold is peeled away and a top layer of PDMS with an inlet and outlet is bonded onto the top of the microchannels to create a water‐tight seal. c) For the experimental setup, magnetic actuation occurred through the controlled movement of a dual‐axis, stepper‐motor‐driven‐belt system. The multipole ring magnet was placed on top of a 3D‐printed part on one of the axes of the belt system. The microfluidic device was held above the magnet. A microscope objective was used to image the microfluidic channel as a sample of blood collected in sodium citrate flowed from the inlet to outlet at a controlled flow rate regulated by a syringe pump (not shown). d) A rotating magnet underneath the microchannels rotates to different angles (θ) to impart motion in the magnetic microposts. Prior to blood flow, the microposts oscillate freely due to actuation from a magnet. The orange dashed outline denotes θ = 0 degrees, and the blue solid outline denotes the position of the micropost in each image frame. Scale bar, 10 µm. e) As blood flows into a channel, platelets aggregate between the block and micropost. The deflection of the micropost toward the block is caused the contraction forces of the platelets. Moreover, the magnetic actuation of the micropost is reduced, which is used to measure plug stiffness. The orange dashed outline denotes the deflected state at θ = 0 degrees. The blue solid outline denotes the position of the micropost due to magnetic actuation in each image frame. Platelets are pseudocolored in green. Scale bar, 10 µm.

When a sample of blood collected in sodium citrate was flowed through the microfluidic channel, the microscale blocks created a local gradient in the shear rate that triggered the formation of a platelet plug (Video S1). The surfaces of the blocks and posts were coated with rat tail collagen type I to support the attachment and activation of platelets. The contractile force (F) produced by platelets within the plug could be determined by optically tracking the deflection of a post (δ) and then by using Hooke's law:

$$F=k_{post}\delta$$ (1)

where k_post is the spring constant of a post (Figure 1d). Once a platelet plug formed, the oscillatory motion of the posts due to magnetic actuation was reduced due to the presence of a platelet‐rich plug between the block and post (Figure 1e).

2.2. Characterization of Magnetic Actuation

We developed a model to describe the movement of a magnetic post based on the rotation of the ring magnet. Specifically, we used an energy minimization model for a flexible, idealized cantilever to find the angle of deflection at the base of a post, which is an approach that has been used previously [55, 57, 60]. Here, the total energy of the system includes two parts: a) the elastic energy of the post based on its deflection and b) the magnetic energy of a cluster of iron particles due to the orientation of its magnetic dipole moment relative to the applied magnetic field. We considered the post to be an idealized cantilever of diameter D that contained an elastic section of length L_e at its base and a magnetic section of length L_m at its top that was subjected to an applied force F_app from the magnetic field (Figure 2a). We noted that the microscale blocks contained iron particles, but their magnetic deflections were considerably less than the posts due to their second moment of area (19,531 vs. 66 µm⁴).

FIGURE 2.

Characterization of magnetic post actuation depends on the amount of magnetic material and magnetic field. a) Under the influence of a magnetic field, a post deflects due to a magnetically applied force, which in turn is imparted to the platelet plug aggregated between it at the microscale block. b) A magnetic field is created with an axially oriented multipole magnet that has four quadrants, where adjacent quadrants have opposite dipole moments. c) To model the deflection δ and deflection angle ϕ of the post, we considered a torque balance equation that took into consideration the applied magnetic field |B| at a field angle ψ. d) The magnitude of the magnetic field |B| relative to the rotation angle of the magnet was measured using a gaussmeter at the position of the magnetic posts in the microfluidic channel. e) The field angle ψ was calculated from the vertical axis based on the measurements. f) Characteristic curves were plotted for the deflection δ of a magnetic post versus the rotation angle θ of the ring magnet based upon the length L_m of its magnetic section. g) Representative data illustrate that the deflection of a magnetic post can be used to fit a characteristic curve based on the model as a means to estimate the magnetic length, that is, L_m = 9.2 µm. h) Histogram of applied force for magnetic posts that were calibrated by their deflection. The mean applied force is 12.9 nN, with a standard deviation of 8.6 nN. The bin size is 3 nN. N = 51 posts.

The multipole ring magnet contained four quadrants where adjacent quadrants had dipoles with opposite orientations, which caused the magnetic field to curve across the interface between the quadrants (Figure 2b). We defined our coordinate system to be centered at the position where the microchannel was placed above the ring magnet and the x‐direction to be oriented along the length of the microchannel. Accordingly, the deflection of the magnetic post δ could be described by an angle ϕ in the xz‐plane due to the magnetic field $\vec{B}$ acting at an angle ψ (Figure 2c). To characterize the magnetic field, we rotated the ring magnet and measured its strength in three orthogonal directions at the position where a microchannel would be placed (Figure S1a). From these measurements, we calculated the strength of the field |B| (Figure 2d) and its angle ψ (Figure 2e).

The magnetic forces that act on an object like a magnetic post involve 1) a torque that rotates the magnetic post to align its dipole moment to an external magnetic field and 2) a force that pulls the magnetic post into the gradient of the magnetic field. Accordingly, we made the following assumptions: first, the contribution to the magnetic energy due to the field strength dominated over the contribution due to the field gradient (Figure S1b), and second, the field strength was below the level needed for saturating the iron particles [55, 60]. With these assumptions, the energy minimization model reduced to a balance between the elastic torque τ _e due to the bending stiffness of a post and the magnetic torque τ _m due to the applied field [57],

$${\tau }_e+{\tau }_m=0$$ (2)

Assuming a magnetic post bends at its base and has a constant radius of curvature, the elastic torque is given by

$${\tau }_e=\frac{\pi E{D}^4}{4L_e}\varphi$$ (3)

where E is the elastic modulus of PDMS (3 MPa). The magnetic torque acting on a post is given by

$${\tau }_m=\frac{V\left(n_r-n_a\right)}{2{\mu }_0{n}_r{n}_a}{\left|B\right|}^2\sin \left(2\left(\varphi -\psi \right)\right)$$ (4)

where V is the volume of iron particles in a magnetic post, n_r and n_a are the radial and axial demagnetization factors defined by treating the shape of a cluster of iron particles as a prolate ellipsoid (Figure S1c,d), and µ₀ is the permeability of free space [60]. By considering the initial concentration of iron particles and the random packing density of identical spheres (64%), we estimated the volume of iron particles V to account for 10% of the total volume of the magnetic section.

The length of the magnetic section L_m impacts the bending stiffness k_post and the magnetically applied force F_app . Since the walls of the microchannels obstructed our ability to obtain a sideview image of the posts, we estimated a value for L_m by analyzing the deflection of each post relative to the applied magnetic field. Specifically, we used our measurements for |B| and ψ to calculate the deflection of a post for different values of L_m (Figure 2f). With these characteristic curves, we could estimate L_m for each magnetic post based upon measurements of its deflection before blood entered the microchannel (Figure 2g). Using Castigliano's method, we derived the applied force F_app based on the deflection δ of a post by assuming F_app acts at the center of L_m and that any bending of the magnetic post occurs only in its elastic section of length L_e :

$$F_{app}=k_{post}\delta =\frac{3\pi E{D}^4}{16\left(4L^3-3L^2{L}_m-{L_m}^3\right)}\delta$$ (5)

where L = L_e + L_m . Using data we collected on the deflection of all magnetic posts in our study, we found that on average, the length of a magnetic section of a post was 7.1 ± 0.8 µm and thus, the average applied force was 12.9 ± 8.6 nN (Figure 2h). The applied force for a few magnetic posts was above 40 nN, which we attribute to a larger magnetic section, that is, L_m > 9.6 µm. For this reason, each magnetic post was characterized prior to flowing a blood sample into a microchannel.

2.3. Mechanical Analysis of Platelet‐Rich Plugs

For each test, a sample of whole blood was injected into the microfluidic channel at a shear rate of 16 000 s⁻¹ for 15 s to promote the rapid adhesion and aggregation of platelets between the microscale blocks and posts (Figure 3a). Before injecting a blood sample, the magnetic posts were actuated five times by rotating the ring magnet to determine L_m , k_{pos t}, and F_app , as mentioned earlier (Figure 3b). After blood entered the microchannel (denoted as time t = 0), we repeated a pattern of five magnetic actuations over a period of 40 s that was followed by a period of 20 s with no actuations. After 15 s, the flow was reduced to a shear rate of 500 s⁻¹ for the remainder of the test to reduce the attachment of additional platelets to the plug.

FIGURE 3.

The magnetic post measures platelet forces and plug stiffening simultaneously. a) Prior to flowing blood into the microchannel, (i) the post is magnetically actuated by an applied force of 40.3 nN while in Tyrode's buffer. Blood enters the microfluidic device at t = 0 at 16 000 s⁻¹ for 15 s to trigger platelet aggregation upon the block and post. The flow of blood is reduced to 500 s⁻¹ for the remaining of the experiment. The plug contracts due to platelet‐generated forces, which causes the post to deflect, as seen at (ii) t = 120 s. Over time, the platelet plug continues to contract, as seen at (iii) t = 300 s. The outline of the plug is a solid green line. The outline of the undeflected position of the post is an orange dashed line. The outline of the deflected position of the post is a blue solid line. Scale bar, 10 µm. b) The centroid of the post in comparison to the block is tracked over time. The dashed line depicts the entry of blood into the microchannel. The displacement of the post reports both the deflection of the post and the oscillatory movement that corresponds with periods of magnet actuation. c) The amplitude of the oscillatory motion of the post during periods of magnetic actuation was isolated from the data. The initial amplitude prior to blood entry was used to calculate the stiffness of the post and the magnitude of the magnetically applied force. After blood enters, the change in amplitude over time illustrates the stiffening of the plug as it develops. d) The change in amplitude between the first period of magnetic actuation and subsequent periods of actuation for individual plugs was used to calculate e) the average plug stiffness. f) The time‐averaged displacement data for individual plugs was used to calculate g) the average magnitude of platelet force. h) Projected area of the plugs was determined from manual tracing their outline to calculate the i) average projected area. Data points and error bars denote the mean and standard error of the mean for data from nine blood donors. ANOVA revealed that values between t = 60 and 300 s are not significant (n.s.). Data represent n = 27 platelet‐rich plugs.

In this example, the displacement of the post increased steadily due to the contraction of platelets within the plug (Figure 3b). We separated the amplitude of the oscillatory motion of a post during periods of magnetic actuation from its time‐average position (Figure 3c). During the second period of magnetic actuation, which is after blood enters the microchannel, we observed that the movement of the post decayed sharply. For subsequent periods of magnetic actuation, the movement was small but detectable, which we attributed to the stiffness of the platelet‐rich plug.

We could determine plug stiffness by comparing the oscillatory movement of a post during a period of magnetic actuation to the period before blood entered the microchannel (t < 0). Specifically, we assumed the spring stiffness of a plug (k_plug ) and the bending stiffness of a post (k_post ) acted as two springs in series, thus

$${\mathrm{k}}_{\mathrm{plug}}={\mathrm{k}}_{\mathrm{post}}\frac{{\mathrm{a}}_{\mathrm{i}}}{{\mathrm{a}}_{\mathrm{t}}}-{\mathrm{k}}_{\mathrm{post}}$$ (6)

where a_i is the initial amplitude prior to blood entering the channel and a_t is the amplitude during a period of magnetic actuation. From this relationship, we defined plug stiffness as

$${\mathrm{E}}_{\mathrm{plug}}={\mathrm{k}}_{\mathrm{plug}}\frac{{\mathrm{L}}_{\mathrm{eff}}}{\mathrm{A}}$$ (7)

where L_eff is the effective length of a plug between the block and post and A is its effective cross‐sectional area, which we estimated to be 80 µm². At each time point in our analysis, we accounted for the reduced distance in L_eff between the block and post due to the steady contraction of the plug.

For this example, we found that plug stiffness increased steadily over time (Figure 3d). For all of the blood samples we tested, we found that plug stiffness reached an average value of 4.7 kPa over a period between 60 and 300 s (Figure 3e). We note that this measurement of plug stiffness is greater than the stiffness of gels made from platelet‐rich plasma [52] and less than the stiffness of individual platelets [31] (Table S1). Therefore, we consider these results and the assumptions in the model of magnetic actuation to be reasonable.

To quantify platelet forces, we analyzed the deflection of the magnetic post for periods without magnetic actuation, for example, t = 60 to 80 s. In accordance with Hooke's law, the deflection was multiplied by the bending stiffness k_post to quantify the contractile force of platelets within the plug (Figure 3f). For this test, we observed that the force of the platelets within the plug was initially zero prior to the first period of magnetic actuation and then increased steadily over time. From all blood samples we tested, we found that platelet plugs measured within our device produced 64.8 nN of force on average for a period between 60 and 300 s (Figure 3g), which is similar to our previous results [13, 14].

Additionally, we measured the projected area of each plug and tracked its size over time (Figure 3h). For this example, the plug increased in area due to aggregation under flow and then decreased in size due to detachment of platelets that were loosely bound to the plug. From all blood samples we tested, we found that platelet plugs reached an average area of 916 µm² over a period from 60 to 300 s (Figure 3i). In summary, we demonstrate that our microfluidic approach using magnetic posts can provide dynamic measurements of the stiffness of a plug along with the strength of its contractile forces and its projected area due to platelet aggregation.

2.4. Stiffening Behavior of Platelet‐Rich Plug Depends on Magnetically Applied Load

Platelets sense and respond to force, so we used our microfluidic device to examine whether magnetically applied forces can influence the biomechanical stiffening of a plug. As mentioned earlier, we noted that during the initial formation of a platelet‐rich plug, there was a sharp decay in the oscillatory movement of the magnetic post, for example, t = 20 to 60 s. As a point of comparison, we observed that when a magnetic force of 40.3 nN was applied (Figure 4a), the decay was more rapid than when a lesser force of 11.8 nN was applied (Figure 4b). We attributed the decay in oscillatory movement to indicate a dynamic change in plug stiffness. Consequently, we fitted an exponentially damped sinusoidal function to the data,

$$a{e}^{-bt}\sin \left(0.25\pi t+c\right)+d$$ (8)

where a, b, c, and d are fitting parameters for amplitude, stiffening rate, phase, and offset, respectively (Figure S2). When we compared the stiffening rate to the strength of the magnetically applied force (Figure 4c), we noted a strong correlation in the data (R² = 0.87). Thus, these results suggest that magnetically applied forces influence the stiffening process within a plug.

FIGURE 4.

Stiffening rate and plug stiffness depend on the applied load. For an applied force of (a) 40.3 nN or (b) 11.8 nN, the dampened oscillatory movement of the magnetic post as the platelets aggregated onto the block and post provided an indication of the stiffening rate, which was determined by fitting a dampened sine wave to each data trace as indicated by a dotted orange line. (c) The stiffening rate versus applied force has a strong positive correlation (R² = 0.87). Similarly, for an applied force of (d) 40.3 nN or (e) 11.8 nN, the amplitude of the oscillatory motion of the posts during the final period of magnetic actuation provides an indication of plug stiffness. (f) Plug stiffness versus applied force has a strong positive correlation (R² = 0.96). Data shown are 27 platelet‐rich plugs from 9 blood donors. The solid orange line represents the linear fit, and the dotted orange line indicates the 95% confidence interval.

To further examine the effect of applied force on plug stiffness, we compared the oscillatory motion of the posts during the last period of actuation, for example, t = 260 to 300 s. Using the same comparison of data, we observed that when a magnetic force of 40.3 nN was applied (Figure 4d), the amplitude of the oscillatory movement was smaller than when a lesser force of 11.8 nN was applied (Figure 4e). Consequently, we compared our measurements of plug stiffness to the applied load for all data and found a high level of correlation (R² = 0.96) (Figure 4f). Taken together, the strength of an applied force has a strong effect on both the initial stiffening rate and the steady‐state stiffness of a platelet‐rich plug. Moreover, the data indicate that there are force‐induced responses in platelets, that is, mechanotransduction, that lead to biomechanical changes in the stiffness of the plug.

2.5. Plug Stiffening Requires Platelet Activation and Myosin‐Based Contraction

Platelets contribute to the stiffening of a clot by generating cytoskeletal forces, so we sought to examine this relationship within platelet‐rich plugs using our microfluidic device. Our analysis revealed a strong correlation between plug stiffness and platelet forces (R² = 0.89) (Figure 5a), affirming that platelet contractility is a critical component in the stiffening of a plug. In contrast, we found a moderate correlation between applied forces and platelet forces (R² = 0.26, Figure S3). Consequently, we performed a multivariate analysis to examine the interrelationships between the three factors: applied forces versus platelet forces versus plug stiffness. We found that the magnitude of an applied force has a positive effect on plug stiffness that was significantly greater than the effect of platelet forces (Figure 5b). Thus, this analysis suggests that externally applied forces contribute to a rise in platelet forces, which subsequently impacts the stiffness of a plug.

FIGURE 5.

Plug stiffness due to applied force depends on platelet activation and myosin activity. (a) Plug stiffness versus platelet forces has a strong positive correlation (R²= 0.89). The line represents the linear fit, and the dotted line indicates the 95% confidence interval. Data shown are from control samples taken from nine donors. (b) A surface fit of the control data indicates a strong correlation of both applied force and platelet force with plug stiffness (R² = 0.94). Using pharmacological inhibitors of platelet activation (0.3 mM acetylsalicylic acid, ASA) and myosin activity (5 µM blebbistatin, BLEB), (c) plug stiffness and (d) platelet forces were markedly diminished as compared to paired control samples. Asterisks denote statistically significant differences (* p < 0.05) using ANOVA followed by Tukey post‐hoc test. The box plot indicates the first and third quartile, the center line is the median, and the whiskers extend from the first or third quartile to the last data point. Data shown are the average per donor for 6 donors (13 platelet‐rich plugs) for CTRL, 3 donors (8 plugs) for ASA (blue), and 3 donors (9 plugs) for BLEB (magenta). (e) Stiffening rate versus applied force for ASA and BLEB has a negative and weaker correlation (R² = −0.03 and 0.26, respectively) as compared to control data (dashed black line from linear fit in Figure 4c). (f) Plug stiffness versus applied force has a positive but modest correlation (R² = 0.41 and 0.73, respectively) as compared to the control data (dashed black line from linear fit in Figure 4f).

To directly examine if platelet activation and platelet forces impact plug stiffness, we treated whole blood samples with acetylsalicylic acid (ASA) to inhibit the generation of thromboxane A2 or blebbistatin to impair myosin activity. Blood samples treated with 0.3 mM ASA or 5 µM blebbistatin produced platelet plugs with platelet forces that were significantly lower than those for untreated blood samples (p‐value = 0.022 for ASA and 0.043 for blebbistatin) (Figure 5c). Moreover, blood samples treated with ASA or blebbistatin produced platelet plugs that were significantly lower in plug stiffness than those from untreated blood samples (p‐value = 0.038 for ASA and 0.051 for blebbistatin) (Figure 5d). For both conditions with inhibitors, we found that the initial stiffening rate not only had a negative relationship with the magnetically applied forces, but there was a weak correlation in the data (R² = −0.03 for ASA and 0.26 for blebbistatin) (Figure 5e). In contrast, we observed a positive relationship between plug stiffness and applied force, as seen previously with control data, but the slopes of the best fit lines were lower, and the correlations in the data were moderate (R² = 0.41 for ASA and 0.73 for blebbistatin) (Figure 5f). Thus, these results indicate that platelet activation and myosin‐based forces are necessary for the mechanotransduction response to applied forces that affect plug stiffening.

3. Discussion

We developed a microfluidic assay that utilizes an array of magnetic block and post structures to assess the mechanobiology of platelets during the early formation of platelet‐rich plugs. Through the application of magnetically applied forces, we observed that platelets exhibited a mechanically activated contractile response, in which the extent of external loading triggered increased platelet‐generated tension that stiffened the plug. Inhibition of myosin activity and thromboxane‐mediated activation attenuated the response, demonstrating that the stiffening process depends on intact biochemical signaling pathways. The proportional increase in plug stiffness with applied force may reflect the extent of mechanosensory structures, like integrins or PIEZO1 ion channels, that could be activated during magnetic actuation [8, 9]. However, the specific molecular mechanisms underlying platelet mechanotransduction in response to magnetically applied forces remain to be determined.

This assay takes advantage of magnetic actuation principles that rely on the use of soft magnetic materials and the rotation of a ring magnet to induce a magnetic torque. We chose to embed iron particles into the micropost, as we and others have done previously [59, 61], rather than cobalt nanowires, magnetite nanoparticles, or nickel shells [55, 56, 57, 62, 63]. We integrated this deposition method into the soft lithography of our microfluidic assay and found it was able to produce a robust number of magnetic posts. In contrast to these previous studies, we chose to rotate a multipole ring magnet underneath the assay because it applied a local field in a compact design, without the need for electromagnetic coils [54]. As modeled previously by Evans et al. [55], our posts behaved similarly to a flexible magnetic rod whose motion is dominated by the strength of a magnetic field and not the field gradient due to the distance between our microfluidic device and the magnet underneath it. Batch‑to‑batch variation in particle loading resulted in modest differences in post magnetization, but this variability proved advantageous by enabling examination of plug stiffening across a range of applied forces.

Our measurements of plug stiffness are on the order of 10 kPa, which corresponds with prior measurements for single platelet assays and tissue‐level elastometry [31, 40, 64, 65]. These measurements are 10‐fold stiffer than platelet‐rich fibrin clots and 100‐fold stiffer than platelet‐free fibrin clots [51, 53], which indicates the contribution due to the contractile forces of platelets. We found that platelet‐rich plugs generate forces in the range of tens to hundreds of nanonewtons, similar to our prior measurements using a nonmagnetic version of this assay [13, 14]. These measurements fall within the range of tens of nanonewtons produced by single platelets and hundreds of newtons by platelet aggregates [11, 12, 14, 15, 16, 17, 18]. Similar to Chen et al. [40], we found that platelet forces and plug stiffness have a strong correlation and that the inhibition of platelet activation and myosin‐generated forces abrogates this response. However, our approach can simultaneously measure platelet forces and plug stiffness during shear‐induced aggregation.

Our results indicate that platelet contraction is vital for a plug to stiffen in response to dynamic loading. However, platelets respond directly to stiffer environments, around the same order of magnitude as our plugs, with enhanced adhesion, activation, and contraction [26, 27, 28, 29, 30]. We measured plug stiffness to be 1.5 kPa on average during the first period of actuation, but the amount of platelet forces for that period was negligibly small. We observed over time that platelet contractility and plug stiffening evolve simultaneously. As such, we cannot determine from the data whether platelets increase their contractile forces in response to a progressively stiffer plug environment, or whether the plug becomes stiffer as a direct consequence of platelet force generation. It is likely that both processes occur concurrently, making their individual contributions difficult to isolate.

One conjecture is that an increase in plug stiffness could arise from enhanced platelet‐to‐platelet adhesion if the oscillating magnetic post compresses adjacent platelets and drives their membranes into closer contact, thereby modestly strengthening the cohesion between platelets within the plug. However, the results with ASA and blebbistatin show that this passive mechanism is not the primary driver. For both inhibitors, the rate of stiffening exhibited a small, negative relationship with applied force, indicating that platelet‐generated forces are needed to accelerate the early phase of plug stiffening. The steady‐state value of plug stiffness with these inhibitors showed a small, positive relationship with applied force, which suggests that externally applied compression may marginally enhance plug stiffness through increased platelet cohesion when contractility is impaired. Taken together, these findings demonstrate that the dominant source of plug stiffening is the active, myosin‐driven forces generated by the platelets themselves.

While this study represents an important first step in measuring platelet mechanobiology in real time, several limitations and future directions merit consideration. The surfaces of the microfluidic device were functionalized with collagen type I to capture and activate platelets, but future studies should explore the incorporation of other extracellular matrix proteins, either individually or in combination, to more closely recapitulate the complex composition of the subendothelial matrix and its differential effects on platelet activation [3, 8, 9]. Additionally, the current experiments were performed using blood samples collected in sodium citrate, which limits the generation of thrombin and the formation of fibrin that are pertinent to secondary hemostasis. Future iterations of this platform could incorporate tissue factor and/or the use of recalcified blood to capture the full continuum from primary to secondary hemostasis, thereby providing a more complete picture of the mechanical evolution of a hemostatic plug.

4. Conclusion

We introduce a novel microfluidic approach that encloses an array of magnetic block‐post pairs to assess the mechanobiology of platelet‐rich plugs during shear‐induced aggregation. Each magnetic post acts as a force transducer and mechanical actuator: its deflection is used to measure platelet forces and its movement in response to a magnetic field is used to assess plug stiffness. We demonstrate that plug stiffness is dependent on myosin‐based platelet forces, specifically in response to the applied force from these posts. By design, this assay can further investigate the mechanobiology of platelet‐rich plugs in the context of new pharmaceutical drugs to assess their ability to affect hemostasis or thrombosis.

5. Experimental Section

5.1. Microfluidic Device Fabrication

Soft lithography was used to create the microfluidic device from a silicon master by a two‐step casting approach. The silicon master was made from SU‐8 photoresist. For the soft lithography process, a negative mold was cast using PDMS (Sylgard 184, Dow Corning). The surface of the negative mold was exposed to plasma (Plasma Prep II, Structure Probe, Inc.) and then treated with the vapor phase of (tridecafluoro‐1,1,2,2‐tetrahydrooctyl)‐1‐trichlorosilane (T2492‐KG, United Chemical Technologies) while under vacuum pressure in a desiccator for 2 h to passivate the surface of the negative mold and prevent bonding to uncured PDMS in subsequent steps of the process. Next, iron particles with a diameter of 800 nm (US Research Nanomaterials, Inc) were mixed into uncured PDMS at a 50:50 ratio by weight and then degassed. The Fe‐PDMS mixture was then poured into negative molds and centrifuged for 3 min at 1000 RCF to drive the iron particles to the bottom of the mold so they would be concentrated at the ends of the microscale posts. The molds were placed above a 10 mT magnet for 1 min to cluster the iron particles together. The excess amount of the Fe‐PDMS mixture on the top of the negative mold was scraped away with a glass slide to increase the optical transparency of the final devices for imaging under a microscope. Uncured PDMS without iron particles was added onto the mold, which was then inverted and placed onto a clean glass slide and cured at 110°C for more than 15 h. After curing, the mold was peeled away to produce the base layer of the microfluidic device, which consisted of open‐topped, microfluidic channels with an array of magnetic block and post structures. A top layer of PDMS was made using a custom‐built aluminum mold with pegs for the placement of silicon tubing (0.040‐inch ID/0.085‐inch OD, HelixMark) that formed the inlet and outlet ports of the microfluidic channels. The top and base layers of the microfluidic device were exposed to plasma (Plasma Prep II, Structure Probe, Inc.) for 10 s and then brought into contact to create a permanent seal between them.

5.2. Surface Coatings

The inner surfaces of the sealed microfluidic channels were coated with rat tail collagen type I (BD Bioscience) by injecting it as a solution at concentration of 200 µg mL⁻¹ in 0.1 M acetic acid into each outlet port and letting it adsorb for 1 h. Afterwards, the microfluidic channels were rinsed by injecting Tyrode's buffer (10 mM 4‐(2‐hydroxyethyl)‐1‐piperazineethanesulfonic acid (HEPES), 138 mM NaCl, 5.5 mM glucose, 12 mM NaHCO₃, 0.36 mM Na₂HPO₄, 2.8 mM KCl, 0.4 mM MgCl₂) into the ports. The devices were stored at 4°C with Tyrode's buffer in the microfluidic channels until the time of the experiments.

5.3. Characterization of Magnetic Field

To measure the strength of the magnetic field, a multipole ring magnet (Torque Polymagnet #1001122, Correlated Magnetics) was attached to a servo motor (Futaba). As the ring magnet was rotated, the strength of the magnetic field in each orthogonal direction was measured using a Hall effect gaussmeter (Model 410, Lake Shore Cryotonics). The probe tip of the gaussmeter was positioned at a 1 mm distance above the magnet and 8 mm distance from its center for measurements of the magnetic field, which corresponded to the location of the microfluidic channel during experiments.

5.4. Characterization of Magnetic Actuation

To characterize the deflection of the magnetic posts due to the change in magnetic field, a microfluidic device was placed above the rotating magnet at the same position where the measurements of the magnetic field were taken. Using a metallurgical microscope (Nikon LV‐100) and a 20× objective, a video was taken at a frame rate of 2 Hz by a camera (Hamamatsu C11440) to record the deflection of the magnetic posts at different rotation angles of the ring magnet. To actuate the magnetic posts during experiments, the ring magnet was mounted on a rotating axle that was situated underneath the microfluidic device and driven by a stepper motor (#1205, Pololu) and a custom‐built pulley system. The rotation of the ring magnet by the stepper motor was controlled at 1/1600 steps and 0.125 Hz by a microprocessor (Arduino Uno v3) with a motor driver (A4988 Allegro). The ring magnet was swept between −10 and 10 degrees of rotation for five times during a period of 40 s, followed by a period of 20 s without any rotations. This pattern was performed six times during the duration of an experiment. The period when the magnet was rotated was when the stiffness of the plug was assessed, while the contractile force of the platelets was measured when the magnet was not rotated.

5.5. Human Blood Collection

Whole blood up was collected from human subjects in sodium citrate tubes (0.109 M/3.2%, 2.7 mL, BD Vacutainer) under a protocol approved by the Institutional Review Board (IRB) at the University of Washington. A blood sample up to 50 mL was collected in multiple tubes and then transferred to a 50‐mL conical centrifuge tube (Falcon Corning). Each blood sample was mixed on a slow rocking agitator for at least ninety minutes prior to testing. To inhibit platelets, acetylsalicylic acid (0.3 mM) or blebbistatin (5 µM, Sigma‐Aldrich) was added to a blood sample and incubated for 10 min prior to testing. In these inhibition experiments, an equivalent volume of either 10 mM HEPES and 2 mM CaCl₂ or dimethyl sulfoxide (DMSO) was added.

5.6. Microfluidic Testing Protocol

Blood was drawn into a 3‐mL syringe (BD Scientific) and then injected into the microfluidic channel using a syringe pump (Harvard Apparatus). We used flow parameters that had been previously shown to produce the rapid formation of platelet plugs between the microscale blocks and posts [13, 14]. Specifically, a blood sample was flowed through a microchannel for a wall shear rate of 16 000 s⁻¹ for the first 15 s to induce the adhesion and aggregation of platelets onto the blocks and posts. Subsequently, the flow rate was reduced to produce a wall shear rate was 500 s⁻¹ for the remaining time of the experiment to reduce the rate of aggregation and avoid occluding the microchannel with platelets.

5.7. Measurement of Platelet Forces, Plug Stiffness, and Plug Area

Using a metallurgical microscope (Nikon LV‐100) encased in a live‐cell chamber (In Vivo Scientific, Inc) to maintain a 37°C environment, the experiments with blood were recorded by a camera (Hamamatsu C11440) using a 20× objective. Video recordings were initiated 30 s prior to when blood entered the channel and continued until the end of the test. There were eight pairs of microscale blocks and posts within the field of view of the camera. After a full video was taken, the distance between the centroid of each post and the edge of its corresponding block was analyzed using a thresholding algorithm in Fiji/ImageJ. This distance was tracked over time and used to either measure the movement of the posts when the ring magnet was rotated or deflection of the posts when the ring magnet was not rotating. There was no flow during the first period when the ring magnet was rotated, so the data from this period was used to calibrate the magnetic actuation of each post. In addition, the projected area of the plug was measurement using manual tracing in Fiji/ImageJ to assess the size of the platelet plug.

5.8. Statistical Analysis

One‐way analysis of variance (ANOVA) with Tukey's post‐hoc test was used to determine the significance between multiple groups, and significance was denoted for p < 0.05. Linear regression analysis and model fitting were performed in MATLAB (Mathworks) using the Curve Fitter toolbox and the least absolute residual option.

Author Contributions

Yoeur‐Man T. Mach: writing – review and editing, data curation, investigation. Nathan J. Sniadecki: writing – review and editing, conceptualization, investigation, funding acquisition, writing – original draft, supervision, data curation, resources, project administration, formal analysis, visualization, methodology. Nakul Sridhar: methodology, writing – review and editing, validation. Nikita Taparia: investigation, writing – original draft, methodology, validation, visualization, writing – review and editing, software, formal analysis, data curation, supervision. Ava M. Obenaus: writing – review and editing, visualization.

Conflicts of Interest

N.J.S. is a founder and has equity in Stasys Medical Corporation.

Supporting information

Supporting File 1: adhm71559‐sup‐0001‐SuppMat.pdf.

Supporting File 2: adhm71559‐sup‐0002‐VideoS1.mp4.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.