Evaluating Stored Platelet Shape Change Using Imaging Flow Cytometry

Abstract

Platelets are routinely stored at room temperature for 5-7 days before transfusion. Stored platelet quality is traditionally assessed by Kunicki's morphology score. This method requires extensive training, experience, and is highly subjective. Moreover, the number of laboratories familiar with this technique is decreasing. Cold storage of platelets has recently regained interest because of potential advantages such as reduced bacterial growth and preserved function. However, platelets exposed to cold temperatures change uniformly from a discoid to a spherical shape, reducing the morphology score outcomes to spheroid versus discoid during cooling. We developed a simpler, unbiased screening tool to measure temperature-induced platelet shape change using imaging flow cytometry. When reduced to two dimensions, spheres appear circular, while discs are detected on a spectrum from fusiform to circular. We defined circular events as having a transverse axis of > 0.8 of the longitudinal axis and fusiform events ≤ 0.8 of the longitudinal axis. Using this assay, mouse and human platelets show a temperature and time-dependent, two-dimensional shape change from fusiform to circular, consistent with their three-dimensional change from discs to spheres. The method we describe here is a valuable tool for detecting shape change differences in response to agonists or temperature and will help screening for therapeutic measures to mitigate the cold-induced storage lesion.

Introduction

Platelets are small discoid blood cells produced by megakaryocytes in the bone marrow at a rate of 10¹¹/day. Platelets circulate for 7-10 days providing vascular integrity and promoting hemostasis at sites of vascular injury, although a plethora of other functions have been described. Patients with acute hemorrhage or thrombocytopenia may receive platelet transfusions to prevent and treat bleeding. For this purpose, platelets are currently stored at room temperature for up to 7 days. In the 1960-1970s, platelets were stored at cold temperatures (4 °C), but this approach was abandoned due to decreased in vivo circulation time.¹ Nevertheless, storing platelets at 4 °C has potential advantages, such as the reduced risk of bacterial growth and possibly better hemostatic function. Overall, these advantages have led to a renewed interest in cold-stored platelets.² Particularly, for patients with acute bleeding during surgery or in a trauma setting, cold-stored platelets could be helpful because long circulation times are not required to promote hemostasis.

Platelet quality during storage is traditionally assessed by a morphology score, described by Kunicki et al. in 1975.³ This method requires extensive training and experience. Still, it is one of the few in vitro markers that correlate well with in vivo recovery and even more so with survival.³ In brief, the score is based on four morphologic characteristics, 1) discoid shape, 2) spheroid shape, 3) presence of dendrites (platelets with pseudopodia and dendritic processes), and 4) presence of ballooned platelets (platelets that are unable to maintain the osmotic gradient). One hundred events are counted and multiplied by an arbitrary factor of 4 for discs, 2 for spheres, 1 for dendrites, and 0 for balloons. Hence, a morphology score of 400 indicates platelets with the most desirable morphology (all discs). The current study is aimed at providing an easier, minimalist and unbiased alternative to the traditional, light microscopy-based morphology score. We report morphology scores from several trials in healthy humans, describe an imaging flow cytometry-based approach, and provide data from proof of principle experiments.

Materials and Methods

Reagents

We utilized the following antibodies: CD61-APC and CD42a-FITC (both BD Biosciences, Franklin Lakes, NJ). Low weight heparin enoxaparin (Lovenox, Sanofi-Aventis, Paris, France).

Mouse colony.

14-18 weeks old (with body mass about 20-24 g) wild-type C57BL6/J mice from Jackson Laboratory Lab (Bar Harbor, ME) were used in this study. All animal procedures were conducted in accordance with approved institutional IACUC protocols.

Preparation of platelets

Whole blood was collected by phlebotomy in ACD-A (15%) from healthy donors in 10ml collection tubes. Murine whole blood was collected by retro-orbital bleeding in a mixture of PBS and low molecular weight heparin (Lovenox, Sanofi-Aventis, Paris, France) in a total volume of 1 ml and a final concentration of low molecular weight heparin. We generated platelet-rich plasma (PRP) by soft centrifugation (200 g) for 5 minutes, and PRP was expelled while sparing the buffy coat. Mouse platelets were stored in 15ml falcon tubes with cell culture lid to allow for gas exchange on a regular platelet agitator with a commonly used speed (72 cycles per minute) to allow for gentle agitation (Helmer, Noblesville, IN). Mouse platelets were stored for up to 48 hours at a concentration of ~5x10⁵/μL and a volume of at least 1ml. For long-term storage (7 days) platelets were collected as PRP by apheresis in 15% ACD-A in licensed storage bags under agitation in the same fashion as is done routinely for clinical samples and as described by us previously.^4,5 The concentration represented the usual clinically used conditions, i.e. the concentration was ~1.5x10⁶/μL and volume the volume ranged between 180-250ml (target platelet yield ~3x10¹¹ platelets per bag). We used an automated blood collection system (Trima Accel, Terumo BCT, Lakewood CO) and stored in 100% plasma.

The platelet preparation of the clinical trial morphology score (Figure 1) is provided in more detail in the original reports of the studies and is identical to the apheresis collection and storage process described above.^4-6 Of note, the Morphology Scores have not been previously published.

Figure 1:

Morphology scores for fresh platelets, 3, 5, 10, 15, 20 day cold-stored, and 5, 7 room temperature-stored platelets. (Fresh (black circles) n=25, Cold-stored platelets, duration: 3 days (black squares, n=12), 5 days (black upward triangles, n=5), 10 days (black downward triangle, n=14), 15 days (grey circles, n=11), 20 days (black diamonds, n=6), RT-stored platelets, duration 5 days ( grey squares, n=9), 7 days (grey squares, n=5). Data show as individual data points, mean, and ± SD. ***p ≤ 0.001.

Storage Condition

Platelets, collected as described above, from human subjects were divided into two equal samples in miniature platelet storage bags with a filling volume of 10ml (BCSI SAFE Sens, Seattle WA). Human platelets were stored at 1.5x10⁶ for human platelets in 100% plasma. Mouse platelets were obtained as described above and stored in falcon tubes with a cell culture cap to allow for gas exchange at a target concentration of 500x10³/μL platelets in 100% plasma. The platelets were either stored at room temperature or at 4 °C. Platelets at room temperature were stored with gentle agitation with a speed of 72 circles per minute (Helmer, Noblesville, IN), while platelets stored at 4° C were stored without agitation in a commercially available blood banking fridge that allows for gas echange (Helmer, Noblesville, IN). For the mouse platelets, fresh and after 1, 24, and 48 hours of storage, the platelets were sampled for the staining for flow cytometry. For the human platelets, fresh and 7 day stored platelets were sampled for the staining for flow cytometry.

For the platelets tested at different temperatures, whole blood was collected in prewarmed 10ml tubes (15% ACD-A) at 37 °C, and was centifugated at 37 °C to generate PRP. PRP was immediately moved to pre-set temperatures including 37, 34, 30, 26, 22, 16, 10 and 4 °C for 1 hour. After this incubation, the samples were fixed and stained for flow cytometry.

Platelet labeling

We used paraformaldehyde as a fixative known for preserving cell shape. For each time point, 3% paraformaldehyde (PFA) solution (1% final) was added directly to the cell medium to avoid any perturbation in cell shape. After 10 minutes of incubation with PFA, the cells were washed twice with PBS and labeled with CD61 – APC and CD42a – FITC for human platelets. After incubation at room temperature in the dark for 10 minutes, the samples were resuspended in 400 μL of PBS.

Imaging Flow Cytometry

Samples were analyzed by using Amnis Imagestream Mk II (Luminextcorp, Austin, TX). Images of cells were then acquired with the Amnis Image Stream with the 60X magnitude objective using the bright field channel, the SSC channel, FITC (channel 2) and APC (Channel 11) by the INSPIRE software for acquisition. Per sample, 20000 events were recorded on the Imaging Flow Cytometry. IDEAS 6.2 software (Luminexcorp, Austin, TX) were used for analysis. First, we identified focused platelets based on bright-field images by the levels higher than 60 on gradient root mean (GRM) square channel 1 (M01) feature to analyze platelet morphology (Figure 2A). We isolated platelets based on CD61 positivity (we gated for CD61 positive cells based on 1e³-2e⁵ MFI, Figure 2B). In the third step, we defined single platelets by using the Area and the Aspect Ratio features (Figure 2C). Lastly, we used Aspect Ratio (circular events with a transverse axis of > 0.8 of the longitudinal axis, and fusiform events with ≤ 0.8 of the longitudinal axis) to analyze different morphologies of stored platelets (Figure 2D). The boundaries of the gates were defined based on the shape of circular versus fusiform dot plots. This was further refined and confirmed with image controls. The percent of the circular and fusiform events were used for statistical analysis.

Figure 2:

Quantification of cold-induced platelet shape change by imaging flow cytometry. (A-D) Gating strategy to differentiate between fusiform and circular platelets. Circular events were defined as having a transverse axis of > 0.8 of the longitudinal axis and the percentage of circular (orange gate) and fusiform platelets (pink gate) was determined (D). After 7days of RT storage, 62% were in the gate for fusiform platelets, and 38% were in the gate for circular platelets (E). After 48h of cold storage, 77% of platelets were in the gate for circular platelets and only 23% were in the gate for fusiform platelets (F).

Healthy human subjects research

The Western Institutional Review Board approved the research (WIRB), and all human participants gave written informed consent. The study was conducted following the Declaration of Helsinki.

Statistical analysis

We reported the results as mean ± standard error of the mean and assessed for statistical significance by ANOVA with Tukey correction for multiple comparisons (Figure 1) or paired Student t-test. The data in Figure 1 were tested for normality and with the exception of the 5 day RT-stored group, and 3 day CSP group, did not have samples to assess normality (7 day RSP, 20 day CSP, 5 day CSP) or were not normally distributed (fresh, 10 day CSP, 15 day CSP). For the rest of the data we assumed normality, but numbers were too low to formally test for it. A P value of ≤ 0.05 was considered significant.

Results

One hallmark of platelets exposed to cold temperature is the change from discoid shape to spherical shape, based on tubulin disintegration and actin barbed end-capping.⁷ Indeed, we have included the Kunicki morphology score in previous cold storage and room temperature storage studies and found it to be uniformly 200 (all spheres), at least for the first 15 days of storage (Figure 1). In contrast, storing platelets at 22 °C increases the score number and variability (Figure 1). To avoid using this complex and subjective method for evaluating morphologic change in cold-stored platelets, we utilized imaging flow cytometry. We based our gating strategy on the premise that spheres can only be detected as circles when reduced to two dimensions, but disks can be seen as fusiform (i.e. spindle-shaped) events or circles and on a spectrum between both shapes. Therefore, we first aimed at selecting a gating strategy that allowed us to differentiate between fusiform and circular events. We used the ratio of the longitudinal to the transverse axis. Circular events were defined as having a transverse axis of > 0.8, while fusiform (spindle-shaped) events were defined as having a transverse axis of ≤ 0.8 of the longitudinal axis. The final gates were based on aspect ratio and intensity threshold (Figure 2A-D). In addition to the bright-field analysis, we added fluorochrome-labeled antibodies to platelet surface markers to exclude the attachment of occasional leukocytes or RBCs to platelets (Figure 2B, Figure 3, A, B). After 48 hours of room temperature storage, we saw a mixed distribution between circular and fusiform events, but 48 hours of cold storage led to mostly circular events (Figure 2E, F).

Figure 3:

Representative samples for fusiform and circular events. Five random events are shown for circular (A) and fusiform (B) human platelets. Representative images from a healthy donor (out of eight tested). Per sample, 20000 events were recorded.

We exposed mouse and human platelets to temperatures between 37 °C and 4 °C for 60 minutes and subsequently examined them in our assay. In both species, lowering the temperature led to a decline in fusiform events and increased circular events, consistent with a transition from fusiform to spherical shape (Figure 4A and B). Lowering the temperature, increased the likelihood of detecting platelets in the circular events gate.

Figure 4:

(A) Human platelets were exposed to temperatures from 37 °C to 4 °C for one hour, fixed, and tested by imaging flow cytometry. n=6 *p ≤ 0.05 (B) Mouse platelets were exposed to temperatures from 37 °C to 4 °C for one hour, fixed, and tested by imaging flow cytometry. n=3, *p ≤ 0.05.

Of note, human platelets show a slightly higher crossover point with more circular than fusiform events than human platelets starting at 16 °C (Figure 4A and 4B). This crossover was seen only at 4 °C for the mouse platelets.

To further study the time-dependency of cold-induced shape change in mouse and human platelets, we exposed mouse platelets to 1, 24, and 48 hours to room temperature or 4 °C in commercially available storage bags. After prolonged storage, room temperature led to an even distribution between circular and fusiform events (Figure 5A). In contrast, cold exposure led to a trend for more circular events after 24 and 48 hours of cold storage (Figure 5A).

Figure 5:

(A). Mouse platelets were stored at room temperature or 4 °C for up to 48 hours and tested as baseline (fresh), and after 1, 24, and 48 hours, and stained and analyzed as baseline (fresh), and after one, 24, and 48 hours, N=4, *p ≤ 0.05. (B). Human platelets were stored at room temperature or 4 °C for up to 7 days hours and tested as baseline (fresh), and after 7 days. The samples were stained and analyzed as baseline (fresh), and after 7 days, N=8, *p ≤ 0.05.

To mimick what is going on in clinical samples, we stored human platelets at room temperature or 4 °C for 7 days. As expected, we observed more circular platelets in cold-stored samples compared to RT-stored samples after 7 days (Figure 5B).

Discussion

Platelets are routinely stored for transfusion to bleeding patients or patients at risk for bleeding. The transfusion community recently re-discovered an interest in cold storage of platelets, intending to transfuse safer and potentially more efficacious platelets. Current research projects aim at preventing shape change and, ultimately, longer circulation times. Although initial studies had suggested that shape change itself is not the primary reason for clearance¹⁰, shape change still represents the hallmark of the cold-stored platelet storage lesion and the prevention thereof, a primary research goal of the transfusion community.¹¹ The platelet quality of room temperature-stored platelets is assessed by the Kunicki morphology score. Our data show that after extended periods of cold storage up to 15 days, the score is uniformly resulted as 200 (all spheres). Hence, we did not note any correlation with in vivo recovery and survival (data not shown). Other tools based on the discoid shape of "healthy" platelets have similar problems, e.g., the "swirling phenomenon" based on the movement and orientation of discoid platelets in plasma is uniformly absent in cold-stored platelets (data not shown).

Moreover, obtaining reliable results with the morphology score requires extensive training, and even after completion of training, the high level of subjectivity inevitably leads to inter-operator and intra-operator variability. Because light microscopy used for the morphology score is not designed to assess three-dimensional objects either, obtaining the morphology score routinely involves physical manipulation of the cell suspension (i.e., slight movement of the coverslip) to characterize the three-dimensional shape of platelets. We developed an imaging flow cytometry-based tool to quantify the degree of cold-induced shape change based on the two-dimensional projection of three-dimensional objects. Our approach omits subjectivity and represents an unbiased tool to quantify the degree of shape change for future research and clinical studies. Assessment of the morphology score is more straightforward with cold-stored platelets than with RT-stored platelets. Still, it requires the identification of spheres and the differentiation between discs and spheres by light microscopy, which requires training. Our proposed method represents an automated and fast approach that allows us to analyze multiple samples quickly. Also, the number of imaging flow cytometers available in academic institutions likely far exceeds the number of individuals with expertise performing Kunicki’s morphology score. Indeed, the search term “Imaging Flow cytometry” in Pubmed shows a continuous increase in publications from 2004-2021 (data not shown), hinting at increasing availability and utilization of this technology. This initial proof of principle study shows how uniformly spherical suspension and discoid cells appear in the gates designed to detect circular and fusiform events. Staining with platelet markers further decreased the chance of including false positive (non-platelet) events. Like other studies before, our study shows a difference in the cold response between mouse and human platelets.¹² One limitation of this technology is that, in its current form, it is useful for the assessment of cold storage-induced shape change, but not for proper assessment of long-term room temperature stored platelets. For these, the four factors outlined above, described by Kunicki et al. are still required.³ In summary, we developed a standardized screening tool to assess and quantify cold storage condition-induced platelet shape change. Our assay will help screening for novel drugs and approaches to circumvent the unwanted side effects of cold storage, including shape change.

Plain language summary.

What is the context?

• Platelets for transfusion are currently stored for 5-7 days at room temperature, increasing the risk for bacterial growth

• Cold storage reduces the risk for bacterial growth but reduces circulation time

• Stored platelet quality can be assessed by the light microscopy-based Morphology Score, first described in the 1970s

• Downsides of the Morphology Score include subjectivity, extensive training, and reduced availability in platelet laboratories.

What is new?

• In this study, we provide data showing that the Morphology score is reduced to a binary spheres versus discs response in cold-exposed platelets

• We developed an imaging flow cytometry-based approach to quantify platelets' response to cold based on the two-dimensional projection of the three-dimensional shapes, i.e., fusiform (discoid) versus circular (discoid and spherical)

• We provide validation of this approach in mouse and human platelets

What is the impact?

• This study provides an easy and unbiased tool for laboratories working on circumventing the cold-induced storage lesion or documenting spherical shape change in general