INTRODUCTION

Pathogenic invasions are very common due to continu- ous increasing in the microbes population worldwide and also the effective mechanism to combat with these infec-

ARTICLE INFORMATION:

*Corresponding Author: bioinfosunil@gmail.com Received 11th May, 2016

Accepted after revision 5th June, 2016 BBRC Print ISSN: 0974-6455 Online ISSN: 2321-4007

Thomson Reuters ISI SCI Indexed Journal NAAS Journal Score : 3.48

©A Society of Science and Nature Publication, 2016. All rights reserved.

Online Contents Available at: http//www.bbrc.in/

tions are needed too. In past evolutionary era antibiotics have proven as strong microbicidal agents that ensure healthy survival of living organisms. Due to exten- sive uses, microorganisms started developing resist- ance against these antibiotics by changing its metabolic

216

pathways affected by these. In general, to kill microor- ganisms antibiotics target the cell wall and protein syn- thesis pathways. During resistance, bacteria modify its biological pathways that enable them escapes the attack of the antibiotics. Newer classes of antibiotics were gen- erated and so as the resistant strains of bacteria, (Anaya 2013 and Mor 2015).

Consequently, the conventional antibiotics proved to be useless to a large extent against these multi-drug resistant strains. This threat resulted in the probe of newer antimicrobial agents, which could prove effective against these resistant strains and then came into picture the naturally occurring antimicrobial peptides (Andreu 1998). These amphiphilic peptides are present in closely every life form as Nature’s antibiotics. These peptides playrole as the first line of defense in case of invertebrates and as the components of innate immunity in case of verte- brates. These are either expressed constitutively or syn- thesized promptly upon induction (Akira 2001). Antimi- crobial peptide is an alternative source of antibiotics was found earlier. AMP is produced by various microorganism such as fungi, bacteria, virus, Protista etc. It was found that plants also have the same properties to be effective against pathogenic microorganisms, (Krishnan 2016).

India have richest biodiversity in terms of flora and fauna. Different plants have medicinal properties to be effective against pathogenic microorganisms. Novel therapeutic modalities are needed due to occurrence of multiple drug resistant strains of pathogenic bacte- ria that lead to serious problem with respect to health perspectives. Mostly, antibiotics affect homeostasis of bacteria by interfering at specific site of action, which require few days for disabling. With these mechanism bacterial growth is inhibited but simultaneously selec- tion pressure for resistant bacteria, (Smith 2007), Ortega‐ Ramirez 2014 and Habiba 2016).

Besides, under such circumstances the bacteria mor- phology is normally preserved and a bacterium that is initially sensitive to the drug can develop resistance through mechanisms such as preventing the antibiotic from binding or entering the organism and producing an enzyme that inactivates the antibiotic or remodeling target molecules.Screening of Antimicrobial peptides (AMPs) from animals & plants are one of the hottest area of research to produce effective molecules as an additional substitutes to the currently used antimicrobial compound(Guralp 2013). In the innate immune system of the organism, these peptides act as effector molecules and represents first-line of defense against various path- ogens. Unlike action of antibiotics, mostly AMPs follows mode of action by disrupting and permeating the micro- bial membrane), which is a mechanism that prevents a target organism from acquiring resistance to the peptide, (Zasloff 2002, Hancock 2015).

Sunil Kumar Suryawanshi and Usha Chouhan

Curcuma longa L., commonly known as turmeric, is a spice used extensively in Eastern cuisine. The root of the plant is normally ground to yield a yellow colored powder, which then used for various applications. Some of the culinary uses of turmeric are as a coloring and flavoring agent and as a food preservative. Turmeric is also known to have medicinal applications dating back several millennia and has been an important component in the ancient Indian Ayurvedic medicines. Some of its documented modern medicinal uses include its wound healing capacity (Biswas 2003) and its use for the treat- ment of inflammation and tumors, (D’incalci 2005 Sahe- brao 2014 and Zorofchian Moghadamtousi 2014).

In this present study we have provided an approach to find out some antimicrobial activity from its peptides using machine learning algorithms such as Artificial Neu- ral Network, Discriminant Analysis, Random Forest and Support Vector Machine available at the interface of the CAMP database. Further, a comparison of physicochemi- cal properties has been made between predicted AMPs and also with patent AMP available at CAMP database.

MATERIAL AND METHODS

In this study, we download 165 protein sequences from NCBI with the hit of keyword “Curcuma longa”. Enor- mous peptide will be generated using these protein sequences by in silico digestion which is further evalu- ate the performance of our methodology. The protein sequences of turmeric proteins were subjected to in- silico proteolysis using Expasy Server(Gasteiger 2005). Specific hydrolytic enzymes action were performed here to determine the release of fragmented peptides from precursor proteins. Availability of peptide cutter tool at expasy server, which have different enzymatic as well as chemical option to digest protein sequences to peptides. In our proposed work, we implemented in silico diges- tion with the enzymatic action only, randomly selected three enzymes i.e., trypsin, protein kinase and pepsin out of 38 available at expasy server.

Digested peptides were subjected to antimicrobial activity with the use of the Prediction Antimicrobial Peptides tool available in the CAMP database. Four different multivariate statistical methods were imple- mented for efficient prediction of AMP: Support Vector Machines (SVM), Random Forest (RF), Artificial Neural Network (ANN) and Discriminant Analysis (DA). Waghu et al has been described the detail of model develop- ment and evaluation. Except ANN, prediction score as a results are presented in the numeric form and pep- tides are classified as AMPs or Non-AMPs. We consider released peptides as AMP, in case when all the four sta- tistical methods predicted as AMP (Waghu 2014). Fur-

Sunil Kumar Suryawanshi and Usha Chouhan

ther, these peptide characterize for its physicochemical properties.

The physicochemical properties of predicted AMPs were calculated with the use of available of free soft- ware and online algorithms. Protein length, peptide mass, hydrophobicity, steric hindrance, side bulk, hydro- pathicity, amphipathicity, hydrophilicity, net hydrogen, charge, pI, boman index, hydropathy index, aliphatic index, instability index, GRAVY (grand average of hydropathicity) are the physicochemical properties of predicted AMP were computed with the use of freely available tools AntiCP is web based prediction server for Anticancer peptides (Tyagi 2013) and APD2: Antimicro- bial Peptide Calculator and Predictor(Wang 2009).

RESULTS & DISCUSSION

IN SILICO PROTEOLYSIS OF C. LONGA PROTEINS

Different techniques are implemented to identify and produce significantly active peptides from different sources of proteins. Naturally, peptides present in the protein sequence which is released by enzymatic action or by the action of enzymes derived from different microorganisms. Mostly proteolytic enzymes such as trypsin, papain, pepsin, chymotrypsin and pancreatic elastase are used to digest proteins to peptides. Further, digested peptides should be refined such as fractiona- tion, separation and purification followed by evaluation of antimicrobial activity. With advent of research and development, now a days the cleavage site of enzymes are known as well as properties of peptides, due to this it is possible to search new AMP with bioinformatics tools (Keil 1992). Whole strategy are based on the computer simulation of proteolysis and application of multivari- ate statistical methods to determine potentially antimi- crobial peptides released from the analyzed proteins. So many tools available online for the proteolysis such as PMAP(Igarashi 2009)and Peptide Cutter(Keil 1986). In this study, 3 enzymes from the peptide cutter were used to simulate the proteolysis of C. longa proteins and pro- duces thousands of different fragments. Prediction of antimicrobial activity were based on specific ranges of 6 to 30 amino acids. As a resultant only 28 peptides out of 1813 non redundant peptides fragments were further evaluated for antimicrobial properties.

PREDICTION OF ANTIMICROBIAL ACTIVITY OF PEPTIDES RELEASED DURING IN SILICO PROTEOLYSIS OF C. LONGA PROTEINS.

Effective tools are available at different database for the prediction of antimicrobial activity of peptides such

as APD (Wang 2004), Bactibase (Hammami 2010), and PhytAMP(Hammami 2009). In this study, prediction tools available in CAMP database were used which is based on various machine learning algorithms such as SVM, RF, ANN and DA. The results accuracy for dif- ferent statistical models ranges from 87 to 93% and its prediction of antimicrobial activity is regardless of length. Firstly, the antimicrobial potential of C. longa protein fragments produced after proteolysis were eval- uated with the use of four statistical models available in the CAMP database. The results were used to manu- ally select peptides where a positive result was reported throughout all the four statistical models (Dziuba 2014). A total of 28 potentially antimicrobial peptides with the prediction scores higher than 0.45 for at least three sta- tistical models were thus selected (Table–2).

PHYSICOCHEMICAL CHARACTERISTICS OF AMPS

Physicochemical properties of 28 predicted AMPs from C. longa were calculated and compared with the AMP patent database available at CAMP. Table 1 represent the comparison of amino acid frequency and amino acid composition (%) between 28 predicted AMPs and AMP patent database. Most frequent amino acids found in AMP are ten which covers more than 66% in the proteo- lytic cleavage of C. longa and 77% of overall amino acid composition i.e., G A V L I K N S T E in patent database. On the other hand, 44 & 23% of amino acid composition covers M P W Q Y C F D R H respectively. The content of amino acids in the proteolytic peptides was calculated and compared with the patent peptide database. The results indicates that frequency of certain amino acids are more common in the AMPs and can influence their biological activity (Bevilacqua 2015).

The Majority of antimicrobial peptides are cationic amphipathic peptides, also containing one or more hydrophobic residues. Hydrophobic amino acids such as I G L accounts for more than 30 percent of ana- lyzed AMPs from C. longa and its play important role in secondary structure formation and interactions with bacterial membrane(Taniguchi 2016). The cationic and hydrophobic characters peptides affect the antimicrobial activity. The analysis of amino acid content of antimi- crobial peptides revealed many fragments with the pre- dominance of one or several amino acids. In comparison with all AMPs listed in the patent database, amino acid such as V & S are less frequently encountered from C. longa. On average, the evaluated peptides in both case have higher content of I, K & G. In both the case, result shows significant frequency of serine to be antimicrobial in nature(Pushpanathan 2013).

As earlier discuss that 165 proteins from C. longa were processed and as a resultant only 28 peptides predicted

to have antimicrobial activity as represented in table

2.Proteolysis of 165 proteins generates 5858 peptides fragments, after removal of redundancy 1813 peptides remain to be non-redundant. These peptides were pre- dicted for AMP through various algorithms such as sup- port vector machine (SVM), random forest (RF), artificial neural network (ANN) and discriminant classifier (DAC). Only 3 out of 4 algorithms provides accuracy in numeric form i.e. above 0.5 and ANN provides only information that whether peptides are AMP or NAMP, then peptide consider as Antimicrobial peptide (AMP). On the basis of prediction evaluation of these peptides only 28 peptides out of 1813 peptides found to be antimicrobial in nature. Due to diversity in these peptide sequences accuracy and physicochemical properties are also varies from pep- tide to peptide. Highest accuracy more than 90 percent among all the three algorithms were found in only one peptide sequences i.e., GAIIGNRKIKLQPHIIIRIID. More than 90 percent accuracy was found among two algo- rithms were found in nine peptides i.e., VWVKPWLHFK, VLGFPSKPIGLFIK, RKRHMY, KKPHRY, IIPKKSIY, GIVC- SAVNSETGEQVAIKKIGNA, GGIRCPLTVVQSRGIGTIISSP,

GAIIGDSKIRLQPHIIKRIIS,LGIQAFQPILVEGSAICLHPL VCK and other remaining 18 peptides have compara- tively lower accuracy below 90 percent.

The selected peptides in our work are potentially antimicrobial AMPs have 6 to 30 amino acids, which corresponds to peptide mass of 117.175 Da to 3427.126 Da. The general classification of AMPs include cationic peptides which can be divided in three subclasses: linear peptides forming helical structures, cysteine-rich open- ended peptides containing single or several disulfide bridges and molecules rich in specific amino acids, such as proline, glycine, histidine and tryptophan. Majority of antimicrobial peptides with positive charges known as cationic antimicrobial peptides (CAPs). In our study, 85.71%, 3.57% & 7.14% peptides are found cationic, anionic and neutral charge in nature.

Statistical analysis of the physicochemical properties of 28 predicted AMP represented in table 3. Some valu- able information inferred from these results in terms of significant correlation. (Chan 2003) Peptide length is sig- nificantly correlated with the peptide mass with 0.967, hydrophobicity with hydropathicity with 0.932, steric

IDENTIFICATION AND PREDICTION THE FOR BIOINFORMATICS OF APPLICATION 220

OMMUNICATIONSC ESEARCHR IOTECHNOLOGYB IOSCIENCEB

Table 2: List of predicted and identified AMP from C. longa and its calculated physicochemical properties.

Chouhan Usha and Suryawanshi Kumar Sunil

IDENTIFICATION AND PREDICTION THE FOR BIOINFORMATICS OF APPLICATION 222

OMMUNICATIONSC ESEARCHR IOTECHNOLOGYB IOSCIENCEB

Table 3: Statistical representation of physicochemical properties of predicted AMP by using Pearson Correlation

Correlations

**. Correlation is significant at the 0.01 level (2-tailed).

*. Correlation is significant at the 0.05 level (2-tailed)

Chouhan Usha and Suryawanshi Kumar Sunil

hinderance & steric bulk with peptide mass with 0.356, Hydropathicity with GRAVY with 0.999, Amphipathic- ity with hydrophilicity with 0.874, Net hydrogen with hydrophilicity with 0.842, charge with pI with 0.780, hydropathy index with aliphatic index with 0.777, insta- bility index with steric hinderance with 0.349. Figure 1 shows graphical representation of various physicochem- ical properties of predicted AMP from C. longa.

CONCLUSION

Peptide drug development is one of the recent approaches which have potential to replace the antibiotics and be effective in multidrug resistance organism. Here, in our work we concluded that 28 shorter fragment of tur- meric protein have potential to be AMP. Synthetically designed these peptide will shows the promising fea- ture to be effective against microorganism progression. Predicted and identified AMP from turmeric which will show positive result in wet lab experiments. Artificial synthesis of peptide requires efficient design, here we provide in silico analyzed curated peptides which needs further wet lab validation before practicing/applying them for therapeutics and industrial application.

ACKNOWLEDGMENTS

The authors are highly thankful to the Ministry of Human Resource Development, Government of India

and Department of Bioinformatics, MANIT, Bhopal, India for providing support in the form of Bioinformat- ics infrastructure facility to carry out this work.

REFERENCES

Akira, S., Takeda, K., Kaisho, T. (2001) Toll-like receptors: critical proteins linking innate and acquired immunity. Nature Immunology, Vol. 2, No. 8,pp. 675-80.

Anaya, L., López M., Ochoa Z., Alejandra (2013), Bacterial resistance to cationic antimicrobial peptides. Critical Reviews in Microbiology, Vol. 39, No. 2,pp. 180-95.

Andreu, D., Rivas, L. (1998) Animal antimicrobial peptides: an overview, Peptide Science, Vol. 47, No. 6,pp. 415-33.

Bevilacqua, M.P., Benitez, D., Deming, T.J., Hanson, J.A., Koz- iol, L. (2015) Composition with high antimicrobial activity and low toxicity. US Patent 20, 150,366,193.

Biswas, T.K., Mukherjee, B. (2003) Plant medicines of Indian origin for wound healing activity: a review. The Interna- tional Journal of Lower Extremity Wounds, Vol. 2, No. 1, pp. 25-39.

D’Incalci, M., Steward, W.P,Gescher, A.J. (2005) Use of cancer chemopreventive phytochemicals as antineoplastic agents. The Lancet Oncology, Vol. 6, No. 11,pp. 899-904.

Dziuba, B., Dziuba, M. (2014) New milk protein-derived pep- tides with potential antimicrobial activity: An approach based on bioinformatic studies. International Journal of Molecular Sciences, Vol. 15, No. 8,pp. 14531-45.

Sunil Kumar Suryawanshi and Usha Chouhan

Gasteiger, E., Hoogland, C., Gattiker, A., Duvaud, S.E., Wilkins, M.R., Appel, R.D., Bairoch, A. (2005) Protein identification and analysis tools on the ExPASy server. Springer.

Guralp, S.A., Murgha, Y.E., Rouillard, J.-M., Gulari, E. (2013) From design to screening: a new antimicrobial peptide discov- ery pipeline. PLoS One, Vol. 8, No. 3, p. e59305.

Habiba, U., Ahmad, M., Shinwari, S., Sultana, S., Shinwari, Z.K., Zafar, M. (2016) Antibacterial and antifungal potential of himalayan medicinal plants for treating wound infections’, Pakistan Journal of Botany, Vol. 48, No. 1,pp. 371-5.

Hammami, R., Hamida, J.B., Vergoten, G., Fliss, I. (2009) PhytAMP: a database dedicated to antimicrobial plant pep- tides. Nucleic Acids Research, Vol. 37, No. suppl 1,pp. D963-D8.

Hammami, R., Zouhir, A., Le Lay, C., Hamida, J.B., Fliss, I. (2010) BACTIBASE second release: a database and tool plat- form for bacteriocin characterization. Bmc Microbiology, Vol. 10, No. 1,p. 1.

Hancock, R.E., Hilpert, K., Cherkasov, A., Fjell, C. (2015) Small cationic antimicrobial peptides. Google Patents.

Igarashi, Y., Heureux, E., Doctor, K.S., Talwar, P., Gramatikova, S., Gramatikoff, K., Zhang, Y., Blinov, M., Ibragimova, S.S., Boyd, S. (2009) PMAP: databases for analyzing proteolytic events and pathways. Nucleic Acids Research, Vol. 37, No. suppl 1, pp. D611-D8.

Keil, B. (1986) Proteolysis Data Bank: specificity of alpha- chymotrypsin from computation of protein cleavages. Protein Sequences & Data Analysis, Vol. 1, No. 1, pp. 13-20.

Keil, B. (1992) Essential substrate residues for action of endopeptidases. Specificity of Proteolysis, Springer, pp. 43-228.

Krishnan, R.J., Nair, S.R. (2016) Preliminary Study on the Anti- bacterial Activity of Six Medicinal Plants against Two Naso- Pharyngeal Pathogens—Streptococccus pyogenes and Pseu- domonas aeruginosa. American Journal of Plant Sciences, Vol. 7, No. 06,p. 907.

Mor, A., Radzishevsky, I. (2015) Novel antimicrobial agents. Google Patents.

Ortega‐Ramirez, L.A., Rodriguez‐Garcia, I., Leyva, J.M., Cruz‐ Valenzuela, M.R., Silva‐Espinoza, B.A., Gonzalez‐Aguilar,

G.A., Siddiqui, M.W., Ayala‐Zavala, J.F. (2014) Potential of medicinal plants as antimicrobial and antioxidant agents in food industry: a hypothesis. Journal of Food Science, Vol. 79, No. 2,pp. R129-R37.

Pushpanathan, M., Gunasekaran, P., Rajendhran, J. (2013) Antimicrobial peptides: versatile biological properties. Inter- national Journal of Peptides, Vol. 2013.

Sahebrao, K.R., Deepak, L.M. (2014) World Journal of Pharma- ceutical Research.

Smith, P.A., Romesberg, F.E. (2007) Combating bacteria and drug resistance by inhibiting mechanisms of persistence and adaptation. Nature Chemical Biology, Vol. 3, No. 9, pp. 549-56.

Taniguchi, M., Ochiai, A., Takahashi, K., Nakamichi, S.i., Nomoto, T., Saitoh, E., Kato, T., Tanaka, T. (2016) Effect of alanine, leucine, and arginine substitution on antimicrobial activity against Candida albicans and action mechanism of a cationic octadecapeptide derived from α‐amylase of rice. Pep- tide Science.

Tyagi, A., Kapoor, P., Kumar, R., Chaudhary, K., Gautam, A., Raghava, G. (2013) In silico models for designing and discov- ering novel anticancer peptides. Scientific reports, Vol. 3.

Waghu, F.H., Gopi, L., Barai, R.S., Ramteke, P., Nizami, B., Idicula-Thomas, S. (2014) CAMP: Collection of sequences and structures of antimicrobial peptides. Nucleic Acids Research, Vol. 42, No. D1,pp. D1154-D8.

Wang, G., Li, X., Wang, Z. (2009) APD2: the updated antimi- crobial peptide database and its application in peptide design. Nucleic Acids Research, Vol. 37, No. suppl 1, pp. D933-D7.

Wang, Z., Wang, G. (2004) APD: the antimicrobial peptide database. Nucleic Acids Research, Vol. 32, No. suppl 1, pp. D590-D2.

Zasloff, M. (2002) Antimicrobial peptides of multicellular organisms. Nature, Vol. 415, No. 6870,pp. 389-95.

Zorofchian Moghadamtousi, S., Abdul Kadir, H., Has- sandarvish, P., Tajik, H., Abubakar, S., Zandi, K. (2014) A review on antibacterial, antiviral, and antifungal activity of curcumin. BioMed ResearchInternational, Vol. 2014.