Cold temperature induces a TRPM8-independent calcium release from the endoplasmic reticulum in human platelets

Platelets are sensitive to temperature changes and akin to sensory neurons, are activated by a decrease in temperature. However, the molecular mechanism of this temperature-sensing ability is unknown. Yet, platelet activation by temperature could contribute to numerous clinical sequelae, most importantly to reduced quality of ex vivo-stored platelets for transfusion. In this interdisciplinary study, we present evidence for the expression of the temperature-sensitive ion channel transient receptor potential cation channel subfamily member 8 (TRPM8) in human platelets and precursor cells. We found the TRPM8 mRNA and protein in MEG-01 cells and platelets. Inhibition of TRPM8 prevented temperature-induced platelet activation and shape change. However, chemical agonists of TRPM8 did not seem to have an acute effect on platelets. When exposing platelets to below-normal body temperature, we detected a cytosolic calcium increase which was independent of TRPM8 but was completely dependent on the calcium release from the endoplasmic reticulum. Because of the high interindividual variability of TRPM8 expression, a population-based approach should be the focus of future studies. Our study suggests that the cold response of platelets is complex and TRPM8 appears to play a role in early temperature-induced activation of platelets, while other mechanisms likely contribute to later stages of temperature-mediated platelet response.


39
The detection of temperature change by nerve receptors in human tissues is a life-preserving function 40 that is highly evolutionarily conserved. Non-neural cells also sense cold, a prime example being blood platelets.

41
Platelets are anucleate blood cells, derived from megakaryocytes in the bone marrow (1). Platelet activation is 5 88 understood. In addition, the molecular entity responsible for cold temperature-induced calcium release from the 89 ER in platelets needs to be further investigated.

91
TRPM8 mRNA in MEG-01 cells. 92 Because platelets are generated by megakaryocytes, we first evaluated TRPM8 expression in MEG01 93 cells, a megakaryocytic cell line. We amplified TRPM8 transcripts by PCR using mRNA isolated from cultured 94 MEG-01 cells using primers designed for human TRPM8 sequences (Figure 1 A) was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made  120 We tested for TRPM8 protein expression in MEG-01 cells by Western blot. We observed a strong band at 121 the expected size of ~130 kDa, corresponding to full-length TRPM8. Interestingly, we also observed two smaller 145 and 35.3% respectively). In addition, we used WS-12-a more specific TRPM8 agonist structurally similar to 146 menthol (22), as well as icilin-a super-cooling agent, which is structurally distinct from menthol (19). In contrast 147 to menthol, WS-12 and icilin, led to a significantly higher calcium influx than vehicle (Figure 1 D and G, Mean ± 148 SEM = 0.12 ± 0.02 and 0.08 ± 0.01, p-value = 0.005 and 0.03, respectively). WS-12 and icilin-responsive cell 149 populations were larger than the vehicle's positivity rate (Figure 1 F, 57.3% and 59%, respectively). Taken 150 together, our data suggest that TRPM8 is functional at low levels in a population of platelet precursor MEG-01 151 cells.

152
TRPM8 protein in human platelets 153 We examined purified human platelets for the presence of TRPM8 messenger RNA. We found evidence 154 for TRPM8 mRNA, but were unable to rule out leukocytes as a source of TRPM8 mRNA (Supplementary Figure   155 2). Therefore, for the detection of TRPM8 protein expression in platelets, we used an antibody specifically . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made recognizing an extracellular epitope of TRPM8. Using Western blot, we detected several bands specific to the 157 TRPM8 epitope, but there was a very low expression of the full-length TRPM8 (Supplementary Figure 3). Next, 158 we examined TRPM8 expression in human platelets by flow cytometry (see Figure 2 A-C for gating strategy).

176
To visualize TRPM8 and gather more information about its specific localization in platelets, we used 177 imaging flow cytometry. Platelets were identified by αIIb integrin staining (CD41). As shown in Figure 2 E, 178 platelets positive for TRPM8 show a punctate staining pattern, which tends to localize to the periphery of the . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted July 19, 2023. ; https://doi.org/10.1101/2023.07.19.549670 doi: bioRxiv preprint platelet. In some instances, the TRPM8-positive foci were localized outside of the platelet outline, seemingly 180 attached to the platelet with a thin undistinguishable protrusion (Figure 2 E, middle row). Interestingly, within   181  the TRPM8-positive platelet population, there was a significantly lower percentage of spheroid cell shape and a   182 significantly higher percentage of discoid cells (Figure 2 F). Thus, a small platelet population is positive for the 183 surface expression of TRPM8.

184
We also sought to visualize TRPM8 channels on platelet footprints attached to glass surface using 185 immunostaining and confocal microscopy. Like our findings by imaging flow cytometry, we observed anti-  determine the level of platelet activation in whole blood. We observed significantly more αIIbβ3 activation in . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made platelets collected into room temperature (22°C) tubes (Mean ± SEM = 13.4 ± 3.2% PAC-1-positive cells, Figure 3 202 A) than in blood collected into pre-warmed tubes (37°C) (Mean ± SEM = 1.4 ± 0.4% PAC-1-positive cells Figure   203 3 A).

222
We hypothesized that inhibition of the TRPM8 channel during the collection at room temperature would 223 lead to decreased platelet activation. It has been previously reported that the temperature threshold of TRPM8 . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made  Figure   228 3 A, Mean ± SEM = 1.4 ± 0.3%, p =0.85). In contrast, when blood was collected at 22 °C, PF 05105679 treatment 229 reduced the percentage of platelets that were PAC-1 -positive compared to vehicle, to the degree that 230 approached statistical significance (Figure 3 A, Mean ± SEM = 7.9 ± 1.3%, p =0.06). The inhibition by PF 05105679 231 at 22 °C failed to reach the baseline, suggesting some level of TRPM8-independent activation by 22 °C 232 temperature during collection. In addition, the TRPM8 inhibitor did not affect the activation of platelets exposed 233 to 4 °C (Figure 3 A, green). These data implicate TRPM8 in αIIbβ3 activation in platelets exposed to room 234 temperature.

235
We also analyzed microaggregates by imaging flow cytometry. Microaggregates are small clusters of 236 platelets, a feature of the pro-aggregatory platelet status, and used as a tool to evaluate the antithrombotic effects 237 of compounds (30,31). In a recent study, the presence of microaggregates was responsible for the wastage of a 238 significant proportion of cold-stored platelet units (32). To evaluate the effects of temperature and TRPM8 239 inhibition on the microaggregate content in our samples, we collected blood into pre-warmed tubes (37°C) 240 containing either vehicle or PF 05105679. After incubation for 5 min, the samples were cooled down to 22°C or 241 4°C for 15 min. . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The temperature-induced platelet shape change is TRPM8-dependent. 246 Platelets change morphology, from discoid to spheroid, upon cold exposure (15,33). We recently 247 developed an unbiased screening protocol for differentiating between the two morphological states using 248 imaging flow cytometry (34). The 2D images of 3D discs and spheres are accurately described as fusiform 249 (spindle-shaped) and circular shapes, but, for simplicity, we will refer to these as discoid and spheroid.

250
We hypothesized that cooling blood from 37°C to 22°C leads to a decrease in discoid events, and an 251 increase in spheroid events, consistent with platelet activation. For samples collected at 37°C, the fraction of 252 spheroid-shaped platelets was 20.9 ± 2.8 % (Mean ± SEM) and discoid cells were 78.4 ± 3.0 % (

265
We also evaluated the effects of the short-term cooling of platelets to 4 °C. Blood collected at 37 °C and 266 then incubated at 4 °C for 15min, evinced a significant increase in the spheroid fraction of platelets ( Figure 3 E . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made   Platelet activation can also be measured by α-granule secretion with a P-selectin (CD62P) antibody (35).
290 P-selectin positivity was not increased upon treatments with TRPM8 agonists compared to vehicle at either 291 temperature (Figure 4 B). In addition, pre-incubation with TRPM8 inhibitor PF 05105679 had no effect (Figure 4 292 B). Furthermore, there was no platelet activation when we incubated platelets with TRPM8 agonists for 1 or 4 293 hours at either 22 °C or 4 °C (Supplementary Figure 4, 5). Therefore, in platelets, activation of TRPM8 by chemical 294 agonists induced neither αIIbβ3 activation nor degranulation of α-granules.

295
Platelet aggregation is TRPM8-dependent in a subpopulation of donors 296 Next, we investigated whether TRPM8 activation affects platelet aggregation. We compared the   330 We observed that TRPM8 inhibition reduces platelet activation, shape change, and aggregation.

331
Therefore, we evaluated the effects of TRPM8 inhibition on aggregation initiated by different signaling 332 pathways. We tested two specific TRPM8 inhibitors, PF 05105679 and AMTB (36), on platelets from room-333 temperature collections. We focused on two distinct pathways for the initiation of platelet aggregation -334 purinergic receptors (activated by ADP), and the immunoreceptor tyrosine-based activation motif (ITAM) . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted July 19, 2023.

346
This indicates that TRPM8 inhibition does not affect the intrinsic SOCE response of platelets. 348 To evaluate the role of TRPM8 for cytosolic calcium influx, we recorded fluorescence from Fura 2-AM 349 loaded platelets using a microplate spectrophotometer. icilin at 22°C or 37°C did not result in a significant change in calcium compared to the vehicle (Figure 6 B).

Platelets do not exhibit TRPM8-dependent calcium influx
. CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted July 19, 2023. ; https://doi.org/10.1101/2023.07.19.549670 doi: bioRxiv preprint Figure 6. Platelets do not elicit TRPM8-  . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made  391 We investigated the effect of fast chilling on platelet intracellular calcium levels using the temperature-  (Figure 7 A). The addition of TRPM8 inhibitor PF 05105679 did not affect the 399 amplitude or shape of the temperature response in platelets in the presence of Ca 2+ (Figure 7 A). Furthermore, 400 when Ca 2+ was omitted, PF 05105679 did not affect temperature response curves (Figure 7 B). These data suggest 401 that acute temperature-induced calcium influx in platelets is not mediated by TRPM8.

Acute cold exposure leads to immediate ER-derived calcium increase independent of TRPM8
. CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted July 19, 2023. ; https://doi.org/10.1101/2023.07.19.549670 doi: bioRxiv preprint Figure 7. Chilling platelets leads to a rapid calcium increase independent of TRPM8. (A-B)   . We conclude that acute response to chilling in platelets is dependent on the calcium released from 422 the endoplasmic reticulum (also referred to as a dense tubular system, DTS).
. CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made  (49) and precedes αIIbβ3 integrin activation and 459 aggregation (50). At the site of superficial vascular injury, blood could be exposed to colder temperatures.

460
Therefore, an immediate, cold-induced shape change could facilitate platelet adhesion to subendothelial matrix 461 proteins such as collagen and thereby play a physiological role in hemostasis. However, we cannot exclude that 462 the PF 05105679 inhibitor has off-target effects on other signaling pathway components upstream of SOCE.

463
We expected TRPM8 to be excitable by chemical agonists as has been reported in other cell types. We 464 tested three different chemical agonists of TRPM8, namely menthol, WS-12, and icilin in multiple assays.
465 Surprisingly, they consistently failed to cause platelet activation. In only one assay, menthol, but not other 466 TRPM8 agonists, lead to a significant increase in aggregation. Menthol is the least specific of the agonists we . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted July 19, 2023. ; https://doi.org/10.1101/2023.07.19.549670 doi: bioRxiv preprint used, and it was shown to have TRPM8-independent responses in other cell types (51). We used different 468 ambient temperature protocols because TRPM8 activation is enhanced by temperature, which may explain why 469 at 22°C (but not at 37°C) the calcium increase to menthol was approaching significance. In addition, the low 470 intensity of TRPM8-positive platelets observed via flow cytometry combined with faint signals on western blot, 471 suggests that there is a small number of copies of TRPM8 per cell. The low prevalence of TRPM8-positive 472 platelets may lead to a small response amplitude of the population measurements, bringing it close to the 473 detection limit of the assays used in this study. Overall, our data indicate that platelet TRPM8 appears 474 unresponsive to chemical agonists, at least with acute stimulation.

475
One of the hallmarks of platelet activation by temperature is the increase in cytosolic calcium. The 476 pathophysiologic mechanism for this has been elusive, but multiple non-exclusive mechanisms have been 477 proposed for this phenomenon, such as an imbalance in the activity of the Ca 2+ -ATPases, the phase transition of 478 lipids, the changes in PLC activity, or the calcium release from the ER. Consistent with the previous report by 479 (15), we observed that platelets exhibit a rapid increase in cytosolic calcium in response to lowered temperatures.

480
Given its role as a calcium-permeable ion channel, we expected TRPM8 to play a critical role in this phenomenon.

481
However, TRPM8 was completely dispensable for the temperature-induced calcium increase. In addition, for 482 the first time to our knowledge, we identified that the rapid increase in intracellular calcium is directly 483 dependent on the ER calcium stores (15,16,40,52,53). Further investigation is needed to identify the molecular 484 entity responsible for the temperature-induced calcium release from the ER.

485
Several prospects arise from our findings. First, further studies are needed to identify the functional role 486 of TRPM8 expression in megakaryocytes. Equally intriguing is the identification of the molecular mechanism 487 for cold sensation at the endoplasmic reticulum level. This would allow to design a pharmaceutical intervention 488 to interfere with this acute response, e.g., in patients with therapeutic hypothermia or to inhibit the cold response . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made of stored platelets. Our study also highlights the need to understand the large variability in TRPM8 expression 490 on a healthy volunteer population basis.

491
In conclusion, TRPM8 contributes to some features of the acute cold response in platelets but is 492 dispensable for acute cold-induced calcium increase. Instead, we identified a critical role of ER-dependent 493 calcium in response to cold. Japan). Images were acquired using SlideBook software (Intelligent Imaging Innovations, Inc., Denver, CO) 530 every 5 seconds with 500 ms exposure. Each recording was no longer than 40 minutes. The data was analyzed 531 using FIJI software. Region of interest (ROI) was drawn around each cell and average pixel intensity over time 532 was measured. Background signal from a similarly sized ROI without cells was subtracted from each cell signal.

Materials and methods
. CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made Individual cell intensity was normalized according to the formula: F/(F max -F 0 ), where F is the average 534 fluorescence of an ROI at any given time, F max is a maximum intensity measured during the calcium-ionophore  ThermoFisher) in the dark imager chamber equipped with a camera. Images were processed using FIJI software.

615
For a line scan quantification, a line was drawn in each lane of the gels, and pixel intensity over distance was 616 measured using the ROI Manager tool of FIJI. Each lane was normalized to the average pixel intensity inside an 617 ROI drawn at the high molecular weight. Finally, measurements from line scans of the blots incubated with the . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made blocking peptide were subtracted from the corresponding lanes of the blot stained with only primary and 619 secondary antibodies.

620
Conventional flow cytometry 621 To determine the levels of platelet activation at different temperatures, whole blood was collected by 622 venipuncture using a vacutainer containing sodium citrate at 37˚C or 22˚C. PRP was isolated as described above 623 and incubated for 30 minutes at the assigned temperature. PRP was adjusted to a platelet density of 300K

629
To determine the levels of washed platelet activation, samples were first prepared as described above.

638
Samples were run on the LSR II flow cytometer (BD Biosciences). For each sample, a total of 100,000 639 events were acquired. Data were analyzed by FlowJo V10 software. Platelets were distinguished from other . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted July 19, 2023. ; https://doi.org/10.1101/2023.07.19.549670 doi: bioRxiv preprint blood cells by using forward (FSC) and side scatter (SSC) and CD61 positive events. The gating strategy for PAC-641 1 and P-selectin was selected based on the binding of the appropriate isotype antibodies.

642
Imaging flow cytometry 643 For TRPM8 positivity we used the same staining protocol as for conventional flow cytometry. We 644 employed a gating strategy that excluded the TRPM8 (FITC) intensity of the unstained (Figure 2 A) and 645 secondary-only control samples (Figure 2 B), to identify the TRPM8-positive population (Figure 2 C) was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted July 19, 2023. ; https://doi.org/10.1101/2023.07.19.549670 doi: bioRxiv preprint