Microfluidic Devices for Studying Coagulation Biology

Abstract

The ability to study the behavior of cells, proteins, and cell-cell or cell-protein interactions under dynamic forces such as shear stress under fluid flow, provides a more accurate understanding of the physiopathology of hemostasis.

This review touches upon the traditional methods for studying blood coagulation and platelet aggregation and provides an overview on cellular and protein response to shear stress. We also elaborate on the biological aspects of how cells recognize mechanical forces and convert them into biochemical signals that can drive various signaling pathways. We give a detailed description of the various types of microfluidic devices that are employed to study the complex processes of platelet aggregation and blood coagulation under flow conditions as well as to investigate endothelial shear-response. We also highlight works mimicking artificial vessels as platforms to study the mechanisms of coagulation, and finish our review by describing anticipated clinical uses of microfluidics devices and their standardization.

Introduction

Blood coagulation is comprised of an intricate sequence of events involving a multitude of molecules designed to stop blood loss following injury, and to promote healing and repair. Activation of the coagulation cascade results in polymerization of fibrin and activation of platelets, leading to the formation of a blood clot that quickly closes the site of injury and mitigates blood loss. The molecular and cellular pathways involved in the coagulation process have been studied extensively, and tests are currently available to evaluate and quantify any aberrations leading to a pathology. However, these traditional methods fall short when examining blood flow-related processes, because they ignore the presence of underlying mechanical forces that are crucial to the normal physiological response [1–3]. Blood flow in the vasculature, as any fluid flow, generates mechanical shear stress on the walls containing the fluid [4]. This wall shear stress has been shown to affect not only cells flowing along the vascular wall but also circulating blood proteins [5]. Evaluation of shear stress can effectively be incorporated into experiments through the use of microfluidic technology. Microfluidic devices use pumps to flow media, or other solutions, through small channels that are specifically engineered to mimic the desired anato-physiological design of their in vivo counterpart. These channels can also be coated with cells, and a stream of media (with or without other cellular components) flowed to accurately recreate the in vivo cell environment [6]. Such a platform allows for the inclusion of all present forces to provide greater insight into cell-cell interaction and protein behavior under native conditions.

Hemostasis and Shear Stress

Normal hemostasis results from a coordinated, balanced, and strictly regulated physiological process involving blood vessels, platelets, coagulation factors, and fibrinolytic proteins. When blood vessels are disrupted by trauma, or the endothelial lining is damaged by the high shear forces produced by blood flowing through the small branches of the arteries and capillaries, vasoconstriction, generation of the platelet plug, and activation of the coagulation cascade ensue, with the viscosity of blood changing markedly as it transforms from a liquid to a solid gel phase [7]. The fibrinolytic system is activated simultaneously with the coagulation cascade to maintain the fluidity of blood and prevent clots from occluding the vessel [8, 9]. While high shear forces can cause endothelial injury, it is important to appreciate that shear/flow rate also plays a critical role in the process of coagulation. For example, thrombin generation and platelet activation are both highly dependent on shear/flow rate, and the degree of platelet activation correlates with both the shear stress magnitude and exposure time [10, 11]. Shear forces deriving from blood flow are also responsible for margination of the platelets, which increases their concentration near the wall and facilitates platelet accumulation at the site of injury [12–14]. von Willebrand factor (vWF), is a plasmatic multimeric glycoprotein that is synthesized by endothelial cells and megakaryocytes and stored in platelet α-granules. Flow dynamics can trigger the reversible self-assembly of vWF and the accumulation of platelet-vWF complexes, which is mediated largely through the interaction between vWF and glycoprotein Ib (GpIb) and/or platelet integrin αIIbβ3 (GP_IIb/IIIa) expressed on the surface of platelets [15]. As a collective result of these alterations and interactions, the adhesion probability of platelets to vWF is strongly and nonlinearly dependent on shear/flow rate [16], enabling platelets to rapidly adhere to collagen present in subendothelial layers upon vascular damage and aggregate to form a plug to halt bleeding [17].

In addition to its effects on platelets, shear induced or any type of vascular damage also triggers activation of the coagulation cascade, which is initiated by two distinct mechanisms: the tissue factor (TF)/ extrinsic pathway, and the contact/intrinsic pathway. The former is triggered when circulating clotting factors are exposed to tissue factor (TF), and the latter when plasma comes in contact with subendothelial collagen and other negatively charged surfaces. TF, also known as tissue thromboplastin, is a protein present in leukocytes and in the subendothelial tissues surrounding blood vessels. Within minutes of vessel wall injury, the exposed TF binds its ligand circulating factor VII (FVII) which, in the presence of phospholipids and Ca²⁺, converts factor X (fX) to factor Xa (FXa) [18–20]. Under normal physiological conditions, the role of the intrinsic pathway in hemostasis is less significant than that of the extrinsic pathway. The initiation of the intrinsic pathway, which is delayed by more than 10 minutes after vascular damage, involves the activation of factor XII (FXII) to fXIIa by kallikrein, and reciprocal activation of plasma prekallikrein (PK) to kallikrein by fXIIa, with high-molecular-weight kininogen (HK) playing the role of facilitating substrate presentation to FXIIa. Generation of fXIIa leads to the activation of FXI to FXIa, a protease which activates FIX. In addition to being activated by FXIa, FIX is also one of the major substrates that is activated by the FVII/TF complex. Another chief player in the extrinsic pathway is the coagulation factor VIII (FVIII), which circulates in the blood in association with vWF. By complexing with FVIII, vWF plays a key role in coagulation by prolonging FVIII half-life in plasma, and by preventing the inactivation of FVIII by FXa and FIXa [21, 22]. Additionally, the activity of vWF is responsive to shear stress, allowing vWF to increase the local concentration of FVIII at sites of vascular injury [22]. Thrombin catalyzes the conversion of FVIII to its active form FVIIIa which, together with its co-factor FIXa, form the membrane-bound intrinsic tenase (FVIIIa-FIXa) complex. This complex, in the presence phospholipids, converts FX to its activated form FXa [18, 23, 24]. Both intrinsic and extrinsic pathways converge in the common pathway, which starts with the formation of FXa. In similarity to the tenase complex, factor V (FV), a glycoprotein structurally and functionally homologous to FVIII, is activated by FXa or by thrombin, and subsequently assembles with FXa to form the prothrombinase complex (FVa-FXa) that cleaves prothrombin into thrombin. The thrombin then cleaves fibrinogen into fibrin, which activates factor XIII alongside thrombin to crosslink fibrin and form a meshwork that patches the damaged area [18].

Coagulation Pathology

Qualitative or quantitative alterations in the levels and/or activity of the coagulation factors cause hemostatic alterations that lead to either bleeding or thrombosis. The most common bleeding disorders are hemophilia A and B. These are X-linked recessive disorders in which mutations in either the FVIII or FIX gene lead to a deficiency in FVIII or FIX, respectively [25]. These FVIII or FIX deficiencies hinder the functionality of the intrinsic pathway of coagulation so that the fibrin clot cannot be quickly formed to stop the hemorrhage [26]. There are many additional rare autosomal recessive genetic disorders that lead to defects in other clotting factors, such as fibrinogen or prothrombin, all of which manifest as bleeding disorders that are clinically similar to the hemophilias [26]. von Willebrand Factor (vWF) is a large multimeric plasma glycoprotein produced by endothelial cells and megakaryocytes that aids coagulation by recruiting platelets. vWF also serves as the carrier and stabilizer of FVIII protein. As a result, absence or alteration of vWF can result in low FVIII levels, producing a bleeding disorder that is very similar to hemophilia A [27, 28]. In addition to genetic lesions, liver damage or disease can also alter hemostasis and produce bleeding disorders because most coagulation proteins are produced within the liver [29].

Clotting disorders are not limited to excessive bleeding; disruptions or changes in the regulation of the coagulation cascade can also result in excessive clotting. For example, activated protein C (APC) and its cofactor protein S (PS) proteolytically inactivate the coagulation proteins FVa and FVIIIa. Therefore, a deficiency in either APC or PS removes the body’s ability to physiologically stop these highly procoagulant factors from generating thrombin, causing thrombosis to occur [30]. There are also FV mutations, such as the Leiden point mutation, that lead to APC resistance, and therefore a greatly increased risk of thrombosis [31, 32].

Traditional methods for studying blood coagulation, platelet aggregation, and other cellular responses to fluid flow.

Parallel plate rheometry and other techniques have previously been used to study blood coagulation, platelet aggregation, and other cellular responses to fluid flow. Parallel plate flow chambers consist of two plates separated by approximately 100 μm, with a syringe pump that pushes liquid through the chamber. While this technology has provided key insights into platelet adhesion and thrombosis, these systems can only provide a limited platform to study shear stress and cannot be easily modified to a wide range of applications. Moreover, PPR has been recognized to suffer from several limitations that include instrument inertia, artifacts from momentum diffusion, viscoelastic waves, and secondary flows, all of which can violate the assumption of homogeneous simple shear deformation and complicate the interpretation of the data generated [33]. To date, these flow chambers have primarily been utilized to investigate platelet interactions with collagen or vWF, and they have not often been used to look at cellular or protein responses to flow [34]. Other studies have simulated shear stress by rolling stainless steel or Teflon cones over seeded cells. Although these techniques provide a mechanical shear that is similar to flow-induced wall shear stress, they are limited since they do not provide a uniform shear stress. Thus, it is difficult to determine the forces being applied on tested cells.

Recently, microfluidic technologies have provided an alternate approach for studying fluid flow in biology [35, 36]. Microfluidic devices provide an easily customizable platform that can subject cells, proteins, platelets, or whole blood to fluid flow and shear stress. Such technologies can easily be tailored to specific purposes, and they represent a significant advance over other current methods of investigating the effects of shear on biological processes. In addition, the ease with which these systems can be customized enables them to be applied to answer a broader range of physiological questions [37–39].

Cell and Protein Response to Shear Stress

Both cells and proteins sense and respond functionally to shear stress to regulate biological processes such as coagulation. Platelets in particular have been extensively studied regarding their role in coagulation, and microfluidic technology has played a key role in advancing the knowledge of platelet mechanics. In order to examine platelet interactions with coagulation-related proteins such as collagen, tissue factor, or vWF, platelets can be circulated in microfluidic devices containing a known layout of these molecules, and platelet adhesion or effectiveness of drug treatments such as anticoagulants determined [40–42]. These techniques enabled the discovery that platelets adhere predominantly to collagen type I and that this adhesion is regulated by the level of shear stress [43–45]. This same technique can also be used to examine how characteristics of the injury surface affect coagulation and thrombosis, as well as to define the kinetics of platelet recruitment, adhesion, and clot contraction. By designing a microfluidic device that incorporated varied spatial distributions of tissue factor, studies were able to demonstrate that the initiation of coagulation requires a large patch of tissue factor or a close distribution of many small patches, since a spread out distribution of small patches with the same total surface area did not initiate coagulation [41]. Atherosclerotic geometries have also been incorporated into microfluidic devices to study how disease states affect thrombosis. Studies using such devices revealed that platelet aggregation and vWF in the outlet of stenotic atherosclerotic plaque geometries exacerbates thrombus formation downstream [46]. The kinetics of platelet recruitment can also be studied with the use of microfluidics. In one such study, platelets were stained with a fluorescent tag and mixed with unstained whole blood. This blood was then flowed through a microfluidic device coated with collagen, and confocal z-stack images were taken at varied time points to track the movement/localization of the labeled platelets over time and thereby define platelet movement within the thrombus and the kinetics of thrombus contraction [47].

Microfluidic Devices to Study Endothelial Shear-Response

Endothelial cells are constantly exposed to blood flow in vivo and have, therefore, been the subject of fairly intense study using microfluidic devices. Microfluidics uses precisely controlled fluid flow in small-scale geometries to process fluids or to study the physical and biological effects of fluid flow. This has become a key technology in studying coagulation, because blood flow is always present in the vasculature and exerts significant effects on the cells and proteins involved in coagulation. Various forms of flow play key roles in both the normal physiology and various pathologies of the circulatory system. For example, laminar flow in large tubular arteries prevents the generation of atherosclerotic plaques, while the turbulent flow present in branching areas of the vasculature creates local gradients of high shear stress, causing these areas to be more susceptible to atherosclerosis [48]. In contrast to the previously discussed parallel plate rheometry, in microfluidics, the effects of the inertial forces are negligible compared to the ones due to viscous forces; thereby, the non-linear term in the Navier-Stokes equation can be neglected. This benefit, combined with their customizability with respect to materials, form factors, and applications make microfluidic devices ideal for studying coagulation, because of the broad variety of conditions that exist in the vascular system. The diameter of vessels in the circulatory system range from 5μm for capillaries up to 10mm for muscular arteries [49] and flow velocities can range from 0.5mm/s to over 100cm/s [50, 51]. These flow rates create physiological shear stresses that can vary from 1 to 6 dyne/cm² in the venous system to over 10 to 70 dyne/cm² in arteries [52], while physiologically relevant shear levels as low as 0.1 dyne/cm² exist in the sinusoids of the liver [53, 54]. Microfluidic devices are unique in their ability to be adapted to mimic this wide variety of physiologically relevant sizes, flow rates, and shear rates, enabling them to fully embody the conditions that are present throughout the vascular system.

Exposing endothelial cells to laminar flow causes them to quickly elongate and align in the direction of flow, showing that shear stress is a critical component of vascular and endothelial remodeling [55–57]. Further mechanistic studies demonstrated that this rapid flow-induced change in endothelial cell alignment and morphology was due to cytoskeletal realignment mediated by the formation and rearrangement of the complex actin network [58]. Shear stress has also been shown to increase endothelial cell proliferation by modulating the cell cycle, increasing the percentage of cells in S, M, and G2 [57]. Microfluidic devices have also been used as a biologically relevant platform to study endothelial production of the potent vasodilator nitric oxide (NO) in endothelial cells. Since adenosine triphosphate (ATP) is a major stimulus for endothelial-derived NO, media containing ATP can be flowed through an endothelial-lined microfluidic device to stimulate NO production and study potential NO-targeting therapeutics [59].

Microfluidic platforms can also be used to model damage and pathology in blood vessels. For example, van der Meer et al. developed a polydimethylsiloxane (PDMS) device harboring a rectangular channel in which three inlets led to a single channel that was seeded with endothelial cells. By perfusing trypsin through the middle channel and media through the outer channels, the authors were able to selectively introduce a wound within the center of the channel, while leaving the edges undisturbed. Time-lapse imaging was then used to study how endothelial cells migrated from the undamaged edges of the channel into the wound, under conditions of constant flow [60].

In other studies, Estrada and colleagues used a microfluidic device to study the disturbed flow profile that occurs in atherosclerotic conditions and assess how this affects the behavior of endothelial cells. Their device that was seeded in flow using a one-way valve that ensured normal laminar flow, and the one-way valve was removed to create retrograde flow alongside, thereby reducing the flow rate and mimicking the aberrant flow patterns seen in atherosclerosis. The endothelial cells under these disturbed flow conditions lost their normal elongated shape and aligned in the direction of flow. In addition, actin was no longer specifically arranged in long, aligned filaments, and β-Catenin cell-cell junctions were decreased. Taken together, these observations affirmed that this model accurately reproduces atherosclerotic conditions [61].

Thrombosis has also been studied by perfusing whole blood through a microfluidic device consisting of a rectangular channel that was coated with collagen and seeded with endothelial cells on each side. Just as occurs in vivo, the presence of the endothelium prevented blood clotting in this device under normal conditions, but pre-activating the endothelium with TNF-α promoted platelet aggregation. These findings confirmed that thrombosis can readily be studied in this device and prompted its use as a clinical diagnostic tool for testing platelet function [62].

Microfluidic devices can also be used to effectively examine endothelial permeability, which allows investigation of the uptake of drugs or other biomolecules. In one example, a device consisting of two flow chambers separated by a porous polyethylene-terephthalate (PET) membrane was seeded with endothelial cells on only one side to enable assessment of permeability. When media was flowed on both sides of the membrane, levels of PECAM-1 and zonula occludens-1 (ZO-1), two regulators of endothelial permeability, were increased [63]. Additionally, when the system was challenged with pro-inflammatory stimuli, PECAM-1 and ZO-1 expression were both severely reduced while expression of ICAM-1, VCAM-1, and E-selectin were increased, showing that endothelial cells grown in this device more closely resemble those in vivo than endothelium grown in traditional static 2D monolayer cultures [63].

Microfluidic Devices to Study Protein Shear-Response

Microfluidic devices can also be extremely helpful for examining the behavior of proteins under flow conditions. As briefly mentioned earlier in this article, vWF is a shear-responsive protein that has been heavily investigated because of its role in platelet adhesion and aggregation. vWF is a multimeric protein that, under static and low shear conditions, is coiled into a small ball-like structure. However, increased shear conditions alter vWF function by triggering its unfolding [64], such that vWF elongates, exposing domains that were previously hidden by the folded protein conformation [65–67]. Specifically, shear stress elongation reveals binding sites that induce adhesion of vWF to collagen, which prompts the formation of a vWF network by self-association and polymerization [68]. The unfolding of vWF unmasks the A1 domain that is present on each multimer of the protein, which binds to GPIbα on platelets [69], allowing vWF to then recruit platelets to the area of vascular damage and aid in platelet adhesion [70]. Not surprisingly, unfolding of vWF occurs at sites of vascular injury, which are characterized by increased shear stress, leading to enhanced vWF activity in these sites where coagulation is needed. Altered folding of vWF has also been shown to have important implications in blood pathologies; for example, larger vWF multimers have been shown to greatly increase the binding of sickled erythrocytes, which could be a potential cause of buildup of erythrocytes leading to thrombosis [71]. Despite numerous studies being done on the response of vWF to shear, one area that has not yet been explored and could yield clinically important data is whether this shear reactivity has any effect on the FVIII that vWF carries. Our group is currently performing microfluidic device-based studies to address this knowledge gap.

Many other proteins are also affected by shear stress in a variety of ways. For example, a 6 base pair, core binding region that is sensitive to shear is present in the promoters of multiple genes that are expressed in endothelial cells [36]. This shear-responsive binding region is bound by nuclear factor-κB (NF-κB) under the influence of shear stress, leading to the shear-dependent upregulation of transcription of platelet-derived growth factor B (PDGF-B) [36]. This shear-induced increase in transcription is prevented by a mutation in this promoter region [35]. Bone morphogenic protein 4 (BMP4) production in endothelial cells has also been shown to be subject to complicated shear-dependent regulation, such that BMP4 is downregulated in laminar shear but is upregulated in oscillatory shear [72]. This complex regulation by shear is theorized to maintain normal levels of BMP4 during healthy flow but enable rapid upregulation of BMP4 when damage to the vessel wall causes a disturbance in the flow. This upregulation of BMP4 induces ICAM-1 expression via an NF-κB-dependent mechanism, ultimately culminating in increased monocyte adhesion to the endothelium [72].

Shear Sensing and Mechanotransduction

The repeated demonstration that cells can respond to shear has prompted mechanistic studies to understand how shear can regulate cellular function/behavior. Mechanotransduction is the process by which cells recognize the mechanical forces that they are experiencing and convert them into biochemical signals that can drive various pathways. The quickest mode of shear mechanotransduction is through ion channels such as the transient potential channels (TRP), a family of ion channels with 6 transmembrane sections, a pore region, and a voltage sensor, and the piezo I channels, which consist of a 3-bladed mechanosensitive propeller surrounding a central ion-conducting core [73–76].

Signaling via traditional membrane proteins is also an important mechanism by which cells sense and respond to shear. Two prominent examples include the caveolae that line the cell membrane of endothelial cells working alongside ion channels to respond to calcium signaling in shear [77] and the activation of the Shc signaling cascade by mechanosensitive tyrosine kinases and integrins in response to shear stress [78, 79]. This Shc-dependent pathway begins with autophosphorylation of focal adhesion kinase (FAK) in response to integrin mechanotransduction and recruitment of c-src, which associates with integrin β1 and FAK [80]. c-src then binds to the SH2 domain of Shc causing tyrosine phosphorylation of Shc, which in turn binds Grb2 and leads to ERK2 activation [81–84]. Activation of the Shc/Grb2/ERK2 pathway ultimately leads to changes in gene expression, highlighting one mechanism by which shear stress can regulate gene expression and subsequent protein synthesis [78, 85]. G-protein coupled receptors (GPCR) are also important membrane-bound proteins that can be activated by shear [74, 77]. GPCR, such as the B2K chameleon protein, undergo a shape change in response to shear stress that can affect their activity. Such changes in the conformation of membrane-bound proteins can lead to shear-induced alterations in the cell’s membrane-tethered cytoskeleton network, like the shear-induced actin rearrangement in endothelial cells we touched upon earlier, which leads to disruptions in the Wnt pathway [86]. The preceding examples highlight just some of the major mechanisms by which cells can respond to the mechanical forces of shear that they experience.

Microfluidic Blood Vessel

The ideal platform to study the mechanisms of coagulation is a vessel that is structurally and biochemically identical to those present in the human body. This has been attempted with the creation of blood vessel-on-a-chip designs that seek to use microfluidics to simulate the structure and function of a blood vessel, including the capability to flow fluid through the “vessel” to study the physiological interactions that occur. The simplest of these systems have been created by molding silicon into a tube-like shape, modifying the surface of the tube with ECM proteins, such as collagen, and seeding with endothelial cells. While a small surface-treated tube can be seeded with cells by simple perfusion, larger structures or devices often require additional techniques such as rotational seeding to fully endothelialize the “vessel” [87]. More sophisticated versions of these devices include branching channels, to simulate the branching network present in capillaries. Such branching can be incorporated into these devices by designing diverging microfluidic channels before perfusing ECM-containing gel to coat the walls and enable subsequent seeding with endothelial cells [88]. More recently, a branching capillary system was created that included micropores between channel branches to improve endothelial permeability [89].

While these relatively simple models are fairly easy to manufacture and can provide key insights into endothelial responses to flow, it is important to realize that blood vessels in the body are multilayered structures that contain a variety of different cell types and multiple structural proteins that all interact to produce the physiology that is characteristic of the vasculature. In an effort to better reproduce a native vessel, Ryu and colleagues created a simplified vascular wall network with multiple cell types by placing loops of fibroblasts around an endothelial coated channel. This arrangement allowed paracrine signaling to drive the cells to produce a “vessel” whose size, structure, and microenvironment were determined entirely by the endothelial cells and fibroblasts in the manufactured channel. This device was also created in such a way that the loops could easily be removed and the vessel placed in a body-on-a-chip system [90]. While this system undoubtedly reproduces a native blood vessel with a far higher degree of physiological accuracy, it is obviously not trivial to produce, and its reliance on cell-derived paracrine factors to drive vessel formation make it difficult to accurately control vessel size and determine shear stress on the vessel wall. Hasan et al. took the preceding model several steps further in an effort to more fully address the cellular features of a native blood vessel by creating a device that incorporated an outer layer of fibroblasts, a middle layer of smooth muscle cells, and an inner layer of endothelial cells, thereby simulating the tunica adventitia, media, and intima that are present in physiological blood vessels. This ambitious objective was accomplished in a stepwise fashion, beginning by flowing a fibroblast-laden hydrogel into a silicon channel and solidifying the hydrogel to coat the channel with fibroblasts. This same process was then repeated to incorporate a smooth muscle cell – hydrogel layer inside the fibroblast layer, and endothelial cells were finally seeded inside the smooth muscle cell – hydrogel layer [91]. The resultant device clearly captures more of the cellular interactions that are present in native blood vessels and thus promises to be a more accurate model for studying vascular physiology in vitro.

Incorporation of a Lymphatic System

The lymphatic system that functions alongside blood vessels in vivo has also been incorporated into a vessel-on-a-chip system by creating 2 channels separated by a PET membrane, with endothelial cells seeded on one side and lymphatic cells seeded on the other. This system was then used to study vessel permeability and cell-cell junctions. Permeability was initially high in this system but decreased with time, until it stabilized at a positive level, demonstrating establishment of appropriate barrier function. This platform also reproduced the perturbations in endothelial permeability that are known to occur in response to various chemical stimuli [92]. While this platform hints at the potential of a device that combines blood and lymphatic vessels, this first device was comprised of only a flat membranous sheet seeded with cells. As such, this platform could likely be improved by creating more complex side-by-side models of blood and lymphatic vessels to produce a more complete and physiologically accurate model.

Clinical Uses of Microfluidics

While most of the studies that have been discussed thus far have focused on understanding cell and protein function, a significant amount of effort in recent years has been put into developing microfluidics systems that can be used for clinical purposes. The use of microfluidics in a clinical setting would advance the management and treatment of patients with coagulation disorders, because it provides a high-throughput testing platform that can be adapted to more closely resemble, in vitro, the conditions present in the blood vessels in vivo, thereby generating a much clearer, more thorough understanding of the patients’ coagulative function. Current clinical tests for clotting disorders, such as the hemophilias, include activated partial thromboplastin time (aPTT) and prothrombin time (PT). The aPTT assay measures clotting time resulting from the intrinsic pathway, whereas PT measures clotting time resulting from the extrinsic pathway [93, 94]. These methods provide an invaluable means of analyzing blood coagulation to diagnose clotting disorders and to assess patient response to therapeutic intervention [95]. However, these methods can be dramatically improved upon with the use of microfluidics. Microfluidics enables the rapid, low cost manufacture of devices that require far less blood volume than traditional methods [96]. For example, devices have been created that incorporate all of the steps necessary to prepare samples and measure aPTT and PT simultaneously in a single analyzer [97, 98]. Conducting both tests simultaneously in this manner reduces both the time required for the technician to run the tests and the amount of blood that is necessary to obtain meaningful and reproducible results. Microfluidic flow devices have also been developed that can effectively quantify vWF levels. Since increased shear stress drives vWF attachment, running these devices at a higher shear rate could provide a better and more sensitive measure than current/traditional methods for quantifying vWF [99, 100]. A point-of-care device has also been created that measures factor Xa. Because factor Xa is common to both the intrinsic and extrinsic pathways, this new device could provide far less variability when compared to aPTT or PT, and would be invaluable to clinicians measuring factor Xa to determine the effect that anticoagulant therapy and thereby optimize treatment [101].

Despite the capabilities of microfluidic devices that have been discussed, there are still key challenges that must be overcome to allow microfluidics to be more widely used. Currently, there are a wide range of materials and coatings that are used in these devices, and a key challenge will be to determine a uniform base material that can be easily crafted into specified dimensions while also being capable of surface modification. Many of the materials that are currently used are not amenable to production scale-up for mass manufacturing, so discovering/developing a material that can be mass produced while also permitting surface modification to create a more relevant microenvironment would open entirely new avenues in the application of microfluidics [102]. Additionally, it will be necessary to determine the geometries, coatings, and cell types that are necessary to fully recapitulate the shear responsiveness of blood vessels. To accomplish this, it will be necessary to conduct further research to determine how integral the effects of circular vessels, additional cell types such as smooth muscle cells, and extracellular matrix coatings are to vascular response to shear. Creating a minimal representative model will allow for easier comparison of findings and replication of results.

Standardization of Clinical Devices and Procedures

Before these types of assays can be an effective clinical tool, there must be standardized devices and protocols to ensure that testing remains consistent across different users and equipment. This is of special importance with microfluidics, because the ease of manufacturing and versatility of these devices has led to wide variation in the devices that are currently being tested. As a first step towards clinical implementation, it will first be necessary to determine the parameters that are variable between devices and decide which parameters will be used for normalization. Blood viscosity is decreased as shear rate and flow velocity increase, so devices must account for this to apply the correct levels of shear stress [103, 104]. Specific recommendations focused on maintaining laminar flow by ensuring sufficient size and limiting shear rate, consistency of coating for each test, and consistency of evaluation based on staining and microscopy have been put forward by the Biorheological Subcommittee [34]. Prior to clinical implementation of microfluidics to coagulation testing/diagnosis, it will also be necessary to establish a rigorous and clearly defined standardized procedure for blood collection and subsequent processing, and to determine a uniform chamber material and dimension as well as the surface treatments to enable and ensure universal testing conditions. It will also be necessary to define standards for imaging machinery and thresholds for responses to have clear parameters to interpret the results of the tests. To set these parameters, it will be necessary to have clear endpoints that can be quantified, such as vWF collagen-binding efficiency for diagnosing von Willebrand disease [105], or clot formation time for diagnosing and assessing the severity of other bleeding disorders, in similarity to the endpoint of the current aPTT or PT testing. However, to generate these necessary parameters and establish the normal values of these endpoints, it will be important to compare these microfluidics-based tests to the traditional clinical tests and to test a large number of people with a wide range of phenotypes. As clinical uses of microfluidics become more prevalent, it will be necessary to build upon these guidelines to ensure the correct conclusions are drawn, given the possibility for small inherent manufacturing variations in devices.

Conclusion

Microfluidics is an important tool to be used to study the biochemical and cellular basis of coagulation in health and disease. It provides a method to quickly and easily produce customizable devices that can administer biologically relevant fluid flow, and thus wall shear stress, on cells. This allows for the recapitulation of physiological conditions that the cells experience in vivo to create a better in vitro model to study coagulation biology. There is no doubt that, in the coming years, the increasing use of microfluidics will allow for new and important findings in the fields of endothelial cell behavior, mechanotransduction pathways, coagulation protein production and behavior, and the testing of novel therapeutics in vitro.

Figure 1.

Shear responsive pathways in coagulation. Endothelial cells are capable of shear mechanotransduction through, caveolae, ion channels, GPCR, and integrins to regulate cell behavior and protein production. Increased shear stress also causes vWF to elongate and bind to platelets aiding in coagulation at the injury site.