Expression and purification of Protease Activated Receptor 4 (PAR4) and analysis with histidine hydrogen deuterium exchange

Protease activated receptors (PARs) are G-protein coupled receptors (GPCRs) that are activated by proteolyis of the N-terminus, which exposes a tethered ligand that interacts with the receptor. Numerous studies have focused on the signaling pathways mediated by PARs. However, the structural basis for initiation of these pathways is unknown. Here, we describe a strategy for the expression and purification of PAR4. This is the first PAR family member to be isolated without stabilizing modifications for biophysical studies. We monitored PAR4 activation with histidine-hydrogen deuterium exchange (His-HDX). PAR4 has 9 histidines that are spaced throughout the protein allowing a global view of solvent accessible and non-accessible regions. Peptides containing each of the 9 His residues were used to determine the t1/2 for each His residue in apo or thrombin activated PAR4. The thrombin cleaved PAR4 had a 2-fold increase (p > 0.01) in t1/2 values observed for four histidine residues (His180, His229, His240, and His380) demonstrating that these regions have decreased solvent accessibility upon thrombin treatment. In agreement, thrombin cleaved PAR4 also was resistant to thermolysin digestion. In contrast, activation with the PAR4 agonist peptide was digested at the same rate as apo PAR4. Further analysis showed the C-terminus is protected in thrombin activated PAR4 compared to uncleaved or agonist peptide treated PAR4. The studies described here are the first to examine the tethered ligand activation mechanism for a PAR family member using biophysically and shed light on the overall conformational changes that follow activation of PARs by a protease.


INTRODUCTION
Protease activated receptors (PARs) are the primary means by which proteases directly initiate intracellular signaling.1 There are 4 PARs (PAR1-4) that comprise a specialized family of G-protein coupled receptors (GPCRs), which have a unique activation mechanism whereby cleavage of the Nterminus generates a tethered ligand. PAR1 was originally identified as the thrombin receptor on platelets, however it is now recognized that PARs are expressed in many cell types and are activated by a growing panel of proteases.2,3 Once activated, PARs signal through Gαq, Gα12/13, Gαi, and arrestin depending on the activating protease and cellular context. 3 It is reasonable to infer that these divergent signaling events are due to allosteric modulation of PARs by cofactors or specific tethered ligands.
However, this has not been tested formally.
PAR4 was identified seven years after PAR1 and, due to its kinetics of activation, was initially considered a low affinity backup thrombin receptor on human platelets. 4 The initial characterization of PAR4 revealed that the N-terminus of PAR4 does not have a hirudin-like sequence, which dramatically lowers the efficiency of PAR4 cleavage by thrombin. [5][6][7][8][9] The net result is a ~10-fold higher threshold for thrombin activation of PAR4 compared to PAR1. Once activated, PAR4 mediates prolonged signaling in platelets, which is required for sustained platelet activation and stable thrombus formation. 10 In contrast PAR1 elicits a rapid, short-lived signal.11,12 Combining these observations, PAR4 has emerged as a therapeutic target for antiplatelet agents that prevent sustained activation by high levels of thrombin that are generated at the site of injury, while sparing PAR1 signaling. 13,14 Initial clinical studies suggest that this therapeutic approach allows the formation of the initial hemostatic plug while preventing unwanted thrombosis.14,15 The activation of PARs via the tethered ligand mechanism was described in 1991 for PAR1.2 To date, there is still limited structural information regarding this unique mode of activation. High resolution crystal structures of PAR1 and PAR2 bound to antagonists have been solved. 16,17 These structural studies reveal the molecular details of the receptor/antagonist interaction, but provide limited information on the tethered ligand mechanism due to modifications required to facilitate crystallization. 18 Specifically, the N-and Ctermini were removed to limit overall flexibility, and T4 lysozyme (PAR1) or BRIL (PAR2) was inserted into intracellular loop 3 (ICL3) to further stabilize the protein. 16 For PAR2, additional stabilizing mutations were introduced. 17 To date, there is no structural information for PAR4. A major roadblock in understanding the structural basis for the tethered ligand-mediated activation of PARs is the lack of structural studies with the native protein to allow the identification of the ligand binding site and the structural dynamics specific to this activation mechanism. Advances in mass spectrometry now allow protein dynamics to be monitored in response to activation or ligand binding. Histidine Hydrogen-Deuterium Exchange (His-HDX) takes advantage of the slow exchange rate of the imidazole C2hydrogen on the histidine residue.19-21 This is in contrast to amide-HDX, which measures the exchange rate of backbone hydrogens. The histidine residues are effectively site-specific probes that monitor changes in the molecular environment at these sites as the protein changes conformation upon activation or ligand binding. PAR4 has nine histidine residues that are spaced throughout the protein, which allows an overall evaluation of the conformational changes using His-HDX. In addition, the slow back exchange allows analysis of proteins in complex mixtures. 19,21 Here, we used histidine hydrogen deuterium exchange (His-HDX) to examine the conformational changes that occur within PAR4 upon activation by thrombin or the PAR4 activation peptide AYPGKF.
Taken together, our results demonstrate that cleavage of PAR4 by thrombin leads to a dramatic rearrangement, which results in a more stable structure with increased resistance to experimental proteolysis. In contrast, the structural stability of PAR4 was unchanged following stimulation by AYPGKF compared to the non-activated receptor. These studies are first to examine the structural rearrangement of PAR4 upon activation and may provide a molecular basis for the differential signaling that is observed from PAR activation peptides.22

Cloning and Cell Culture
The cDNA for BRIL-PAR4 was codon optimized for expression in Sf9 cells and synthesized by GenScript.
The cDNA for Gαq (catalog # GNA0Q00000) was purchased from cDNA Resource Center (Bloomsburg, PA). The cDNA for Gβ1 (catalog # 67016) and Gγ2 (catalog # 67018) were purchased from Addgene (Cambridge, MA). All constructs were cloned into pFastBac and expressed in Sf9 cells using the Baculovirus Expression Vector System (Invitrogen, Carlsbad, CA) grown in ESF 921 media from Expression Systems (Davis, California). The recombinant bacmids were transfected into Sf9 cells using Cellfectin II Reagent for production of baculovirus for BRIL-PAR4, Gαq, Gβ1, or Gγ2 that was subsequently used for protein expression in Sf9 cells using manufacturer's protocol (Invitrogen).
To map the C-terminal epitope of the C-20 antibody, we transiently expressed wild type PAR4 and two C-terminal truncation mutants at amino acid 358 (RAGLFQR*) or 373 (KASAEGG*) in HEK293 cells.24 Transfections were done using Lipofectamine 2000 from Thermo Fisher Scientific (catalog # 11668019) following manufacturer's protocol. Cells were harvested at 48 hr post transfection and lysed in RIPA buffer.

Protein Purification
Membranes of Sf9 cells expressing BRIL-PAR4 were prepared by resuspending cell pellets in lysis buffer (25 mM Tris pH 8, 300 mM sucrose, and 5 mM EDTA plus protease inhibitors (cOmplete tablets-EDTA free by Roche-Basel, Switzerland). Cell suspensions were lysed in a microfluidizer followed by centrifugation at 3000g for 10 min and 14,000g for 45 min to remove cell debris. Finally, membranes were pelleted at 100,000g for 60 min. All centrifugations were done at 4 C. Once the membranes were obtained, they were solubilized at 4 C for 2 hours in 0.5% DDM, 20 mM HEPES pH 7.5, 50 mM NaCl, 0.1% Cholesteryl hemisuccinate (Sigma, St. Louis, MO), 20% glycerol and 1 mM DTT, plus proteinase inhibitors. The BRIL-PAR4 for Histidine Hydrogen-Deuterium exchange was purified by taking advantage of the C-terminal 1D4 epitope (Figure 1) for affinity chromatography with 1D4-coupled sepharose beads.

Histidine Hydrogen Deuterium Exchange (His-HDX) coupled with Mass Sepctrometry
Purified recombinant PAR4 was analyzed with His-HDX as previously described.19,21 Briefly, apo or thrombin cleaved PAR4 were incubated in 0.1% LMNG, 20mM HEPES pH7.5 made with D2O containing 50mM NaCl, 0.02% Cholesteryl hemisuccinate, 5% glycerol and 1 mM DTT at 37 C for 72 hr. The samples were run on SDS-PAGE followed by staining with Coomassie blue. The electrophoresis and the subsequent staining were carried out at 4 C to minimize the deuterium to protium back exchange. The bands corresponding to PAR4 were excised, cut into small pieces, and divided into 2 tubes. The samples were processed for in gel digestion with trypsin or chymotrypsin for 1 hr at RT. Following digestion, the trypsin and chymotrypsin digests were pooled, dried in a Speed Vac concentrator, redissolved in 0.1% formic acid, and then analyzed by LC/MS/MS using LTQ-Orbitrap XL mass spectrometer as described previously. 20 The peptides were resolved and the half-life (t½) for H/D exchange was calculated as previously described. [19][20][21] The data are from three independent experiments.

Limited Proteolysis with thermolysin
Purified recombinant apo-PAR4 or thrombin cleaved PAR4 samples were treated with thermolysin (0.01 µg/reaction) for 0, 2, 5, 10, 20, 60 or 90 minutes, reactions were stopped with the addition of 50 mM EDTA and Laemmeli sample buffer containing 2-mercaptoethanol (5%). Samples were run on 12% SDS-PAGE and stained with QC Colloidal Coomassie blue (BioRad, catalog # 161-0803)). The rate of thermolysin digestion of PAR4 was quantified using the Li-Cor Odyssey imaging system (Li-Cor, Lincoln, NE). To follow the stability of the N-and C-termini, thermolysin digestion was followed by Western blotting using the antibody 2D6 to monitor the N-terminus and the antibody C-20 to monitor the C-terminus.

G protein activation assay
The functionality of PAR4 was verified with a G protein activation assay that monitors the increase of intrinsic tryptophan fluorescence of the G subunit as the trimeric complex dissociates to Gαq and Gβγ.25 The fluorescence increase was monitored at 300 nm excitation and 335 nm emission at 20 C. The reaction contained PAR4 (50 nM), Gαqβγ (500 nM), and GTPγS (Catalogue # 10220647001) Roche, (Basal, Switzerland) (300 µM) and was initiated with the addition of thrombin (100 nM). Fluorescent data was acquired for 500 sec. Control experiments were done as above without Gαqβγ complex or without thrombin.

Expression and purification of human PAR4.
The goal of this study was to examine the structural rearrangements of PAR4 following activation by thrombin using histidine hydrogen deuterium exchange (His-HDX). Since we were interested in the conformational changes of PAR4 as a result of activation, the recombinant PAR4 was free of stabilizing modifications that are typical for structural studies with GPCRs (Figure 1).18,26,27 Human PAR4 was expressed as a apocytochrome b562 (BRIL) fusion protein with epitope tags on the N-and C-termini ( Figure 1A and B). The addition of BRIL to the N-terminus of PAR4 increased expression >20-fold over our original constructs that did not contain it (data not shown). The signal sequence from the influenza hemagglutinin is removed during trafficking and processing the recombinant PAR4 in the Sf9 cells leaving an N-terminal FLAG tag that was used for affinity purification. 28 The 1D4 antibody epitope (TETSQVAPA) was added to the C-terminus to facilitate a second affinity purification strategy. Following purification, FLAG and BRIL were removed from PAR4 with tobacco etch virus (TEV) protease. The TEV recognition sequence was engineered such that the Gly residue at the P1 site was incorporated into the PAR4 sequence as Gly18; the first amino acid of the mature PAR4 protein (Figure 1A). 4 The final purified recombinant human PAR4 migrated as a single band at the expected 37 kDa on reduced SDS-PAGE as indicated by Coomassie Blue staining ( Figure 1C). Thrombin removed 2.9 kDa (Gly18-Arg47) and the resulting activated PAR4 migrated at the expected 34 kDa (Figure 1C). Native SDS-PAGE confirmed that the purified PAR4 did not form aggregates following purification ( Figure 1C). The functionality of the purified PAR4 was confirmed with an in vitro G-protein activation assay that monitors the increase in intrinsic protein fluorescence of the Gα subunit.25 Stimulation of recombinant PAR4 with thrombin (100 nM) resulted in a rapid increase in fluorescence indicating activation of the G-protein complex ( Figure   1D). Importantly, there was no increase in fluorescence in the absence of the Gαβγ complex or absence of thrombin ( Figure 1D).

Histidine Hydrogen-Deuterium exchange (His-HDX)
His-HDX monitors the rates of hydrogen-deuterium exchange at the C2 position on the imidazole ring of histidine residues (Supplemental Figure 1A). [19][20][21] The exchange is slow at this position, allowing for long time scale measurements using histidine residues as site-specific probes. PAR4 has 9 histidine residues that are well spaced throughout the protein (Figure 2A), which provides a global view of solvent accessible and non-accessible regions. Following incubation in D2O for 72 h, unique peptides that contained each of the 9 histidine residues were generated by trypsin and chymotrypsin digestion (Supplemental Table 1). These peptides were used to determine the exchange rate (t½) for each histidine residue in non-activated or thrombin activated PAR4 (Table 1 and Figure 2B). The histidine exchange rates for PAR4 can be categorized into three groups; 1) a long half-life that did not change upon activation, 2) a moderate half-life that did not change upon activation, and 3) a moderate half-life that was prolonged upon PAR4 activation. The range of responses allows both a global view of PAR4 activation as well as site specific conformational changes.
The ligand binding site of PAR4 has not been determined experimentally. However, based on comparisons with PAR1, ECL2 is predicted to play a role.16,29 Four histidines (His136, His229, His240, and His306) are located in extracellular loops (ECL) of PAR4. His136 is located in ECL1 (shown in light green, Figure 2C) had a half-life of 7.52  1.07 d in the non-activated state and did not change upon activation by thrombin (8.88  3.19 d, p = 0.52) ( Table 1 and Figure 2B). Similarly, His306, which is located in ECL3 (light green Figure 2C), also did not show a change in t½ upon activation, 8.05  2.00 d versus 10.59  3.41 d, p = 0.32. In contrast, His229 and His240, both located in ECL2, had a longer t½ upon activation.
His229 has ~2-fold shift from 4.87  0.02 d to 9.25  1.64 d (orange to light green, Figure 2C). His240 is also significantly prolonged from 4.13  0.29 to 6.28  0.43 (orange to yellow, Figure 2C). These data suggest that the N-terminal tethered ligand of PAR4 is interacting with ECL2 upon activation to mediate signaling.
The t½ values for His159 and His269 (shown in purple, Figure 2C) were >40 d in both states (non-activated and thrombin activated) indicating that these were not exposed to solvent and did not change in response to thrombin. The long half-life of His159 is expected as it is buried within transmembrane helix 3 (TM3) (Figure 2A and C). His269 is located at the base of TM5. Homology models with other GPCRs indicate that His269 is surrounded by other TM helices (Figure 2A and C), which likely contributes to its prolonged half-life in both states.
The cytoplasmic surface of GPCRs mediate signaling through interactions with G-proteins and arrestins.
Two histidines, His180 and His280, are located in the intracellular loops (ICL) and His380 is located at the Cterminal tail. The t½ values for His180 increased from 8.42  0.61 to 12.88  2.37 d, p = 0.01 ( Figure 2B) and indicated from light green to green in Figure 2C. In contrast, His280 did not change upon activation ( Figure 2C, shown in green). His380 is located 6 amino acids from the C-terminus. The t½ nearly doubled from 3.11 0.29 to 5.54  0.55 d, p = 0.003 upon activation suggesting that this region of PAR4 dramatically changes following activation (Figure 2B and C).
In summary, the His-HDX experiments identified regions important for both receptor activation on the extracellular loops and signal transduction in the intracellular face. The increase in half-life for His229 and His240 demonstrates that ECL2 changes its molecular environment upon activation. A role for ECL2 in PAR4 activation is consistent with other PARs based on sequence homology. Finally, the His residues that have a longer half-life (t½) of H/D exchange following cleavage by thrombin are located throughout PAR4 indicating that PAR4 undergoes a global rearrangement and becomes more compact and potentially more rigid upon activation.

Thrombin activated PAR4 is resistant to proteolysis.
Resistance to proteolysis has long been used as a measure of structural stability.30,31 To validate the results from our studies using His-HDX, we compared the overall structural changes in PAR4 following activation with limited proteolysis by the low specificity metalloproteinase thermolysin. Initial experiments determined that 20 nM thermolysin was required to digest PAR4 to completion in 60 min (data not shown).
We next analyzed the stability of PAR4 or thrombin activated PAR4 by monitoring the thermolysin reaction over time (1-90 min) via SDS-PAGE followed by Coomassie blue staining. A representative gel from three independent experiments is shown in Figure 3A. The gels where quantitated by measuring the disappearance of the two bands at ~37 kDa using a Li-Cor Odyssey imaging system. Non-activated (apo) PAR4 was digested to completion in 60 min with a half-life (t½) of 11.9  1.0 min and first order rate constant of 0.060  0.005 min-1. The error is reported as the standard deviation from three independent experiments. Thrombin activated PAR4 did not show appreciable proteolysis at 90 min. Taken together, these data support the results from the His-HDX studies that PAR4 globally becomes structurally more compact upon cleavage by thrombin.

Mapping the proteolysis.
The His-HDX and initial limited proteolysis experiments show that PAR4 becomes more compact and protected from proteolysis upon activation. We next determined which region of PAR4 was becoming protected by monitoring the thermolysin digest by Western blotting with antibodies directed to the N-or C-terminus of PAR4. The mouse monoclonal antibody 2D6 binds to the N-terminus at (G74WVPTR78), which is near the start of transmembrane helix 1 (Supplemental Figure 3). 23 The goat polyclonal antibody (C20) maps to the last 12 amino acids of PAR4 (Supplemental Figure 3). These epitopes were monitored simultaneously using secondary antibodies labeled with distinct infrared probes and the Li-Cor Odyssey System. The 2D6 epitope was lost at 90 min in apo-PAR4 (Figure 4A, green). The C-terminus was also progressively cleaved to generate a profile of proteolytic products that reached a maximum at 120 min (Figure 4A, red). The 2D6 epitope was lost at 60 min for thrombin cleaved PAR4 indicating that the tethered ligand may lead to an exposed region in the N-terminus that is accessible to thermolysin. In contrast, the C-terminus was resistant to proteolysis indicating that it is less accessible to thermolysin.
These data are in agreement with the His-HDX data showing that His380 has a longer t1/2 showing that it less accessible H/D exchange and protected from proteolysis.

PAR4 activation by AYPGKF
PARs can be activated independent of cleavage using soluble peptides that mimic the tethered ligand sequence.32,33 However, the specific response from the agonist peptides differs from that of the endogenous tethered ligand.32,34 We next determined the sensitivity of PAR4 to thermolysis digestion following stimulation with 1 mM of the PAR4 agonist peptide (AYPGKF), which is ~5-fold higher than the EC50. The t½ of PAR4 stimulated with AYPGKF was not different from that of apo-PAR4 (Figure 5A) indicating that the agonist peptide does not induce the same conformational changes in PAR4 as thrombin. Finally, monitoring the N-and C-termini revealed that these were also unchanged compared to apo-PAR4 (Figure 5B and Figure 4A).

DISCUSSION
In this report we provide two advances toward the understanding of protease activated receptor (PAR) activation mechanism. First, we describe an expression and purification strategy for full-length PAR4 without mutations or insertions to stabilize the protein. This is the first PAR to be expressed and purified with its native primary sequence that is fully activatable by an endogenous protease agonist. The second major advance builds on our purification scheme by using histidine hydrogen deuterium (His-HDX) coupled with mass spectrometry to examine the conformational dynamics of PAR4 following activation by thrombin. We compared the H/D exchange rates for each of the 9 histidines that are distributed throughout PAR4 and determined that the protein undergoes a global rearrangement following cleavage of the N-terminus by thrombin. The His-HDX data were validated by comparing the rates of proteolytic degradation by thermolysin. Cleavage of the N-terminus led to increased stability and resistance to proteolysis compared to inactive (apo) PAR4. In contrast, stimulation with the PAR4 activation peptide AYPGKF did not increase the stability of PAR4 over the apo state. Taken together, our experiments show that PAR4 undergoes a global rearrangement and becomes more stable following activation by thrombin.
Finally, these studies are the first to examine PAR4 structurally and, more generally, to examine the structural dynamics of PAR activation by the tethered ligand in response to proteolytic activation.
PARs are important mediators that link extracellular proteases to intracellular signaling.1,35 PAR4 is the fourth member in the family that was identified and cloned in 1998.4 PAR4 has been primarily studied in the context of platelet signaling, which has led to efforts targeting PAR4 therapeutically.14,15,36 Further, polymorphisms that lead to sequence variants impact platelet reactivity in response to PAR4 agonists. 37,38 Beyond platelets, PAR4 is also expressed on many cell types and human tissues where it has critical roles in vascular biology, inflammation, and cancer.39-41 Despite these numerous physiological studies, the molecular basis of PAR4 activation upon enzymatic cleavage is largely unknown. Unlike most of other receptors whose ligands are free and diffusible, the endogenous ligand of PAR4 is tethered to the protein.
Therefore, PAR4 activation is likely synchronized with a global structural rearrangement of the receptor, which recruits the corresponding G-proteins that elicit downstream signaling, however this has not been directly examined experimentally.
The central question we set out to address is the structural dynamics of PAR activation in response to cleavage of the N-terminus using PAR4 activation by thrombin as a representative. Structural studies with GPCRs have largely focused on X-ray crystallography. 42 The high resolution structures are valuable in providing detailed molecular contacts within the receptor and with specific ligands. 16 These modifications are at odds with determining the conformational changes of PARs following activation by a protease. To overcome this obstacle, we produced full length human PAR4 free of stabilizing sequences as a BRIL fusion protein ( Figure 1A and B) in Sf9 insect cells using baculovirus. 26,18,27 Our construct was designed to remove the FLAG epitope and BRIL sequence with TEV resulting in a final protein that begins with the first amino acid of mature human PAR4. Importantly, we show that the recombinant PAR4 is functional in response to the endogenous agonist thrombin ( Figure 1). Our approach can be adapted to express other PAR family members. We anticipate that this will allow biophysical and structural experiments to determine the unique aspects of the tethered ligand for PAR1, PAR2, and PAR3.
Based on mutagenesis and domain swapping studies with PAR1, it is predicted that ECL2 is critical for PAR activation.29 Given that PAR1 and PAR4 have only 34% sequence identity overall and even lower in ECL2, each receptor needs to be empirically tested. The stable modifications of His residues during H/D exchange allow site specific monitoring as a protein changes conformations (Supplemental Figure  1).19-21 PAR4 has two histidine residues located in ECL2 (His229 and His240). Both of which had an increased t½ following activation by thrombin ( Figure 2) suggesting ECL2 also has a role in PAR4 activation. Higher resolution methods are needed to determine if these are specific interactions with the tethered ligand, or if the longer t½ is due to general steric hinderance in this region. The t½ of His380 in the C-terminal tail increased 1.8-fold, Table 1 and Figure 2, suggesting that the C-terminus changes conformation and becomes protected from the solvent. These data are consistent with an increased stability and resistance of the C-terminus to proteolytic digestion by thermolysis ( Figure 4B).
The slow H/D exchange of histidine imidazole ring, thus also its slow back exchange from D to H, allows long working times that provide the opportunity to examine proteins in complex mixtures by employing strategies that enrich for the protein of interest, such as affinity purification or immunoprecipitation, before the mass spectrometry. The ability to use these extra steps provide opportunities to examine PAR4 in its native membrane where it forms homo and heterodimers that impact it rates of activation and signaling.24,45-48 While advantageous for these applications, the slow His H/D exchange also creates limitations. The long exposure required for the H/D exchange on histidines greatly depends on the stability of the protein of interest. In addition, the site specific information is dependent on the number and location of His residues within the protein, however it is possible to engineer His residues at particular sites to probe regions of interest. Finally, the site specific information is useful, but the overall coverage of the protein is limited. Going forward, we can use our purification strategy to analyze PAR4 using other MS methods. By monitoring the exchange rate between the amide hydrogen at the backbone of the protein and the solvent deuterium, amide-HDX will allow us to monitor a more detailed PAR4 conformational change with a full sequence coverage and under a more physiological relevant time scale.
PARs can also be activated experimentally using soluble peptides derived from the N-terminus. For example, GYPGQV will activate PAR4 by mimicking the tethered ligand at the N-terminus following cleavage.32,33 However, GYPGQV is a poor PAR4 agonist, which prompted peptide library screenings for optimizing potency that AYPGKF as an alternative.32 The activation peptides have been essential for elucidating PAR4 specific signaling in platelets and other cells that also express PAR1, which is also activated by thrombin.49 However, the cellular response from the agonist peptides are not identical to thrombin suggesting that activation by the tethered ligand induces a structurally distinct conformation. 22 We analyzed PAR4 stability by its resistance to thermolysin proteolysis following stimulation with the PAR4 agonist peptide AYPGKF. In contrast to thrombin, stimulation of PAR4 with AYPGKF (1 mM) does not increase the stability of PAR4, compare Figure 3 to Figure 5. Our His-HDX data support the hypothesis that activation by thrombin leads to a dramatic structural rearrangement of PAR4 resulting in resistance to thermolysis digest. The difference in PAR4 response is likely due to the distinct mechanisms by which the tethered ligand and the soluble activation peptide interact with the receptor.
In summary, the data presented here provide a strategy for expressing PAR4 and monitoring its activation by the tethered ligand using His-HDX. Going forward, it is now possible to further refine the purification scheme to provide a native environment using styrene maleic acid co-polymer lipid particles (SMALPS) or membrane nanodisc for applying the rapidly developing cryo-EM technologies to determine the nature