Yue Fn,Zhipeng Yu,,*,Wenzhu Zho,Long Ding,Fuping Zheng,Jinrong Li,Jingo Liu

a College of Food Science and Engineering,Bohai University,Jinzhou,121013,China

b Beijing Advanced Innovation Center for Food Nutrition and Human Health,Beijing Technology and Business University(BTBU),Beijing,102488,China

c College of Food Science and Engineering,Northwest A&F University,Yangling,712100,China

d Lab of Nutrition and Functional Food,Jilin University,Changchun,130062,China

ABSTRACT This study aimed to identify novel ACEI peptides from Larimichthys crocea titin using in silico approaches and to clarify the molecular interaction mechanism.The hydrolyzed peptides of titin were compared with known ACEI peptides in the AHTPDB and BIOPEP-UWM database.Furthermore,peptides were evaluated for their solubility,ADMET properties,△G(kcal/mol)values,and in vitro ACEI activity.Molecular mechanism of ACE-peptide was performed by molecular interactions and binding orientation study.The results revealed that IC50 values of Trp-Ala-Arg(WAR)and Trp-Gln-Arg(WQR)were(31.2±0.8)and(231.33±0.02)μmol/L,respectively.The docking interactions result suggested that ACE-WAR and ACEWQR complexes have same binding site, including the residues LYS511, TYR520, TYR523, HIS353, and HIS513.Molecular docking of two tripeptides WAR and WQR with ACE studies predicted their binding site and clarified the interaction between ACE and its inhibitors.The molecular docking data are consistent with the ACE inhibitory activity of the studied peptides.The results showed that Larimichthys crocea titin may be a valuable source for developing nutraceutical food.

Keywords:ACE peptide Identification In silico approaches ADMET Molecular docking

1.Introduction

Hypertension is a chronic arterial hypertension sign and causes many diseases, such as heart failure, loss of vision, and kidney disease.Angiotensin-converting enzyme(ACE)is an exopeptidase and belongs to peptidase family M2 (clan MA(E)).ACE converts decapeptide angiotensin I to octapeptide angiotensin II,which is a potent vasoconstrictor and inactivates the vasodilator bradykinin,leading to arterial constriction[1].Therefore,the excessive action of ACE is believed to be one of the major factors contributing to the development of hypertension disease.Thus, the inhibition of ACE is a novel therapeutic approach for antihypertension [2,3].It has proven that peptides derived from various food proteins exert bioactive effects.Compared with amino acids and proteins, peptides could be more easily absorbed.In addition,30%-50%portion of dietary nitrogen is absorbed in the form of small peptides[4,5].To the best of our knowledge, the number of angiotensin-converting enzyme inhibitory (ACEI) peptides have been isolated from food proteins,such as RVPSL from egg[6],MAA from whey protein[7],WDDME from fish[8-10],LSW from soy[11],and ASPYAFGL from mushroom[12].Larimichthys crocea is an economically important marine fish with particular high protein content; however, there is no report of ACEI peptides from L.crocea titin.Titin is part of the skeletal muscle and the third most abundant protein in muscle,after myosin and actin.Titin also known as connectin which protein sequences from L.crocea have also been reported[13,14].The titin from L.crocea may be a good source of ACEI peptides.

When preparation of ACEI peptides from food proteins, one of the most common technique is enzymatic hydrolysis [6], but it exists some limitations in time saving and cost effective benefits.The drawback of this approach included activity reduction of hydrolysates in enzyme inactivation process with high temperature, high-throughput but relatively low-resolution and so on [15,16].During the past decades, more efforts were directed towards the enzymatic hydrolysis,separation and identification of peptides.On the premise of saving the time and costs of purification and identification,virtual screening methods narrow down the number of compounds to be actually tested in biological assays[17].Compared with the classical enzymatic hydrolysis, simulations hydrolysis has been utilized to elucidate the peptide sequences and position of cleavage sites in parent protein once using selection of enzymes and chemicals [18].The application of in silico tools can accelerate the process and cut down costs for discovery of novel peptides, and illuminate the action mechanism.In silico analysis as a bioinformatics approach, is a useful method for predicting potential cleavage sites of protein sequences and screening bioactive peptides effectively.It has been reported that tripeptides could be absorbed directly from the digestive tract into the blood circulatory system[19].Therefore,tripeptides with high ACEI activities have attracted much attention.

The objective of the present study was to:(i)develop the simulations enzymatic hydrolysis of L.crocea titin using in silico approach;(ii)predict the absorption,distribution,metabolism,excretion,and toxicity (ADMET) properties of selected peptides; (iii) clarify the molecular mechanism of the interactions between potent tripeptides and ACE using Discovery Studio (DS) 2017 R2 Client; (iv)investigate the in vitro ACEI activity of peptides from the titin hydrolysates.

2.Materials and methods

2.1.Materials

Rabbit lung ACE (A6778), hippuryl-histidyl-leucine (HHL) and hippuric acid (HA) were obtained from Sigma-Aldrich (St.Louis,MO).Acetonitrile, methanol and trifluoroacetic acid (TFA) were high performance liquid chromatography (HPLC)-grade and purchased from Fisher Scientific (Waltham, MA, USA).All other chemicals used were of analytical grade.

2.2.Proteins screened for ACEI peptides

The titin from L.crocea including 17943 amino acid sequences has been taken from the national center for biotechnology information (NCBI), accession is KKF25250 [20].The titin was selected based on:(i)important component of L.crocea muscle protein;(ii)the availability of sequence information for the parent proteins.

2.3.In silico digestion

The digestion process was simulated using ExPASy Peptide-Cutter program,available at http://web.expasy.org/peptide_cutter,accessed on August 27th2017.The titin was entered in the form of a sequence, in one-letter amino acid code.ExPASy PeptideCutter predicts potential cleavage sites cleaved by proteases in titin sequences.The proteolysis simulation process of titin was performed by two typical enzyme of pepsin (pH 1.3, EC: 3.4.23.1)and trypsin (EC: 3.4.21.4).In addition, the peptides which were selected were compared with known ACEI peptides in bioactive peptides of BIOPEP-UWM database(http://www.uwm.edu.pl/biochemia/index.php/en/biopep)and anti-hypertensive inhibiting peptide database(AHTPDB)(http://crdd.osdd.net/raghava/ahtpdb/)[21,22],accessed on August 29th2017.In this study,‘ACE inhibitor’was searched by‘a(chǎn)ctivity’in BIOPEP-UWM database,which bioactive peptides was 3790 in current number.

2.4.In silico bioactive prediction

PeptideRanker server is used to predict the probability of these peptides being.The peptide is supposed to be active as its expected value exceeds 0.5 with PeptideRanker program.PeptideRanker ranks peptides by the predicted probability that the peptide will be bioactive [23] (http://bioware.ucd.ie/～compass/biowareweb/Server_pages/peptideranker.php),which was accessed on September 2nd2017.

2.5.In silico toxicity and solubility prediction

The program ToxinPred (http://crdd.osdd.net/raghava//toxinpred/) [24], a unique in silico method to predict toxicity of peptides,was implemented(5th September 2017).In batch submission,ToxinPred tool allows to identify highly toxic or non-toxic peptides from large number peptides submitted.In this study,the peptides in single letter cade were paste and all the physicochemical properties to be displayed.Moreover,an off line computational tool was adopted to investigate the solubility of peptides at http://www.innovagen.com/proteomics-tools (September 8th2017).In peptide property calculator, peptide sequences was inputted, the results will show the full application in a popup window.

2.6.ADMET analysis

ADMET properties prediction plays a vital role in identifying novel peptides with enhanced pharmacokinetic profiles [25].Therefore, we decided to estimate ADMET properties of virtually selected tripeptide before performing the costly experimental assays.In silico ADMET properties including human intestinal absorption (HIA), blood-brain barrier (BBB) penetration, and cytochrome P450(CYP 450)2D6 inhibition prediction,were computed by admetSAR.The prediction of HIA parameter of peptides was well-absorbed compounds of importance for identification potential peptides candidate[26].The BBB penetration is to avoid central nervous system (CNS) side effects used get through CNSactive compounds a very important pharmacokinetic parameter[27].In the current work, the process of ADMET was to analyses using a computational tool admetSAR, which was available at http://lmmd.ecust.edu.cn/admetsar1/(September 13th2017).The admetSAR operates with codes of peptides should be translated into simplified molecular input line entry specification (SMILES)by ChemSketch 12.0[28,29].

2.7.Screening for ACEI peptides

SwissDock as a web server was used to dock peptides and target proteins (http://www.swissdock.ch/), which was accessed on September 17th2017 [30].The structure of peptides was constructed using DS 2017 R2 Client.The crystal structure of ACE from Homo sapiens(1A1B.pdb)was obtained from the RCSB protein date bank(PDB),which was accessed at https://www.rcsb.org/on 17th September 2017.In this study, the ligand use MOL2 format and target use PDB format.To prepare all files necessary to perform docking with SwissDock, we use the program DS 2017 R2 Client.We enter an amino acid sequence and save as sybyl MOL2 files using DS 2017 R2 Client.The target obtained online from protein date bank and save as PDB files.The docking score was analyzed for selecting the potential ACE inhibitors.

2.8.Synthesis peptides

The peptide sequences were purchased from Shanghai Top Biological Technology Corporation(Shanghai,China)and synthesized using fluorenylmethyloxycarbonyl chloride (FMOC) protected amino acids synthesis methods[31].

2.9.ACEI activity assay

ACEI activity was measured by reverse phase-HPLC, using the method described by Liu et al.[6].A reaction mixture in a volume of 60 μL contained 100 mmol/L borate buffer (pH 8.3), 4 mmol/L HHL,300 mmol/L NaCl,and 10 milliunit ACE.The isocratic mobile phase consisted of 25% acetonitrile in deionizer water (V/V) with 0.5%TFA.An aliquot of 10 μL of the reaction mixture was analyzed on a Shimadzu C18 column.HHL and HA were detected at 228 nm,as the mobile phase was controlled at 0.5 mL/min.

2.10.Molecular mechanism of ACE-peptide

Molecular docking was conducted to calculate CDOCKER interaction energy value and to investigate the molecular mechanisms between the tripeptides and ACE, aimed to explore the binding affinity, binding mode, and molecular interaction.Molecular mechanism of ACE-peptide interaction was analyzed by molecular docking.In the DS 2017 R2 Client Macromolecules module,structure of the tripeptides was constructed.The energetically optimized of tripeptides were generated with the CHARMm force field using the“Minimize Ligands”and“Prepare Ligands”of Small Molecules module.(The parameters are all default).The crystal structure of the ACE (1O86.pdb, Resolution: 2.0 ?, EC: 3.4.15.1)was obtained from RCSB PDB(https://www.rcsb.org/)[32].Prior to docking,ACE was prepared using the Prepare Protein protocol and a binding site was created using the Define and Edit Binding Site protocol (default parameters were used).The docking runs were carried out with a radius of 9 ? with coordinates x: 40.4974, y:34.8829,and z:44.3814.

3.Results and discussion

3.1.Process of digestion and selection of peptides

L.crocea titin in NCBI with the protein sequence of 17943 amino acids was selected to evaluate potential ACEI peptides.Titin was simulated digested by trypsin and pepsin using the PeptideCutter server.Following simulated hydrolysis,the generated peptides were selected with reasonable molecular weight and absorption properties.PeptideRanker was applied to predict the relation between amino acid composition of the peptides and bioactivity.For every peptide, PeptideRanker corresponding scores the more closer to 1(between 0 and 1),the more confident that the peptide is bioactive.In fact,we chose a higher threshold which was 0.72 to reduce the number of false positives that all tripeptides scores[23].The sequence of 52 tripeptides was selected in this study, subsequently; active score in each predicted tripeptide sequence were showed in Table 1.Hence, the amino acid composition of tripeptides were selected by active score could further be evaluated as the main potential target for the anti-hypertensive activity.

3.2.Evaluated properties of selected tripeptides

Following simulated hydrolysis, the generated peptides were selected for good water solubility,no-toxic and good ADMET properties.ToxinPred was useful for discriminating toxic peptides from non-toxic peptides.The selected tripeptides were non-toxic under the toxicity category,which may be considered as promising ACEI peptides for controlling hypertension and development as a food ingredient.Solubility of tripeptides were estimated using the peptide property calculator,and 20 tripeptides,i.e.,CGR,DNF,DQF,FYK,GGR,GMR,GWK,MWK,PPR,WAR,WCR,WGK,WIR,WMK,WMR,WQR,WTR,WVR,WYR,and YDF,have been selected for good water solubility (Table 1).It is well known that good water solubility is the basis of oral bioactive peptides.Oral bioavailability of peptide is also considered as a significant parameter for the development of bioactive molecules into bioactive peptides.The results indicated that tripeptides with aromatic amino acids Trp at the N-terminus appear to have higher rates of good water solubility in this study.

In the research process of identification of food-derived bioactive peptide, it was found that ADMET were the important parameters.Hence,ADMET analysis was a reliable theoretical tool to evaluate these aspects such as absorption and metabolism.The ADMET prediction of peptides(i.e.,CGR,DNF,DQF,FYK,GGR,GMR,GWK,MWK,PPR,WAR,WCR,WGK,WIR,WMK,WMR,WQR,WTR,WVR,WYR,and YDF)was performed using admetSAR online.The absorption factor of peptides can be shown as BBB and HIA per-meability.The important requirements for a successful CNS-active compounds is to penetrate cross the BBB penetration.Good BBB probability shows good absorption of the peptide to the CNS,which suggested that the peptides are effective and possess good quality [33].The molecular weight of CGR, DNF, DQF, FYK, GGR, GMR,GWK,MWK,PPR,WAR,WCR,WGK,WIR,WMK,WMR,WQR,WTR,WVR, WYR, and YDF were calculated as 334.42, 394.41, 408.44,456.57, 288.34, 362.48, 389.49, 463.63, 368.46, 431.52, 463.58,389.49, 473.61, 463.63, 491.64, 488.58, 461.55, 459.58, 523.62,and 443.48 Da,respectively.Therefore,these low molecular weight peptides may cross the BBB with lipid mediated free diffusion[18].The BBB probability of the peptides (i.e., GMR, GWK, WAR, WCR,WGK,WQR,WVR,and WYR)also indicated ranges from 0.5161 to 0.7857.

Table 2 AdmetSAR operates with SMILES codes of good ADMET properties peptides and SwissDock computed binding energy details.

Fig.1.The docking resulted for the interaction of WAR with ACE(PDB:1O86),interactions with residues are shown in different colors.(a)Predicted 3D structure of ACE-WAR complex.(b)2D diagram of the ACE-WAR molecular interactions.(C)The 3D of WAR hydrogen bonds surface plot at the binding site.

Although the oral route is the most convenient and cheapest route of human body, the intestinal wall can present a significant barrier to peptides entry.HIA probability shows the peptides excellent permeability and absorptive property.The HIA probability of the peptides (i.e., GMR, GWK, WAR, WCR, WGK, WQR, WVR, and WYR)indicated ranges from 0.6327-0.9408.Based on the results in promising information,8 tripeptides(i.e.,GMR,GWK,WAR,WCR,WGK, WQR, WVR, and WYR) were neither CYP450 substrate nor CYP450 inhibitor, suggesting less compounds interactions in the course of metabolism.ADMET prediction results suggested that these peptides were potential for orally ACEI peptides.It is proved that the GMR,GWK,WAR,WCR,WGK,WQR,WVR,and WYR possessed good properties of HIA and BBB(Table 2).ADMET prediction results suggested that these predicted peptides were potential for orally ACEI peptides and should be further evaluated.

Fig.2.Docked interactions of ACE-WQR,interactions with residues are shown in different colors.(a)Predicted 3D structure of ACE-WQR complex.(b)2D diagram of the ACE-WQR molecular interactions.(C)The 3D of WQR hydrogen bonds surface plot at the binding site.

Molecular docking studies provided binding pose of ACEpeptide and predicted the interactions of ACE-peptide binding.The crystal structures of GMR,GWK,WAR,WCR,WGK,WQR,WVR,and WYR were utilized for the preliminary screening study of molecular docking by SwissDock server, and the △G value was -8.46,-8.07,-10.40,-8.42,-8.16,-9.57,-7.60,and-8.67 kcal/mol,respectively (Table 2).By energy minimizing, the existence of hydrogen bond between ACE and peptide will be more stable.Then selected peptides of WAR and WQR were the best-ranked which had much lower interaction energy.

3.3.In vitro ACE-inhibitory activity of peptides

After using SwissDock to discover bioactive peptides in silico methods, peptides would be experimented with in vitro to identify the bioactivity.ACEI activity of synthetic peptides WAR and WQR samples were analysis by RP-HPLC.The IC50value of WAR and WQR determined by an RP-HPLC method was(31.2±0.8)and(231.33±0.02)μmol/L, respectively.Finally, WAR and WQR from titin were compared with ACEI peptides of AHTPDB and BIOPEPUWM (data available at BIOPEP-UWM database on June 2018)database.It has confirmed that the peptides WAR and WQR have not been reported in previous literature.

According to the comparison of ACEI activity, it shows clear that the activity of WAR was better than that of WQR.Tripeptide WAR displayed the potent ACEI activity.Meanwhile, WAR was found to a process a stronger ACEI than the peptides TLS(IC50=102.1 μmol/L) from sweet sorghum grain protein [34],PGPLGLTGP (IC50=95 μmol/L) and QLGFLGPR (IC50= 148 μmol/L)from skate skin [35], AQGERHR (IC50=47.1 μmol/L) from freshwater zooplankton [36].The result indicated that Arg located C-terminal and Trp located N-terminal in tripeptides played an important role in the ACEI activity in this study.Many previous researches have proven that the N-domain and C-domain of ACEI peptides play an important role in ACE site.The N-terminal amino acids Trp of WYLHYA from casein proteins exhibited higher ACEI activity, with IC50value of (16.22±0.83) μmol/L in vitro [37].TNGIIR showed potent ACEI activity with the IC50value of 70 μmol/L, may due to the presence of amino acid Arg in C-terminal[38].The results demonstrated that WAR exhibited ACEI potency.It may prove that L.crocea titin is the potential source as the precursors of ACE inhibitor.

3.4.Molecular docking analysis of ACEI peptides and ACE

In order to understand the interactions between the target(ACE)and peptides, docking of the WAR and WQR into the binding site of ACE was performed, respectively.The active sites of ACE contained 12 amino acid residues which was GLN281,GLU411,HIS513,HIS383, GLU384, HIS387, TYR523, HIS353, GLU162, TYR520,LYS511,and ALA354[39].The complex of ACE-WAR and ACE-WQR binding sites was predicted by the DS 2017 R2 Client (Figs.1a and 2a ).The ACE-peptide was stabilized by the bonding energies.The docking of stabilized poses of ACE-WAR and ACE-WQR were obtained, and CDOCKER ENERGY values were -93.322 and-98.5874 kcal/mol, respectively, which indicate a good binding affinity between both of two peptides with ACE.

WAR and WQR were found able to bind the ACE active residues and therefore they could restrain the ACE activity.WAR interacts with 9 hydrophilic residues and 3 hydrophobic residues of ACE (Fig.1b), and WQR interacts with 8 hydrophilic residues and 3 hydrophobic residues of ACE (Fig.2b).Therefore, WAR contacts with ACE by more residues and gets more stable conjunction [40].ACE-WAR formed 12 hydrogen bonds with amino acids ALA354, HIS353, HIS513, TYR523, GLU376, TYR520, LYS511,ASP453, GLU384, and HIS513 in ACE (shown in Fig.1b).WAR formed pi-alkyl and metal-acceptor interactions with VAL518 and Zn701 of ACE, respectively.WAR could establish 12 hydrogen bonds, and was able to interact only with 10 ACE key residues,suggesting its non-competitive inhibitory way[41].WQR formed 9 hydrogen bonds with ASP453, LYS511, TYR520, TYR523, HIS353,HIS513, and ALA354 residues in ACE (shown in Fig.2b).WQR formed a metal-acceptor (Zn701) and a pi-pi T-shaped interaction (HIS387) with ACE.WQR also had the attractive charge interactions with LYS511 and GLU376.Meanwhile,three pi-donor hydrogen bond interactions participated in the binding of WQR with ACE (ALA356, GLU384, and HIS513).Both of two peptides formed multiple hydrogen bonds with ACE residues, as well as these residues may enhance ACE ability to immobilize peptides more stabilized which may lead to higher inhibitory ability[42].The results demonstrated that WAR could establish the higher number of hydrogen bonds with residues of ACE, while WQR could interact with the more numbers key residues in the ACE active site.It was noteworthy that ACE-WAR and ACE-WQR had same binding site, including the residues LYS511, TYR520, TYR523, HIS353, and HIS513.The molecular docking data are consistent with the higher ACE inhibitory activity of WAR when compared to WQR,which was attributed to the higher number of hydrogen bonds[43].

4.Conclusion

This study has supported that in silico approach was effective to predict in vitro ACEI peptides from L.crocea titin hydrolysates.According to the results of in silico analyses, the tripeptides with no-toxic, good water solubility and good ADMET properties were selected.Two novel tripeptides WAR and WQR were identified and confirmed to exert potent ACEI activities.The IC50values of tripeptides WAR and WQR were(31.2±0.8)and(231.33±0.02)μmol/L,respectively.The docking interactions of ACE-WAR and ACE-WQR have same binding site, including the residues LYS511, TYR520,TYR523,HIS353 and HIS513.Compare with WQR,WAR could establish the highest number of hydrogen bonds with ACE residues,and might be considered as good candidates for ACE inhibitor.Molecular docking of two tripeptides WAR and WQR with ACE studies predicted their binding site and clarified the interaction between ACE and its inhibitors.Thus, this result is predictable on the basis of molecular docking.Further research should be conducted on the model of the quantitative structure-activity relationship (QSAR)ACE binding to ACEI peptides.

Declaration of Competing Interest

The authors declare that they have no competing interests.

Acknowledgement

This paper was supported by The National Natural Science Funds of China(No.31901635).