Platelet functional testing via high-throughput microtiter plate-based assays

Abstract

Platelets play a critical role in hemostasis and thrombosis; therefore, in vitro assays that measure platelet reactivity are fundamental tools to gain insight into these physiological processes, diagnose platelet disorders, and develop antithrombotic therapies. However, conventional platelet assays such as aggregometry, the clinical gold standard for assessing platelet function, are low-throughput and require specialized equipment. Since platelets have a finite life span ex vivo, processes to miniaturize and multiplex assays allow a much broader overview of platelet function in significantly less time than conventional assays. Several groups have developed simplified, high-throughput approaches to quantify platelet activation with standard laboratory equipment to lower the barrier of entry to study platelet biology. This article describes a panel of optimized and validated high-throughput microplate assays to comprehensively assess platelet functionality, independently or in combination, to increase throughput and reduce costs. Specifically, following stimulation of platelets, a plate reader can be used to measure light transmission aggregation via absorbance, dense granule secretion based on ATP-dependent luminescence generation, and cytosolic calcium levels with a cell-permeant fluorescent Ca²⁺-sensitive dye. Additionally, platelets are an easily accessible component of the blood that share signaling pathways with other cells, making them ideal for high-throughput drug screens. The highly adaptable and complementary assays presented in this article can be used to decipher the molecular mechanism underlying platelet activation or identify novel inhibitors.

Basic Protocol 1:

Microtiter plate-based light transmission aggregometry

Basic Protocol 2:

Measuring dense granule secretion in high-throughput microplate assays

Basic Protocol 3:

High-throughput microplate assays to measure calcium mobilization in platelets

Support Protocol 1:

Platelet isolation and enumeration

INTRODUCTION

Platelets are small, anucleated cellular fragments derived from the cytoplasm of megakaryocytes that play a pivotal role in hemostasis and thrombosis¹. The formation of a platelet-rich hemostatic plug or pathogenic thrombus is a complex process that requires coordinated platelet activation in response to multiple stimuli. In order for the hemostatic plug or thrombus to grow and stabilize following vascular injury, platelets must adhere to the subendothelial extracellular matrix, secrete their granules, and form platelet-platelet interactions. While specialized assays have been developed to selectively study each of these processes, aggregometry, which measures platelet-platelet interactions, remains the gold standard for diagnosing patients with platelet disorders and is the most commonly used assay to study platelet reactivity. The development of aggregometry revolutionized platelet biology by helping researchers gain insight into thrombus formation and develop antiplatelet therapy. However, many limitations are associated with platelet functional assays including conventional aggregometry. The main challenge is the need for expensive specialized and dedicated equipment, like an aggregometer, that requires an experienced operator. Additionally, running a full agonist panel on a single donor with conventional assays requires a large sample volume and can be time-consuming and labor-intensive due to the limited number of samples that can be run simultaneously. This is especially consequential because platelets have a finite life span ex vivo, depending on the protocol used to prepare the platelets, due to the continuous decrease in platelet activity over time. Therefore, current techniques allow only a limited number of samples to be assayed reproducibly.

To address the limitations of conventional techniques, several platelet assays have been adapted to analyze platelet function in a microtiter plate, which allows a more in-depth high-throughput analysis of platelet function with a lower volume of platelets, in significantly less time than conventional methods^2–4. This article will discuss detailed protocols for measuring platelet aggregation, dense granule secretion, and calcium (Ca²⁺) mobilization in microtiter plates independently or combined. The simplest and most cost-effective of these assays is microtiter plate light-transmission aggregometry, as described in Basic Protocol 1, in which an agonist is added to platelets in a 96-well plate. As platelets activate in response to the agonist, they form clumps known as aggregates, which allows more light transmission. This can be quantified by measuring absorbance with a plate reader. Microplate readers and shakers are common pieces of laboratory equipment, and are therefore accessible to most labs. The elimination of the need for an aggregometer greatly reduces the initial setup costs for this assay.

Additionally, a multi-mode plate reader capable of measuring luminescence and fluorescence can be used to perform microplate-based platelet dense granule secretion and Ca²⁺ mobilization in real-time or as dedicated end-point assays⁵. Similar to conventional lumi-aggregometry, Basic Protocol 2 describes methodologies to measure dense granule secretion in high-throughput microplate assays. Aggregation and luminescence can be read in parallel from the same sample in a 96-well plate with the simple addition of commercially available luciferase/luciferin mixtures either prior to or immediately after agonist stimulation. Furthermore, the rise in cytosolic Ca²⁺, which is essential to most platelet functions such as aggregation and granule secretion, can be easily monitored in the presence of Ca²⁺-sensitive dyes. Basic Protocol 3 provides a detailed description of high-throughput microplate assays to measure Ca²⁺ mobilization in platelets. Lastly, Support Protocol 1, Platelet isolation and enumeration, offers practical advice on handling platelets, highly sensitive cells that are susceptible to preanalytical processes and incidental activation. Together the miniaturized platelet functional assays retain their sensitivity, and provide a more accessible, highly adaptable platform to acquire large amounts of data or perform multiplexed drug screening.

STRATEGIC PLANNING

Improper whole blood collection or sample handling can significantly alter platelet responsiveness to agonists or lead to inadvertent agonist-independent spontaneous platelet activation. Although preanalytical procedures have been kept to a minimum in these protocols, several critical preanalytical variables should be considered and rigorously standardized before testing platelet function to ensure accurate and consistent results. Furthermore, verify blood donors have no history of bleeding disorders and have refrained from using substances known to modulate platelet activation, including abstaining from COX inhibitors, non-steroidal anti-inflammatory drugs, or aspirin for 7 days prior to the blood draw.

Blood Collection

The act of drawing blood is a potential source of artifactual platelet activation. To reduce the most common sources of platelet activation, tissue factor contamination, and red blood cell lysis, we recommend the following; 1) ensure a clean needle stick with a large bore needle 21 gauge or larger, 2) discard the first vacutainer to capture any tissue factor generated as part of the initial needle stick as well as remove the dead space in the line to ensure proper fill volume, and 3) confirm smooth blood flow during collection. Vacutainers should be inverted immediately after collection to ensure the blood is adequately mixed with the anticoagulant. Without the addition of an anticoagulant whole blood would clot within minutes of collection⁶. However, the addition of anticoagulants that chelate Ca²⁺ or inhibit thrombin introduce a number of challenges, reviewed in-depth^7,8. For experiments with platelets suspended in plasma typically known as platelet rich plasma (PRP), we recommend collecting blood in sodium citrate, a weak Ca²⁺ chelator, while for isolating platelets, acid citrate dextrose (ACD) or sodium citrate can be used.

Sample Handling

The proper handling of the sample following blood collection is critical to obtain robust platelet reactivity data. For instance, time between blood draw and sample preparation should be minimized to reduce spontaneous platelet activation. Prolonged storage of platelets at cold temperatures can activate platelets⁹; therefore, whole blood, PRP, and platelets should be kept at room temperature or 37 °C. Furthermore, store samples in polypropylene tubes and avoid unnecessary agitation prior to use. Platelets should be allowed to rest for 30 minutes following resuspension to recover. Below we will discuss methods to validate your method of platelet isolation and handling to ensure spontaneous activation is kept to a minimum. All platelet functional tests should be performed within 4-6 hours of the blood draw, depending on how the platelets were prepared and stored. Alternatively, a time-course can be performed with submaximal concentrations of agonists to determine when platelets lose >10% of their initial activity based on your isolation protocol.

Basic Protocol 1 – Microtiter plate based light transmission aggregometry

This protocol describes light transmission aggregometry performed in a microtiter plate to assess agonist-dependent platelet-platelet interactions. Briefly, platelets will be aliquoted into the appropriate wells of a half-area clear bottom 96-well plate. After adding an agonist, the plate is shaken for 5 - 10 minutes at 37° C, and the absorbance is read on a plate reader. A simple previously published equation can be used to calculate the final percent of aggregation for each sample¹⁰. A diagram depicting how the aggregation formula was derived is shown in Figure 1A,B. For most assays, platelet activation should be measured in a dynamic, submaximal range of agonist concentrations. See Table I for suggested concentrations of common agonists (Table I); however, the investigator should optimize agonist concentrations (Figure 1D). PRP may be substituted for platelets in the protocol, so long as platelet poor plasma (PPP) is used as the blank rather than Tyrode’s, and absorbance is measured at 595 nm instead of 405 nm (Figure 1D). Platelets should be resuspended at 3.0 x 10⁸ platelets/mL. If a hematological analyzer is unavailable platelet count can be adjusted based on optical density at 405 or 800 nm (Figure 1E). For aggregation assays with PRP, it’s unnecessary to adjust the platelet count^3,11,12.

Figure 1: Principles of light-transmission aggregometry.

A) Left panel: Graphical illustration depicting light transmission through cuvettes representing 1) resting platelet-rich plasma (PRP) to measure the nominal light transmittance at baseline, 2) agonist-stimulated PRP with platelet aggregates allowing an increase in light transmission, and 3) a blank used to measure the maximum light transmission possible. Right panel: Depiction of a conventional aggregation tracing of agonist-stimulated platelets to illustrate how the percent aggregation is calculated based on the change (Δ) in the optical density (OD) between the sample and baseline divided by the difference in the OD between baseline and blank, which represents the maximum amount of change possible in the OD. B) The complete mathematic formula used to calculate the percent of aggregation from the OD of the sample, the OD of the baseline (0% aggregation), and the blank (100% aggregation). C) Image of wells containing platelets stimulated with increasing concentrations of protease-activated receptor-4-activating peptide (PAR4-AP) and shaken for 5 minutes following the procedures described in Basic Protocol 1. The OD and aggregation (%) are provided for visual comparison for each well. D) A spectral scan of wavelengths ranging from 300-800 nm was used to determine the wavelength that maximizes the change in OD (red dotted line) between 100 μL of the PRP and platelet-poor plasma (PPP) or between platelets and Tyrode’s buffer. An OD of 1.0 corresponds to 10% light transmission; therefore, while lower wavelengths of light provide a larger maximum change in OD between PRP and PPP, they should be avoided due to the increased chance of error caused by readings above an OD of 1. E) There is a linear relationship between the concentration of platelets and the OD of the sample in a 96-well plate at all wavelengths tested between 405 and 800 nm. For comparison, the OD of 50 μL of platelets at 405 nm (blue line) and 800 nm (black line) was reported for platelet concentrations ranging from 0.5 - 6.0 x 10⁸ platelets/mL.

Table I.

common platelet agonists

Agonist	Cat. no.	Company	Suggested dose response	Receptor	Notes
Protease-activated Receptor 1 activating peptide (PAR1-AP)	Custom Order	GL Biochem	0.156, 0.3125, 0.625, 1.25, 2.5, 5, 10 and 20 μM	PAR1	One letter amino acid peptide sequence SFLLRN-NH2
Protease-activated Receptor 4 activating peptide (PAR4-AP)	Custom Order	GL Biochem	18.75, 25, 37.5, 50, 75, 100, 150, and 200 μM	PAR4	One letter amino acid peptide sequence AYPGKF-NH2
Thrombin, α -Thrombin (Factor IIa)	HT 1002a	Enzyme Rearch Lab	0.156, 0.3125, 0.625, 1.25, 2.5, 5, 10 and 20 nM	PAR1/PAR4	Cleaves fibrinogen to fibrin, use the tetrapeptide GPRP to block fibrin cross-linking when using high concentrations of thrombin
Collagen, native collagen fibrils (type I) from equine tendons	Cat no. 385	Chrono-Log	0.156, 0.3125, 0.625, 1.25, 2.5, 5, 10 and 20 μg/mL	GPVI, α2β1	Liable in physiological solution
Adenosine diphosphate (ADP)	A2754	Sigma	0.156, 0.3125, 0.625, 1.25, 2.5, 5, 10 and 20 μM	P2Y12, P2Y1	Isolated platelets may be refractory to ADP
Epinephrine	E4250	Sigma	0.78, 1.56, 3.125, 6.25, 12.5, 25, 50, 100 μM	α2a	Isolated platelets respond poorly to epinephrine
Arachidonic acid (AA)	90010	Cayman Chemical	7.8, 15.6, 31.25, 62.5, 125, 250, 500, 1000 μM	TPα/β	Should be use to access whether donor was on a COX inhibitor
anti-CD9 antibody (clone ALB6)	IM0117	Beckman Coulter Inc.	0.156, 0.3125, 0.625, 1.25, 2.5, 5, 10 and 20 μg/mL	FcγRIIA	Slow kinetics compared to other agonists
Fibrinogen	F3879	Sigma	100–250 μg/mL	αIIbβ3	The stimulation of washed platelets with a weak agonist may require the addition of exogenous fibrinogen to see aggregation

Materials:

• Platelets or platelet-rich plasma (see Support Protocol 1) Tyrode’s buffer (see recipe)

• Platelet agonist(s) (see Table I)

• Bovine serum albumin (BSA) (Thermo Fisher Scientific Inc, cat. no. BP9706) iloprost (CAS No. 78919-13-8; Cayman Chemical, cat. no. 18215)

• Cangrelor (CAS No. 163706-36-3; Cayman Chemical, cat. no. 22086)

• 96-well half area clear flat bottom plates, not treated/nonsterile (Corning, cat. no. 3695) Orbital plate shaker (Eppendorf ThermoMixer^® FP or equivalent)

• Plate reader (BioTek Synergy H1 Hybrid Multi-Mode Reader or equivalent) Multichannel pipettes (10 and 200 μl)

• 96-well conical (V) bottom polypropylene plate (USA Scientific, cat. no. 1833-9600) Microplate Sealing Film (USA Scientific, cat. no. #2920-0000)

• Computer running data analysis software (e.g., GraphPad Prism)

1. Prepare working stocks of each agonist in a separate V-bottom 96-well plate such that 10 μL of each working stock can be used to achieve the desired final concentration in a volume of 100 μL.

   If agonists will not be used immediately the 96-well plate can be sealed and stored at 4 °C.

   Most agonist dose-responses are steep; therefore, 2-fold serial dilutions or lower are recommend.

   A wide range of agonist volumes are suitable to initiate aggregation; however, we recommend adding 10% of the final volume. Larger volumes will decrease the OD of the sample while lower volumes, less than 2 μL, may lead to an increase variability in the data due to pipetting error, especially if using a multichannel pipette.

2. Aliquot 90 μL of platelets (See support protocol 1) into the wells of a 96-well half area clear flat bottom plate.

   If applicable, treat platelets with pharmacological inhibitor or vehicle control prior to aliquoting them into individual wells. Aggregation can be performed in conjunction with either dense granule secretion or calcium mobilization using this same protocol with only slight modification. Chrono-lume, or Fluo4-AM can be added to platelets at this step and luminescence and fluorescence can be read with absorbance at Step 6. The use of Chrono-lume or Fluo4-AM is discussed below in detail in Basic Protocol 2 and 3, respectively.

   Lower volumes of platelets can be used; however, to prevent the spontaneous aggregation observed at lower sample volumes, treat the wells with 75 μL of 10% BSA (w/v) in Tyrode’s for 15 minutes at 37 °C, and wash twice with 100 μL of Tyrode’s before adding platelets (Figure 2A). Additionally, slightly more expensive microtiter plates coated with a proprietary nonbinding surface (Corning; cat no 3881) can be used to reduce spontaneous binding if required.

   Avoid introducing air bubbles into the well, as they may interfere with absorbance readings.

3. Transfer 10 μL agonist(s) from the 96-well V-bottom plate to wells containing platelets with a multichannel pipette.

   Equivalent results have been achieved by adding platelets to wells containing agonist with a multichannel pipette.

4. Shake the plate at 1000 rpm for 5 minutes at 37 °C.

   Previous studies have shown the OD of aggregated platelets is stable for at least 15 minutes¹³.

   Some agonists such as anti-CD9 antibodies may require a longer time to initiate platelet aggregation.

   Orbital plate shakers have unique orbit shapes or diameters and may need to be optimized for shaking speed.

5. Aliquot 90 μL of platelets + 10 μL of Tyrode’s buffer into an unused well to use as unstimulated resting platelet suspension representing 0% aggregation and 100 μL of Tyrode’s buffer into another well to use as a buffer only control representing 100% aggregation.

6. Read the absorbance of the plate at 405 nm.

   Other wavelengths have been used to successfully quantify platelet aggregation. See spectral scan of PRP and platelets (Figure 1D). The OD of platelet concentrations below 6.0 x 10⁸ platelets/mL are linear at wave-lengths from 405 to 800 nm (Figure 1E). However, ODs greater than one should be avoided as they correspond to < 10% light transmittance and may contribute to increased variability in readings.

7. Calculate the percent of aggregation. Detailed methods to calculate results are discussed below. Aggregation data is typically plotted as a concentration-response curve (Figure 2C–E) and is often used to test the effect of pharmacological inhibitors (Figure 3).

   Percent of aggregation = ((OD of 0% aggregation control – OD of sample)/(OD of sample 0% aggregation control – OD of buffer control))*100

Figure 2: Microtiter plate light-transmission aggregometry.

A) 5% or less spontaneous, agonist-independent aggregation is recommended for most assays. Increasing the sample volume from 50 to 100 μL decreases the amount of spontaneous aggregation in either PRP or platelets; experimental details to measure spontaneous aggregation are provided in Support protocol 1. Blocking wells with 10% bovine serum albumin (BSA) for 15 minutes prior to the addition of platelets reduced spontaneous platelet aggregation. B) The spontaneous aggregation observed with 50 μL of platelets was as effectively blocked by coating wells with BSA as treatment of platelets with iloprost (20 nM), a potent platelet inhibitor, or RGDW (4 mM), a small peptide that blocks the interaction of αIIbβ3 with fibrinogen. C) To demonstrate the reproducibility of microtiter-based aggregometry, six technical replicates were performed with platelets stimulated with increasing concentrations of protease-activated receptor-4-activating peptide (PAR4-AP), and the optical density (OD) at 405 nm is reported. D) Six-point concentration-response curves (0.156 – 20; see the graph for units) were performed with platelets and common agonists, anti-CD9 antibodies (n = 7), collagen (n=14), ADP (n=4) and thrombin (n=12). E) Concentration-response curves were performed with PRP and standard agonists, anti-CD9 antibodies (n = 5), collagen (n=4), and ADP (n=5). Data represent mean ± SD.

Figure 3: Microtiter plate light-transmission aggregometry with pharmacological inhibitors.

A) Platelets were stimulated with 0.5 or 5 μg/mL of collagen in the presence of increasing concentrations of ibrutinib, a Bruton’s tyrosine kinase inhibitor. Data represent mean ± SD. B) Platelets were incubated with increasing concentrations of a single-chain variable fragment (clone IV.3) that is known to block the function of FcγRIIA and then stimulated with anti-CD9 antibodies, a FcγRIIA-dependent agonist. Data represent mean ± SD. C) ADP and epinephrine-mediated aggregation of platelets treated with cangrelor, a P2Y₁₂ receptor inhibitor. Concentration-response curves of platelets stimulated with D) thrombin, E) collagen, and F) anti-CD9 antibodies in the presence of iloprost (20 nM), cangrelor (1 μM) or vehicle control (DMSO), n =4; Data represents mean ± SD.

Basic Protocol 2 – Measuring dense granule secretion in high-throughput microplate assays

Introductory paragraph:

Following stimulation, platelets initiate several positive feedback mechanisms, including the secretion of dense granules. Dense granules contain small molecules including ADP, ATP, and serotonin that elicit platelet activation. This protocol uses a microtiter luciferin-luciferase bioluminescence assay to measure ATP release from agonist-stimulated platelets as a marker for dense granule secretion. The method is based on the consumption of extracellular ATP by the enzyme luciferase, which oxidizes its substrate luciferin to oxy-luciferin. The decay of oxy-luciferin emits visible light that can be quantified by a plate reader (Figure 4A). The luciferin and luciferase are provided at an excess, so the rate-limiting step in luminescence generation is the amount of ATP released. Thus, the luminescence generated is proportional to the amount of ATP secreted by the platelets. Importantly, luminescence is linear when measuring the exogenous addition of ATP to platelets at the concentrations tested up to 40 μM (Figure 4B).

Figure 4: Measuring dense granule secretion in high-throughput microplate assays.

A) Graphical illustration depicting dense granule secretion from stimulated platelets. The presence of a luciferin-luciferase mixture can convert the extracellular ATP released from dense granules into light via a chemical reaction. B) The luminescence (relative luminescence units; RLU) generated by the addition of exogenous ATP (0.078-40 μM) to platelets incubated with Chrono-lume is linear (n=4, Data represents mean ± SD) C) ATP secretion from platelets stimulated with increasing concentrations of collagen (0.078-20 μg/mL) was measured at various gains on a mutli-mode plate reader. The RLU increases as the gain increases, but the percent of ATP secretion measured relative to the maximum RLU for each gain remains constant. D) Platelets were stimulated with a range of thrombin or collagen concentrations, and dense granule secretion was measured as a single-end point read. E) Lumi-aggregation, in which parallel aggregation and luminescence readings are taken on the same well, was performed with PAR4 stimulated platelets treated with RGDW (4 mM), iloprost (20 nm), aspirin (100 μM) or vehicle-only control (ethanol 0.1%; Ctrl). F) Six technical replicates were performed with 50 μM PAR4-AP to determine the reproducibility of real-time ATP secretion measurements. G) Real-time measurement of ATP secretion following PAR4 stimulation of platelets. Data can be presented as representative tracings plus maximum secretion or composite tracings (n=3, Data represents mean ± SD). H) Real-time ATP secretion monitored in whole blood stimulated with anti-CD9 antibodies in the presence of iloprost (20 nM; ilo; n=3), cangrelor (1 μM; can; n=6) or vehicle control (Ctrl; n =6), Data represents mean ± SD.

Most modern plate readers allow users to adjust the gain of luminescence. The relative luminescence units (RLU) of a sample increases as the gain increases, but the percent change compared to maximum release is constant (Figure 4C). This assay can be performed as either a real-time or static end-point assay with aggregation, lumi-aggregometry (Figure 4D–G). Platelets incubated with Chrono-lume can be stimulated and shaken for 5 minutes and a single luminescence read can be performed (Figure 4D). As noted above, lumi-aggregometry can be performed in a 96-well plate simply by adding 1 volume of Chrono-lume to 24 volumes of platelets prior to aliquoting platelets into individual wells (i.e., add 40 μL of Chrono-lume to 960 μL of platelets then aliquot 90 μL of mixture to each well) (Figure 4E). Then follow the same protocol for 96-well plate aggregation in Basic Protocol 1, but measure both absorbance and luminescence at the end of plate shaking. The high sensitivity of luminescence detection can also be used to measure dense granule secretion in whole blood or PRP, which limits preanalytical procedures and allows testing dense granule secretion under more physiological conditions¹⁴ (Figure 4H).

Materials:

• Platelets (see Support Protocol 1) Tyrode’s buffer (see recipe) Platelet agonists (see Table I)

• Chrono-Lume reagent (ChronoLog Corp., cat. no. 395) ATP standard (ChronoLog Corp., cat. no. 387)

• 96-well half area white/clear flat bottom (Corning, cat. no. 3883)

• 96-well Black-walled Half Area Clear Flat Bottom (Corning, cat. no. 3880)

• Plate reader (BioTek Synergy H1 Hybrid Multi-Mode Reader or equivalent) Multichannel pipettes (10 and 200 μl)

• 96-well conical (V) bottom polypropylene plate (USA Scientific, cat. no. 1833-9600) Microplate Sealing Film (USA Scientific, cat. no. #2920-0000)

1. Prepare working stocks (see Table 1) of each agonist in a separate conical (V) bottom 96-well plate such that 5 μL of each working stock can be used to achieve the desired final concentration in a final volume of 50 μL.

   If agonists will not be used immediately the 96-well plate can be sealed and stored at 4 °C.

   Most agonist dose-responses are step; therefore, 2-fold serial dilutions or lower are recommend. See Table I for suggested concentrations of common agonists.

   While spontaneous aggregation was observed in end-point assays with 50 μL of platelets (Figure 2), in real-time dense granule secretion assays we didn’t observe an increase in baseline ATP release overtime using 50 μL (Figure 4G). Therefore, we choose to use 50 μL for the real-time ATP secretion assays to maximize the number of conditions that could be tested with the platelets and converse reagents.

2. Add 2 μL Chrono-lume, a luciferin-luciferase mixture, to 43 μL platelets and incubate for 5 minutes in the wells of a 96 well half area white or black-walled, clear flat bottom plate.

   To decrease the amount of pipetting and ensure Chrono-lume is equally distributed in each well, we recommend adding Chrono-lume to the platelets and then aliquoting 45 μL of the platelet/Chrono-lume mix into the plate. If using an antiplatelet agent with an incubation time of less than 10 minutes, add Chrono-lume to all platelets first and then aliquot and treat the platelets with the inhibitor or vehicle control.

   When using PRP or whole blood add 5 μL Chrono-lume to 40 μL of sample and perform protocol as described.

3. Program a kinetic read on a Bio-Tek Synergy HT or equivalent multimode plate reader to perform steps 4-7.

4. Read baseline luminescence prior to agonist stimulation.

   To optimize the gain some machines provide an autogain function. If this function is not available, we recommend selecting a gain that maximizes the change in luminescence between an unstimulated sample and a maximally stimulated sample, 10 nM thrombin or highest concentration of agonist used for the experiment. Once selected, the gain should be kept the same for all experiments in the same study.

   Baseline luminescence can vary between donors and even between different preparations of platelets from the same donor depending on the quality of platelet preparation¹⁵.

5. Eject plate and transfer 5 μL agonist to platelets from 96-well v-bottom plate with a multichannel pipette.

   pipette up and down 3 times to mix agonist, being careful not to introduce bubbles into the sample.

6. Read plate luminescence post agonist stimulation.

7. Shake plate continuously at 800 rpm for 5 minutes while reading luminescence every 30 seconds.

   Shaking is not required for all agonists, such as thrombin, PAR1-AP and PAR4-AP. However, some agonist such as collagen and anti-CD9 antibodies will have a marked increase in granule secretion when platelets are shaken. Therefore, when using certain agonists, the number of samples that can be run simultaneously is limited by the time required to read those samples without infringing on the shaking time between reads. The more samples there are, the longer it will take to read them, and the less time the platelets will have to shake between reads. To increase the number of samples that can be run simultaneously, plate readers can be set to “sweep mode” or the interval between reads can be increased to 45-60 seconds.

8. Report data as relative luminescence units.

   Due to variability in background ATP secretion and inter-individual variability, normalizing data as percent of maximum can reduce variation in the data. If absolute quantification of ATP release is required, prepare a standard curve of ATP (0-40 μM) and mix 5 uL of ATP dilution with 43 μL platelets + 2 μL Chronolume. Read the ATP standard as an end-point read before or after measuring real-time ATP secretion.

Basic Protocol 3 – Microtiter plate based calcium mobilization

Introductory paragraph:

The rise in intracellular Ca²⁺ is an early second messenger that is common and essential to most platelet functions, including platelet aggregation and granule secretion. Upon the activation of platelets basal cytosolic Ca²⁺ levels go from ~100 nM to low micromolar levels depending on the type and amount of agonist used. This protocol describes the use of a non-ratiometric, fluorescent Ca²⁺ sensitive dye, Fluo-4, to measure rises in intracellular Ca²⁺ following agonist stimulation. Ca²⁺ mobilization is transient following stimulation with many agonists; therefore, it is best to measure changes in Ca²⁺ mobilization in real-time (Figure 5A–C). However, for drug screens Ca²⁺ mobilization can be measured as a single-end point by simply adding 1 μl Fluo-4 AM per 499 μl of platelets and performing Basic Protocol 1.

Figure 5: High-throughput microplate assays to measure calcium mobilization in platelets.

A) Six technical replicates were performed with 100 μM PAR4-AP to determine the reproducibility of real-time Ca2+ mobilization. Data were reported as mean fluorescence intensity (MFI). B) Real-time intracellular Ca2+ mobilization in response to thrombin (5 nM), PAR1-AP (5 μM), or control (ctrl). Data represent mean ± SD, n=3. C) A single end-point fluorescence measurement was made four minutes after stimulation with thrombin (5 nM), PAR1-AP (5 μM), PAR4-AP (200 μM), or control (ctrl), using Fluo-4 AM treated platelets. D) The rise in cytoplasmic Ca2+ was measured in real-time following stimulation with increasing concentrations of PAR4-AP (n=3, Data represent mean ± SD).

Additionally, calcium can be multiplexed with dense granule secretion for real-time measurements (Figure 6A–D) or as an end-point assay with aggregation and dense granule secretion (Figure 6E).

Figure 6: Multiplexing platelet functional assays in microplates.

A) Calcium mobilization and B) ATP secretion were monitored in real-time following stimulation of washed platelets with anti-CD9 antibodies in the presence of iloprost (20 nM; ilo), cangrelor (1 μM; can) or vehicle control (Ctrl). n=3, data represents mean ± SD. Platelets were stimulated with thrombin, and C) Calcium mobilization and D) ATP secretion was monitored in real-time in the presence of iloprost (20 nM; ilo), cangrelor (1 μM; can) or vehicle control (Ctrl). n=3, data represents mean ± SD. E) Aggregation, dense granule secretion and calcium mobilization were measured in collagen stimulated platelets as a single end-point after 4 minutes of stimulation. n=3, data represents mean ± SD.

Materials:

• Platelets (see 11) Tyrode’s buffer (see recipe) Platelet agonist (Table I)

• Fluo-4 AM, (Invitrogen, cat. no. F14201)

• Calcium Chloride (CaCl2; Invitrogen, cat. no. F14201)

• 96-well Black-walled Half Area Clear Flat Bottom (Corning, cat. no. 3880) Plate reader (BioTek Synergy H1 Hybrid MultiMode Reader or equivalent) Multichannel pipettes (10 and 200 μl)

• 96-well conical (V) bottom polypropylene plate (USA Scientific, cat. no. 1833-9600) Microplate Sealing Film (USA Scientific, cat. no. #2920-0000)

1. Prepare working stocks of each agonist in a separate conical (V) bottom 96-well plate such that 5 μL of each working stock can be used to achieve the desired final concentration in a final volume of 50 μL.

   If agonists will not be used immediately the 96-well plate can be sealed and stored at 4 °C.

   Most agonist dose-responses are steep; therefore, 2-fold serial dilutions or lower are recommend. See Table I for suggested concentrations of common agonists.

2. Add 1 μL Fluo-4 AM (1 μg/mL in DMSO), a fluorescent Ca²⁺ sensitive dye, to 498 μL platelets and incubate for 10 minutes at 37°C. Immediately before aliquoting platelets add 1 μL of 1 M CaCl₂ for a final concentration of 2 mM.

   If using an antiplatelet agent with an incubation time of less than 10 minutes, add Fluo-4 AM to all platelets first and then aliquot and treat the platelets with the inhibitor or vehicle control.

3. Aliquot 45 μL of Fluo-4 AM treated platelets into the wells of a 96-well half area black-walled clear flat bottom plate.

4. Program a kinetic read on a Bio-Tek Synergy HT or equivalent multimode plate reader to perform steps 5-7.

5. Read baseline fluorescence (Ex/Em 494/506) prior to agonist stimulation.

   To optimize the gain some machines provide an autogain function. If this function is not available, we recommend selecting a gain that maximizes the change in fluorescence between an unstimulated sample and a maximally stimulated sample without oversaturating the signal. Once selected the gain should be kept the same for all experiments in the same study.

6. Eject plate and transfer 5 μL agonist to Fluo-4 AM treated platelets from 96-well v-bottom plate with a multi-channel pipette.

   Pipette up and down 3 times to mix agonist, being careful not to introduce bubbles into the sample.

7. Read plate fluorescence post agonist stimulation.

8. Shake plate continuously at 800 rpm for 4 minutes while reading fluorescence every 20 seconds.

   Shaking is not required for all agonists, such as thrombin, PAR1-AP and PAR4-AP. However, some agonist such as collagen and anti-CD9 antibodies will have a marked increase in calcium mobilization when platelets are shaken. Therefore, when using certain agonists, the number of samples that can be run simultaneously is limited by the time required to read those samples without infringing on the shaking time between reads. The more samples there are, the longer it will take to read them, and the less time the platelets will have to shake between reads. To increase the number of samples that can be run simultaneously, plate readers can be set to “sweep mode” or the interval between reads can be increase to 45-60 seconds.

9. Report data as mean fluorescence intensity.

   Due to inter-individual variability, normalizing data as percent of maximum can help reduce variation.

Support Protocol 1 – Whole blood collection, preparation of PRP and isolation of platelets

This protocol provides a detailed procedure for the isolation of PRP or platelets from human whole blood. PRP and platelets can be separated from other cellular components of blood by centrifugation.

Materials

• 10x acid citrate dextrose (ACD) (see recipe) Apyrase (50 U/mL) (Sigma, cat. no. A6410)

• Prostaglandin E1 (PGE1; CAS 745-65-3; 50 μg/mL; Cayman Chemical, cat. no. 13010) Tyrode’s buffer (see recipe)

• sodium citrate vacutainer (Becton, Dickinson and Company, cat. no. 363083) ACD vacutainer (Becton, Dickinson and Company, cat. no. 364606)

• Benchtop centrifuge (Eppendorf 5810 or equivalent)

• Transfer pipette (Thermo Fisher Scientific Inc, cat. no. 13-711-7M) 15 mL polypropylene conical tubes

• 50 mL polypropylene conical tubes

• Automated hematology analyzer (Drew Scientific, Hemavet 950FS or equivalent)

Preparation of platelet-rich plasma (PRP)

1. Draw up to 50 mL of blood from the antecubital vein of a consenting donor with a 21 gauge needle into 5 mL sodium citrate vacutainers.

   Sodium citrate is the anticoagulant of choice for isolating PRP to use in functional assays. ACD or sodium citrate can be used for the isolation of platelets.

   The final yield of 3 x 10⁸ platelets/mL in Tyrode’s buffer from one mL of whole blood usually is between 0.3-0.8 mL depending on the donor’s platelet count, which can range from 150,000-450,000 platelets/μL in whole blood.

2. Centrifuge citrated whole blood in vacutainers at 200 x g for 10 minutes at room temperature with low acceleration set to “2” and brake set to “0” (off) on an Eppendorf 5810.

   The centrifugation time is dependent on the volume of blood in the vacutainer being spun. The time listed is the total centrifugation time including acceleration for a 5 mL vacutainer. If using 8.5 mL vacutainers the total centrifugation time should be increased to 15 minutes.

3. Transfer platelet-rich plasma to a clean 15 or 50 mL polypropylene conical tube with a plastic transfer pipette, being careful not to disturb the buffy coat or red blood cell layers.

   To decrease red blood cell contamination leave ~10% PRP remaining in tube.

4. Centrifuge remaining whole blood in vacutainers at 2000 x g for 10 minutes at ambient temperature with acceleration set to “9” and brake on to obtain PPP. Transfer PPP to a clean polypropylene conical tube, being careful not to disturb the buffy coat and red blood cell layers.

   Only perform this step to collect platelet-poor plasma (PPP) as a blank if you are performing aggregometry with PRP. If using washed platelets, the remaining whole blood can be discarded.

Isolation of platelets from PRP

• 5.
  Add 1 volume of 10x ACD to 9 volumes of PRP then add apyrase to a final concentration of 0.02 U/mL and/or PGE₁ at 50 ng/mL to PRP and mix gently by inverting three times.

  PGE₁ is a prostacyclin (IP) receptor agonist that activates protein kinase A (PKA) and elicits a strong antiplatelet effect^16,17. Alternatively, prostacyclin (PGI₂) can also be used at a final concentration of 0.5 μM instead of PGE₁ but is more liable than PGE₁.

  Apyrase is an enzyme that degrades ATP or ADP released from platelet dense granules during platelet processing.

  We find the addition of both PGE₁ and apyrase results in less chance of spontaneous platelet activation during isolation, but ACD plus either PGE₁ or apyrase alone can be used to isolate platelets.

• 6.Centrifuge PRP at 2000 x g for 10 minutes at room temperature with acceleration set to “2” and brake set to “2”.
• 7.
  Remove the platelet-poor plasma (PPP) supernatant, and gently resuspend the platelet pellet(s) in Tyrode’s buffer.

  For new donors, resuspend platelets in 0.2 mL of Tyrode’s buffer per mL of whole blood collected. The amount of platelets isolated from a donor remains relatively consistent over time. Therefore, for repeat donors, resuspend their platelets in 80% of their previous total platelet volume, assuming the same amount of blood was drawn.

• 8.
  Count platelets using a hematology analyzer (Hemavet 950FS or equivalent)

  If a hematology analyzer isn’t accessible, the platelet count can be adjusted to a standard OD using a plate reader. The OD of 50 μL of 3.0 x 10⁸ platelets/mL is 0.455 at 800 nm or 0.745 at 405 nm (Figure 1D). Although OD of platelets can vary between donors based on a number of intrinsic platelet factors such as their size, and granule content, the use of OD to estimate platelet count is accurate enough to obtain reproducible data.

  $$\text{Platelet count}(\times {10}^8\text{platelets}/\text{mL})=({\text{OD}}_{800}-0.06276)/0.1309$$

  $$\text{Platelet count}(\times {10}^8\text{platelets}/\text{mL})=({\text{OD}}_{405}-0.07582)/0.1419$$

• 9.
  Test platelets for the amount of spontaneous, agonist-independent aggregation to examine the quality of the platelet preparation. Aliquot 90 μL of platelets and 10 μL of Tyrode’s buffer into the wells of microtiter plate, perform in duplicate. Aliquot 100 μL of Tyrode’s into a well as a blank to measure maximal light transmittance.

  The inadvertent activation of platelets during preparation is often undetectable until the platelets are shaken, which allows the platelets to come into contact with each other.

  Avoid introducing air bubbles into the well, as they may interfere with absorbance readings.

• 10.Read the optical density (OD) of wells at 405 nm on a plate reader.
• 11.
  Shake plate at 1000 rpm for 10 minutes at 37 °C.

  Aggregation can be performed at room temperature without a significant change in platelet responsiveness.

• 12.Re-read OD of wells at 405 nm on a plate reader.
• 13.
  Calculate the percent of spontaneous aggregation.

  Percent of spontaneous aggregation = ((OD of sample pre-shake – OD of sample post-shake)/(OD of sample pre-shake – OD of blank))*100

  Ideally, spontaneous aggregation should be less than 5% to use platelets for further functional assays.

REAGENTS AND SOLUTIONS

10x acid citrate dextrose buffer (ACD)

• 2.5% (w/v) sodium citrate tribasic

• 1.5% (w/v) citric acid

• 2.0% (w/v) D-glucose

  Use deionized distilled water. Pass through 0.22 μm filter before use (Stericup Vacuum Filtration System; Sigma, cat. no. S2GPU10RE). Store for up to 6 months at room temperature.

Tyrode’s buffer

• 10 mM HEPES

• 11.9 mM sodium bicarbonate

• 127.2 mM sodium chloride

• 5.0 mM potassium chloride

• 0.4 mM sodium phosphate monobasic

• 1.0 mM magnesium chloride hexahydrate

• 5.0 mM D-glucose

• 0.35% bovine serum albumin (BSA)

  Use deionized distilled water. Adjust pH to 7.4 with 1 M NaOH. Pass through 0.22 μm filter before use (Stericup Vacuum Filtration System; Sigma, cat. no. S2GPU10RE).

  Store for up to 3 months at 4° C. Bring to room temperature before use. Store for up to 1 week at room temperature.

COMMENTARY:

Background Information:

The formation of platelet-rich clots at the site of vascular damage in humans has been recognized since the late 1800s⁸. However, without the ability to methodically study platelet activation in vitro, the molecular and physiological mechanism underlying the accumulation of platelets at the site of vascular injury remained unknown. In the 1960s, Gustav Born developed a method to quantify platelet reactivity in vitro after determining that light transmittance was proportional to the number of platelets in suspension, now commonly referred to as aggregometry^18,19. The aggregometer was a significant breakthrough clinically for diagnosing individuals with platelet disorders and was instrumental in developing antiplatelet therapies. Furthermore, the advent of aggregometry, coupled with later advances in flow cytometry and microfluidic devices, revolutionized the study of platelet function leading to the discovery of crucial platelet agonists, their cognate surface receptors, and the signaling pathways they elicit. Although hemostasis and thrombosis are not identical processes, a similar sequence of events occurs in both. The formation of a platelet-rich hemostatic plug or pathogenic thrombus can be divided into two separate but overlapping stages: 1) platelet-matrix adhesion, including the tethering, activation, and spreading of platelets on subendothelial extracellular matrix components predominantly mediated by adhesion receptors, and 2) platelet-platelet interaction including cohesion, typically referred to as platelet aggregation and stabilization. Both platelet aggregation and stabilization are driven by the interaction of integrin α_IIbβ₃ with fibrinogen or another multivalent ligand, and reinforced by platelet-derived soluble agonists, including dense granule secretion, reviewed in depth^20–22.

Platelet-rich plugs grow in a well-defined hierarchal structure due to the distribution and concentration of agonist²³. While platelet adhesion receptors interact with numerous subendothelial extracellular matrix components following vascular injury, collagen, a GPVI and integrin α₂β₁ agonist, is often used for in vitro platelet activation due to its potency compared to other matrix components. The most potent soluble platelet agonist is thrombin, a serine protease activated as part of the coagulation cascade, which is restricted to the site of vascular injury and results in a core of highly activated, densely packed platelets. Around the core is an outer shell of loosely packed platelets that protrudes into the intravascular space and is chiefly driven by the soluble platelet-derived mediators, including dense granule secretion of small molecules like adenosine diphosphate (ADP) and formation of thromboxane A2 (TXA₂)^20–23. The selection of agonists for in vitro assays is usually done either based on their physiological role in the formation of hemostatic plug or to test specific intracellular signaling pathways. Usually, both immunoreceptor tyrosine-based activation motif (ITAM)-containing receptor and GPCR agonists are selected. When studying a novel pharmacological inhibitor or the role of a protein in platelets, a panel of agonists responsible for each of these processes can be used to characterize which signal transduction pathways are altered.

The primary limitation to conventional platelet functional assays is the need for specialized equipment, such as an aggregometer. Additionally, these assays are often time-consuming and labor-intensive due to their low throughput. To address these limitations, microtiter plate functional assays were developed²⁴. Microtiter plate assays retain their ability to detect changes in platelet reactivity in the presence of antiplatelet therapies and are being validated as clinical diagnostic tools^3,5,25. Further platelet functional tests have been adapted to microtiter plate assays, including dense granule secretion and calcium mobilization. For library drug screens, ultra-high throughput versions of these protocols that use 384-or 1536-well plates have been developed^26,27. The assays presented in this article have numerous applications in platelet research and drug discovery. These protocols are and either real-time or end-point data measurements can be collected, allowing a cross-comparison of a large number of agonists with multiple functional readings.

Critical Parameters:

The preanalytical variables, including the collection, processing, and storage of samples prior to running the functional assay, are sensitive to user manipulation, and are the most critical factor to optimize in order to achieve robust results²⁸. Different isolation protocols can significantly change the responsiveness of platelets to an agonist. In an extreme example, the release of ADP from platelet dense granules during isolation can cause the desensitization of ADP receptors resulting in platelets that are refractory to ADP stimulation, while other preparations have robust responses to ADP²⁹. Based on an international survey of experts, there is a general, but not universal, consensus on the best practices for preanalytical processing of samples prior to platelet functional tests^30,31. The lack of standardization in laboratory practices can in part be attributed to the inability to decipher what is an artifactual platelet response and what is true platelet reactivity. For example, studies have reported that platelets stored at ambient temperature have an increase in reactivity compared to platelets stored at 37°C³², while another study attributed the increased reactivity of platelets stored at ambient temperature to being an artifact caused by preactivation during storage³³. Standardization has been further hampered by conflicting studies on key variables which have been discussed in-depth by others^{29–31,34,35}. Briefly, the following parameters are often contested, including 1) choice of the anticoagulant, 2) blood collection method (vacutainers, syringe, drip), 3) storage temperature (ambient or 37°C)³², 4) components of Tyrode’s buffer (BSA, divalent cations, apyrase) and 5) platelet concentration²⁸.

The blood collection and sample handling techniques highlighted in Support Protocol 1 and the Strategic planning sections have been optimized to provide quiescent platelets at baseline that robustly respond to most agonists and can be inhibited by known pharmacological antagonists. However, the response of platelets prepared this way to ADP is much less responsive than ADP stimulation of PRP (Figure 2D, E). Furthermore, the endorsement of this protocol does not invalidate the use of other protocols. Researchers should be conscious of the advantages and limitations that exist for isolation protocols and strive to maintain consistency, and provide detailed methodologies on platelet processing in their publications.

The use of either platelets or PRP are suitable for most of the assays described in this article. Certain advantages and disadvantages exist for using either platelets or PRP. Platelet isolation requires more preanalytical handling than PRP; therefore, washed platelets take longer to acquire and are more susceptible to user manipulation. However, because platelets are resuspended in a buffered medium, they are stable for longer than PRP. PRP requires more assay reagents including Fluo-4 AM and Chrono-lume than platelets. Furthermore, pharmacological inhibitors are often bound by plasma proteins, lowering the effective free concentration of antagonist available to bind to platelets. Hence, the concentration of an inhibitor used in PRP is a good approximation to the amount of inhibitor required in whole blood. However, if the assays are being performed with a rare or expensive reagent, the use of isolated platelets might be better to test its functions. A multi-valent adhesive ligand capable of crosslinking α_IIbβ₃ on adjacent platelets, predominantly fibrinogen, in the medium is required for platelet aggregation. Plasma contains 2-4 mg/mL of fibrinogen to support aggregation mediated in PRP. With washed platelets, enough fibrinogen is released from platelet α-granules upon stimulation to support aggregation with most agonists. However, weak agonists (ADP, serotonin, or epinephrine) that activate α_IIbβ₃ without triggering α-granule secretion will not initiate aggregation without the supplementation of exogenous fibrinogen (100-250 μg/mL), see Table I. α-thrombin is part of the coagulation cascade responsible for converting fibrinogen to fibrin that also causes platelet activation through protease activated receptors (PARs). The important for clot stabilization, can facilitate the incorporation of unactivated platelets into clumps, which cannot be distinguished from aggregation by optical density and thus may give an artificially higher read than otherwise expected. A small peptide, Glycine-Proline-Arginine-Proline (GPRP; 1-5 mM) can inhibit fibrin crosslinking³⁶. Thus, the use of GPRP is recommended with higher concentrations of α-thrombin with platelets but is a requisite to use α-thrombin in PRP.

It’s essential to use the appropriate concentration of antagonist and agonist to achieve robust results. Interindividual differences have been observed in platelet activation to threshold concentrations of all agonists³⁷. Therefore, rather than using a single agonist dose, a concentration-response curve with a range of agonists is recommended to provide broad phenotype of platelet reactivity under the conditions being tested. Once diluted in a physiological buffer, some agonists, such as collagen, are labile, and a noticeable decline in platelet activation is observed after 30 minutes. As seen in Figure 3A, higher concentrations of agonists require increased amounts of antagonist to achieve inhibition, and a high enough concentration of agonist may bypass the need for a particular protein to achieve full aggregation. Therefore, pharmacological inhibitors should be tested with submaximal agonist concentrations. Furthermore, pharmacological inhibitors often have off-target effects at higher concentrations; thus, we recommend performing a dose-response of antagonist and using the lowest concentration of inhibitor that caused maximum inhibition, examples provided in Figure 3.

Additionally, a number of factors will influence the extent of aggregation, for example increasing shaking time or speed will enhance the amount of aggregation seen for a given agonist concentration. We have obtained similar lumi-aggregation data using a Thermo Scientific LP Vortex Mixer or an Eppendorf ThermoMixer FP. However, since different shakers may have different orbital diameters, mixing patterns, or slight differences in actual shaking speeds, concentration response curves should be established for each orbital plate shaker used. Real-time kinetics of platelet aggregation in a 96-well plate reader are often slow and require more agonist for the same effect due to the decrease in shaking speeds in most plate readers and the stoppage time required to read the plate¹³. Therefore, while real-time aggregation data can be collected in a plate reader for most experiments the added kinetic data is not impactful.

Platelet counts exists in a linear relationship with optical densities in the microtiter wells (Figure 1E). Besides platelet concentration, the OD of a given sample is dependent on the shape, size and granularity of the platelets which may all contribute to the overall changes in OD observed following the addition of agonist. Aggregation monitors bulk changes in platelet aggregation, one important limitation is its inability to detect small or microaggregates. Another important reminder is that OD is on a log scale; therefore, an OD of 1 corresponds to 10% light transmittance while an OD of 2 corresponds to just 1% light transmittance. Therefore, to obtain robust data OD readings under 1 are preferred for aggregation.

Troubleshooting:

Understanding Results:

Techniques capable of monitoring platelet reactivity in vitro are invaluable to better understand the molecular mechanism regulating platelet activation and the physiological processes underlying hemostatic plug or clot formation. The assays in these protocols are primarily designed to evaluate pharmacological agents known or hypothesized to inhibit platelet activation. To this end, we recommend using multiple agonists that signal through GPCRs and immunoreceptor tyrosine-based activation motif (ITAM)-containing receptors. The pathways leading from calcium mobilization to aggregation or dense granule secretion have redundant and distinct signaling components; therefore, these 3 assays were selected to help broadly characterize pharmacological inhibitors. Importantly, there is positive predictive value in using platelet functional assays to identify novel pharmaceutical antiplatelet therapies, which makes these assays a great tool to screen pharmacological libraries. However, caution should be used in directly extrapolating the findings of in vitro experiments to the role a drug may have on physiological hemostatic plug or thrombus formation. Complementary studies, including ex vivo whole blood microfluidic and in vivo mouse models, should be performed with any novel inhibitors identified by these techniques⁷. Lastly, the assays are readily adaptable to address other experimental questions. After data collection, the plate can be stored to perform additional biochemical assays with the samples, including western blot or ELISA.

There are limitations to conventional and 96-well-based platelet aggregation. Due to the feed-forward amplification mechanisms inherent to platelets, once a threshold level of activation has been reached with most agonists, roughly >30% aggregation will proceed to full >70% aggregation. Therefore, for most agonists, aggregation is typically a binary response of either low aggregation (< 30%) or full aggregation. In composite data, at threshold concentrations of agonists, a typical dose response more often reflects the proportion of responders and non-responders. Using single end-point reads to monitor final aggregation allows the simultaneous evaluation of multiple platelet signaling pathways (Figures 2D and 2E). However, with some weak agonists such as ADP, platelets can aggregate to greater than 30% and then disaggregate back to near baseline levels. Therefore, the final aggregation may not reflect the achieved maximal aggregation. In conventional aggregometry, immediately following the addition of most agonists, there is a paradoxical increase in OD that has often been attributed to platelet shape change but may also reflect the formation of small platelet aggregates³⁸.

A dose-response to assess an agonists’ dynamic, submaximal activation concentrations is recommended to detect inhibition of platelet activation. For most applications, a dose-response starting with a concentration that results in <20% aggregation should be performed with an additional 3 doses in the linear or submaximal response and 2-4 potent agonist concentrations above the concentration known to cause the maximal response. To determine the potency of investigational new drugs, we recommend comparing them to known antiplatelet therapies under similar conditions. In these studies, we used iloprost (20 nM), a potent platelet inhibitor, cangrelor, a medium potency P2Y₁₂ receptor inhibitor, and aspirin, a COX inhibitor with mild inhibitory effects in the 96-well plate format (Figures 3 and 4E). Vehicle controls are always required when studying inhibitors to ensure they do not have inhibitor activity. Ideally, DMSO and EtOH should be kept to ≤ 0.1%.

The luminescence generated by platelets treated with Chrono-lume is proportional to the amount of ATP released from dense granules (Figure 4B). The amount of luminescence is typically reported as maximum relative luminescence units since the luminescence generated by each plate reader experiment can vary. Interestingly, differences in ATP secretion can be observed between higher agonist concentrations with the same amount of aggregation. Similar to aggregation, there is inter-donor variability in granule secretion. Therefore, to reduce variability between donors, background levels of luminescence can be subtracted, and the data can be reported as fold-change relative to control, or ATP can be quantified as nmol/1x10⁸ platelets by using an ATP standard curve. Furthermore, we have used microtiter-based ATP secretion in whole blood to help screen for the appropriate amount of inhibitor in whole-blood microfluidic devices. Since other cellular components of blood are present, it is essential to note that even if platelet-specific agonists are used, other cellular sources may contribute to the levels of released ATP.

A hierarchy of platelet functions depends on the concentration of intracellular Ca²⁺, with increasing threshold levels of Ca²⁺ inducing different platelet functions, including shape change, integrin activation, secretion, and procoagulant activity²⁰. Using non-ratiometric fluorescent dyes such as Fluo-4 AM, Ca²⁺ mobilization is typically reported as maximum fluorescence intensity. Ratiometric Ca²⁺ sensitive dyes such as FURA-2 AM, can be used for absolute quantification of calcium. The baseline fluorescence intensity levels are dependent on the time and temperature used to incubate Fluo-4 AM with platelets. Different types or concentrations of agonists can elicit transient (PAR1-AP) or prolonged (thrombin) Ca²⁺ mobilization (Figure 5B). This is important because some agonists like PAR1-AP are unsuitable for measuring calcium mobilization as an end-point read (Figure 5C). It should be noted that for most agonists, the maximum Ca²⁺ mobilization will not be captured, and the assays are probably best used to detect broad changes in Ca²⁺ mobilization. Even in real-time assays, the peak Ca²⁺ release for thrombin occurs in seconds, especially at high concentrations, and the max may be missed because of the time it takes from the addition of agonist until the first read. The maximum Ca²⁺ mobilization occurs typically within the 5 minutes following agonist stimulation. However, there is a lag time from adding some agonists, such as anti-CD9 antibodies or collagen, before a response commences. When performing Ca²⁺ mobilization assays with these agonists, peak Ca²⁺ release is more easily observed.

Time Considerations:

Preparation of platelets from whole blood takes approximately 50 minutes. Consenting and collecting whole blood from donors typically takes 15 minutes or less. The centrifugation step to obtain PRP from whole blood takes about 20 minutes, and isolating platelets from PRP takes an additional 15 minutes to pellet and resuspend the platelets. Preparation of platelets takes a total hands-on time of about 10 minutes. Following the resuspension of platelets, they should be allowed to rest for 30 minutes to recover. To save time during the recovery step, count platelets and, if necessary, begin treatment with antiplatelet agents. If you are new to preparing platelets, you should test platelets for spontaneous activation prior to preparing the antagonists or agonists. However, if you are confident in your platelet isolation ability, then preparing reagents needed for experiments while platelets are resting can save time. Preparing reagents, inhibitors, or agonists can take 15-20 minutes. Serial dilutions of agonists are typically prepared in a 96-well plate and transferred to platelets with a multichannel pipette.

Once you have platelets or PRP, each experiment can take an additional 15-30 minutes to perform, depending on the number of samples and type of experiment being performed. For example, lumi-aggregometry performed with an inhibitor would take 5-10 minutes for antagonist treatment, 5-10 minutes to add agonist and stimulate platelets, and 5 minutes to set up and read the plate. Therefore, the total time to isolate platelets and perform lumi-aggregation will take approximately 2 hours. Real-time dense granule secretion and calcium mobilization assays have lower throughput for some agonists and may take ~20 minutes per run with 5-10 minutes to treat platelets with antagonists and 5-10 minutes to measure platelet activation.

Table II.

Troubleshooting platelet activation assays.

Problem	Possible Cause	Solution
Platelet pellet will not resuspend	Not enough ACD, PGE1 or apyrase were used during the centrifugation of PRP.	Ensure appropriate amount of ACD, PGE1 and apyrase were used.
Spontaneous activation of washed platelet	Inappropriate sample processing or nonspecific platelet interaction with the surface of the 96-well plate.	Coat 96-well plate with BSA to prevent surface activation.
Spontaneous activation of PRP	Inappropriate blood draw or sample processing	Review best practices for blood draw and handling in the strategic planning section.
No platelet activation to AA	Donor took a COX inhibitor (aspirin or NSAID) within the last seven days.	The sample should not be used as results will be invalid. During donor recruitment remind donors to not take aspirin and at time of blood collection confirm the donors have not taken a COX inhibitor.
Greater than 100% aggregation	PPP or Tyrode’s blank may be contaminated	Re-spin PPP to get rid of any potential cellular contaminants
Inconsistent OD reads	Air bubbles in well	Dip a pipette tip in alcohol (ethanol, methanol or isopropanol) and then touch bubbles.
Unstable Flou-4 AM baseline in unstimulated platelets	Fluo-4 AM concentration may be too high.	Decrease amount of Fluo-4 AM used.