SARS-CoV-2 Omicron variant infection affects blood platelets, a comparative analysis with Delta variant

Abstract

Introduction: In November 2021, the SARS-CoV-2 Omicron variant of concern has emerged and is currently dominating the COVID-19 pandemic over the world. Omicron displays a number of mutations, particularly in the spike protein, leading to specific characteristics including a higher potential for transmission. Although Omicron has caused a significant number of deaths worldwide, it generally induces less severe clinical signs compared to earlier variants. As its impact on blood platelets remains unknown, we investigated platelet behavior in severe patients infected with Omicron in comparison to Delta.

Methods: Clinical and biological characteristics of severe COVID-19 patients infected with the Omicron (n=9) or Delta (n=11) variants were analyzed. Using complementary methods such as flow cytometry, confocal imaging and electron microscopy, we examined platelet activation, responsiveness and phenotype, presence of virus in platelets and induction of selective autophagy. We also explored the direct effect of spike proteins from the Omicron or Delta variants on healthy platelet signaling.

Results: Severe Omicron variant infection resulted in platelet activation and partial desensitization, presence of the virus in platelets and selective autophagy response. The intraplatelet processing of Omicron viral cargo was different from Delta as evidenced by the distribution of spike protein-positive structures near the plasma membrane and the colocalization of spike and Rab7. Moreover, spike proteins from the Omicron or Delta variants alone activated signaling pathways in healthy platelets including phosphorylation of AKT, p38MAPK, LIMK and SPL76 with different kinetics.

Discussion: Although SARS-CoV-2 Omicron has different biological characteristics compared to prior variants, it leads to platelet activation and desensitization as previously observed with the Delta variant. Omicron is also found in platelets from severe patients where it induces selective autophagy, but the mechanisms of intraplatelet processing of Omicron cargo, as part of the innate response, differs from Delta, suggesting that mutations on spike protein modify virus to platelet interactions.

Keywords: COVID-19; LC3B; autophagy; omicron variant; platelets; spike.

Figure 1

Platelet phenotype in severe COVID-19 patients infected with the Omicron variant. Representative Western blotting analysis of IFITM3 levels in platelets from 8 healthy donors (Healthy), 11 severe COVID-19 patients with the Delta variant (Delta), and 9 severe COVID-19 patients with the Omicron variant (Omicron) is shown (A). As an index of the purity of the washed platelet preparation, the amount of remaining leukocytes was 1 leukocyte/1000 platelets. The quantification of the IFITM3/actin ratio is shown. ## p <.01 represents the comparison of Delta patients vs. Omicron patients; *** p<.001 and ****p <.0001 represent the comparison of healthy donors vs. Delta or Omicron patients, based on the nonparametric Mann-Whitney test. Capillary Western Assay analysis of phospho- and total IRAK4 in platelets from 8 healthy donors (Healthy), 11 severe COVID-19 patients with the Delta variant (Delta), and 9 severe COVID-19 patients with the Omicron variant (Omicron) is shown (B). The quantification of the phospho-IRAK4 (p-IRAK4)/Total protein ratio and the IRAK4/Total protein ratio is shown in graphs. Significance denoted as ## p<.01 and ### p<.0001 represents the comparison of Delta patients vs. Omicron patients. Significance denoted as ***p <.001 and ****p <.0001 represents the comparison of healthy donors vs. Delta or Omicron patients based on the nonparametric Mann-Whitney test. Detection of SARS-CoV-2 in platelets from severe COVID-19 patients with the Omicron variant was performed using transmission electron micrographs revealing the presence of characteristic crown-like structures resembling SARS-CoV-2 particles (C). Images shown are representative of those obtained from the 6 patients with the Omicron variant (2 seronegative, 4 seropositive) analyzed by TEM. Viral-like particles (white arrows) were identified based on specific criteria, including the size of the viral particle (100 nm ± 20%) as described (22). Quantitative analysis was performed on transmission electron micrographs (186 from n = 6 healthy donors [31 per donor], and 486 from n = 6 patients with the Omicron variant [80 per patient]). The percentage of platelets from healthy controls and patients with the Omicron variant containing viral-like particles were quantified (D). Results are mean ± SEM; each circle represents an individual sample (full circle: seropositive, empty circle: seronegative). Significance denoted as **** p<.0001 represents the comparison of healthy donors vs. Omicron patients based on the nonparametric Mann-Whitney test. The number of vesicle-like structures per cross-section (E), and the vesicle-like surface/platelet cross-section surface ratio were quantified using electron micrographs of complete platelet cross-sections. Significance denoted as ***p <.001 represents the comparison of healthy donors vs. Omicron patients based on the nonparametric Mann-Whitney test. The pink areas in the graphs represents the mean +/- SEM we previously found in platelets from severe patients with the Delta variant (22).

Figure 2

Ultrastructure of platelets in severe COVID-19 patients infected with the Omicron variant. Transmission electron micrographs of platelets from patients infected with the Delta or the Omicron variants and controls (A) were imported into QuPath software and regions of interest were annotated. Once the selection and annotation were complete, the entire platelet was displayed in green while the vacuoles were shown in red. The pixel size was determined by using the scale bar provided on the electron micrograph. The brush tool was used to draw and annotate the regions of interest. The software determined the position of the centroids allowing to measure the distance between the center of the platelet and the vesicle centroids. The position of 1683 vesicles for the Delta variant (300 platelets from 10 patients), 3705 vesicles for the Omicron variant (486 platelets from 6 patients) and 1790 vesicles for controls (129 platelets from 4 healthy donors) were determined (B). The presence of SARS-CoV-2 spike protein within platelets from Delta or Omicron infected patients was analyzed by super-resolution confocal microscopy (C), and the positions of spike-positive vesicles within platelets was determined. The percentage of platelets exhibiting labeling at the periphery of the cytoplasmic membrane was quantified (D). Significance denoted as #p <.05 and ###p <.0001 represents the comparison of Delta vs. Omicron patients based on the nonparametric Mann-Whitney test. Significance denoted as **** p <.0001 represents the comparison of healthy donors vs. delta patients based on the nonparametric Mann-Whitney test.

Figure 3

Platelets from severe COVID-19 patients infected with the Omicron variant have autophagic structures. Representative TEM sections of platelets from severe COVID-19 patients infected with the Omicron variant highlighting structures typical of elongation membrane (EM), autophagosome (AP), and autophagolysosome-like (AL) are shown (A). The percentage of platelets containing autophagy structures is presented. Each circle represents an individual (full circle: seropositive, empty circle: seronegative) and results are mean ± SEM (B) along with the frequency of occurrence of the different types of structures (C) in platelets from 6 healthy donors and 6 severe COVID-19 patients with the Omicron variant. Representative Western blotting analysis of LC3B-I and LC3B-II levels in platelets from 6 severe COVID-19 patients with the Omicron variant (O1 to O6) and 9 healthy donors (H1, H2) are shown (D). The exposure time was adjusted in order to efficiently distinguish the LC3B type I and type II. The quantification of the LC3B-I/actin ratio and LC3B-II/actin ratio for healthy donors (n=9), and patients with the Omicron variant (n=6) is shown (E). Significance denoted as **p <.01 and ***p <.001 represents the comparison of healthy donors vs. Omicron patients, based on the nonparametric Mann-Whitney test. The pink areas in the graphs represents the mean +/- SEM we previously found in platelets from severe patients with the Delta variant (22).

Figure 4

Omicron spike protein subcellular distribution in platelets from severe COVID-19 patients. The intraplatelet localization of the spike protein was investigated using immunofluorescence and super-resolution confocal microscopy with the Airyscan module. A specific anti-spike antibody was used. Its colocalization with LC3B, a marker of autophagosomes (A), EEA1, a marker of early endosomes (B), and Rab7, a marker of late endosomes (C), was analyzed. Representative images from 6 different patients with severe COVID-19 are shown. Quantification was performed by analyzing 30 to 40 platelets from each of the 6 patients with the Omicron variant (4 seronegative, 2 seropositive). Pearson’s correlation coefficient and Manders’ coefficient calculations are presented (D, E). Significance denoted as *p <.05 and ****p <.001 represents the comparison between the colocalization of spike protein with LC3B vs. the colocalization of spike protein with Rab7. Significance denoted as #### p <.001 represents the comparison between the colocalization of spike protein with EEA1 and the colocalization of spike protein with Rab7. The pink areas in the graphs corresponds to the mean +/- SEM we previously found in the Delta variant (22).

Figure 5

Healthy platelet signaling pathways activated by Omicron and Delta spike proteins. Washed platelets from 7 healthy donors were incubated in the presence or absence of 5µg/mL of the Delta or Omicron variants spike proteins for 1, 10 and 30 minutes at 37°C without shaking. Phosphorylation of the signaling proteins AKT (A), P38MAPK (B), LIMK1/2 (C), and SLP76 (D) was monitored over time using phosphoflow cytometry (31). As a comparison, washed platelets were also stimulated with 50 µM TRAP or 9 µg/ml CRP for 10 minutes at 37°C without shaking. Results are presented as the MFI ratio of spike protein-stimulated platelets to resting platelets (mean ± SEM, n= 7). (E) The heatmap summarizes the evolution of phosphorylation intensity for each protein as a function of platelet stimulation time. Statistical significance is indicated as *p<.05, according to the nonparametric Wilcoxon test for comparison of vehicle vs. Delta or Omicron spike proteins. Additionally, #p<.05 represents the comparison of Delta vs. Omicron spike protein using the nonparametric Wilcoxon test. Significance denoted as §§§p <.001 indicates comparison of vehicle vs. agonists, according to the nonparametric Wilcoxon test.