Flow cytometric comparison of platelets from a whole blood and finger-prick sample: impact of 24 hours storage

Abstract

In this study, we investigate the validity and laboratory utility of flow cytometry when analyzing platelet activation by studying CD41, CD42b, CD62P and CD63. We compare flow cytometry results from citrated whole-blood and finger-prick samples directly after collection and also after storing both a finger-prick and whole-blood sample for 24 hours. Citrated whole-blood and finger-prick samples were taken from three healthy individuals on two occasions, and a total of 60 000 cells were analyzed for each of the four phycoerythrin-labeled monoclonal antibodies. Half of each sample was analyzed immediately after sampling while the other half was kept in the fridge at 6°C for 24 hours before analysis. No significant difference was found between the sampling methods or the period of time before analysis. Results therefore suggest that an appropriately prepared finger-prick sample can be used for platelet function analysis, and samples can be stored for 24 hours in the fridge at 6°C before analysis.

Keywords:

• Flow cytometry
• Platelets
• Whole blood
• Finger-prick
• Storage

Introduction

Flow cytometry is a powerful technique, and its importance in research and clinical investigation has long been established.¹ Markers used in flow cytometry may give us valuable information regarding cellular activity and since 1989 it has been deemed an essential tool for the investigation of platelets.² Flow cytometry thus provides a numerical technique, which is both objective and quantitative, to assess platelet function.³

Several surface glycoproteins (GPs) are found on the membrane of platelets and flow cytometry has been used to a great extent in the immunophenotyping of these entities. The study of platelet function, physiology, and their interaction with other cells has been advanced by investigating the recognition of these surface GPs by specific monoclonal antibodies (MoAbs). Not only can the GPs on the surface of the platelet membrane be detected by MoAbs, but molecules transferred to the surface from the internal platelet granules can be identified.³ This is of particular interest in the study of platelet activation.

The application of a panel of MoAbs is preferred for flow cytometric analysis of platelets, as different flow cytometric probes reveal different characteristics of platelet function.⁴ CD41 as well as CD42b are frequently used as platelet identifiers, as they are only present on platelets and not any other circulating blood cell.^2,5,6 Alterations in antigenic determinants can also be investigated and MoAbs are used to label epitopes specifically expressed on the platelet membrane. Studies of extracorporeal circulation have mainly used CD62P and CD63 to investigate platelet activation.⁷ Therefore, flow cytometry can be employed to establish the amount of activated and non-activated platelets.³

Whole blood can be obtained from either blood drawn in blood tubes with added anti-coagulants (e.g. citrate tubes) or by sampling blood from a finger-prick. These methods have been used previously by Wall et al. to investigate platelet dense granule release and uptake. Both whole blood collected in sodium citrate tubes and 20 µl of blood from a ‘fingerstick’, as they referred to it, without added anticoagulant were used. After collection, the samples were immediately diluted with Hanks balanced salt solution and a fluorescent marker, mepacrine, was added.⁸

In research where a sample population needs to be followed over a prolonged period of time, or over consecutive days, it is not practical to frequently draw blood. It is much more acceptable for the participants to donate a finger-prick sample. The question also arose whether storage times would influence results. Therefore, in this study we investigate the repeatability of flow cytometry results when using blood drawn in citrate tubes and samples from finger-pricks – either used immediately, or after samples were stored for 24 hours.

Materials and methods

Blood collection

Blood was collected from three healthy, control male individuals. Ethical approval was obtained from the University of Pretoria, Human Ethics Committee, and all participants completed informed consent forms. The participants were non-smokers, did not use any chronic medication and did not have a history of thrombotic disease. None of the participants were using aspirin or aspirin analogues. Firstly, 5 ml of blood was drawn into a citrate tube (0.5 ml of sodium citrate (3.8%) for 4.5 ml of blood) for the determination of the platelet gate. Another 5 ml of blood was drawn into a citrate tube for the subsequent experiments. Sodium citrate is an anticoagulant usually used for flow cytometric analysis of whole blood.^4,9 A volume of 20 µl of blood was collected from finger-pricks. For each concentration of citrate analyzed, a separate finger-prick was taken to ensure that any possible platelet activation after the first sampling was eliminated. Procedures were done in duplicate.

Sample preparation

Platelet-rich plasma preparation

Whole blood was left to stand for 45 minutes to obtain platelet-rich plasma (PRP) by self-sedimentation. Twenty microliters of PRP was transferred into a 1-ml aliquot of Isoflow^TM Sheath Fluid (Beckman Coulter, Krefeld, Germany). Isoflow^TM Sheath Fluid, an isotonic fluid used exclusively for flow cytometry, is specifically formulated to ensure low particle and fluorescence backgrounds to guarantee superior signal-to-noise ratio measurements. This mixture of PRP and Isoflow^TM was stained with 20 µl (the quantity of product sufficient to stain, as prescribed by Beckman Coulter) of one of the following probes separately: CD41-PE (clone P2), CD42b-PE (clone SZ2), CD62P-PE (clone CLB-Thromb/6), and CD63-PE (clone CLBGran/12).

Whole-blood preparation

Experiment 1

From the 5 ml of citrated blood, 20 µl was transferred into 1-ml aliquots of Isoflow^TM Sheath Fluid. The 20 µl of blood was chosen as to be able to compare the whole-blood sample to that of the 20 µl of blood collected from the finger-prick. Hundred microliters of the mixture was placed into four flow tubes. The remainder of the blood mixture was placed in the fridge at 6°C. Each tube was stained with the 20 µl of one of the following probes separately: CD41-PE, CD42b-PE, CD62P-PE, and CD63-PE (Beckman Coulter). The flow tubes containing CD62P-PE and CD63-PE were also stained with 20 µl of CD41-FITC.

Experiment 2

Twenty microliters of the 5 ml whole blood drawn in the citrate tube was placed into 1-ml aliquots of sheath fluid. Hundred microliters of the mixture was placed into four separate flow tubes. Each tube was stained with 20 µl of CD41-FITC and 20 µl of one of the following probes separately: CD41-PE, CD42b-PE, CD62P-PE, and CD63-PE (from Beckman Coulter).

Twenty-four hours after sampling, the remaining blood mixture, placed in the fridge at 6°C overnight, was prepared in the same manner as described above.

Experiment 3

From the 5 ml of citrated blood, 20 µl was transferred into 1-ml aliquots of sheath fluid containing various concentrations of thrombin namely (i) 5 µl thrombin, (ii) 10 µl thrombin, and (iii) 20 µl thrombin. This volume of blood (20 µl) was chosen as to be able to compare the whole-blood sample to that of the 20 µl of blood collected from the finger-prick. 1 Hundred microliters of the mixture was placed into four separate flow tubes. Each tube was stained with 20 µl of CD41-FITC and 20 µl of each of the following probes separately: CD41-PE, CD42b-PE, CD62P-PE, and CD63-PE (from Beckman Coulter).

Finger-prick preparation

Experiment 1

For the finger-prick sample collection, three separate finger-pricks were done with a lancet. Twenty microliters of blood was collected from each finger-prick and immediately placed in 1-ml aliquots of sheath fluid containing various volumes of citrate (3.8% sodium citrate), namely (i) no citrate, (ii) 5 µl of citrate, and (iii) 10 µl of citrate. Hundred microliters of the blood, sheath fluid, and citrate combination was transferred to four individual flow tubes from each sample. The remaining blood, sheath fluid, and citrate combination was placed in the fridge at 6°C. Twenty microliters of the following probes were added to each tube separately: CD41-PE, CD42b-PE, CD62P-PE, and CD63-PE (from Beckman Coulter). The flow tubes containing CD62P-PE and CD63-PE were also stained with CD41-FITC.

Experiment 2

The 20 µl samples of blood collected from three separate finger-pricks were immediately placed in 1-ml aliquots of sheath fluid containing various volumes of citrate (3.8% sodium citrate), namely (i) no citrate, (ii) 5 µl of citrate, and (iii) 10 µl of citrate. Hundred microliters of the blood, sheath fluid, and citrate combination was transferred to four individual flow tubes. Each tube was stained with 20 µl of CD41-FITC and 20 µl of one of the following probes: CD41-PE, CD42b-PE, CD62P-PE, and CD63-PE (from Beckman Coulter).

Twenty-four hours after sampling, the remainder of the blood, sheath fluid, and citrate mixture, placed in the fridge at 6°C overnight, was prepared in an identical manner as described above.

Flow cytometric analysis

The flow cytometer was calibrated and standardized before use with fluorochrome-labeled bead (FlowCheck^TM Fluorospheres, Beckman Coulter). Samples stained with the different probes, were incubated at room temperature in the dark for 20 minutes before analyzed by a flow cytometer (FC 500, Beckman Coulter). Forward scatter (FS) and 90° side scatter (SS) were displayed on logarithmic scales. Two platelet gates were set. The first gate was set according to the morphological characteristics of platelets in the PRP sample (FS and SS) while the second gate was set according to CD41-FITC fluorescence, a platelet-specific marker (PSM). For the duplicate procedures 60 000 platelets were counted and analyzed per phycoerythrin-labeled monoclonal antibody (including CD41-PE, CD42b-PE, CD62P-PE, and CD63-PE). Instrument configuration and settings are shown in Table 1. The fluorescence of the different antibodies was plotted on 256-channel log histograms. The acquired Listmode data were analyzed with CyflogicTM software (CyFlo Ltd, Finland). The results were expressed in arbitrary units as mean channel fluorescence intensity (MCFI).

Table 1. Instrument configuration and settings

Acquisition setup – discriminator
FS = 1
SS = OFF
FL1 = OFF
FL2 = OFF
FL3 = OFF
FL4 = OFF
FL5 = OFF
AUX = OFF
Acquisition setup – detectors	Volt	Gain
FS	816	1
SS	251	5
FL1	553	1
FL2	639	1
FL3	677	1
FL4	450	1
FL5	577	1
AUX	300	1
Compensation	None

Statistical analysis

Results from the flow cytometric analysis were compared by using the paired two-sided Student's t-test. MCFI results are represented as mean ± standard deviation. A P value of ≤ 0.05 was considered significant.

As three similar participants were chosen, the MCFI observations for the three participants can be regarded as being independent and identically distributed random variables for each of the different blood-preparation techniques. Furthermore, as the MCFI for each participant was calculated as the mean fluorescence of a large sample of platelets (10 000 platelets per individual per sample), the well-known Central Limit Theorem assures us that the Normal distribution is a close approximation for the distribution of the MCFIs. This allows us to make use of the paired two-sided Student's t-test to compare the results from the flow cytometric analysis at a 5% level of significance.

Results

Gating strategies

Two platelet gates were set, according to the scatter characteristics and specific platelet marker utilized. The first platelet gate was set according to the FS and SS properties of platelets found in the PRP obtained from self-sedimentation. The PRP contained only platelets and no other blood cells. Therefore, a pure platelet population could be distinguished without additional MoAb staining. Fig. 1 represents a subsequent analysis where whole blood (containing all blood cells) was analyzed by the PRP gate.

Figure 1. PRP gate. Platelet gate set according to FS and SS of PRP. This dot plot shows the PRP gate when a whole-blood sample was analyzed. The other populations outside the PRP gate represent the other blood cells found in the whole-blood sample.

Fig. 2 shows the second platelet gate set according to CD41-FITC fluorescence, a PSM on a dot-plot displaying FS on the y-axis and FL1 fluorescence on the x-axis. CD41 is used as a reliable PSM. It enables the differentiation of platelets in a whole-blood sample.² The CD41-FITC marker was used in conjunction with the mentioned phycoerythrin-labeled monoclonal antibodies (CD41-PE, CD42b-PE, CD62P-PE, and CD63-PE). For the PSM gate, a clear distinction can be made between the unactived platelets (to the lower right of the diagram) and the activated platelets (to the upper right part of the diagram).

Figure 2. PSM gate. Platelet gate set according to FL1 of CD41-FITC PSM. Activated platelets (AP) are also incorporated in the PSM gate. The other populations outside the PRP gate represent the other blood cells found in the whole blood sample.

An overlay of the two platelet gates is shown in Fig. 3. In Fig. 3A it is shown that the PRP gate only corresponds to the non-activated platelets in the population while Fig. 3B shows the activated platelets are incorporated in the PSM gate but not the PRP gate. Therefore, from Fig. 3 we can deduce that the gating strategy is very important if all parameters are to be fully evaluated.

Figure 3. Overlay of platelet gates. The PRP sample not containing CD41-FITC was analyzed by use of the PSM gate while the whole-blood sample that did contain CD41-FITC was analyzed by the use of the PRP gate. (A) PRP only corresponds to the non-activated platelets of the PSM-gate. (B) When the activated platelets are represented on the dot plot of the PRP gate, these cells are not incorporated in the PRP gate but are clustered on the upper right part of the dot plot.

Both gating strategies were used for analysis of the different phycoerythrin-labeled monoclonal antibodies used. The PRP gate showed a significantly lower MCFI value compared with the PSM gate for all instances.

Comparisons

To determine the integrity of the different sampling methods, whole-blood samples were compared with finger-prick samples containing various volumes of citrate immediately after sampling. The volumes of citrated added to the sheath fluid before sampling was (i) no citrate, (ii) 5 µl of citrate, and (iii) 10 µl citrate. The comparison of whole-blood samples with various volumes of thrombin added directly after sampling was done to determine the activation of platelets. Thrombin volumes added to whole-blood samples before analysis was (i) 5 µl, (ii) 10 µl, and (iii) 20 µl. The influence of storage time was compared for whole blood and finger-prick samples analyzed 24 hours after sampling.

The results from experiments 1 and 2 were pooled and each of the mentioned phycoerythrin-labeled markers were used to analyze specific characteristic of platelets gated by CD41-FITC. For each marker, 10 000 platelets were counted and analyzed per person adding up to a total of 60 000 platelets for the duplicate procedures.

The following results, obtained immediately after sampling, were compared with determine any significant changes. Table 2 displays results from whole-blood sample containing no thrombin (non-activated whole blood sample) compared with finger-pricks containing various concentrations of citrate. Table 3 displays results from whole-blood samples containing various concentrations of thrombin.

Table 2. Analysis immediately after sampling: whole-blood sample and finger-pricks containing various concentrations of citrate

MoAb	WBNo thrombin	FPNo citrate	FP5 µl citrate	FP10 µl citrate
PRP gate
CD41-PE	24.04 ± 2.71	17.74 ± 0.38	23.02 ± 3.12	24.36 ± 4.27
CD42b-PE	29.68 ± 3.37	28.17 ± 1.32	30.81 ± 2.85	32.54 ± 3.17
CD62P-PE	79.94 ± 2.15	78.16 ± 4.61	77.20 ± 3.32	76.99 ± 2.16
CD63-PE	49.54 ± 2.09	47.01 ± 0.72	48.61 ± 1.09	45.88 ± 4.88
PSM gate
CD41-PE	47.72 ± 6.04	70.51 ± 1.24	64.13 ± 2.52	52.82 ± 9.45
CD42b-PE	35.07 ± 3.97	47.58 ± 2.35	40.39 ± 1.74	36.44 ± 3.55
CD62P-PE	115.49 ± 5.95	100.97 ± 3.24	99.15 ± 0.47	107.55 ± 4.86
CD63-PE	63.52 ± 2.53	54.86 ± 7.26	49.12 ± 6.35	57.31 ± 5.94

Results represented as mean ± SD of MCFI (n = 60 000 platelets analyzed for each of the sampling procedures).

WB, whole blood; FP, Finger prick.

Table 3. Analysis immediately after sampling: whole bloodsamples containing various concentrations of thrombin to evaluate activation

MoAb	WBNo thrombin	WB5 µl thrombin	WB10 µl thrombin	WB20 µl thrombin
PSM gate
CD41	47.72 ± 6.04	54.46 ± 18.28	53.90 ± 18.70	71.24 ± 5.72
CD42b	35.07 ± 3.97	37.75 ± 6.66	39.26 ± 6.82	40.11 ± 6.48
CD62P	115.49 ± 5.95	148.57 ± 19.97	155.92 ± 25.96	174.92 ± 9.40
CD63	63.52 ± 2.53	76.07 ± 13.55	72.73 ± 9.00	84.13 ± 12.49

Results represented as mean ± SD of MCFI (n = 60 000 platelets analyzed for each of the sampling procedures).

WB, whole blood.

Table 4 shows the results of immediate analysis compared with results of analysis after 24 hours. Whole blood analyzed immediately was compared with the whole-blood sample stored for 24 hours. The finger-prick containing no citrate analyzed immediately was compared with the same sample stored for 24 hours. The same procedures were performed for the other finger-prick samples, containing 5 and 10 µl citrate, respectively, immediately analyzed after collection and the aliquot stored for 24 hours.

Table 4. Analysis 24 hours after sampling

MoAb	WBNo thrombin	FPNo citrate	FP5 µl citrate	FP10 µl citrate
PRP gate
CD41-PE	25.99 ± 1.84	46.32 ± 0.44	48.93 ± 8.39	28.07 ± 1.64
CD42b-PE	37.88 ± 6.60	38.79 ± 0.38	36.40 ± 7.59	39.76 ± 0.81
CD62P-PE	81.32 ± 0.84	75.98 ± 4.76	72.87 ± 0.82	73.53 ± 2.28
CD63-PE	45.63 ± 1.95	42.84 ± 6.05	46.63 ± 0.50	47.25 ± 1.51
PSM gate
CD41-PE	51.41 ± 1.58	70.51 ± 1.24	64.13 ± 2.52	60.82 ± 3.41
CD42b-PE	32.01 ± 0.31	47.58 ± 2.35	40.39 ± 1.74	39.13 ± 2.62
CD62P-PE	118.69 ± 4.61	100.97 ± 3.24	99.15 ± 0.47	104.59 ± 4.12
CD63-PE	62.86 ± 1.17	54.86 ± 7.26	49.12 ± 6.35	53.98 ± 7.42

Results of the whole blood (WB), first finger-prick containing no citrate, the second finger-prick (FP) containing 5 µl of citrate and the third FP containing 10 µl of citrate analyzed 24 hours after the sample was taken. Results are represented as mean ± SD of MCFI (n = 60 000 platelets analyzed for each of the sampling procedures).

Whole blood vs. finger-prick

There were no significant difference between the whole blood samples and the finger-prick samples containing 10 µl of citrate (P > 0.05). However, significant changes were shown for the other finger-prick samples containing no citrate and 5 µl of citrate separately (P < 0.05).

Both gating strategies were used for analysis of the different phycoerythrin-labeled monoclonal antibodies used to evaluate the different sampling methods. The PRP gate showed a significantly lower MCFI value compared with the PSM gate for all instances.

Non-activated vs. activated

Whole-blood samples were activated by adding different volumes of thrombin to separate samples. The whole-blood sample containing no thrombin (thus, the non-activated sample) was compared with whole-blood samples containing 5, 10, and 20 µl of thrombin. A significant difference was found between the non-activated whole-blood sample and the activated whole-blood samples (P < 0.05). This was also true when the activated whole samples were compared with the finger-prick sample containing 10 µl citrate. The MCFI for all phycoerythrin-labeled monoclonal antibodies (CD41-PE, CD42b-PE, CD62P-PE, and CD63-PE) increased with the increase in thrombin added to the whole-blood samples.

Immediate analysis vs. 24 hours

No significant difference was found for the MCFI of whole blood analyzed immediately compared with whole blood analyzed 24 hours after sampling (P > 0.05). This was also true for the finger-prick samples with 10 µl of citrated added before sampling. The immediate analysis of the finger-prick samples with no citrate and 5 µl of citrate added separately, showed significant differences from the same samples analyzed 24 hours after sampling (P < 0.05).

Discussion

The current investigation confirms that both a whole-blood sample and a finger-prick can be used interchangeably for flow cytometric analysis of platelets. Storage time of up to 24 hours in a fridge at 6°C also does not influence the platelet activation in the blood sample. Aspects such as the strategy employed for gating the sample, the specific flow cytometer instrument used, the preparation of the samples, the specific probes used as well as the storage time will be discussed in greater detail.

Gating strategies

Flow cytometry enables the researcher to detect several parameters from a single sample. Different probes or markers can simultaneously be analyzed by using multicolor flow cytometry.¹⁰

The function of ‘gating’ or isolating particular cell clusters in flow cytometry enables classification and investigation of platelets in the mixed populations of cells like found in whole blood.³ Two parameters are employed to facilitate platelet gating namely (1) forward light scatter which arranges cells according to size, and (2) the use of a platelet-specific antibody like CD41 or CD42b.³

Multicolour flow cytometry has the advantage of gathering more information from a single sample therefore minimizing the sample size and decreasing preparation effort. This technique also decreases variation since fewer sample preparation is required and is an effortless method to study platelet population.¹⁰

The reliability of the population of platelets in a whole-blood sample is determined by the gating strategy employed.^11–13 Van Velzen and colleagues have recently investigated the effect of different platelet-gating strategies. They stated that the changes in scatter characteristics after platelet activation will differ from the morphology of non-activated platelets and therefore a pure platelet population will not be selected. They asserted that utilizing a platelet antigen like CD41 or CD42b, which is commonly present on the platelet membrane, is preferred above the use of the FS/side scatter strategy. By using a PSM not only eliminate coincidence but also excludes complexes platelets form with monocytes and decreases possible contamination.¹⁰

This research of van Velzen et al. was done for a fixated whole-blood sample. In this investigation similar results were found for an unfixed whole-blood sample. As the PRP gate only shows the unactived platelets in the absence of CD41-FITC, it is important to firstly stain the sample with CD41-FITC and secondly to set the platelet gate according to the FL1 information obtained from CD41-FITC, i.e. the PSM gate. This indicates the importance of multicolor gating for optimal analysis of platelets.

Instrumentation

Earlier flow cytometric analysis has been performed on FACScan cytometer (Becton Dickinson, San Jose, CA, USA). Platelet microparticles¹⁴ and magnetic labeling of platelets¹⁵ have been studied on the FC500 (Beckman Coulter). Erythrocytes,¹⁶ dysfunctional T regulatory cells,¹⁷ and stem cells¹⁸ among others have also been analyzed on the FC500 (Beckman Coulter). This is the first study conducted on the FC500 (Beckman Coulter) to investigate platelets in an unfixated whole-blood and finger-prick samples as well as the effect of storage time on the platelets.

Blood preparation for flow cytometry

Whole blood was drawn in citrate tubes (0.5 ml of sodium citrate (3.8%) for 4.5 ml of blood). For the finger-pricks, citrate had to be added to the sheath fluid to prevent coagulation of the blood sample. Weil¹⁹ stated that if blood is mixed with a sodium citrate solution in appropriate portions, the blood will not clot since the calcium salts are not accessible for coagulation. Weil¹⁹ added that blood mixed with sodium citrate can be kept in the fridge at 6°C for several days, and only slight changes in cell structure can be observed after 1 week of storage.¹⁹ The results indicate that a volume of 10 µl of citrate added to the sheath fluid aliquot is adequate to prevent coagulation of a finger-prick sample.

Typically, platelets have been studied after preparing PRP or washing the platelets.^20,21 These separation procedures lead to in vitro activation of platelets due to the formation of artifacts. The pioneer use of whole blood for flow cytometric studies by Shattil et al.¹¹ was just the beginning for important improvements for the relevance of flow cytometry for clinical application.⁴ The flow cytometric analysis of whole blood holds many advantages for the study of platelet activation. This method has been used for over 20 years to investigate platelet function. Ault et al.²² used the MoAbs specific for GPIb (CD42b) and GMP140 (CD62P) to measure platelet aggregation and release reaction in a whole-blood sample. Activation-dependent variations in multiple surface receptors can be determined in a whole-blood sample. In vitro platelet activation is also prevented, as there is minimal manipulation of the samples and it will decrease the possible loss of platelet subpopulations.⁴

Probes

A Pubmed search revealed few references in recent literature that discuss the use of CD41-PE, CD42b-PE, CD62P-PE, CD63-PE, and CD41-FITC in whole-blood and finger-prick samples as well as the particular storage methods employed in this study. The current literature showed that flow cytometry has been employed to study platelet function testing in apheresis products, where all the above-mentioned probes were used in conjunction with secondary antibodies.⁷ Other studies include flow cytometric analysis of platelet count,²³ platelet reactivity,²⁴ platelet function in children,²⁵ and the importance of the sampling site in patient about to undergo surgery;²⁶ these studies used only one or two of the above-mentioned probes. All the mentioned studies were performed on a FACScan cytometer (Becton Dickinson).

The most recent investigation of platelet activation was done by van Velzen et al. in 2012. They incorporated all four the mentioned antibodies; however, the conjugated fluorochromes differed from the probes used in this study and the samples were fixated before analysis.¹⁰

Sampling method

No significant difference was found between the whole-blood sample and the finger-pricks containing 10 µl of citrate. There was a significant difference between the whole-blood sample and the finger-pricks containing no citrate and with 5 µl of citrate added. This indicated that the method of sampling, both the whole blood drawn in the citrate tube as well as the finger-prick samples, was adequate. It also shows that 10 µl of citrate is the preferred concentration to be added to the sheath fluid before adding blood from a finger-prick. The whole-blood sample and each of the finger-prick samples were done on separate days. This was done to eliminate the possible effect of platelet activation after the first sample.

Activation of platelets

The activated whole-blood samples showed significantly higher MCFI values when compared with the non-activated whole-blood sample and the finger-prick sample containing 10 µl of citrate. We can therefore deduce that the whole-blood sampling containing no thrombin, as well as the finger-prick with 10 µl of citrate added, did not contain an activated platelet population.

Once receptor activation occurs, the internal platelet granules are secreted and the cytoskeleton will be rearranged leading to signal transduction.^27,28

Upon activation, platelets exhibit elevated levels of specific activation markers on the platelet surface for example CD62P²⁹ and CD63.³⁰ Both these markers are dominant immunologic indicator of platelet activation.^30,31

CD62P, also referred to as P-selectin, is an activation-dependent MoAb most extensively used in the study of platelet granule membrane proteins.³² It is a constituent of a resting platelet's α-granule membrane.³¹ It will only be expressed on the surface membrane once the contents of the α-granule are secreted.^33–35 For this reason, the MoAb specific for CD62P will not bind to resting platelets, only to degranulated, activated platelets.⁴ Once activated, platelets also rapidly transfer CD63 from the lysosome-like granules to the platelet surface through the surface canalicular system.²

Van Velzen et al. also found the expression of CD62P and CD63 to be increased when platelets were exposed to thrombin activation in a fixated sample.¹⁰

Period between sampling and analysis

CD41 is a calcium-dependent compound of GPIIb/IIIa. The interactions between cells and the cells with the matrix are mediated by GPIIb/IIIa.^36,37 CD42b is a receptor for the von Willebrand factor, which plays a critical role in the adhesion of platelets to the wall of injured blood vessels.³⁸

CD41 and CD42b are reliable PSMs. These MoAbs are used to differentiate between platelets and ‘debris’ or fragments.²

As the expression of CD41 and CD42b in the whole-blood samples was similar for immediate analysis of the sample and analysis 24 hours after sampling, we can infer that these platelets identified by CD41 and CD42b expression were not activated after sampling and remained inactive for at least 24 hours. Michelson and associates indicated on several occasions that the binding of CD42b to resting platelets is noticeably increased compared with its binding with activated platelets.^39–41 The decreased expression on the surface of activated platelets possibly results from the translocation of this complex to the surface-connected canalicular system membranes.^41,42 This shows that CD42b could be a sensitive marker of in vivo activation of platelets.

Although platelet storage has been associated with increased expression of CD62 in the past,³ our results indicate that both an unfixated whole-blood and a finger-prick sample can be stored in the fridge at 6°C if prepared appropriately. The time of analysis, whether it is immediately after the sample was take or 24 hours after sampling, appears to not have an influence on the expression of CD41 and CD42b. However, as the participants were all young and healthy individuals, the effect of age and disease will have to be investigated for this method.

Limitations and possible applications of study

Before conclusions can be drawn from these results, potential limitations need to be addressed. The small sample size, the use of only four MoAbs and the analysis at two time points can possibly constrain the findings of this study. A greater sample size, using different MoAbs and analyzing the samples over a greater time period or at more regular intervals can bring greater insight.

However, although the sample size is limited, the results indicate that the platelets are not additionally activated by this particular methodology. Therefore, these methods may be successfully implemented in clinical studies with antiplatelet agents.

Conclusion

Firstly, the importance of gating strategy for a unfixated sample was established. If all platelets, activated and non-activated, are to be taken into consideration, then the PSM gate should be employed and all samples should be stained with CD41-FITC and not only with the specific phycoerythrin-labeled monoclonal antibodies (including CD41-PE, CD42b-PE, CD62P-PE, and CD63-PE).

Secondly, we can conclude that an unfixated whole-blood and a finger-prick sample are identical with regard to platelet function. Both sampling methods showed little activation when compared with activated whole-blood samples. Therefore, a whole-blood sample and a finger-prick sample can be used interchangeably for flow cytometric analysis of platelets. This is advantageous for research where a sample population needs to be examined over consecutive days or an extending time interval. In these cases, when it is not practical to repeatedly draw blood, a finger-prick will be sufficient.

And lastly, samples can be stored for 24 hours after sampling. Sheath fluid is a sufficient medium for storing unfixated whole-blood and finger-prick samples in the fridge at 6°C for 24 hours, provided that a sufficient amount of citrate is added to the sheath fluid for a finger-prick sample. A volume of 10-μl citrate provides sufficient anticoagulant action for a 20-μl aliquot of finger-prick blood. Both whole-blood samples and finger-prick samples can be kept in the fridge at 6°C for 24 hours before analysis as no platelet activation occurs. This will aid studies where analysis cannot immediately be performed due to traveling distance from flow cytometer or time constraints.

Acknowledgements

The authors acknowledge Peter Smith for his assistance with the statistical analysis.