Acetylcholine inhibits platelet activation and regulates hemostasis

Platelets are key mediators of thrombosis. Many agonists of platelet activation are known, but there are fewer identified endogenous inhibitors of platelets, such as prostacyclin and nitric oxide (NO). Acetylcholinesterase inhibitors such as donepezil can cause bleeding in patients, but the underlying mechanisms are not well understood. We hypothesized that acetylcholine is an endogenous inhibitor of platelets. We measured the effect of acetylcholine or analogues of acetylcholine upon human platelet activation ex vivo. We characterized expression of components of the acetylcholine signaling pathway in human platelets. We tested the effect of a subunit of the acetylcholine receptor, CHRNA7, on acetylcholine signaling in platelets. Acetylcholine and analogues of acetylcholine inhibited platelet activation, as measured by P-selectin translocation and GPIIbIIIA conformational changes. Conversely, we found that antagonists of the acetylcholine receptor such as pancuronium enhance platelet activation. Furthermore, drugs inhibiting acetylcholinesterase such as donepezil also inhibit platelet activation, suggesting that platelets release acetylcholine. We found that NO mediates acetylcholine inhibition of platelets. Human platelets express members of the acetylcholine signaling pathway including CHRNA2, CHRNA7, CHRNB1, and ACHE. Platelets from mice lacking Chrna7 are hyperactive when stimulated by thrombin and resistant to inhibition by acetylcholine. Furthermore, acetylcholinesterase inhibitors prolonged bleeding in wild-type mice. Knockout mice lacking Chrna7 subunits of the acetylcholine receptor display prolonged bleeding as well. Our data suggest that acetylcholine is an endogenous inhibitor of platelet activation. The cholinergic system may be a novel target for anti-thrombotic therapies.


28
Platelets are key mediators of thrombosis. Many agonists of platelet activation are 29 known, but there are fewer identified endogenous inhibitors of platelets, such as prostacyclin 30 and nitric oxide (NO). Acetylcholinesterase inhibitors such as donepezil can cause bleeding in 31 patients, but the underlying mechanisms are not well understood. We hypothesized that 32 acetylcholine is an endogenous inhibitor of platelets.

33
We measured the effect of acetylcholine or analogues of acetylcholine upon human 34 platelet activation ex vivo. We characterized expression of components of the acetylcholine 35 signaling pathway in human platelets. We tested the effect of a subunit of the acetylcholine 36 receptor, CHRNA7, on acetylcholine signaling in platelets. Acetylcholine and analogues of 37 acetylcholine inhibited platelet activation, as measured by P-selectin translocation and GPIIbIIIA 38 conformational changes. Conversely, we found that antagonists of the acetylcholine receptor 39 such as pancuronium enhance platelet activation. Furthermore, drugs inhibiting 40 acetylcholinesterase such as donepezil also inhibit platelet activation, suggesting that platelets 41 release acetylcholine. We found that NO mediates acetylcholine inhibition of platelets. Human Introduction

66
Platelet activation is crucial for hemostasis and thrombosis (1-3) . A variety of agonists 67 activate platelets in vivo, including thrombin, collagen, and ADP (4-8). An equally important 68 aspect of platelet biology is inhibition of activation, limiting excess thrombosis which can 69 otherwise lead to stroke or pulmonary embolism. Endogenous platelet inhibitors include factors 70 released from endothelial cells such as nitric oxide and prostacyclin (9-12).  Figure 1A).

205
We next explored the effect of acetylcholine on platelet activation. Acetylcholine inhibits TRAP 206 activation of human platelets in a dose responsive manner by over 25% of maximal stimulation 207 ( Figure 1B), and acetylcholine inhibits platelet activation over a range of TRAP doses ( Figure   208 1C).

209
We tested the effect of acetylcholine signaling upon platelets stimulated with different 210 agonists, including: TRAP, which activates the thrombin receptors PAR1 and PAR4; ADP, 211 which activates the ADP receptor P2Y12; U44619 which activates the thromboxane receptor TP;

212
and convulxin, which activates the collagen receptor GPVI. Although carbachol inhibits TRAP 213 activation of human platelets, carbachol has no effect on platelet activation by other agonists 214 ( Figure 1D).

215
The above data show that acetylcholine inhibits alpha-granule release. Next we tested 216 the effect of acetylcholine signaling on other aspects of platelet activation, namely dense granule 217 secretion and GPIIbIIIa conformational changes. We found that the acetylcholine analogue 218 carbachol decreases dense granule exocytosis measured by release of ATP ( Figure 1E) and 219 inhibits GPIIbIIIa activation measured by FITC-fibrinogen binding ( Figure 1F). Furthermore, 220 endogenous acetylcholine has the same effect (as shown when the acetylcholine esterase 221 inhibitor pyridostigmine is added) ( Figure 1E). Taken together, these data suggest that stimulation of the acetylcholine receptor inhibits 223 platelet activation.  While acetylcholine signaling inhibits platelet activation, the potential source of 250 acetylcholine in vivo remains unclear. We hypothesized that platelets release acetylcholine 251 which inhibits platelet activation in an autocrine or paracrine manner. We treated platelets with 252 the acetylcholinesterase inhibitor pyridostigmine bromide prior to activation. We observed that 253 inhibition of acetylcholinesterase (AChE) decreases platelet activation (Figure 2A). This is 254 consistent with the idea that pyridostigmine bromide inhibits acetylcholinesterase, increasing the 255 amount of acetylcholine released by platelets which is available to signal through the 256 acetylcholine receptor. We then confirmed that pancuronium bromide, which antagonizes the 257 acetylcholine receptor, enhances platelet activation ( Figure 2B). We tested these effect of these 258 compounds on platelet GPIIbIIIa activation using FITC-fibrinogen, and observed that agonism of 259 acetylcholine receptors inhibits, and antagonism of acetylcholine receptors enhances binding 260 ( Figure 2C).

261
Patients who take donepezil may have an increased risk of bleeding (14,16,17). Since

262
donepezil is an acetylcholinesterase inhibitor, we hypothesized that donepezil inhibits platelet 263 activation. To test this hypothesis, we treated platelets with donepezil hydrochloride and then 264 stimulated them with TRAP. Donepezil inhibits platelet activation ( Figure 2D). These data are 265 consistent with the hypothesis that endogenous acetylcholine released from platelets inhibits 266 platelet activation.

279
Platelets were isolated and treated with 10 nM carbachol, 100 uM pyridostigmine or 100 nM

290
Platelets express NOS3 (25). We proposed that nitric oxide mediates acetylcholine inhibition of 291 platelets. In order to test our idea, we treated human platelets with an inhibitor of nitric oxide 292 synthase, L-nitroarginine methyl ester (L-NAME), and then treated with carbachol and stimulated 293 with TRAP. We observed that carbachol inhibits platelets, but NOS inhibition blocks the effects 294 of carbachol ( Figure 3A). To confirm that acetylcholine signaling triggers NO synthesis in 295 platelets, we measured carbachol stimulation of cGMP, a messenger downstream of NO.

296
Carbachol increases cGMP levels in human platelets, and the effect of carbachol is blocked by 297 the NOS inhibitor L-NAME ( Figure 3B). The inhibitory effect of NO was further tested with a 298 range of L-NAME doses. We found that L-NAME inhibits the effects of acetylcholine on platelets 299 in a dose-dependent manner ( Figure 3C). Since calcium signaling can regulate NOS activation, we explored a calcium signaling pathway in platelets. First, carbachol increases intracellular 301 calcium levels in platelets ( Figure 3D). Second, the calcium chelator BAPTA blocks the ability of 302 carbachol to inhibit platelets. Finally, calmodulin is important for acetylcholine inhibition of 303 platelet activation ( Figure 3F). Taken together, our data suggest that NO mediates acetylcholine 304 inhibition of platelets via a calcium-calmodulin dependent mechanism.

325
Human platelets express mRNA encoding acetylcholine receptor subunits We next analyzed human platelets expression of genes involved in acetylcholine 327 signaling and metabolism. Using qPCR of human platelet RNA, we found that human platelets 328 express mRNA for a subunit of the acetylcholine receptor CHRNA7 and acetylcholinesterase 329 ACHE ( Figure 4A). Since our qPCR data suggest that human platelets express CHRNA7 330 mRNA, we then evaluated if platelets express CHRNA7 protein using a well characterized 331 natural toxin, alpha-bungarotoxin (-BT), which selectively binds to CHRNA7 subunits. We 332 found that FITC labeled -BT binds to human platelets, and this binding is specific since the 333 binding can be competitively inhibited by excess non-labeled -BT ( Figure 4B). These data 334 confirm that platelets express functional CHRNA7 subunits on their surface. The acetylcholine receptor subunit CHRNA7 inhibits platelet activation

344
We next tested the hypothesis that acetylcholine alters platelet activation by signaling 345 through the CHRNA7 subunit of the acetylcholine receptor complex. We treated platelets from 346 Chrna7 (WT) or Chrna7 (KO) mice with thrombin and measured their activation by examining P-347 selectin exposure and fibrinogen binding to the integrin GPIIbIIIA. We observed that Chrna7(KO) 348 mice are more sensitive to activation by thrombin indicated by P-selectin exposure ( Figure 5A),

349
and that they reach maximum changes in integrin GPIIbIIIa conformation at lower doses of 350 thrombin than WT controls ( Figure 5B). This is consistent with the idea that endogenous 351 acetylcholine signaling inhibits platelets. We also tested the ability of Chrna7 to mediate acetylcholine inhibition of platelet 353 activation. Acetylcholine inhibits activation of platelets from Chrna7(WT) mice but fails to inhibit 354 activation of platelets from Chrna7(KO) mice ( Figure 5C). The acetylcholine receptor agonist 355 carbachol also inhibits activation of platelets from wild-type mice but not from Chrna7(KO) mice 356 ( Figure 5C).

358
Acetylcholine signaling regulates hemostasis in mice

364
For a genetic approach, we compared Chrna7(WT) and Chrna7(KO) mice. Chrna7(KO) 365 mice have shorter bleeding times than wild-type mice in a tail transection model of hemostasis

367
For a pharmacological approach, we treated mice with donepezil, an AChE inhibitor, and 368 then measured time to cessation of tail bleeding. Donepezil prolongs the bleeding time of mice 369 ( Figure 5E). We also confirmed that the acetylcholine receptor agonist carbachol increases 370 bleeding time ( Figure 5E).

371
These data suggest that endogenous acetylcholine inhibits platelet activation and 372 hemostasis in vivo.

393
The major finding of our study is that acetylcholine inhibits platelet activation.

394
Acetylcholine signals through the acetylcholine receptor, increasing NO levels, and inhibiting 395 platelet activation. Acetylcholine inhibits activation of platelets from humans and mice by over 396 15%. Acetylcholine signaling is important in vivo, since mice lacking the acetylcholine receptor 397 subunit Chrna7 have shorter bleeding times. Finally, an acetylcholinesterase inhibitor drug 398 used in humans that is associated with bruising also inhibits activation of human platelets and 399 prolongs bleeding in mice. Taken together, our results suggest that acetylcholine is an 400 endogenous inhibitor of platelet activation.

401
We showed that CHRNA7 is necessary for acetylcholine inhibition ( Figure 5)

425
We found that acetylcholine inhibits platelet activation in vitro by about 15% ( Figure 1B).

426
Carbachol, an analog of acetylcholine, has a much stronger effect upon platelet activation,

427
inhibiting P-selectin translocation by over 90% (Figure 1A and 5A). Thus exogenous agonists 428 like carbachol have a powerful effect upon platelet activation, but endogenous agonists such as 429 acetylcholine have a more modest inhibitory effect on platelet activation.

430
Our work extends prior research on cholinergic signaling in platelets. Others have shown 431 that agonists of AChR decrease human platelet activation ex vivo as measured by GPIIbIIIa 432 conformational changes and by aggregation (19). We show that acetylcholine itself inhibits 433 platelet degranulation. ( Figure 1B). Others have shown that platelets from mice lacking Chrna7 434 have increased aggregation when stimulated by ADP ex vivo (20). We extend these data,

435
showing that these platelets do not respond to acetylcholine and have increased exocytosis 436 ( Figure 5C). Finally, we add to the current literature by showing that acetylcholine and its 437 analogs modulate hemostasis in vivo.

438
Our study has several limitations which suggest future studies. We have not yet defined 439 the composition of the acetylcholine receptor on platelets, and we have not identified the role of 440 all acetylcholine subunits in mediating platelet inhibition. Another limitation is that we have 441 indirect evidence that platelets store acetylcholine in their granules, since acetylcholinesterase 442 inhibitors boost platelet inhibition, but we have not directly measured acetylcholine inside platelet 443 granules.

444
Our studies have pharmacological relevance to humans. We show that donepezil 445 inhibits platelet activation ex vivo at a concentration between 5 -50 uM ( Figure 2D). This increase bruising by about 2% more than placebo (18). Our data support our proposal that 450 drugs that target acetylcholine esterase can promote bleeding in humans, and may explain why 451 donepezil is associated with hemostatic abnormalities in humans.

452
Our study also has therapeutic implications for the management of thrombosis. Our data 453 suggest that drugs targeting acetylcholine receptor subunits might inhibit thrombosis.

454
Furthermore, our data suggest that drugs increasing acetylcholine signaling will increase the risk 455 of bleeding and bruising in patients.