Super-Resolution Immunofluorescence Imaging of Platelet Granules

Introduction

Platelets play a central in hemostasis by facilitating thrombus growth at sites of vascular injury. During activation, platelets secrete a highly diverse cocktail of proteins from their granules, thereby enhancing the hemostatic response. How platelets organize their granules in a resting state as well as during activation and secretion is still poorly understood. We have developed a platform for quantitative analysis of platelet granules employing super-resolution microscopy. Using structured illumination microscopy (SIM), individual alpha- and dense-granule can be separated in platelets. Together with morphological analysis and other quantitative imaging data that can be processed in automatic analysis workflows using ImageJ software, this method yields novel insights in granule organization and has implications for both platelet biology and translational studies on patients with platelet defects.

Methods

Platelet were isolated from citrated whole blood from consenting healthy donors and select patients. Samples were immediately fixed in 2% paraformaldehyde, washed, and loaded on poly-L-lysine coated high-resolution coverslips. We performed indirect immunostainings for various antigens, including alpha-tubulin, GP1b, Von Willebrand Factor (VWF) and platelet factor 4. Samples were imaged on an Elyra PS1 (Zeiss) with a SIM module using five phases and five rotations. After image acquisition, samples were reconstructed using Zen Software. Reconstructed images were analyzed by ImageJ plugins combined with in house written macro's which allowed for assessment of platelet morphology (volume, surface and shape), granule morphology (counts, volume, surface, shape) and granule metrics (smallest distance between granules and closest distance to membrane). We analyzed platelets from five healthy donors for general variation as well as from selected patients with platelet defects, which allowed us to validate our analytic workflow. Statistics was performed using GraphPad Prism 8.

Results

We have first assessed the variation in granule content in healthy donors. Segmentation of single resting platelets was done on cells stained for GP1b or alpha-tubulin, while alpha granule content of platelets was assessed by staining for VWF. These analyses revealed that the granule number (defined as unique VWF-positive spots of fluorescence using an optimized threshold) was normally distributed around the average of 16.78 per platelet. While there is considerable variation in alpha granule number (SD = 7.641), the sensitivity of our analysis is retained by analyzing hundreds of platelets per donor (SEM = 0.6923). We then tested whether our analysis could pick up subtle differences in platelet granules. For this purpose, we analyzed platelets from a type 2A Von Willebrand's Disease (VWD) patient with a C1190R mutation, leading to defective multimerization of VWF. Morphological analysis of SIM images revealed that the number of VWF-positive alpha granules was similar when compared to controls (16.7±0.7; p>0.05). Alpha granules in this patient were smaller (granule volume = 42.9±0.9 vs. controls 207.7±9.2; p<0.0001) and irregularly shaped (granule sphericity = 0.47±0.002 vs. controls 0.72±0.01; p<0.0001). As a control, we stained for platelet factor 4. There were no differences in granule number (18.3±0.4 vs. controls 19.5±0.5; p>0.05), nor granule volume (p=0.78). Altogether, this suggests that there is less VWF located in the alpha-granules of the VWD patient which we confirmed by Western blot and confocal microscopy. Our results imply that the packaging of VWF into alpha granules is impaired in VWD patients carrying the C1190R mutation.

Discussion

We have developed an automated image analysis workflow that allows for evaluation of platelets and their granules in patients with VWD. Our findings show the utility of unbiased, high throughput analyses of platelets and platelet granules for the diagnosis of hemostatic disorders.