Differences in PAR-2 activating potential by king crab (Paralithodes camtschaticus), salmon (Salmo salar), and bovine (Bos taurus) trypsin

Background Salmon trypsin is shown to increase secretion of the pro-inflammatory cytokine interleukin (IL)-8 from human airway epithelial cells through activation of PAR-2. Secretion of IL-8 induced by king crab trypsin is observed in a different concentration range compared to salmon trypsin, and seems to be only partially related to PAR-2 activation. This report aim to identify differences in the molecular structure of king crab trypsin (Paralithodes camtschaticus) compared to salmon (Salmo salar) and bovine trypsin (Bos taurus) that might influence the ability to activate protease-activated receptor-2 (PAR-2). Results During purification king crab trypsin displayed stronger binding capacity to the anionic column used in fast protein liquid chromatography compared to fish trypsins, and was identified as a slightly bigger molecule. Measurements of enzymatic activity yielded no obvious differences between the trypsins tested. Molecular modelling showed that king crab trypsin has a large area with strong negative electrostatic potential compared to the smaller negative areas in bovine and salmon trypsins. Bovine and salmon trypsins also displayed areas with strong positive electrostatic potential, a feature lacking in the king crab trypsin. Furthermore we have identified 3 divergent positions (Asp196, Arg244, and Tyr247) located near the substrate binding pocket of king crab trypsin that might affect the binding and cleavage of PAR-2. Conclusion These preliminary results indicate that electrostatic interactions could be of importance in binding, cleavage and subsequent activation of PAR-2.


Background
Trypsin is a known activator of protease-activated receptor (PAR)-2 [1,2]. Receptor activation by proteases is achieved by proteolytic cleavage of the N-terminal sequence. This cleavage unmasks a new amino terminus that serves as a tethered ligand that binds to conserved regions in the body of the receptor, resulting in the initiation of signal transduction [3]. Exogenously applied synthetic peptides based on the sequence of the tethered ligand are also capable of activating PARs by directly binding to the body of the receptor. To date, four PARs have been cloned and characterized; PAR-1, PAR-2, PAR-3, and PAR-4. PARs have emerged as important receptors in airway inflammation and allergy, and PARs are expressed in all cell types that participate in the inflammatory response of the lung; epithelial cells, mast cells, macrophages, infiltrated neutrophils and eosinophils, fibroblasts, smooth muscle cells, endothelial cells, lymphocytes, and neurons [4,5].
Our previous work has confirmed that purified salmon trypsin increase secretion of the pro-inflammatory cytokine interleukin (IL)-8 from human airway epithelial cells through activation of PAR-2 [6]. Secretion of cytokines from the airway epithelium contributes to an inflammation response and can be induced by both endogenous and exogenous proteases [3]. Based on the knowledge that occupational airway symptoms are frequently presented by workers handling different species of fish and crustaceans [7][8][9][10][11][12][13][14][15], we have tested several types of seafood trypsins in our cell based assays. This in order to investigate possible initiation of signal transduction connected to inflammation processes in human airway epithelial cells [6,16] Molecular size, conformation and electrostatic potential will influence on a molecule's ability to bind and interact with signalling partners. To bind tightly, the ligand must possess a shape and a charge distribution that are complementary to the target receptor [17,18]. In the molecular complex formed, attractive van der Waals and electrostatic (charge-charge) interactions are made across the binding interface. While the van der Waals interactions are relatively non-specific and small in magnitude, the electrostatic (charge-charge) interactions are highly specific and act over a significantly longer range. A small chemical change as conversion of one amino acid from Lto D-form or substitution of amino acids can inactivate the molecule, as the receptor may fail to bind the altered form or bind it less efficiently. The nature of the interaction between two signalling partners will influence upon downstream signalling pathways following molecular interaction leading to activation and transmission of the molecular signal (full or partial agonists) or inactivation without signal transduction (antagonists) [19].
We decided to explore the observed divergence between king crab and fish trypsins further. Purified salmon, sardine, bovine and king crab trypsins were evaluated by their ability to hydrolyze a chromogenic substrate (DL-BAPNA) and molecular modelling was executed of salmon, bovine and king crab trypsins for comparison of molecular structure and possibly identification of important amino acids in the trypsin -PAR-2 interaction.

Protein determination
The protein concentration was determined with a Nano-Drop ND-1000 spectrophotometer (Thermo Scientific) using the Protein A280 determination module.
The following day the benzamidine purified trypsins were applied to a 1 ml Resource Q ion exchange column equilibrated with 25 mM TrisHCl, 10 mM CaCl 2 , pH 7.5 and the enzymes were eluted with 1 M NaCl using a 7.5% gradient for 10 fractions (total of 5 ml) followed by a linear gradient rising to 100% in 20 fractions (total of 10 ml). Fractions corresponding to the observed peaks were tested for enzymatic activity and pooled before dialysis as described previously.

SDS-PAGE
After purification the trypsins were run on a sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) (4-12% NUPAGE; Invitrogen) and stained with SilverQuest TM Silver Staining Kit (Invitrogen) for verification of their purity.

Protease activity determination
The enzymatic activity of the purified trypsins was determined by a serine protease assay where the hydrolyzation of a chromogenic substrate (DL-BAPNA) was measured spectrophotochemically by the increase in absorbance at 405 nm at room temperature or 37°C for the length of 10 min. The substrate was diluted in substrate buffer (25 mM TrisHCl, 10 mM CaCl 2 , 2% (v/v) DMSO, pH 8.1) and used at a final concentration of 0.5 mM. The activity was measured in a total volume of 250 μl (10 μl of enzyme and 240 μl of diluted substrate) in clear, 96 well trays with flat bottom (BD Falcon, NJ, USA). The results were expressed as units (U)/ml [23], and one unit of activity was defined as 1 μmol substrate hydrolyzed per minute using an extinction coefficient of 8800 M -1 cm -1 [24]. The calculations were made using the following formula: Unit : dA/dt × 1/(ε × optical path length × 10 6 ) × V final , where dA/dt = rate of absorbance change and ε = extinction coefficient.

Molecular modelling
A homology model of king crab trypsin (Paralithodes camtschaticus; TREMBL accession code: Q8WR10_ PARCM) was built by using ICM3.5 [25] and the crystal structure of Atlantic salmon trypsin (pdb code: 1hj8) as a structural template. The homology model of king crab trypsin, the structures of Atlantic salmon trypsin (pdb code: 1hj8) and bovine trypsin (pdb code: 1s0r) were superimposed on the thrombin-PAR-4 structure (pdb code: 2pv9; [26]). The backbone of the catalytic site residues in the trypsins were superimposed with the corresponding residues of the PAR-4-thrombin complex. A model of the N-terminal fragment of human PAR-2 (Gly 28 -Val 53 ) was build and superimposed on top of the corresponding PAR-4 segment in the crystal structure of the thrombin-PAR-4 complex. Electrostatic potentials were calculated on the molecular surfaces of the model and structures of the trypsins.

Effect of temperature on enzymatic activity
The effect of temperature on the enzymatic activity of four different trypsins (bovine, salmon, sardine, and king crab) was examined using the substrate DL-BAPNA at room temperature and 37°C. As depicted in Figure 2, the total enzymatic activity of all trypsins decreased with lowered protein concentration. Comparing the enzyme activities at room temperature showed that salmon and sardine trypsin exhibited higher enzyme activity (11.3 and 8.34 mU/μg, respectively), as compared to king crab and bovine trypsin (2.26 and 2.38 mU/μg), using enzyme solutions with a protein concentration of 90 μg/ml. All trypsins showed an increased enzymatic activity when increasing the temperature to 37°C (Figure 2). The purified king crab trypsin showed the most pronounced increment with a 29% rise in enzymatic activity at the highest protein concentration, compared to 19.3% in the bovine trypsin, 18.1% in the purified sardine trypsin, and 15.2% in the purified salmon trypsin. The mean rise in enzymatic activity was 20.6% for the purified king crab trypsin, 17.7% for the bovine trypsin, 15.9% for the purified sardine trypsin, and 9.6% for the purified salmon trypsin. As for the total enzymatic activity, the degree of increment was also reduced with lowered protein concentration for all trypsins.
Molecular modelling shows differences in the protein structure of the king crab trypsin compared to salmon and bovine trypsin The surface of king crab trypsin has a large area with strong negative electrostatic potential (Figure 3a). The surface of bovine trypsin has smaller areas with strong negative potential, in particular around Asp 171 at the catalytic site, and also areas with strong positive electrostatic potential (Figure 3c).
Comparison of the homology model of king crab trypsin with the structures of Atlantic salmon and bovine

Discussion
A recent study by Ramachandran and co-workers [28] reports the ability of PAR-2 to exhibit functional selectivity where the proteolytically revealed tethered ligand (TL) sequence(s) and the mode of its presentation to the receptor (tethered vs. soluble) can confer biased signalling. Thus, PAR-2 can signal to multiple pathways that are differentially triggered by distinct protease-revealed tethered ligands. Differential signalling depending on the activating ligand (termed "agonist-biased signalling" or "functional selectivity") are also reported for other G protein-coupled receptors [29][30][31][32]. The exact factors that lead to activation of different signal pathways during binding of agonist to and cleavage of PAR-2 are unknown. Molecular modelling of the human PAR-1 has revealed various electrostatic, steric, and hydrophobic interactions between receptor and the antagonist used in docking studies [33]. Some residues identified as main points for electrostatic interactions have previously been reported as important in site directed mutagenesis studies for PAR-1 function and activity (Asp 256 and Glu 347 ) [34]. It is tempting to speculate that functional selectivity is a result of different capacity in the agonists to bind at these various interaction points.
Furthermore, the positive Arg 5 in the agonist peptide is from receptor chimera studies in PAR-1 suggested to interact with the negative Glu 260 in the second extracellular loop of PAR-1 during receptor activation [35,36]. Since these amino acids are conserved in PAR-2 an Arg 5 -Glu 232 interaction might operate in recognition of the PAR-2 agonist peptide SLIGRL by the receptor. Changing this residue in the PAR-2 agonist peptide (the positive Arg 5 in SLIGRL to a neutral alanine or a negative glutamatic acid creating SLIGAL or SLIGEL) markedly reduces the peptides' potency to cause intracellular Ca 2+ signalling [37]. The Al-Ani study [37] indicate that changes in the net charge of interacting amino acids influence on activating capacity; a result that may be due to interference with electrostatic interactions.
Differences in IL-8 secreting potential and NF-κB activation have been identified for salmon and king crab trypsin. Both effects are coupled to PAR-2 activation, but only partly for the king crab trypsin [6,16]. Molecular modelling shows that the surface of king crab trypsin has a large area with strong negative electrostatic potential compared to the smaller areas of bovine and salmon trypsins. In addition, these latter trypsins also display areas with strong positive electrostatic potential, a feature lacking in king crab trypsin. Because of the lack of a full amino acid sequence of the sardine trypsin molecule we were not able to do any modelling of and comparison with sardine trypsin.
The modelling of bovine, salmon, and king crab trypsin suggests that at least 3 divergent positions are located near the substrate binding pocket and might affect the binding of substrate to PAR-2: 1) The negative Asp 196 in king crab trypsin corresponds to the neutral Met 175 in salmon trypsin and the neutral Gln 155 in bovine trypsin.
2) The positive Arg 244 in the king crab trypsin corresponds to the negative Glu 221 in salmon trypsin and the neutral Gln 199 in bovine trypsin.
3) The neutral Tyr 247 in king crab trypsin corresponds to the neutral Asn 224 in salmon trypsin and to the positive Lys 202 in bovine trypsin.
It is possible that the positive Arg 36 and/or positive Lys 34 of PAR-2 may interact differently with the binding pocket in the three trypsins. Because of differences in the electrostatic potential it is possible that PAR-2 might bind weaker to king crab trypsin than to other trypsins due to repulsive interactions between the positive Lys 34 / Arg 36 in PAR-2 and the positively charged Arg 244 in king crab trypsin. This residue corresponds to a negative amino acid (Glu 221 ) in salmon trypsin and a neutral amino acid (Gln 199 ) in bovine trypsin. Zhang and co-workers [38] have recently documented that long-range electrostatic interactions presumably play an important role in aligning the PAR-2 N-terminal polypeptide with the activating protease (factor VIIa (FVIIa)) domain during binding and subsequent activation of PAR-2. By molecular simulations they show that positive amino acids in the proximity of the cleavage site of PAR-2 (Arg 31 , Lys 34 , and Arg 36 ) are located close to negatively charged residues on the binding pocket surface of FVIIa, whereas the negatively charged Asp 43 and Glu 56 are close to positively charged FVIIa residues. Although cell-based control studies were not conclusive (antibodies blocking possible PAR-2 interacting residues in the FVIIa molecule resulted in inhibited tissue factor (TF)-FVIIa signalling through PAR-2, while charge reversal mutations in FVIIa (positive Arg 62 to negative Glu 62 , and positive Arg 84 to negative Glu 84 ) did not significantly inhibit PAR-2 activation), we can not eliminate the possibility that electrostatic interactions in specific regions guide substrate orientation under physiological conditions. However, the specific FVIIa Arg 62 -PAR-2 Glu 56 interaction seems not to be essential for PAR-2 activation by TF-FVIIa.
The measurement of enzymatic activity at different concentrations both at room temperature and 37°C using DL-BAPNA as substrate yielded two separate but similar patterns among the four types of trypsins tested, with bovine and king crab trypsin in the lower area and the fish trypsins in the higher area. During the purification process the bovine trypsin behaved like the fish trypsins with regards to the binding capacity, giving rise to an inconsistency as it had similar enzymatic activity as the king crab trypsin. On the other hand, binding of trypsins to the N-terminal end of PAR-2 for cleavage and subsequent receptor activation might behave different from the binding and cleavage of the substrate in the serine protease assay, giving rise to differences in agonist potential in cell based assays.
These preliminary results indicate that electrostatic interactions can be of importance in binding, cleavage and subsequent activation of PAR-2, and that difference in electrostatic charge in residues at key interacting positions may result in altered potency of the agonist in question. However, more extensive molecular modelling of the entire PAR-2 together with docking studies of different agonists and functional binding assays (Isothermal Titration Calorimetry, Fluorescence Resonance Energy Transfer) to independently quantify the binding capacity of all the trypsins to PAR-2 peptide would be highly interesting. Mutations of key amino acids or generation of chimeric trypsins to investigate the structure and activity relationship, along with cell based assays, are necessary to identify essential residues that might influence upon functional selectivity.

Conclusion
The results indicate that electrostatic interactions could be of importance in binding, cleavage and subsequent activation of PAR-2 and differences in electrostatic charge in residues at key interacting positions may result in altered potency of the agonists in questions.

Availability of supporting data
The models constructed in this study are available from the authors upon request.