Abstract
This letter describes the further deconstruction of the known PAR4 inhibitor chemotypes (MWs 490 – 525 and with high plasma protein binding) to identify a minimum PAR4 pharmacophore devoid of metabolic liabilities and improved properties. This exercise identified a greatly simplified 2-methoxy-6-arylimidazo[2,1-b][1,3,4]thiadiazole scaffold that afforded nanomolar inhibition of both activating peptide and γ-thrombin mediated PAR4 stimulation, while reducing both molecular weight and the number of hydrogen bond donors/acceptors by ~50%. This minimum PAR4 pharmacophore, with competitive inhibition, versus non-competitive of the larger chemotypes, allows an ideal starting point to incorporate desired functional groups to engender optimal DMPK properties towards a preclinical candidate.
Keywords: PAR4, Platelet Aggregation, Miniumum Pharmacophore, Structure-Activity Relationship (SAR)
Graphical abstract
Protease Activated Receptor 4 (PAR4) is a G Protein-Coupled Receptor (GPCR) essential for the thrombin-induced procoagulant effect on platelets, and as such, has garnered a great deal of interest as a target for anti-platelet therapy to treat thrombosis without bleeding.1-3 Historically dominated by antibody therapy, small molecule PAR4 antagonists are only now emerging as in vivo tool compounds and clinical candidates.1-4 Until recently, the PAR4 antagonists in the primary literature suffered from poor DMPK properties and a lack of activity upon γ-thrombin activation.4-7 Our lab recently devulged a new series of small PAR4 inhibitors 1 with an improved DMPK profile and weak activity upon γ-thrombin activation.4-7 Bristol-Myers-Squibb (BMS) also disclosed a novel series of PAR4 inhibitors, represented by 2 (an analog of BMS986120), with exquisite potency against both activating peptide (AP) and γ-thrombin mediated PAR4 stimulation, but with virutally no free drug levels (rat and human fu <0.001).8 Both 1 and 2 (Figure 1) are high molecular weight compounds (490-510), with high clogPs (>4) and the noted high plasma protein binding.4,8
Figure 1.
Structures of reported PAR4 antagonists 1 and 2, and the minimum pharmacophore of 2, fragment 3, and plans to further optimize 3 into a more desirable fragment core.
Thus, we deconstructed 2 in an attempt to identify a minimum pharmacophore that retained potent PAR4 inhibition against both AP and γ-thrombin that could then be optimized with more favorable DMPK properties. This exercise led to the discovery that the most basic core of 2, a 6-(benzofuran-2-yl)-2-methoxyimidazo[2,1-b][1,3,4]thiadiazole 3 as a potent PAR4 inhibitor (PAR4 AP IC50 = 1.69 nM, PAR4 γ-thrombin IC50 = 58.8 nM), as the minimum pharmacophore (MW = 271) of 2.9 However, the potential liabitlities of an unsubstitiuted benzofuran, as in 3, raised metabolic stabilty concerns. Thus, in this Letter, we describe efforts to survey alternative 6-position substituents on the 2-methoxyimidazo[2,1-b][1,3,4]thiadiazole core to identify a minimum PAR4 pharmacophore that could then be further optimized with functional groups that would engender desirable DMPK properties.
In the initial SAR campaign, we elected to survey functionalized aryl moeities in the 6-position and assess if these simple analogs were sufficient to elicit PAR4 inhibition. Analogs 4, or 3, could be readily accessed in two steps from commercial materials (Scheme 1), and enabled the evaluation of multiple regions of the heterobiarylcore.10 Here, commercial α-bromo ketones 5 were condensed with 5-bromo-1,3,4-thidiazol-2-amine 6 to provide 2-bromo-6-arylimidazo[2,1-b][1,3,4]thiadiazoles 7 in 65-86% yield. An SNAr reaction with methoxide under mild conditons delivered analogs 4 in yields ranging from 72-80%.
Scheme 1.
Synthesis of aryl analogs 4. Reagents and Conditions: (a) CH3CN/IPA, 80 °C, 18 hr, 65-86%; (b) NaOMe, DCM/MeOH (3:1), 1 hr, rt 72-80%.
SAR for select analogs 4 are highlighted in Table 1, all with molecular weights less than 325. The most basic analog, 4f with an unsubstiuted phenyl ring, was a PAR4 antagonist with a PAR4 IC50 against the AP of 757 nM. Simple substitution of the phenyl ring afforded robust SAR in terms of PAR4 inhibition against the AP, with a significant right-shift in PAR4 potency upon activation with γ-thrombin. In general, substitutions at the 4-postion of the phenyl ring afforded the most potent analogs. Moreover, lipophilic, electron-withdrawing moieties proved optimal. For example, the 4-CF3 derivative (4c) was a 15.6 nM PAR4 inhibitor against the AP, and displayed an IC50 of 348 nM against γ-thrombin activation. This result is quite significant considering the reduced, minimum pharmacophore and acitivies relative to 1 and 2. Likewise, the 4-OCF3 congener (4n) was a 27.7 nM PAR4 inhibitor against the AP, and displayed an IC50 of 246 nM against γ-thrombin activation. The 4-Cl analog (4k) proved also to inhibit PAR4 upon AP activation (IC50 = 39.5 nM) as well as γ-thrombin (IC50 = 621 nM). Interestingly, the diflouromethoxy congener 4v, of 4n, remained potent against the AP (IC50 = 83.4 nM), but loss of a single fluorine atom led to no activity against γ-thrombin activation. Finally, substituents in the 2-position were generally detrimental to potency, and electron donating moities, such as 4-OCH3 (4g) and 4-OH (4l), led to significant diminution in potency.
Table 1.
Structures and activities of analogs 4.
| |||
|---|---|---|---|
| Cmpd | R | PAR4-AP PAC-1 IC50 (nM)a (pIC50±SEM) | PAR4 γ-thrombin PAC-1 IC50 (nM)a (pIC50±SEM) |
| 4a | 2-CF3 | 1,330 (5.88±0.06) | ND |
| 4b | 3-CF3 | 36.3 (7.44±0.10) | 1310 (5.88±0.24) |
| 4c | 4-CF3 | 15.6 (7.81±0.07) | 348 (6.46±0.06) |
| 4d | 2,6-diF | 1,300 (5.89±0.05) | ND |
| 4e | 2,4-diF | 175 (6.76±0.03) | ND |
| 4f | H | 757 (6.12±0.07) | ND |
| 4g | 4-OCH3 | 592 (6.23±0.11) | ND |
| 4h | 4-CH3 | 57.2 (7.24±0.04) | ND |
| 4i | 2-Cl | 437 (6.36±0.02) | ND |
| 4j | 3-Cl | 72.8 (7.14±0.07) | >10,000 (>5) |
| 4k | 4-Cl | 39.5 (7.40±0.04) | 621 (6.21±0.15) |
| 4l | 4-OH | >10,000 (>5) | ND |
| 4m | 4-CN | 937 (6.03±0.06) | ND |
| 4n | 4-OCF3 | 27.7 (7.57±0.09) | 246 (6.64±0.19) |
| 4o | 4-SO2CH3 | 1,200 (5.92±0.27) | ND |
| 4p | 3,5-diCF3 | 52.3 (7.28±0.04) | 5620 (5.25±0.08) |
| 4q | 2-F | 526 (6.28±0.04) | ND |
| 4r | 3-F | 675 (6.17±0.14) | ND |
| 4s | 4-F | 406 (6.39±0.03) | ND |
| 4t | 3,4-diCl | 28.5 (7.55±0.06) | 1460 (5.84±0.39) |
| 4u | 4-NO2 | 207 (6.68±0.04) | ND |
| 4v | 4-OCHF2 | 83.4 (7.08±0.05) | 2540 (5.60±0.21) |
| 4w | 3-CF3,4-F | 68.8 (7.16±0.05) | 1090 (5.96±0.15) |
| 4x | 2-NO2 | 4840 (5.32±0.24) | ND |
| 4y | 3-NO2 | >10,000 (>5) | ND |
Concentration response curves (CRCs) for 4c, 4n and 4k against AP and γ-thrombin mediated activation are shown in Figure 2A-F, and all three proved to be competitive inhibitors of PAR4 (Figure 2G-I). This is in sharp contrast to 1, which is a non-competitive inhibitor of PAR49, and 2 which has a mixed competitive/non-competitive profile.4 These results are exciting as until now small molecule tools with diverse modes of pharmacolgy and PAR4 antagonism to potentially assess in vivo effects of PAR4 inhibition have not existed.
Figure 2.
Molecular pharmacology profile of 4c, 4n and 4k. A-C) PAR4 antagonist CRCs (n=3) against 200 μM PAR4-AP showing equivalent inhibition of both PAC-1 and P-selectin. n=3, mean±SEM; D-F) PAR4 antagonist CRCs (n=3) against 100 nM γ-thrombin showing equivalent inhibition of both PAC-1 and P-selectin. n=3, mean±SEM; G-I) Progressive fold-shift experiments showing a parallel right-ward shift of the CRC (competitive mode of PAR4 inhibition) with 4c, 4n and 4k. Schild EC50 log DR-1 versus log [antagonist] plot slopes: 4c (0.95±0.09), 4k (1.09±1.3), 4n (0.94±0.08). 4,9
While PAR4 activity against AP and γ-thrombin, coupled with a novel competitive mode of inhibition generated enthusiasm for these low molecular weight PAR4 antagonists 4c, 4n and 4k, their in vitro DMPK profiles were suboptimal, and reminiscent of 1.4,8,9 While cLogPs were acceptable (3.2 to 3.7), experimental LogPs were >4 and PSAs were <40. These physiochemical properties correlated with high plasma protein binding (human and rat fus between 0.006 to 0.014) and moderate to high intrinsic clearance in hepatic microsomes (human CLheps 13.6 to 19.9 mL/min/kg and rat CLheps 46.3 to 58.3 mL/min/kg).
These data led to second and third generation libraries aimed at replacing the 6-aryl moiety with heterocycles and surveying alternative ethers and amine substituents for the 2-methoxy group in 4c, 4n and 4k. Overall, heterocyclic replacements for the 6-aryl moiety (e.g., 2-, 3- and 4-pyridyl, thienylthiazolyl) lost 10-to 50-fold activity relative to the unsubstituted phenyl comparator 4f. Alternatives for the 2-methoxy group were equally steep. 2-Ethoxy congeners displayed activity (IC50s 600-900 nM), but lost ~20-fold activity relative to 4c, 4n and 4k. Larger, branched and cyclic ethers lost all PAR4 inhibitory activity. In an attempt to improve the physiochemical properties of this series, we performed SNAr reactions on cores 7 with various primary and secondary amines. SAR was steep, with the vast majority of 2-amino congeners studied possessing no PAR4 inhibitory activity. As shown in Figure 3, only a single N(CH3)2 analog 8 was active (PAR4-AP IC50 = 3.45 μM), whereas the NHCH3 derivative was inactive. None of the non-2-OCH3 derivatives displayed any PAR4 activity against γ-thrombin mediated activation. Clearly, the 2-methoxy moiety is an essential element of the PAR4 pharmacophore, at least with the context of the imidazo[2,1-b][1,3,4]thiadiazole bicyclic scaffold.
Figure 3.
Structures and PAR4 activity of 2-amino congeners 8 and 9.
We envisioned that deletion of the 2-hydroxy methyl moiety and transposition of the imidazo[2,1-b][1,3,4]thiadiazole bicycle from the 3- to the 2-position, while simultaneously truncating the N-benzyl moiety to a simple N-Me, would afford a small molecule that aligns with fragment 3 (Figure 4A). Indeed, this proved successful, generating indole 5, a 20 nM PAR4 inhibitor against AP (~9-fold more potent than 1) with a 25% reduction in molecular weight. While activity against γ-thrombin with 5 was more potent and efficacious (IC50 = 1.0 μM) than that of 1, an ~50-fold difference in potency was noted between inhibition of PAR4-AP and γ-thrombin mediated stimulation (Figure 4B), yet 5 retained selectivity versus PAR1 (IC50 > 10 μM) and off-target effects on the collagen receptor were eliminated (Figure 5). Although 5 maintains γ-thrombin stimulated antagonism within four-fold of 4n, fragments 4 represented the best path forward towards improved PAR4 inhibitors based on physiochemical properties and γ-thrombin ligand efficiency metrics (5 LE = 0.33 vs. 4n LE = 0.43) with competitive inhibition.
Figure 4.
Identification and pharmacological profile of a minimum pharamacophore of 1. A) Strategy to mimick the minimum PAR4 pharmacophore 3 of 2 within the indole series, and identification of 5. B) Pharmacolgoical profile of 5 against PAR4 γ-thrombin (100 nM) PAC-1 (IC50 = 1.0 μM, pIC50 = 6.01±0.05), PAR4 γ-thrombin P-selectin (IC50 = 1.0 μM, pIC50 = 6.00±0.04), PAR4-AP PAC-1(IC50 = 21 nM, pIC50 = 7.68±0.04), PAR4-AP (200 μM) P-selectin (IC50 = 20 nM, pIC50 = 7.70±0.03) and PAC-AP (PAC1 and P-selectin, IC50s >10 μM). n=3, mean±SEM.
Figure 5. Identification of a minimum pharmacophore of 1 devoid of off-target effects.
Washed human platelets were incubated with the indicated concentration of antagonist (1,5) for 20 min prior to activation with either A) 20 μM PAR1-AP or B) 10 μg/ml collagen I (Coll I). Platelets were allowed to aggregate for 10 min. Shown are representative tracings of three independent experiments.
We previously reported on the potency and selectivity of example 1 noting >100 fold difference in IC50 values comparing PAR1-AP and PAR4-AP.4 However, when used at micromolar concentrations that are effective against the tethered ligand (γ-thrombin mediated activation) significant off-targets against collagen I induced aggregation were noted (Figure 5). This was surprising considering the weak effects on the primary anti-target PAR1. This was also an alarming revelation since collagen I-mediated platelet activation is intimately involved in hemostasis and thrombosis. Therefore, before focusing solely on the simplified pharmacophores 4, we elected to re-evaluate 14 in an effort to further minimize the key components required for PAR4 inhibition and address off-target effects. Fragment 3 is devoid of collagen receptor off-target effects as are fragment inhibitors 4, suggesting the hydroxymethyl moiety of 1 might be suspect in this series.
The PAR4 IC50s from the flow cytometry assays were encouraging, but would these novel, competitive PAR4 antagonists (4c, 4k, and 4n) inhibit human platelet aggregation? As shown in Figure 6, a standard platelet aggregation assay11 demonstrated excellent correlation and consistent data with the three competitive PAR4 inhibitors. Here, platelet aggregation was significantly inhibited at doses (316 nM to 1 μM against AP and 1 μM to 3.16 μM against γ-thrombin) of the PAR4 antagonists that abolished PAC1 binding. This was a pivotal finding, as this is the first demonstration that competitive PAR4 inhibitors are efficacious in this pharmacodynamics assay.
Figure 6.
Activity of fragments 4 in platelet aggregation. Washed human platelets were incubated with the indicated concentration of antagonist for 20 min prior to activation with either 40 nM γ-thrombin A (4c), C (4k), E (4n) or 200 μM PAR4-AP B (4c), D (4k), F (4n). Platelet aggregation was monitored for 10 min at which point final reported % aggregation values were taken. Shown are the means±SEM for at least three independent experiments. Values were compared to controls and significance determined by paired t-test *p value<0 05 **p value<0 005 ***p value<0 0005
Furthermore, we wanted to assess the ability to inhibit platelet aggregation with 5 (Figure 7), as there was a large disconnect between γ-thrombin and AP. Here, platelet aggregation was significantly inhibited at doses (316 nM to 1 μM against AP and only at 10 μM against γ-thrombin) of the PAR4 antagonist 5 that abolished PAC1 binding. Thus, excellent correlation once again between the FACS and aggregation assays, but poor correlation between activity against PAR4-AP and γ-thrombin induced PAR4 activation. γ-thrombin cleavage of PAR4 generates an intramolecular tethered ligand which is covalently attached to the receptor, while PAR4-AP is a soluble surrogate for the tethered ligand. It is impossible to measure the concentration of the tethered ligand proximal to the binding pocket because it is covalently attached to the receptor. Theoretically, an infinite concentration of ligand is available due to its inability to diffuse away from the receptor. Thus, one would predict that it would take higher concentrations of an orthosteric antagonist to compete with the tethered ligand as compared to the soluble AP. Indeed, we have extensively documented this phenomenon. If this was the sole distinguishing factor between γ-thrombin and PAR4-AP, IC50s should correlate. It should be possible to predict γ-thrombin IC50s based on PAR4-AP IC50s. However, this is not the case. Proportional differences between γ-thrombin and PAR4-AP IC50s are highly variable, which is why we chose to track SAR with both ligands. Based on the data presented here we can't explain this lack of correlation, however it is important to keep in mind that one is an enzyme (γ-thrombin) and the other is a peptide (PAR4-AP). Thrombin interacts with PAR412,13 and its heterodimer partner PAR114 proximal to but outside the binding pocket. γ-thrombin binding to PAR4 outside the binding picket, or the reorganization of the N-terminus after cleavage could induce conformational states in the receptor that the AP is incapable of inducing since it interacts only with binding pocket. This could create two distinct receptor states and therefore two distinct lines of SAR for antagonist. In support of this hypothesis our lab has previously published on functional selectivity between PAR1-AP and thrombin induced endothelial cell signaling.15
Figure 7.
Activity of 5 in platelet aggregation. Washed human platelets were incubated with the indicated concentration of antagonist 5 for 20 min prior to activation with either A) 40 nM γ-thrombin or B) 200 μM PAR4-AP Platelet aggregation was monitored for 10 min at which point final reported % aggregation values were taken. Shown are the means±SEM for at least three independent experiments. Values were compared to controls and significance determined by paired t-test. ***p-value<0.0005.
In summary, we identified a greatly simplified 2-methoxy-6-arylimidazo[2,1-b][1,3,4]thiadiazole scaffold that affords nanomolar PAR4 inhibition of AP as well as γ-thrombin, while reducing both molecular weight and the number of hydrogen bond donors/acceptors by ~50%. This simplified core represents the minimum pharmacophore for PAR4 inhibition in a competitive manner. While initial analogs did not address the DMPK liabilities of 1 and 2 (though fu was slightly improved), the most basic core, 4h, can serve as a sub-micromolar lead from which to build in more optimal DMPK and physiochemical properties. Similarly, 5 was identified as a minimum PAR4 pharmacophore of indole series 1, but with an unexpected 50-fold loss in activity at γ-thrombin, yet nanomolar potency against the AP. Finally, we disclose the first example demonstrating that competitive PAR4 inhibitors significantly inhibit human platelet aggregation. Further optimization efforts around these fragments are in progress, as well as the evaluation of alternate phenyl bioisosteres beyond the imidazo[2,1-b][1,3,4]thiadiazole bicyclic scaffold, and these results will be reported in due course.
Acknowledgments
We thank the NIH for funding (NS081669 and NS082198.) We also thank William K. Warren, Jr. and the William K. Warren Foundation who funded the William K. Warren, Jr. Chair in Medicine (to C.W.L.).
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- 1.French A, Hamilton JR. Br. J. Pharmacol. in pess. [Google Scholar]
- 2.Gieseler F, Ungefroren H, Settmacher U, Hollenberg MD, Kaufmann R. Cell. Commun. Signal. 2013;11:86. doi: 10.1186/1478-811X-11-86. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Duvernay M, Young S, Gailani D, Schoenecker J, Hamm HE. Mol. Pharm. 2013;83:781–792. doi: 10.1124/mol.112.083477. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Temple KJ, Duvernay MT, Young SE, Wen W, Wu W, Maeng JG, Blobaum AL, Stauffer SR, Hamm HE, Lindsley CW. J. Med. Chem. 2016;59:7690–7695. doi: 10.1021/acs.jmedchem.6b00928. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Lee F-Y, Lien J-C, Huang J-L, Tsai S-C, Teng C-M, Wu C-C, Cheng F-C, Kuo S-CJ. Med. Chem. 2001;44:3746–3749. doi: 10.1021/jm010001h. [DOI] [PubMed] [Google Scholar]
- 6.Young SE, Duvernay MT, Lindsley CW, Hamm HE. PLOS One. 2013;8:e65528. doi: 10.1371/journal.pone.0065528. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Wen W, Young SE, Duvernay MT, Schulte ML, Nance KD, Melancon BJ, Engers J, Locuson CW, Wood MR, Daniels JS, Wu W, Lindsley CW, Hamm HE, Stauffer SR. Bioorg. Med. Chem. Lett. 2014;24:4708–4713. doi: 10.1016/j.bmcl.2014.08.021. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.a Halford B. C&E News. 2016;94:22–23. [Google Scholar]; b Banville J, Remillard R, Ruediger EH, Deon DH, Gagnon M, Dube L, Guy J, Priestly ES, Posy S, Maxwell BD, Wong PC. 2013 WO2013163279A1.; c Priestly ES, Posy SL, Tremblay FT, Martel A, Marinier A. 2013 WO2013163241A1.; d Lawrence MR, Miller MM, Seiffert DA, Posy SL, Wong PC, Banville J, Ruediger EH, Deon DH, Martel A, Tremblay F, Guy J, Lavallee J-F, Gagnon M. 2013 WO2013163244A1.
- 9.Duvernay MT, Temple KJ, Young SE, Wen W, Wu W, Maeng JG, Blobaum AL, Stauffer SR, Lindsley CW, Hamm HE. Mol. Pharmacol. in press. doi: 10.1124/mol.116.106666. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Example experimental (4h). A mixture of commercially available 2-bromo-1-(p-tolyl)ethan-1-one (2.35 mmol) and 5-bromo-1,3,4-thiadiazol-2-amine (3.52 mmol) were dissolved in CH3CN / IPA (1:1; 9.4 mL) in a microwave vial that was sealed and heated to 80 °C for 18 hr. Next, the vial was placed in a Microwave for 30 minutes at 150 °C. The solvent was evaporated and the mixture was re-suspended in DCM (20 mL), washed with saturated NaHCO3 (20 mL), brine, dried over magnesium sulfate and filtered. The organic layer was concentrated under reduced pressure onto celite and loaded onto a 3 inch silica plug and purified using Teledyne ISCO Combi-Flash system (solid loading, 100% DCM, 10 min run) to afford 2-bromo-6-(ptolyl)imidazo[2,1-b][1,3,4]thiadiazole (1.70 mmol) in 72% yield. LC-MS [M+H] = 294/296, RT = 1.174; 1H NMR (400 MHz, CDCl3) δ 7.98 (s, 1H), 7.68 – 7.70 (d, J = 8.16 Hz, 2H), 7.21 – 7.23 (d, J = 8.03 Hz, 2H), 2.38 (s, 3H). A solution of 2-bromo-6- (p-tolyl)imidazo[2,1-b][1,3,4]thiadiazole (1.02 mmol) in a mixture of DCM/MeOH (3:1; 20 mL) was treated at 22°C with a 25 wt.% solution of NaOMe (2.04 mmol) in MeOH and the reaction was stirred for 1 hr at room temperature. Upon completion as determined by LCMS, the reaction mixture was quenched by addition of 1% HCl (10 mL) followed by addition of saturated NaHCO3 (10 mL). The aqueous layer was extracted with DCM (3 × 15 mL), dried over magnesium sulfate and filtered. The organic layer was concentrated under reduced pressure onto silica gel and purified using Teledyne ISCO Combi-Flash system (solid loading, 12G, 100% DCM, 15 min run) to afford 2-bromo-6-(ptolyl)imidazo[2,1-b][1,3,4]thiadiazole, 4h (0.737 mmol) in 72% yield. LC-MS [M+H] = 246, RT = 1.113; 1H NMR (400 MHz, CDCl3) δ 7.78 (s, 1H), 7.66 – 7.68 (dt, J = 8.21 Hz, 2H), 7.19 – 7.21 (d, J = 8.03 Hz, 2H), 4.18 (s, 3H), 2.37 (s, 3H).
- 11.Platelet aggregation Platelets were diluted to 2.0×108/mL and aliquoted into glass cuvettes. PAR4 antagonists were allowed to equilibrate with platelets for 20 min prior to stimulation with the indicated agonists. Aggregations were recorded for 10 min on a model 700 Optical Lumi Aggregometer (Chrono-log, Havertown, PA).
- 12.Nieman MT, Schmaier AH. Biochemistry. 2007;46:8603–8610. doi: 10.1021/bi700597p. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Nieman MT. Biochemistry. 2008;47:13279–13286. doi: 10.1021/bi801334s. [DOI] [PubMed] [Google Scholar]
- 14.Arachiche A, Mumaw MM, de la Fuente M, Nieman MT. J. Biol. Chem. 2013;288:32553–32562. doi: 10.1074/jbc.M113.472373. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.McLaughlin JN, Shen L, Holinstat M, Brooks JD, Dibenedetto E, Hamm HE. J. Biol. Chem. 2005;280:25048–25059. doi: 10.1074/jbc.M414090200. [DOI] [PubMed] [Google Scholar]