Introduction

In silico methods and their utility is widely applied in protein and genome sequence analysis (Ashokan and Pillai 2008; Shivakumar et al 2007; King-Haw Ling et al 2007; Yuri et al 2003).The repository of protein sequences and genome databases are increasing exponentially day by day. Computational tools provide researchers to understand physicochemical and structural properties of protein. A large number of online tools and servers are available from different sources for making prediction regarding the identification and structure of proteins. The various parameters like sequence length, number of amino acids and the physicochemical properties of a protein such as molecular weight, atomic composition, extinction coefficient, isoelectric point, GRAVY, aliphatic index, instability index etc. can be computed by various computational tools for the prediction and characterization of protein structure. The amino acid sequence provides most of the information required for determining and characterizing the molecule‟s function, physical and chemical properties. Marine mussels produce and secrete specialized protein adhesives that allow them to attach themselves to various substrates in aqueous and marine environments. The protein adhesives secreted by marine mussels overcome the harsh conditions and adhere tightly to wet surfaces by using the byssus (which consists of a bundle of threads) secreted from the foot of the mussel (Autumn et al 200, Chisolm and Kelley 2001, Walte 2002). At each end of the byssal thread is a plaque protein containing water-resistant protein glue that enables the plaque to anchor itself to wet solid surfaces. The primary mechanism for this strong protein adhesiveness is believed to be derived from naturally occurring L-DOPA amino acid residues, a derivative of tyrosine, inherent in the protein adhesive molecule. The L-DOPA amino acid residues (L-3, 4-dihydroxyphenylalanine) present in the protein function by cross-linking with each other via an oxidative conversion to their ortho quinone functionality (Waite 1999). This cross-linking results in strong, stable adhesive bonds. Mussel adhesive proteins (MAP) form permanent and strong adhesive bonds that are also flexible underwater (Wate 1999). Mussel adhesive proteins also form strong underwater bonds to such difficult to adhere to substrates as glass, Teflon®, metal and plastics (Crispe et al 1985). In addition, the biodegradable properties of MAP make them environmentally friendly for a number of applications. In particular, MAP can be used as a medical adhesive due to their lack of toxicity and immunogenicity in humans (Lee et al 2007, Staz et al 2005, Dalisin et al 2003).Synthetic adhesives (e.g. acrylics, cyanoacrylates, epoxies, phenolics, polyurethanes, and silicones) have largely displaced natural adhesives in the automotive, aerospace, biomedical, electronic, and marine equipment industries over the past century. Marine mussel adhesive protein is a formaldehyde-free natural adhesive that demonstrates excellent adhesion to several classes of materials, including pure metals, metal oxides, polymers, and glasses (Doraiswamy et al 2007). The available research on mussel adhesive protein shows that most of the works are focused on purification, identification and application of the protein. Thus the present investigation is concentrated on the characterization of mussel adhesive protein using computational tools and servers in six different species of mussel.

Materials and methods

Mussel adhesive protein (MAP) sequences were retrieved from the manually curated public protein databank SwissProt (Bockman et al 2003). SwissProt is scanned for the key word mussel adhesive protein, result yield 73 adhesive protein sequences. From this we retrieved 6 different sequences by random selection and have organized a non-redundant data set (Table1). The MAPs were retrieved in FASTA format and used for analysis.

Computational tools and servers

Results and Discussion

The result of the primary analysis suggests that most of the MAPs are hydrophobic in nature due to the presence of high non-polar residues (Table 2 and 3).

Figure 1

Table 1: Mussel adhesive sequence retrieved from TrMBL database

Figure 2

Table 2: Amino acid composition (in %) of Mussel adhesive protein using CLC sequence viewer tool.

Figure 3

Table 3: Hydrophilic and hydrophobic residues contents in mussel adhesive protein

The primary analysis also suggests that the MAPs contain one signal peptide in Mytilus coruscus, Mytilus galaprovincialis and Mytilus sp.JHX 2002 (Table 4). The presence of two cystein residues in MAP of Mytilus californianus (0.26%) indicates the absence of disulphide bridges (“SS”bonds). The MAPs is lack of tryptophan a non-polar aromatic amino acid and high concentration of tyrosine and lysine is common in all MAPs. The mol.wt, pI, number of positive and negative charged residues, EC, II, AI and GRAVY are depicted in the table 5. The average molecular weight of MAPs is 83694 Da. Isoelctric point is the PH at which the surface protein is covered with charge but net charge of protein is zero. At pI proteins are stable and compact. The computed pI of proteins are greater than seven (pI >7) indicates that MAPs are basic in character and it ranges between pI 9.84-pI 10.20. The computed isoelectric point (pI) will be useful for developing buffer system for purification by isoelectric focusing method. ProtParam computes the extension coefficient (EC) for a range of (276nm, 278nm, 219nm, 280nm and 282 nm) wave length, 283 nm is favorable because proteins absorb strongly there, while other substances commonly in protein solution do not. Extinction coefficient of MAPs at 280 nm is ranging from (30950-24855 M-1 cm-1) which indicates the presence of high concentration of tyrosine and lysine and not due to cystein, because cystein is very low in all the MAPs selected (Zero in Mytilus edulus). This indicates that MAPs cannot be analyzed using UV spectral methods. The computed protein concentration and extinction coefficient helps in the quantitative study of protein-protein and protein – ligand interactions in solution. The biocomputed half life of most of the MAPs is 30 hrs in all selected MAPs indicating the stability of the MAPs. The instability index (II)

Figure 4

Table 4: Various components of Mussel adhesive protein

Figure 5

Table 5: parameters computed using Expasy’s Protparm tool

Figure 6

Table 6: Trans membrane region identified by SOSUI server

Figure 7

Table 7: Disulphide (SS) bond pattern of pair predicted by CYS_REC

(Primary structure) and identified rasmol (using 3D structure modelled)

Figure 8

Table 8: PDB template (first two hits with maximum % of identified)

Obtained using BLASTP search against the protein data bank

Figure 9

Table 9: Validation parameters computed for the built 3D structure

Figure 10

Table 10: Criteria for a good (Model) 3D structure

Figure 11

Table 11: Signal peptide predicted in Mussel Adhesive protein

Figure 12

Table 12: Prediction of Localization of signal peptide

Figure 13

Table 13: Predicted Phosphporylation site in Serine, threonine and tyrosine amino acid in mussel adhesive protein sequence

Figure 14

Table 14: identified signal peptides in MAPs

Figure 15

Figure 16

computed by Expasy‟s ProtParam showed all the MAPs selected are stable with II greater than 30 (II>30). The aliphatic index (AI) which is defined as the relative volume of a protein occupied by aliphatic side chain (A, V, I and L) is regarded as a positive factor for increasing of thermal stability of globular proteins. The lower AI of BAA09850,

Figure 17

BAA09851, AAL87245 and Q25434 are indication of more flexible structure when compared to other MAPs. The very high AI of CAA38294 and ABC84184 infer these are stable at wide range of temperature .GRAVY index of MAPs are range from -0.835 to - 3.384. The very low GRAVY of all the selected MAPs infers that

Figure 18

these MAPs could results in a better interaction with water. The secondary structures predicted with the help of SOPMA (Data not presented) infer that the MAPs have rich lysine and tyrosine and contain mostly β- helices. In Myitlus californianus ά-

Figure 19

helices predominant and shows 30 times more than other MAPs (Fig1a-1f). The server SOSUI classifies the MAPs of Mytilus galprovincialis as a soluble protein. The SOSUI server has identified one transmembrane region in BAA09851 (Mytilus galaprovincialis) (Table 6). The transmembrane region is rich in hydrophobic amino acids and it is also well documented by Kyte and Doolittle (1982). Mean hydrophobicity profile (Fig 2) in which all the points are above the “0.0” line. The helix of BAA09851 is visualized using EMBOSS PepWheel (Fig 3). The tools CYS_REC recognizes the presence of two cystein in ABC84184 and predicted no “SS” bond pattern of pairs in the protein. Though cystein is very low the protein attains the stability because of the extensive hydrogen bonds and various turns and helices. The position of the cystein recognized by CYS_REC and Rasmol in the ABC84184 was modelled using PDB template (Table 8) selected from the hits obtained through BLASTP against PDB, and the modelled structure was evaluated. According to the evaluative analysis the Ramchandran plot and other parameters (Table 9) were within the standard acceptable limits for 3D structure (Table 10) modelled using the PDB template AIXF84.The cystein identified using the 3 dimensional structure of MAP ABC84184 is shown in the figure 4.The lacks of correlation between the position identified by CYS_REC and Rasmol might be due to the fact that CYS_REC identified cystein position from the primary structure. We speculated that the cystein position identified by Rasmole might be correct. The signal peptide predicted by SignalP server (Table11) shows that BAA09851, AAC87249 and ABC84184 sequence contain signal peptide with strongest prediction value one, indicating the localization in secretory pathway .Signal peptides are cleaved off the protein once its final destination has reached. The cleavage site is located in between 24-25th position in all the selected MAPs sequences except ABC84184, where it is between 19 and 20th position, and in CAA38294 the signal peptide is absent. The presence of signal peptide is reflected from the S-score (signal peptide score) (Fig 5). The sequences of signal peptide in MAPs are depicted in the table 12. The properties of the amino acids that constitute signal peptide region of a protein are the significant factors determining interaction with the protein transport system, hence the destination to which that protein is delivered. Different classes of signal peptide are used to specify different cellular placement. The protein which lack signal peptide will maintain in the cytoplasm. Thus the analyzed MAP can be classified into three groups based on the reliability class (RC) predicted by TargetP server. MAPs of BAA09850 belong to one class with poor RC (i.e. 5), BAA09851, AAL87249 and AB184184 belong to another class with strongest prediction of signal peptide (i.e. 1), and CAA38294 and Q25434 belong to separate class with RC two. The NetPhos server predicted phosphorylation site of serine, threonine and tyrosine (Table 13). The highest number of phosphorylation site is present in CAA38294 for serine and tyrosine, and BAA09850 for threonine. Phosphorylation usually occurs onserine, threonine in eukaryotic protein and rarely in tyrosine. Interestingly tyrosine is the most phosphorylated amino acid in MAPs except in ABC84184. This indicates that MAPs are relatively easy to purify using antibodies (Burnett and Kennedy 1954). Antibodies bind to and detect phosphorylation induced conformational changes in protein (Burnett and Kennedy 1954).

Conclusion

The biomimetic, mussel adhesive proteins have been chosen mainly to study their physicochemical properties, primary and secondary structure, by using computational tools and servers. Primary structure analysis reveals that MAPs are hydrophobic in nature, not contain any disulphide bridges, and contain signal peptide in some MAPs. Physicochemical characterization studies give good idea about the properties of such as pI, CE, AI, GRAVY and II that are essential and vital in providing data about the proteins and their properties. Secondary structure analysis predicts that the MAPs contain more β- helices than ά- helices. The signal peptide predicted showed that the peptide sequence located in secretory pathway and the cleavage site lies invariably between 24th and 25th position of the sequence. The phosphorylation site prediction showed tyrosine is the most phosphorylated in all the MAPSs and hence MAPS can be purified by using antibodies.