The rapid change of shear rate gradient is beneficial to platelet activation

Abstract

Fluid shear plays a key role in hemostasis and thrombosis, and the purpose of this study was to investigate the effect of shear gradient change rate (SGCR) on platelet reactivity and von Willebrand factor (vWF) activity and its mechanism. In this study, we developed a set of microfluidic chips capable of generating different shear gradients and simulated the shear rate distribution in the flow field by COMSOL Multiphysics software. Molecular markers of platelet activation (P-selectin, activated GPIIb/IIIa, phosphatidylserine exposure, and monocyte-platelet aggregate formation) were analyzed by flow cytometry. Platelet aggregation induced by shear gradient was studied by a microfluidic experimental platform, and plasma vWF ristocetin cofactor (vWF: RCO) activity was investigated by flow cytometry. The expression of p-Akt was studied by Western blotting. The results showed that the faster the SGCR, the higher the expression of platelet p-Akt, and the stronger the platelet reactivity and vWF activity. This indicates that fluid shear stress can activate platelets and vWF in a shear gradient-dependent manner through the PI3K/AKT signal pathway, and the faster the SGCR, the more significant the activation effect.

Plain Language Summary

What is the context?

• Recent studies have shown that fluid shear stress plays a key role in platelet activation and thrombosis. However, its mechanism and effect have not been fully elucidated.

• The development of microfluidic chip technology enables people to study platelet function in a precisely controlled flow field environment.

• Previous studies have shown that the PI3K-AKT signal pathway may be a mechanically sensitive signal transduction pathway.

What is new?

• In this study, we designed a microfluidic model with different narrow geometry, and controlled the injection pump to perfuse fluid at the same flow rate, so that the platelets flowing through the model experienced the flow field environment of different shear gradients.

• We studied the activities of platelets and von Willebrand factor in different flow fields and explored their signal transduction pathways.

What is the impact?

• Our results suggest that vascular stenosis does increase platelet activity and the risk of thrombosis. However, its ability to activate platelets is not only related to the peak shear rate and shear time, but also closely related to the decreasing rate of shear gradient. Even if the peak shear rate at the stenosis is the same, the faster the shear rate decreases, the higher the reactivity of platelets and von Willebrand factor, which may be mediated by the PI3K-AKT signal pathway. This study not only helps clinicians to judge the risk of thrombosis in patients with atherosclerosis or percutaneous coronary intervention, but also helps us to better understand the mechanism of shear-induced platelet activation.

Keywords:

• Cardiovascular disease
• microfluidic chip
• PI3K/AKT signal pathway
• platelet reactivity
• shear gradient change rate
• von Willebrand factor activity

Introduction

Platelet adhesion and aggregation are the basis of normal hemostasis and pathological thrombosis. Studies have shown that platelet activation and aggregation occur through the synergistic signaling of exposed subendothelial matrix proteins (vascular factors), biochemical activators [thrombin, Adenosine diphosphate(ADP), and Thromboxane A₂(TXA₂)] and hemodynamics.¹ Although the role of extracellular matrix and chemical factors has been widely studied, the key hemodynamic parameters and their effects on platelet reactivity and plasma proteins have not been fully elucidated. Areas with limited blood flow, stenosis, or mural thrombosis have long been thought to produce local shear force gradients.^2,3 When local vessel geometry is formed due to vessel wall damage or atherosclerotic plaque, the platelets flowing through it undergo complex physiological or pathological changes, including intracellular calcium influx signals, surface integrin exposure, and cytoskeleton reorganization and contraction.⁴ Studies have shown that the shear gradient produced by the stenotic part can activate platelet GPIb and bind to the vWF fixed on the collagen surface, promoting thrombosis downstream of the stenosis.^5–7 Under the action of fluid shear force, the morphology of vWF is stretched and the A1 domain is exposed.⁸ The binding of the platelet membrane GPIb receptor to vWF initiates downstream events.⁹

The development of microfluidic technology enables people to study the interaction between cells and protein molecules in a precisely controlled flow field. More and more studies focus on the effect of shear gradient on platelet and vWF activity and its mechanism.^10,11 Published fluid simulations have shown that plaque geometry can produce different local flow pressures, velocities, and shear gradients.¹² Recent evidence suggests that PI3K/Akt signaling pathway plays a key role in platelet activation induced by non-physiological shear stress.^13,14 However, what is not clear is the effect of shear gradient change rate (SGCR) produced by different geometric structures on platelet activity and its mechanism.

To study this problem, we have designed and fabricated microfluidic chips with different narrow angles to generate flow fields with different SGCRs. Through the microfluidic analysis platform, we observed the morphology of platelet aggregation in different flow fields. We studied the effects of different gradient flow fields on platelet reactivity (P-selectin, activated GPIIb/IIIa, and monocyte-platelet aggregation formation) and vWF activity by flow cytometry. We quantified the expression of p-Akt by immunoblotting. Our data show that SGCR can activate platelets and vWF through the PI3K/AKT signal pathway, and the faster the SGCR, the higher the platelet reactivity and vWF activity.

Materials and methods

Materials

Bovine serum albumin (ORTHO, USA), Adenosine diphosphate (TOP SCIENCE, USA), ristocetin (Caymen, USA), collagen fiber (Shandong Telikangxin Medical Technology Co., Ltd.), Calcein AM (Invitrogen, USA), CD41a Monoclonal Antibody (HIP8), FITC/CD61 (Integrin beta 3) Monoclonal Antibody (VI-PL2), PerCP-eFluor 710/CD62P (P-Selectin) Monoclonal Antibody (AK-4), PE/PAC-1 Monoclonal Antibody (PAC-1), FITC/CD14 Monoclonal Antibody (ebioscience, USA), Alprostadil (TOP SCIENCE, USA), PBS (Gibco, USA), paraformaldehyde (Biyuntian Biotechnology Company), erythrocyte lysate (BD Biosciences, USA), microscope slides (Jiangsu Shitai Laboratory Equipment Co., Ltd.), anti-phospho-Akt (Ser473) and anti-GAPDH antibodies (Cell Signaling Technology, USA),TGX-221[PI3K Signaling pathway inhibitors (MedChemExpress, USA)], Sylgard 184 polydimethylsiloxane (Dow Corning, USA)

IX71 inverted fluorescence microscope (Olympus, Japan), plasma cleaner (PDG- 32 G-2, Germany), thermostatic incubator, BC6800 automatic blood cell analyzer (mindray, China), vacuum venous collection tube (Shandong Weigao Technology Co., Ltd.), RSP01-CS two-way push-pull precision injection pump (Jiashan Ruichuang), COMSOL Multiphysics 5.4 software (COMSOL, Sweden), Coreldraw software (Corel, Canada), streampix 5.0 software (Norpix, Canada), centrifuge (China Pingfan Technology Co., Ltd.), ImageJ software (NIH, USA), 2 ml syringe (Shandong Weihai Technology Co., Ltd.), GraphPad Prism (GraphPad Software, USA), ChemiDoc XRS+ (BIO-RAD, USA), flow cytometer (Beckman Coulter, USA)

Processing and fabrication of microfluidic chips

The fabrication method of the microfluidic chip is the same as before.^15,16 There are three types of microfluidic chips used in this study, and their contraction and expansion angles are 30°, 60°, and 90°, respectively. The chips mainly consist of a sample cell, microchannels, and outlets. The length, width, and height of the microchannel are 7 mm, 700 um, and 70 um, respectively, and the diameters of the sample cell and outlet are 7 mm and 1.5 mm, respectively. There is a narrow area in the center of the microchannel with a narrowing degree of 80% (based on the research of Nesbitt et al.¹⁷) and a narrowing area length of 0.5 mm (Figure S1A). The operating schematic of the analysis system is shown in Figure S1b. In short, the injection pump is set as the suction mode to make the blood sample flow in the microchannel with a set shear rate, and the platelet aggregation image is recorded in real time through a fluorescence microscope or a phase contrast microscope. A physical diagram of the analysis system is shown in Figure S1c. According to Poiseuille’s law, the relationship between the input shear rate and the velocity of the injection pump can be expressed as: γ = 6Q/a²b. where γ (s⁻¹) denotes the shear rate, Q (ul/min) denotes the flow rate, a (mm) represents the microchannel depth, and b(mm) represents the microchannel width.

Simulation analysis of the flow field in the microchannel

It is reported that the average wall shear rate of the artery is about 1500 s⁻¹, and we set the flow rate of the injection pump to 52 ul/min to make the shear rate of the input wall of the microfluidic channel reach 1500 s^−1.18 The microchannel structure was designed by COMSOL Multiphysics software and computational fluid dynamics (CFD) analysis was performed on the wall shear rate and velocity distribution within the microchannel. It is assumed that the density and viscosity of whole blood are 998.2 kg m⁻³ and 0.0038 Pa s, respectively, and the fluid properties are stable and incompressible laminar flow. The independence and convergence of the mesh are checked by 256 000 threshold cells. The wall shear rate is calculated using the continuity equation and the Navier-Stokes equation.

Blood collection

The method of blood collection is consistent with our previous research.¹⁹ Blood samples were collected from 18 healthy adult volunteers randomly recruited between 2021 and 2022 by the Yongchuan Branch of Chongqing Blood Center. The inclusion criteria were as follows: the volunteers reported no history of medication, operation and alcohol abuse within one month, and the blood cell pressure, platelet count and coagulation index were all within the normal reference value.The study was approved by the Ethics Committee of Yongchuan Hospital, Chongqing Medical University (approval number: 20170318–12), and all subjects signed an informed consent form. The whole blood was anticoagulated with 3.2% sodium citrate at 1:9 (v/v) and used immediately after collection.

Platelet phosphatidylserine exposure study

The exposure of phosphatidylserine (PS) on the surface of platelet membrane can bind coagulation factor V to it, and promote the calcium-dependent binding of coagulation factor X, leading to the assembly of prothrombinase complex and accelerating the generation of thrombin. In order to better determine the effect of rapidly increasing and decreasing shear rate gradient on platelet activation, we detected the exposure level of PS on platelet surface by flow cytometry. In short: before blood perfusion, 5% BSA was used to block the microchannel to reduce nonspecific reaction. Then, the whole blood was perfused into the microchannel and the effluent was collected, and 5ul was added into the centrifuge tube containing 0.5% BSA. Immediately, 5 ul percp-efluor710-labeled anti-CD61 monoclonal antibodies and FITC-labeled anti-annexin v monoclonal antibodies were added respectively for fluorescent staining. After 15 min of incubation, 500 ul PBS was added to the centrifuge tube and analyzed immediately. ADP-activated resting whole blood was used as a positive control. A blue laser (488 nm) and a red laser (625 nm) were used as the excitation light. The filter configuration of the fluorescence detector were APC-A7OO 712/25 and FITC 525/40.

Platelet activation study

Figure S2 shows a simplified flow chart of the experimental design. The expression of platelet surface activation markers (CD62P, GPIIb/IIIa) was assessed by flow cytometry under different shear conditions.The method is consistent with our previous research.¹⁹ Briefly, effluent flowing through microchannels was collected, fluorescently stained, and fixed with paraformaldehyde.Resting whole blood and ADP activated whole blood were used as negative and positive controls. In order to better confirm the role of shear stress in platelet activation, we used whole blood samples with PI3K specific inhibitor TGX-221 as control. The gating strategy for the analysis was to identify platelet populations by forward scattering (FS) and anti-CD61 monoclonal antibodies, and then to exclude platelet aggregates. Finally, the expression levels of CD62P and activatedGPIIb/IIIa on the surface of individual platelets were determined.

Monocyte-platelet aggregation (MPAs) studies

For quantitative analysis of MPAs,5 ulof shear-induced whole blood was collected, and 5 ul of FITC-labeled HIP8, PE-Cyanine7-labeled anti-CD14 monoclonal antibodies were added respectively. Incubation at room temperature and dark room for 15 minutes.Finally, 1 ml of 1% paraformaldehyde was added for fixation. The unstimulated sample served as a negative control and the ADP-activated sample (10 uM) served as a positive control. A blue laser (488 nm) was used as the excitation light. The filter configuration of the fluorescence detector were FITC 525/40 and PC7 780/60.

Platelet aggregation morphology study

The morphology of platelet aggregation was studied by microfluidic chip technology. Before the study, the surface of the three microchannels was coated with type I collagen fibers (type I, 50 ug/ml). After incubation at 37°C for 1 h, the uncoated area was blocked with 5% BSA. After blocking for 1 h, the microchannels were washed with PBS. Calcein AM is a living cell fluorescent dye, which can penetrate the cell membrane and enter the cell, and then is cleaved by cell lactonase to form Calcein, which emits strong green fluorescence. When the blood is flowing, the platelets can react with the collagen fibers, making them adhere to the bottom, while other blood cells will be washed away with the blood flow, as a result, fluorescently labeled platelets are left at the bottom of the channel.The fluorescently labeled whole blood sample was added into the sample pool, and the blood was controlled to flow in the microchannel at a set shear rate by using a syringe pump. The aggregation behavior of platelets on the microchannel was recorded by using a camera controlled by Streampix 5.0 software at 1 frame/1s (objective × 20), and the observation area was selected from the shaded part in Figure S1a. 150 sequential fluorescence images were obtained after recording for 2.5 min. When the blood in the storage pool was exhausted, normal saline was added to flush the microchannel. The microscope was switched to brightfield and the morphology of platelet aggregation downstream of the stenosis was observed and recorded using phase contrast.

vWF: Ristocetin (vWF: RCO) assay

The reactivity of platelets to vWF was assessed by a modified vWF: Ristocetin (vWF: RCO) assay developed by Chen and Daigh.²⁰ The principle of the assay is that platelet GPIb protein is induced by ristocetin to bind specifically to vWF, resulting in a “platelet-vWF-platelet” binding form. Each vWF multimer has multiple binding sites to the platelet GPIb-IX-V complex. When platelets labeled with HIP8 (GPIIb) and VI-PL2 (GPIIIa) are bound to a single vWF multimer, a micromerger is formed. The mircomerger are detected by flow cytometry as a double positive event for CD41 and CD61. The degree of micromer formation is related to the binding affinity of vWF to platelets; if the molecular weight of the vWF multimers is too low to bind at least two platelets, a double-positive event will not occur. The brief steps are: freshly collected whole blood is centrifuged (1000 r × 10 min) and the upper layer of platelet-rich plasma (PRP) is collected. Equal amounts of PRP labeled with anti-CD41a and anti-CD61 monoclonal fluorescent antibodies, respectively, were taken and incubated at room temperature and protected from light for 15 min. Subsequently, the unbound fluorescent dye was removed from the supernatant by centrifugation (2000 g × 5 min). Platelets were washed using saline and finally resuspended by adding saline and adjusting the density to 50 000 cells/uL. 10 ul of shear-induced PRP was collected and 10 ul was added to a solution containing 50 ul HIP8-labeled platelets + 50 ul VI-PL2-labeled platelets + Ristocetin (1.5 mg/ml) and incubated for 45 min at 37°C with shaking. Experiments were performed in Blank (10 ul PBS added to the above system) as a negative control. The experiments were performed to quantify the aggregated platelets by the percentage of double-positive events. The higher the percentage of double-positive events, the higher the platelet aggregation rate and therefore the stronger the reactivity of platelets to vWF.

Platelet count

Platelets were counted by BC6800 automated blood cell analyzer. Platelet counts were performed separately on whole blood flowing through microchannels with different degrees of stenosis, and the experiments were performed with untreated whole blood and ADP (10 uM) induced whole blood as negative and positive controls.

Western blotting test

Whole blood was centrifuged (1000 r × 10 min) to obtain PRP. The PRP was divided into groups and different experimental conditions were imposed. The same amount of PRP was collected and centrifuged (800 g × 10 min) to remove the supernatant. If there were red blood cells in the precipitation, the red blood cell solution was used to split, and the tyrodes solution was used to wash and resuscitate the platelets. RIPA lysate containing protein hydrolase and phosphorylated protein hydrolase inhibitor was added to washed platelets and cracked on ice for 30 minutes. After the platelets were completely lysed, the upper protein suspension was obtained by centrifugation at 4°C (12000 r × 10 min). According to the BCA kit, the BSA standard curve was calculated, and the total protein concentration of each group was adjusted to 2 mg/ml. Proteins were separated by sodium dodecyl sulfate-polyacrylamide gel (SDS-PAGE) and transferred to precut nitrocellulose (PVDF,0.22 um) membrane. The nonspecific sites were blocked with a blocking solution containing 5% skim milk powder, and the membrane was incubated overnight with the primary antibody against phospho‐Akt (p‐Akt) (Ser473)/or anti-GAPDH (1:1000 dilution), and the second antibody was used to couple the primary antibody and chemiluminescence groups. The immune response was detected by enhanced chemiluminescence (ECL) and a gel imaging system. The amount of protein uptake between groups was measured by the expression of GAPDH, and the relative expression of the protein was expressed by the ratio of phosphorylated protein to GAPDH.

Statistical analysis

The data were analyzed by IMP SPASS Statistics 21.0 software. All experiments were repeated at least three times and the values were expressed as mean ± standard deviation (deviation ± SD). All the data were tested by the gaussian distribution test and homogeneity of variance test. One-way ANOVA and Turkey posttest were used to compare the differences between groups, and P < .05 indicated that the differences were significant.

Results

SGCR in microchannel increased with the increase of model narrowing angle

Fluid shear force can regulate the physiological function of platelets and play a key role in mediating cell-cell interaction. Therefore, it is necessary to accurately define the change of fluid shear rate in different stenosis models. To determine the shear flow field within the microchannel, detailed computational fluid dynamics (CFD) simulations were performed. Figure 1A shows the wall shear rate distribution for the constriction model with 80% stenosis and 30°, 60°, and 90° constriction and expansion angles, respectively. We found that in all three models, the shear rate increases with channel narrowing (color gradually shifts from dark blue to dark red), with a maximum shear rate of approximately 8300 s⁻¹ in the narrow channel and the other non-stenotic regions in line with the input shear rate (1500 s⁻¹). To determine the history of shear rate changes experienced by platelets flowing through the microchannel, We analyze the particle tracks that flow through narrow microchannel (Figure 1B). The results show that in all models, the shear rate undergoes a rapid increase and a rapid decrease after reaching a peak. The rate of decrease in shear rate is fastest in the 90° stenosis model and slowest in the 30° stenosis model (Figure 1B). This difference may be the key reason for platelet activation and aggregation.

Figure 1. Flow field simulation analysis. (A) Wall shear distribution on different narrow models. (B) Shear rate variation history on the flow line through the narrow channel.

The exposure level of PS on platelet membrane increased with the increase of SGCR

Platelet activation is at the center of blood coagulation, and the production of thrombin requires great changes in morphology and membrane topology accompanied by platelet activation. Platelet PS was initially thought to provide a negatively charged membrane surface that could bind factors Va and Xa and assemble an active prothrombin complex on the platelet surface. Based on this key role, platelet PS exposure is considered to be a key event to control blood coagulation. In this study, we quantitatively analyzed platelet PS exposure mediated by different SGCRs. The results showed that platelet PS exposure increased with the increase of SGCR (Figure S3a). Compared with the 30° stenosis model, the high SGCR produced by the 90° stenosis model significantly increased the PS exposure level (Figure S3B)(P < .05). This suggests that SGCR can preactivate platelets in a gradient-dependent manner.

The expression levels of platelet P-selectin and activated GPIIb/IIIa increased with the increase of SGCR

P-selectin and activated GPIIb/IIIa expression are common markers to characterize the level of platelet activation. In this study, we hope to analyze the ability of different shear microenvironments to induce platelet activation by flow cytometry. In addition, in order to improve the specificity of shear stress induction, we used TGX-221 treatment group (PI3K specific inhibitor) as control. The experimental steps are as follows:In short, we collected whole blood flowing through blocked microchannels to characterize the effects of different SGCR on platelet activation. Since platelets cannot form initial adhesion on the surface of BSA, shear-activated platelets will not remain in the microchannel in the form of thrombus, so that activated platelets can be retained to the maximum extent in the collection solution. In order to avoid continuous activation of platelets induced by shear, the samples were stained and fixed as soon as they were collected. Figure 2A demonstrates the expression of platelet surface P-selectin and activated GPIIb/IIIa.Compared to the resting state, platelet P-selectin and activated GPIIb/IIIa expression levels in the flow-through 30° stenosis model were 15.05% and 9.0%, respectively, with significantly higher activation levels (P < .05)(Figure 2B,C). In addition, when the contraction and expansion angles of the stenotic channel increased from 30° to 90°, the expression levels of both were 31.38% and 18.28%, respectively, with significant differences. The results of TGX-221 treatment group showed that the whole blood treated with specific inhibitor could not be activated by shear stress (Figure S4) (P > .05). Through the above experiments, we cautiously believe that platelets can be activated in a shear gradient-dependent manner.

Figure 2. Platelet activation. (A) Scatterplot distribution of platelet activation. (B-C) P-selectin and activated GPIIb/IIIa expression statistics. Platelet activation was measured as a percentage of positive events. In the experiment, resting platelets and ADP-activated platelets were used as negative control and positive control respectively, and 30°, 60° and 90° respectively represented platelets induced by different SGCRs.All groups were compared with the 30° stenosis model. P < .05 means the difference is significant, * stands for P < .05, *** stands for P < .0005, **** stands for P < .00005. The microchannel was blocked with BSA before the blood sample was perfused.

The formation of monocyte-platelet aggregation was positively correlated with the rate of shear gradient change

MPAs is a bridge between inflammation and thrombosis. Monocyte-platelet aggregation in circulation is a sensitive and powerful indicator of platelet activation and inflammation. After platelet activation, the exposed P-select is bound by monocyte P-selectin glycoprotein ligand-1 resulting in the formation of MPAs, which plays an important role in promoting blood hypercoagulability and thrombosis and participating in the regulation of inflammatory reaction. In this study, we analyzed the effects of different shear gradient change rates on the formation of MPAs. The degree of monocyte-platelet aggregation was studied by flow cytometry (Figure 3A,B). Shear-induced whole blood was collected similarly to the platelet activation part of the experiment.In the study, resting platelets and ADP-activated platelets were used as negative and positive controls. Monocyte-platelet aggregates (MPAs) were assessed by flow cytometry gating on CD14+ monocytes and measuring the proportion of CD41+/CD14+ double-positive events, as well as the anti-CD41-MFI of those events. We found that the 30° stenosis channel induced a significantly higher degree of aggregation (P < .05) with 18.79% MPAs formation compared to the negative control (0.57%). Compared to the 30° stenosis channel, the 60° stenosis channel further elevated the formation level of MPAs, but the difference was not significant; with the shear change rate further elevated, the formation of MPAs induced by the 90° narrow channel was 41.18%, which significantly increased the degree of aggregation; the formation of MPAs induced by ADP was also significantly elevated (P < .05) (Figure 3B). The results showed that the formation of monocyte-platelet aggregation was positively correlated with the change rate of shear gradient.

Figure 3. Monocyte-platelet aggregations(MPAs) formation. (A) Scatterplot distribution of MPAs. (B) Statistical analysis of MPAs formation. Platelet activation was measured as a percentage of positive events.All samples were compared with samples flowing through the 30° narrow microchannel. In the experiment, resting platelets and ADP-activated platelets were used as negative control and positive control respectively. P < .05 means the difference is significant, * stands for P < .05. The microchannel was blocked with BSA before the blood sample was perfused.

The area of platelet aggregation increased with the increase of SGCR

Platelet aggregation occurs in the microfluidic channel under the influence of gradient shear force. In order to obtain the most direct evidence of platelet aggregation induced by different shear microenvironments, we monitored this reaction in real time by fluorescence and phase contrast microscopy. The effect of shear change rate on platelet aggregation morphology was investigated by microfluidic microarray technology. The shaded area in Figure S1A is the study area for this experiment, in which the shear flow field in the microchannel undergoes a rapid reduction from the peak shear (~8300 s⁻¹) to the input shear rate (1500 s⁻¹). Figure 4A shows the fluorescence aggregation images of whole blood after 150 s of flow on different narrow-angle chips and the local bright-field view after PBS washout. We found that the platelet aggregates downstream of the 30° narrow chip were small, and their area were relatively uniform, about 310 um²; with the increase of shear change rate, the number of platelet aggregates downstream of the 60° chip gradually increased, and the length and width of the aggregates gradually increased; when the shear change rate was further increased, the platelets downstream of the 90° chip stacked heavily and formed thick aggregates, and their maximum area reached 2770 um² (Figure 4B). The experimental results directly reflect the effects of different SGCRs on the behavior and morphology of platelet aggregation in the flow state, and the disc-shaped clustered platelet morphology provides a strong support for shear-induced platelet aggregation.

Figure 4. Morphological study of platelet aggregation. (A) Platelet aggregation downstream of different stenosis models. (B) Statistical analysis of aggregate area. The scale bar is in the lower left corner of the figure. P < .05 means the difference is significant, ** stands for P < .005, **** stands for P < .00005. The surface of the microchannel was covered with type I collagen fibers to provide the thrombus surface needed for platelet aggregation.

Platelet aggregation induced by ristocetin

Ristocetin can induce specific binding of platelet GPIb protein to vWF. In order to better reveal the relationship between different shear microenvironments and GPIb-vWF binding pathways, we made the following analysis: The reactivity of vWF to platelets was evaluated by ristocetin-induced platelet aggregation (Figure 5). When platelets labeled by HIP8 and VI-PL2, respectively, were mixed in the absence of plasma, double-positive platelets (platelet aggregates) were almost negligible (<3%); when plasma was added to the reaction system after shearing in the 30° stenosis chip, the double-positive platelet expression rate was 12.82%, which significantly elevated the level of platelet aggregation compared to the negative control; when the shear change rate gradually increased (60° and 90° stenosis models), the double-positive platelet expression increased to 15.64% and 21.11%, respectively, compared to the 30° model, and the degree of platelet aggregation was further increased in a shear deceleration rate-dependent manner (P < .05) (Figure 5). The experimental results in this part are consistent with the platelet aggregation observed under the microscope in real time, suggesting that the shear gradient-induced platelets can aggregate in a vWF-GPIb-dependent manner.

Figure 5. VWF:RCO test.(A) Representative vWF and platelet aggregation.(B) Results statistics. Resting platelets were used as negative control, and all samples were compared with 30° stenosis model. The double positive area in the upper right quadrant of the scatter plot represents the platelets that bind to vWF. P < .05 means the difference is significant, * stands for P < .05, ** stands for P < .005. The microchannel was blocked with BSA before the sample was perfused.

Platelet count decreased after flowing through narrow microchannels

In order to make the results of shear-induced platelet aggregation more solid and convincing, we confirmed the experimental results from many different angles, such as activation, aggregation and so on. In this part of the study, we collected the effluent flowing through the narrow microchannel and counted the platelets. Platelets were counted by BC6800 automated blood cell analyzer and the results were presented as mean ± standard deviation. The results showed that the number of platelets in the effluent decreased in varying degrees.Compared with the platelets flowing through the 30° narrow microchannel, the number of platelets flowing through the 90° microchannel decreased significantly (P < .05) .Combined with the previous experimental evidence, we reasonably speculate that different SGCRs can induce different degrees of platelet activation, platelet aggregation after activation, and finally reduce the number of single platelets in the effluent.

The expression of platelet p-Akt increased with the increase of SGCR

Further exploration of the molecular mechanism of platelet activation induced by shear gradient is helpful for us to better understand the key role of fluid shear stress. It is reported that PI3K-AKT pathway is an important molecular pathway that mediates biomechanical transduction. To study the mechanism of platelet activation and aggregation mediated by SGCR, we analyzed the expression of p-Akt under different treatment conditions. In this study, resting platelets and platelets flowing through straight microchannel was used as control group, and platelets induced by different SGCR were used as experimental group. In order to confirm the effect of shearing on platelet activation, we added a specific inhibitor of PI3K (TGX-221) to the study. The results showed that the p-Akt expression level of 90° was significantly higher than that of unsheared samples, and the specific inhibitor (TGX-221) could significantly inhibit its expression (Figure 6A). At the same time, the expression of p-Akt induced by different SGCR was also different. The larger the SGCR, the higher the expression level of p-Akt (Figure 6B). This suggests that SGCR may mediate platelet activation and aggregation through the PI3K/AKT signal pathway.

Figure 6. Western blotting results.(A) Expression of platelet P-Akt,Akt and GAPDK.(B) the relative expression of P-Akt. In the experiment, resting platelets were used as negative control, the expression of GAPDH was used to measure the sample loading between groups, and the relative expression of P-Akt was measured by P-Akt/Akt. P < .05 means the difference is significant, * stands for P < .05, *** stands for P < .0005. The microchannel was blocked with BSA before the sample was perfused.

Discussion

Platelets play a key role in thrombosis in healthy and diseased blood vessels and have been widely studied for decades.²¹ Recently, hemodynamics has been considered to be a key factor affecting platelet activation and aggregation.^22,23 It has been reported that the shear gradient produced at the site of vascular stenosis can promote thrombosis in the stenotic area, and this process may be mediated by the binding of von Willebrand factor (vWF) to platelet membrane protein GPIb.^5,24

In this study, we developed a group of microfluidic chip models with different stenosis angles (80% stenosis, 30°, 60° and 90°, respectively). Combined with the microfluidic chip analysis system and finite element simulation, we build a hardware platform that can control the flow of fluid in the target shear rate range (Figure S1). Whole blood samples with different shear gradients were collected, the expression of molecular markers related to platelet activation (P-selectin, activated GPIIb/IIIa, PS exposure and MPAs formation) was analyzed by flow cytometry. In resting platelets, P-selectin is located on the inner membranes of platelet granules. After platelet activation, P-selectin was transferred from intracellular granules to the outer membrane.²⁵ GPIIb/IIIa is a heterodimer formed by the combination of calcium-dependent α-subunit (GPIIb) and β-subunit (GPIIIa). When platelets are at rest, the affinity between GPIIb/IIIa and its ligands (fibrinogen and vWF) is low. Agonist-induced platelet activation can trigger the signal pathway from inside to outside, resulting in conformational changes in GPIIb/IIIa.²⁶ Unlike Rahman and Hlady,²⁷ which found that transiently increased shear rates (4860 s⁻¹ and 11 560 s⁻¹) can effectively activate platelets, we found that even if the transiently increased peak shear rates are the same (about 8300 s⁻¹), different shear change rates can lead to varying degrees of platelet activation (Figure 2). Our results showed that there were almost no activation molecular markers on the surface of resting platelets. However, after being induced by different shear gradients, the levels of platelet activation (P-selectin and GPIIb/IIIa) increased significantly, and the degree of increase was related to SGCR. The faster the SGCR, the higher the degree of platelet activation. Compared with the 30° stenosis model, the shear of the 90° stenosis model significantly activated platelets (P < .05). At the same time, our specific inhibition test results showed that different SGCRs could not activate TGX-221 treatment group (P < .05), which showed the specificity of shearing activation. In the PS exposure study, we obtained similar experimental results. The PS exposure level of the shear treatment group was significantly increased (P < .05). Compared with the low SGCR, the high SGCR produced by the 90° stenosis model significantly increased the PS exposure level (P < .05). The two groups of markers corroborated each other, solidly proving the influence of SGCR on platelet activation.

Recent studies have shown that the formation of MPAs may be an early inflammatory indicator of atherosclerosis. When platelets are activated, P-selectin binds to the constitutively expressed P-selectin glycoprotein ligand-1 (PSGL-1) on the monocyte membrane, resulting in MPAs formation.²⁸ In this study, we found that when platelets were resting, there is almost no MPAs. The results of flow cytometry showed that compared with the negative control, the MPAs induced by the 30° stenosis chip were significantly increased, and the formation level of MPAs was also significantly increased when the stenosis angle increased from 30° to 90°. The results of platelet count showed that the platelet count decreased in varying degrees after shear induction and ADP activation (Table I). This shows that platelet aggregation occurs immediately after platelet activation, and the platelet after aggregation loses its original volume and sends out a huge pulse signal when it passes through the detection circuit of the automatic blood cell analyzer so that it can not be detected as a platelet. Combined with the results of flow cytometry, and blood cell count, we have good reason to believe that shear gradient induces platelet aggregation, whether it is the mutual aggregation between platelets or the binding between platelets and leukocytes, and the faster the SGCR, the more platelet aggregation. At the same time, the results further support the conclusions of the activation study.

Table I. Platelet counts (×10⁹). Platelet count was detected by Uncorrected Fisher’s LSD test. All samples were compared with those perfused by the 30° stenosis model (30°), n = 3, P < .05 indicates a significant difference.

Sample	N	P	30°	60°	90°
Mean±SD	189.3 ± 12.1	24.33 ± 6.0	139.0 ± 17.4	123.3 ± 13, 7	100.3 ± 11.5
P	0.0180	.0035		0.0189	0.0341

It has been reported that platelet aggregation is dominated by GPIIb/IIIa-dependent interaction at the low shear rate (600–900 s⁻¹), while GPIbα and GPIb/IIIa play a major role in platelet aggregation at a moderate shear rate (1000–10000 s⁻¹). However, with the increase in shear rate, the vWF-dependent interaction becomes more and more important. Therefore, the combination of vWF and GPIbα plays an important role in thrombosis.^22,29 Our microfluidic experiments showed that platelet aggregation occurred in different degrees downstream of stenosis in different models. With the further increase of the contraction angle and expansion angle, that is, the SGCR gradually increased, and the platelet aggregation downstream of the stenosis also gradually increased. When the stenosis angle is 90°, the platelets downstream of the stenosis are arranged in clusters in the shape of a disc, and the maximum aggregate is 120 × 23 um, with a plane area of about 10 times that of the downstream aggregate of 30° stenosis (Figure 4). Nesbitt and Westein⁵ found that low levels of calcium influx can maintain the stable aggregation of discoid platelets independent of shape change and degranulation. Under the promotion of soluble vWF, platelets that suddenly undergo shear acceleration form transient aggregates without pre-activation or shape change by binding to fixed vWF. The unstable interaction between vWF and GPIbα leads to rapid platelet translocation to the downstream region with a much lower shear rate, and then calcium-dependent signal transduction leads to the reorganization of filamentous platelet junctions into globules, resulting in the stabilization of discoid platelet aggregates. Our study supports this conclusionTo further confirm the contribution of vWF, we collected the plasma after splicing and detected the activity of vWF by flow cytometry. The results showed that with the increase of SGCR, the binding ability of vWF to platelets gradually increased, indicating that the activity of vWF also gradually increased (Figure 5). The results support the conclusion that SGCR induces increased platelet reactivity and discoid platelet aggregation.

Studies have shown that phosphatidylinositol 3-kinase (PI3K) plays an important role in shear-induced platelet activation.³⁰ Akt (also known as protein kinase B) is a serine/threonine-specific protein kinase and the main downstream factor of PI3K.³¹ Previous studies have shown that PI3K/Akt signaling pathway plays a role in shear-mediated vWF-GPIb interaction and subsequent platelet activation.³² To understand the effects of different SGCR on platelet activation pathways, we analyzed the expression level of p-Akt. The results showed that shear induction significantly increased the expression level of p-Akt compared with the negative control (unsheared PRP) (Figure 6). After treatment with PI3K inhibitors, the expression level decreased significantly in the sheared group, but not in the unsheared group. The results support the previously published conclusions.¹⁴ In addition, we also confirmed that SGCR can activate platelets through the PI3K/AKT signal pathway. The faster SGCR is, the stronger the expression of platelet p-Akt is.

There are also some shortcomings in our research. The microfluidic chip model involved in the experiment is a rectangular channel structure rather than a cylindrical blood vessel, which may lead to some differences in shear rate values. In addition, the model only allows blood to be exposed to high shear at a single time, while platelets in the body are sheared each time they pass through a narrow blood vessel. We hope that in the future, by improving the device, it can more truly simulate the process of blood circulation in the body.

Conclusion

In summary, we developed a set of microfluidic chips with different contraction and expansion angles to provide a flow field environment with different shear gradients. Through a series of experiments, we found that a rapid change in shear gradient activates platelets GPIbα through the PI3K/AKT signal pathway, and the faster the rate of shear change, the higher the reactivity of platelets to vWF, the larger the area of platelet aggregates downstream of the stenosis, and the greater the platelet reactivity (P-selectin, activated GPIIb/IIIa, MPAs formation, and PS exposure). Our study further refines the effect of fluid shear on platelet activation and aggregation function and contributes to a better understanding of shear-induced platelet adhesion and aggregation mechanisms.

Author contributions

Tiancong ZHANG, Xuemei GAO, Xiaojing HUANG, and Xuanrong HUAN performed the experimental research and data analysis. Tiancong ZHANG, Ling LIU and Yuan LI wrote and edited the manuscript. Cui HE contributed to the study design, data analysis, and writing and editing of the manuscript. All authors have read and approved the final manuscript and, therefore, have full access to all the data in the study and take responsibility for the integrity and security of the data.

Author statement

No work in this manuscript has been reviewed or published in other journals.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

Please contact the corresponding author for detailed data.

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/09537104.2023.2288679

Additional information

Funding

This work was supported by the National Natural Sciences Foundation of China [11702047], the Special Project of Science and Technology Innovation for People’s Livelihood Security of Chongqing [cstc2017shmsA130009], the Chongqing Medical Research Program [2017MSXM079], and the Postdoctoral Research Project of Chongqing [Xm2017082].