Int. J. Mol. Sci. 2014, 15(6), 11082-11099; doi:10.3390/ijms150611082

Article
The Structure and Dynamics of BmR1 Protein from Brugia malayi: In Silico Approaches
Bee Yin Khor , Gee Jun Tye , Theam Soon Lim , Rahmah Noordin and Yee Siew Choong *
Institute for Research in Molecular Medicine, Universiti Sains Malaysia, Minden, Penang 11800, Malaysia; E-Mails: kbyin89@gmail.com (B.Y.K.); geejun@usm.my (G.J.T.); theamsoon@usm.my (T.S.L.); rahmah8485@gmail.com (R.N.)
*
Author to whom correspondence should be addressed; E-Mail: yeesiew@usm.my; Tel.: +604-653-4837; Fax: +604-653-4803.
Received: 28 January 2014; in revised form: 25 March 2014 / Accepted: 4 June 2014 /
Published: 19 June 2014

Abstract

: Brugia malayi is a filarial nematode, which causes lymphatic filariasis in humans. In 1995, the disease has been identified by the World Health Organization (WHO) as one of the second leading causes of permanent and long-term disability and thus it is targeted for elimination by year 2020. Therefore, accurate filariasis diagnosis is important for management and elimination programs. A recombinant antigen (BmR1) from the Bm17DIII gene product was used for antibody-based filariasis diagnosis in “Brugia Rapid”. However, the structure and dynamics of BmR1 protein is yet to be elucidated. Here we study the three dimensional structure and dynamics of BmR1 protein using comparative modeling, threading and ab initio protein structure prediction. The best predicted structure obtained via an ab initio method (Rosetta) was further refined and minimized. A total of 5 ns molecular dynamics simulation were performed to investigate the packing of the protein. Here we also identified three epitopes as potential antibody binding sites from the molecular dynamics average structure. The structure and epitopes obtained from this study can be used to design a binder specific against BmR1, thus aiding future development of antigen-based filariasis diagnostics to complement the current diagnostics.
Keywords:
Brugia malayi; BmR1 protein; protein structure prediction; epitope prediction; molecular dynamics simulation

1. Introduction

Lymphatic filariasis (commonly known as elephantiasis) has infected 120 million people worldwide especially in developing and under-developed countries and approximately 1.3 billion people in 81 countries are at risk of infection. Brugia malayi, one of the causative agents of lymphatic filariasis, falls under the category of nematodes that infects human and animals. Infection occurs when the thread-liked parasitic filarial parasites are transmitted to humans through infected mosquitoes and develops into adult worms in human lymphatic vessels. In the year 2000, the WHO initiated a Global Program for Elimination of Lymphatic Filariasis (GPELF) with two main strategies: to stop the spread of infection (interrupting transmission) and to alleviate the suffering of the affected population (controlling morbidity) [1].

The availability of an easy on-site lymphatic filariasis diagnostic test which is rapid, affordable and accessible for disease management and therapy is one of the important factors for the elimination of lymphatic filariasis [2,3]. For brugian filariasis, one of the available diagnostics is the rapid immunochromatography detection of IgG4 antibody (Brugia Rapid). It is based on the recombinant antigen (BmR1) expressed from Bm17DIII gene (GenBank: AF225296) [2,3,4]. Studies showed that the BmR1 antigen is highly specific and sensitive for detection of IgG4 antibody in brugian filariasis [2,3,4]. Despite IgG4 being a good indicator of lymphatic filariasis, a sensitive and specific antigen-based detection test would also be an important alternative, which is currently not available. This antigen-based diagnostic would be a more direct test for active lymphatic filariasis infection. Thus, the sensitivity and specificity of BmR1 antigen makes it a promising candidate for development of an antigen-based detection test.

In addition, identification of epitopes as possible antibody binding sites is important in leading to the development of such a test. The three-dimensional (3D) structure of the BmR1 protein is essential in order to identify its epitopes, however, it has yet to be solved experimentally. Protein structure prediction is therefore the only way to predict the structure of BmR1 from its amino acid sequence. Here, in silico studies were carried out to construct the predictive structure of BmR1 and predict possible antibody epitopes of BmR1. Our results showed that an ab initio method built the best model. A total of three epitopes have been identified from the average structure of molecular dynamics simulation.

2. Results and Discussion

2.1. Results

The molecular weight of the BmR1 protein (206 amino acids) is 25 kDa. BLASTp results against non-redundant protein identified neither protein family nor conserved domains. Functional annotation by CD Search returned the same result as the BLASTp search. InterProScan, SMART and Pfam results have showed that the residue from 45–148 is a domain of unknown function 148 (DUF148). Pfam also indicated a match to baculovirus polyhedron protein, PEP C terminus from residues 116–160. Secondary structure prediction showed that BmR1 sequence consists of 9–11 α-helices and 1–3 β sheets (Table 1).

Table Table 1. Secondary structure prediction by PSIPRED [5], Jpred3 [6], SSpro 4.0 [7] and PORTER [8]. Secondary structure calculation of average MD structure performed by STRIDE [9].

Click here to display table

Table 1. Secondary structure prediction by PSIPRED [5], Jpred3 [6], SSpro 4.0 [7] and PORTER [8]. Secondary structure calculation of average MD structure performed by STRIDE [9].
1112131
SequenceM I K M N E K Y V KE L I L L L L F A MI Y T S L E S N C EF W I E D D F H P F
PSIPRED- - - H H H H H H HH H H H H H H H H HH H H H H H H - - BB B B - - - - - - -
Jpred3- - - - - H H H H HH H H H H H H H H HH H H - - - - - - BB B B - - - - - - -
SSpro- - - - - H H H H HH H H H H H H H H HH H H H H H H - - -B B B B - - - - - -
PORTER- - - - - H H H H HH H H H H H H H H HH H H H H H H - - -H H H - - - - - - -
STRIDE- - - - H H H H H HH H H H H H H H H HH H H H H H H H H HH H H H H - - - - -
41516171
SequenceV P K S E E A R E EY C G F F K E M N LS R N E L M D T I RK W A S K Y G V L E
PSIPRED- - - - H H H H H HH H H H H H H - - -- H H H H H H H H HH H H H H H H H H H
Jpred3- - - - - - H H H HH H H H H H H - - -- H H H H H H H H HH H H - - - - H H H
SSpro- - - - H H H H H HH H H H H H H H H -- H H H H H H H H HH H H H H - - H H H
PORTER- - - - H H H H H HH H H H H H H H - -- H H H H H H H H HH H H H H H H H H H
STRIDE- - H H H H H H H HH H H H H H H H - -- - - - H H H H H HH H H H H H - H H H
8191101111
SequenceQ F D N Y V D E E LR Y E N M V Y D I FK D K V N S T C G SE K I K R T L F E I
PSIPREDH H H H H H H H H HH H H H H H H H H HH H H H - - - - - HH H H H H H H H H H
Jpred3H H H H H H - - - -- - - - H H H H H HH H - - - - - - - -H H H H H H H H H H
SSproH H - - - - H H H HH H H H H H H H H HH - - - - - - - - HH H H H H H H H H H
PORTERH H H H H H H H H HH H H H H H H H H HH H H H - - - - - HH H H H H H H H H H
STRIDEH H H H H H - - - HH H H H H H H H H HH H H H H H H - - -- H H H H H H H H H
121131141151
SequenceT D L L T D R D T AQ Q T I Q T K I D EI I N N L N E R E RM E L T Q L W A I L
PSIPREDH H H H - - H H H HH H H H H H H H H HH H H - - - H H H HH H H H H H H H H H
Jpred3H H H H - - - H H HH H H H H H H H H HH H H H - - - - - HH H H H H H H H H H
SSproH H H H - - H H H HH H H H H H H H H HH H H H H H H H H HH H H H H H H H H H
PORTERH H H H - - H H H HH H H H H H H H H HH H H H H H H H H HH H H H H H H H H H
STRIDEH H H - H H H H H HH H H H H H H H H HH H H H - - H H H HH H H H H H H H H H
161171181191
SequenceG E E A I I E A Q DK F E N G N S I W EA V E N T T Q T D NF K S E I V K D N D
PSIPREDH H H H H H H H H HH H H - - - H H H HH H H - - - - - - -H H H H H H - - - -
Jpred3- - H H H H H H H HH H - - - - - H H HH H H - - - - - - -- H H H B - - - - -
SSproH H H H H H H H H HH - - - - - - B B BB B B - - - - - - -- - H H H - - - - -
PORTERH H H H H H H H H HH H H - - - H H H HH H H H - - - - - -- H H H H H - - - -
STRIDE- H H H H H H H H HH H H H - - - H H HH H H H - - - - - -- - - H H H H H H H
201
SequenceK I L I S N
PSIPREDB B B - - -
Jpred3B B B B - -
SSproB B B B - -
PORTER- B B - - -
STRIDE- - - - - -

H = α-helices; B = β-strand.

Table Table 2. Secondary structure calculation by STRIDE [9] on the models built by MODELLER 9v9 [10], QUARK [11], Robetta [12,13], Rosetta [14], I-TASSER [15] and Bhageerath [16].

Click here to display table

Table 2. Secondary structure calculation by STRIDE [9] on the models built by MODELLER 9v9 [10], QUARK [11], Robetta [12,13], Rosetta [14], I-TASSER [15] and Bhageerath [16].
1112131
SequenceM I K M N E K Y V KE L I L L L L F A MI Y T S L E S N C EF W I E D D F H P F
MODELLER9v9- - - - - - H H H -H H H H H H H H H HH H H H H H - - - -- - - - - - - - - -
QUARK- - - - - H H H H HH H H H H H H H H HH H H H H H H - - -- - - - - - - - - -
Robetta- - H H H H H H H HH H H H H H H H H HH H H H H H H - - -- - - - - - - - - -
Rosetta- - - - H H H H H HH H H H H H H H H HH H H H H H H H H HH H H H H - - - - -
I-TASSER- - - - - - H H H HH H H H H H H H H -- - H H H H H H H HH H - - - - - - - -
Bhageerath- - - - - H H H H HH H H H H H H - - -- - - - - - - - - -- - - - - - - - - -
41516171
SequenceV P K S E E A R E EY C G F F K E M N LS R N E L M D T I RK W A S K Y G V L E
MODELLER9v9- - - - H H H H H HH H H H H H H H H HH H H H H H H H H HH H H H H H - H H H
QUARK- - - - H H H H H HH H H H H H H H - -- H H H H H H H H HH H H H H H - H H H
Robetta- - - H H H H H H HH H H H H H H H - -- H H H H H H H H HH H H H H H - H H H
Rosetta- H H H H H H H H HH H H H H H H H - -- - - - H H H H H HH H H H H - - H H H
I-TASSERH H H H H H H H H HH H H H H H H H - -- H H H H H H H H HH - - - H H H H H H
Bhageerath- H H H H H H H H HH H H H H H H H H HH H H H - - - - - -- - H H H H H H - H
8191101111
SequenceQ F D N Y V D E E LR Y E N M V Y D I FK D K V N S T C G SE K I K R T L F E I
MODELLER9v9H H H H H H H H H HH H H – H H H H H HH H H H H H - - - HH H - - H H H H H H
QUARKH H H H H H H H H HH H H H H H H H H HH H H H H - - - - HH H H H H H H H H H
RobettaH H H H H H H H H HH H H H H H H H H HH H H H H H - - - HH H H H H H H H H H
RosettaH H H H H H H H H HH H H H H H H H H HH H H H H H - - - HH H H H H H H H H H
I-TASSERH H H H H H H H - -- H H H H H H H H HH H H H H - - - - -H H H H H H H H H H
BhageerathH H H H H H H H H HH H H H H H H H H -- - H H H H H H H HH H H H H H H - - -
121131141151
SequenceT D L L T D R D T AQ Q T I Q T K I D EI I N N L N E R E RM E L T Q L W A I L
MODELLER9v9H H H H H - H H H HH H H H H H H H - HH H H H H H H - H HH H H H H H H H H H
QUARKH H H H - - H H H HH H H H H H H H H HH H H H - - H H H HH H H H H H H H H H
RobettaH H H H H – H H H HH H H H H H H H H HH H H H - - H H H HH H H H H H H H H H
RosettaH H H H H H H H H HH H H H H H H H H HH H H H - - H H H HH H H H H H H H H H
I-TASSERH H H H - - H H H HH H H H H H H H H HH H H - - - H H H HH H H H H H H H H H
Bhageerath - - - - - H H H H HH H H H H H H H H H- - - - - - - - - -- - - - H H H H H H
161171181191
SequenceG E E A I I E A Q DK F E N G N S I W EA V E N T T Q T D NF K S E I V K D N D
MODELLER9v9H H H H H H H H H HH H H H H H H H - HH H - - - - - - - -- - H H H - - - - -
QUARKH H H H H H H H H HH H H H - - - H H HH H H - - - - H H HH H H H H H H - - -
RobettaH H H H H H H H H HH H H H - - - H H HH H H - H H H H H H- - - - H H H H - -
Rosetta- H H H H H H H H HH H H H - - - H H HH H H H - - - - - -- - - H H H H H H -
I-TASSER- - - - - H H H H HH H H - - - - - H HH H H H H H H H - -H H H H H H - - - -
BhageerathH H H H H H H H H HH H H H - - - - - -- - - - - - - - - -- - - - - - - - - -
201
SequenceK I L I S N
MODELLER9v9- - - - - -
QUARK- - - - - -
Robetta- H H H - -
Rosetta- - - - - -
I-TASSER- - - - - -
Bhageerath- - - - - -

H = α-helices; B = β-strand.

For automated protein structure prediction, only CPHmodels 3.0, QUARK, Bhageerath, Robetta and I-TASSER were able to predict the structure of BmR1. Secondary structure calculations by STRIDE [9] were done for the predicted structures to obtain the secondary structure information are showed in Table 2. The predicted structure from automated protein structure prediction, MODELLER 9v9 and Rosetta were further evaluated for backbone conformation and compatibility (Table 3). Overall, the evaluation data indicated that model built by Rosetta was the best structure with 97.5% of residues in the most favored region from the Ramachandran plot, 71.0% VERIFY 3D and 97% ERRAT scores.

Table Table 3. Model validation of structures predicted via comparative modeling, threading, ab initio method or combination of the approach.

Click here to display table

Table 3. Model validation of structures predicted via comparative modeling, threading, ab initio method or combination of the approach.
NameApproachRamachandran Plot [17]VERIFY3D [18] (%)ERRAT [19] (%)
Residues in Most Favored Region (%)Residues in Disallow Region
MODELLER 9v9 [10]Comparative modelling87.3-16.458.6
CPHmodels 3.0 [20]Comparative modelling79.6-48.393.3
QUARK [11]Ab initio92.9-83.695.5
Robetta [12,13]Ab initio97.5-73.991.4
Rosetta [14]Ab initio98.5-71.097.0
I-TASSER [15]Threading & ab initio80.7Ser 6155.696.0
Bhageerath [16]Ab initio84.3Met 20
Ser 24
Ile 142
Asp 102
Asn 176
33.361.4

A 5 ns molecular dynamics (MD) simulation was performed with structure predicted by Rosetta. The root mean square deviation (RMSD) with respect to its starting structure shows that BmR1 protein is stable after 1000 ps with an average RMSD of 2.9 ± 0.5 Å (Figure 1A). RMS fluctuation (RMSF) is approximately 4.5 Å at the C- and N-terminal and less than 2.5 Å for other residues (Figure 1B). Value for radius of gyration is also equilibrated after 1000 ps with the average of 20.2 ± 0.2 Å (Figure 1C). Average MD structure was created based on this analysis (1001 to 5000 ps). Secondary structure calculation on the average MD structure has showed similar patterns with the secondary structure prediction from PSIPRED [5], Jpred3 [6], SSpro 4.0 [7] and PORTER [8] in Table 1 and the initial structure (Rosetta) in Table 2, except the beta sheets regions (residue 30–33 and 201–204).

Ijms 15 11082 g001 200
Figure 1. Analysis on (A) RMSD (B) RMSF and (C) Radius of gyration of BmR1 protein during molecular dynamics simulation.

Click here to enlarge figure

Figure 1. Analysis on (A) RMSD (B) RMSF and (C) Radius of gyration of BmR1 protein during molecular dynamics simulation.
Ijms 15 11082 g001 1024

The average MD structure was further analysed by Prosa II Z-score [21] to access the quality of the results. The Z-score value for average MD structure was −6.77, which is within the range observed for native set of proteins of same size (Figure 2). If the Z-score of a model structure is located outside the range of typically native proteins found by X-ray and NMR, it indicates an erroneous structure [21]. In addition, ANOLEA [22] was used to evaluate the packing quality of the modelled structure based on the non-local atomic interactions [23]. The ANOLEA program is able to assess the global quality of protein, observe local error and consequently gives the energy value for each amino acid of a protein [22,24]. Figure 3 shows the pseudo-energy profile for each amino acid. It appears that the high-energy zones (positive ANOLEA values; Figure 3) are located at the loops. From the result, most of the amino acids fall into the negative ANOLEA values (Figure 3), which indicates that the energy of amino acids is in a favorable state.

Ijms 15 11082 g002 200
Figure 2. Prosa II Z-score plot of BmR1 protein. The Z-score for modelled BmR1 protein is represented as a black dot.

Click here to enlarge figure

Figure 2. Prosa II Z-score plot of BmR1 protein. The Z-score for modelled BmR1 protein is represented as a black dot.
Ijms 15 11082 g002 1024
Ijms 15 11082 g003 200
Figure 3. The packing quality of average MD structure analysed by ANOLEA [22]. High-energy amino acids show positive ANOLEA values (red bar) while low energy amino acids are with negative ANOLEA values (green bar).

Click here to enlarge figure

Figure 3. The packing quality of average MD structure analysed by ANOLEA [22]. High-energy amino acids show positive ANOLEA values (red bar) while low energy amino acids are with negative ANOLEA values (green bar).
Ijms 15 11082 g003 1024

STRING [25] and PROSITE [26] were unable to analyse the function of BmR1. ProFunc [27], which requires protein 3D structure as input, showed that BmR1 has 23% sequence identity to PDB id 3KNT [28] and 3FHF [29]. PDBeMotif [30] and Motif Scan from ExPASy [31] showed that there are two N-glycosylation sites (residues 59–62 and 105–108), three protein kinase C phosphorylation sites (residues 68–70, 110–112 and 125–127) and six casein kinase II phosphorylation sites (residues 23–26, 27–30, 61–64, 116–119, 125–128 and 136–139) in the BmR1 protein.

Surface representation of built average MD structure with predicted epitopes and binding sites is depicted in Figure 4A. Predicted linear epitopes (by Ellipro [32], FBCPred [33], AAP [34], BCPred [33] and Bepipred [35]) were overlapped with predicted conformational epitopes (by Ellipro [32] and DiscoTope-2.0 [36]). Based on these analyses, at least five servers predicted that residues 37–49, 104–112 and 193–197 (Figure 4B) are epitopes. Although epitope prediction for linear and conformational epitopes showed comparable results, the sequence 193–197 was not selected as potential epitopes as there are only 5 residues [37]. Protein binding site prediction of BmR1 protein by ProBis [38] showed structurally conserved sequences, located at 45–48, 70–73 and 126–143. Based on the result from COACH server, residues 18, 66, 72, 76, 126, 128, 130, 133, 135, 138–140, 143, 147 and 149 are predicted as potential protein-ligand binding sites. Therefore, regions 125–148 was taken into consideration as a potential binding site. As a conclusion, we report three potential epitopes (sequences 37–49, 104–112 and 125–148).

Ijms 15 11082 g004 200
Figure 4. (A) Surface representation of built structure of average MD BmR1 structure with predicted potential epitopes (residues 37–49, 104–112 and 125–148); (B) Predicted epitopes from Ellipro [32], DiscoTope-2.0 [36], FBCPred [33], AAP [34], BCPred [33] and Bepipred [35]. Ellipro-C represents predicted conformational epitopes and Ellipro-L represents predicted liner epitopes.

Click here to enlarge figure

Figure 4. (A) Surface representation of built structure of average MD BmR1 structure with predicted potential epitopes (residues 37–49, 104–112 and 125–148); (B) Predicted epitopes from Ellipro [32], DiscoTope-2.0 [36], FBCPred [33], AAP [34], BCPred [33] and Bepipred [35]. Ellipro-C represents predicted conformational epitopes and Ellipro-L represents predicted liner epitopes.
Ijms 15 11082 g004 1024

2.2. Discussion

Bm17DIII gene product of B. malayi in this study is the BmR1 protein with 206 amino acids. The structure prediction for BmR1 is challenging for several reasons. First, no putative conserved domains and functional annotation were identified. Conserved domains are important for elucidating the protein’s function; Secondly, the sequence identity of the BmR1 protein with available structures in PDB is less than 30%. This increases the probability of errors in predicted models, such as errors in side-chain packing, distortions and shifts in correctly aligned regions, errors in regions without a template, errors due to misalignment and incorrect templates [39].

Protein structures can be modeled via comparative method (both comparative modeling and threading) that depends on known protein structures or by ab initio, that relies on amino acid sequences. To date, comparative modeling is the most successful and accurate method as evolutionarily related proteins usually share a similar structure (sequence identity > 30%) [40,41]. However, searching for homologous proteins is difficult when the sequence identity is low or known as the “twilight-zone”, where the sequence identity falls between 10%–30% [42]. When confronted with this problem, threading or ab initio is an alternate method to obtain the protein structure [43]. In this study, comparative and ab initio methods were performed in order to obtain the most accurate structure. For comparative modeling, a total of 250 structures were generated by MODELLER 9v9 [10]. After subsequent steps of secondary structure restraints and loop refinement, the best model was selected based on high DOPE score and low MODELLER objective function (molpdf) as lower values of the molpdf indicate more accurate models. The molpdf measures how well the model satisfies the input spatial restraints [44]. The optimized model from comparative modeling (MODELLER 9v9 [10]) was evaluated (PROCHECK Ramachandran plot [17], VERIFY3D [18] and ERRAT [19]). A Ramachandran plot showed that even though the residues in the most favorable region are 87.3%, the VERIFY3D (16.43%) and ERRAT score (58.6%) were low (Table 3). VERIFY3D score of a satisfactory predicted model is expected to have score more than 80% and a value of ~95% ERRAT score indicates high resolution (approximately 2.5–3.0 Å) [19,45]. Due to the unsatisfactory evaluation result, ab initio approach by Rosetta and nine web servers were employed to obtain the structure for BmR1 protein. For automated protein structure prediction, SWISS-MODEL [46], 3D-JIGSAW [47], ESyPred3D [48], Geno3D [49] failed as there is no suitable templates which are similar to known structure. Out of nine servers, only CPHmodels 3.0 [20], QUARK [11], Robetta [12,13], I-TASSER [15] and Bhageerath [16] were able to predict the structure of BmR1. Evaluation results for the structures obtained were compared (Table 3).

The percentage of residues in the allowed regions was expected to be more than 90% for a good model. Results from a Ramachandran plot showed more than 90% of the residues built from Robetta, and Rosetta and QUARK were in the most favorable region. Structure obtained by Rosetta showed that 98.5% of the residues were in allowed regions and none were in the disallowed region in the Ramachandran plot. Even though the VERIFY3D score of Rosetta (71.01%) is slightly lower than the model by Robetta (73.91%) and QUARK (83.57%), the overall quality factor (ERRAT) value is the highest (97%) and is within the accepted range. Therefore, the structure built by Rosetta was the best amongst others (Table 3). The G-factor score, which indicates the overall normality of a model, obtained from Rosetta is 0.48. This shows that the molecular geometry of the structure is stereochemically reasonable. The overall quality factor (ERRAT) value is 97% thus concluding that the built model would have a resolution of not more than 3 Å. A VERIFY3D value of 71.0% showed that the quality of predicted structure was expected to be satisfactory. Thus, the overall model validation showed that the structure predicted by Rosetta was reasonable.

When the sequence identity between two proteins is less than 30%, it is difficult to discriminate between related or non-related protein. In this case, the secondary structure of the protein could provide valuable information for the detection of related protein 3D structure. Secondary protein structure is important in analysing tertiary protein structure predictions as it represents the local conformation of amino acids into regular structures. Proteins with low sequence similarity are also likely to have higher similarity in their secondary structure information if they are from the same structural class [50]. Thus, secondary protein structure is utilized as a feature for tertiary protein structure prediction [51,52]. Secondary protein structure predictions based on protein sequence by PSIPRED [5], Jpred3 [6], SSpro 4.0 [7] and PORTER [8] showed in Table 1 were compared to the secondary protein structure calculation using STRIDE [9]. Secondary structure calculation of all the predicted structures (QUARK [11], Robetta [12,13], Rosetta [14], Bhageerath [16], I-TASSER [15]) was compared in Table 2. CPHmodels 3.0 [20] was not included as the protein generated was less than 206 amino acids. All predicted protein structures have similarity in their secondary structure except for structure predicted using Bhageerath. In this study, results from secondary structure prediction of the BmR1 sequence and the STRIDE calculation of the average MD structure showed that the protein consists of 9 helices. Beta sheet was not detected for both the starting structure and average MD structure may be due to the lower reliability value of prediction accuracy for the beta sheet. The reliability value obtained for the beta sheet is within the scores of 5 or lower (the value ranges from 0–10 with higher value indicating better reliability).

Based on the secondary protein structure, the 25 kDa BmR1 protein is a helix-rich protein. Fatty acid and retinol (Vitamin A)-binding (FAR) protein, which appear to be confined to nematodes, are relatively small in size (~20 kDa) and rich in alpha-helices [53]. The similarities showed a potential relationship between BmR1 protein and FAR protein. In addition, multiple sequence alignment (MSA) of BmR1 protein with 4 different parasitic FAR proteins of nematodes by T-Coffee [54] showed approximately 28% similarity. Brugia malayi (Bm-FAR-1), Brugia pahangi (Bp-FAR-1), Wuchereria bancrofti (Wb-FAR-1) and Loa loa (Ll-FAR-1) were chosen for MSA as a distance-based analysis of FAR protein. Garofalo et al., [53] showed that the FAR proteins were from two main clusters but only FAR proteins from the same cluster as B. malayi was further studied here. Studies suggested that the FAR protein may play a crucial role in the life cycle, development and reproduction of nematodes and infection [55,56]. The FAR protein scavenges fatty acids and retinols from the host for the survival of the parasite [55,56]. Nematodes require fatty acids and retinol for lipid biosynthesis and assembly of macromolecular structures. However, they are unable to synthesize those metabolites by themselves thus making this a logical possibility for survival [57]. FAR protein not only helps nematodes in obtaining lipid from its host but also to infect the host and inhibit host defense mechanism [55]. The suggested relationship of BmR1 protein and FAR protein remains relevant as there is no data showing the function of BmR1 protein yet. In 2008, Moreno and Geary analysed the excretory-secretory products (ESP) of adult female, adult male and microfilariae (Mf) of B. malayi [58]. A total of 76 proteins in Mf, including recombinant antigen R1 (similar in sequence to BmR1) were analysed. From the analysis with Blast2GO [59] recombinant antigen R1 remained as a protein with no annotated function but was the most abundantly expressed in Mf. A year later, Bennuru and colleagues also failed to resolve the function of recombinant antigen R1 [60]. The function of BmR1 remain unknown in our analysis with Blast2GO (data not shown).

Protein function analysis from ProFunc [27] showed 23% of similarity to 3KNT and 3FHF, which are classified as N-glycosylase. There are two N-glycosylation sites detected in BmR1 protein at residue 59–62 and 105–108. Some FAR proteins have a casein kinase II phosphorylation site [56,61]. Analysis results from PDBeMotif [30] and Motif Scan [31] showed that BmR1 protein contains a site (residues 61–64) that is conserved in known FAR proteins [62].

The structure from Rosetta was used as a starting structure for a 5 ns MD simulation in the presence of water. In order to evaluate the stability of the built BmR1 protein, the dynamics of the protein was studied. RMSF analysis from MD trajectories showed that the most flexible residues are located at both C- and N-terminal of the protein. The radius of gyration and RMSD of the protein area plateaued after 1000 ps, showing no significant changes to the overall structure. The average structure with a MD simulation time of 1001 to 5000 ps was evaluated, showing good quality secondary structure and overall packing (Figure 1).

Regions predicted by at least five out of seven servers were selected as epitopes. Epitope prediction was carried out to identify the binding site of an antigen, which is usually located on the surface, loops and turns of an antigen. In 1986, Novotny and colleagues proposed that surface exposure of protein was the reason for the contact with antigen- combining sites [63]. In the same year, Barlow and colleagues found a good correlation between epitopes and protein regions protruding from protein’s globular surface [64]. Residues must be located on the surface of antigens in order to be recognized and accessible for interaction by antibodies [65]. Thus, surface exposure and accessibility was taken into consideration for epitope prediction (e.g., CEP [65], DiscoTope-2.0 [36], LEPS [66]). Surface accessibility was considered when predicting potential epitopes in both Ellipro [32] and DiscoTope-2.0 [36]. For Ellipro [32], a protrusion index (PI, Table S1) was given to each residue. PI was defined as percentage of protein atoms enclosed in the ellipsoid. Regions with high protrusion index values were identified as potential conformational epitopes [32]. In DiscoTope-2.0 [36], overall prediction scores which included surface measures were analysed. Linear and conformational epitope predictions were also carried out here. However, the final selection of potential epitopes depends on the number of residues to form an epitope. The sequence 193–197 was not taken into consideration as it contains only 5 residues. A study showed that the total number of amino acid residues per epitope ranged from 9 to 22 residues for antibodies [37]. The predicted epitope of sequence 37–49 also included residues 45–48 which were predicted as the structurally conserved residues and location of putative binding sites by ProBis [38] making it a sound predicted epitope. ProBis and COACH predicted sequence 125–148 as a potential epitope. From these data we suggest 3 potential epitopes sequences 37–49, 104–112 and 125–148. These epitopes may thus lead to the generation of designer antibodies specific to BmR1 protein.

3. Methods

3.1. Sequence Analysis

The amino acid sequence of BmR1 antigen was retrieved from GenBank (accession number AF225296). ProtParam from Expert Protein Analysis System (ExPASy) Proteomics Server [31] was implemented to calculate the protein molecular weight. The protein has a total of 206 amino acids and was subjected to Basic Local Alignment Search Tool (BLAST) on the NCBI server against non-redundant protein sequences to determine protein family. Templates identification for comparative modeling was performed using protein-protein BLAST (BLASTp) against RCSB Protein Data Bank (PDB) with default parameters. Secondary structure prediction on BmR1 was performed by PSIPRED [5], Jpred3 [6], SSpro 4.0 [7] and PORTER [8]. Function annotation and identification of the conserved domain of BmR1 were carried out using databases such as Conserved Domain Search Service (CD Search) [67,68], InterProScan [69], SMART [70] and Proteins Families database (Pfam) [71]. For protein functional analysis, ProFunc [27], STRING [25], PROSITE [26], PDBeMotif [30] and Motif Scan from ExPASy [31] were used. Linear epitope prediction on BmR1 protein sequence was carried out by FBCPred [33], AAP [34], BCPred [33] and Bepipred [35]. T-Coffee (Tree-based consistency objective function for alignment evaluation) [54] was used for multiple sequence alignment.

3.2. Structural Prediction and Evaluation

The structure of BmR1 was modelled via comparative modeling, threading and ab initio approaches. In comparative modeling by MODELLER 9v9 [10], a total of 250 initial models were generated from multiple templates (PDB id: 2G3Y, 2IE8 [72], 3QOE [73], 1V32, 2E87, 2R3V [74] and 3I4Q), followed by secondary structure restraints and loop refinement using MODELLER 9v9 [10]. DOPE score and molpdf were used to evaluate the models. Best structure with high DOPE score and low molPDF was chosen. Protein prediction by ab initio approach was carried out by Rosetta [14]. Fragment libraries of three- and nine-residue were generated by Robetta fragment server [12] and those models were assembled by fragment insertion [75]. Five hundred initial structures were created using AbinitioRelax command with the three- and nine-residue fragments as input. AbinitioRelax is the combination of ab initio folding and refinement by Rosetta full-atom force field (Relax) [14]. Automated server (SWISS-MODEL [46], 3D-JIGSAW [47], ESyPred3D [48], Geno3D [49], CPHmodels 3.0 [20] QUARK [11], Bhageerath [16], Robetta [12,13] and I-TASSER [15]) with default parameters were also employed to predict the structure of BmR1.

All predicted structures were sent for secondary structure calculation using STRIDE [9] and were evaluated with PROCHECK Ramachandran plot [17], VERIFY3D [18] and ERRAT [19]. Ramachandran plot was obtained for backbone conformation evaluation [17]. VERIFY3D was used to determine the compatibility of an atomic model (3D) with its own amino acid sequence (1D) [18]. A higher score indicates high quality of a structure. ERRAT is to analyse the statistics of non-bonded interactions between different atom types.

3.3. Minimization and Molecular Dynamics Simulation

The protein model with the best validation value obtained from protein structure prediction subsequently underwent energy minimization by the Sander module from AMBER11 [76] with AMBER ff03 force field. The protein was solvated with TIP3P water in a truncated octahedron periodic box with 10 Å distance from the edge of the box. The solvated system was neutralized by 19 sodium ions (Na+). The system has a total of 37,223 atoms. The protein was restrained during the first stage of minimization with 300 kcal/mol restraint force. For stage 2 of the minimization, the entire system was minimized. The temperature of the system was gradually heated to 300 K over 20 ps. The system was subsequently equilibrated at 300 K over 60 ps during the NVT equilibration. Temperature was controlled by Langevin thermostat [77]. Finally, a total of 5000 ps MD simulation at 300 K and 1 atm was carried out. SHAKE algorithm [78] was turned on throughout the MD simulation to constrain bonds involving hydrogen.

Secondary structure calculation of the average MD structure was performed by STRIDE [9]. The structure was validated through PROCHECK Ramachandran Plot [17], VERIFY3D [18], ERRAT [19], Prosa II Z-score [21] and ANOLEA [22]. Conformational epitope prediction by Ellipro [32] and DiscoTope-2.0 [36] was performed on the average structure from MD simulation. The protein binding site of the BmR1 protein was predicted by ProBis [38] and COACH [79].

4. Conclusions

In this study, BmR1 structure predicted via ab initio method (Rosetta) produced a quality and reliable structure. Furthermore, the average structure obtained from molecular dynamics simulation also showed overall good secondary and 3D packing. A total of three potential epitopes were identified leading to the possibility of future designer antigen-based detection test and -specific binders capable for therapy using BmR1 as the targeted antigen.

Supplementary Files

  • Supplementary File 1:

    Supplementary Information (PDF, 561 KB)

  • Acknowledgments

    This work was supported by Exploratory Research Grant Scheme (ERGS) (203/CIPPM/6730058) and Higher Institutions Centre of Excellence (HICoE) Grant (311/CIPPM/44001005) from the Malaysia Ministry of Higher Education. Thanks also due to MyBrain from Ministry of Education for the scholarship for Bee Yin Khor.

    Author Contributions

    Bee Yin Khor performed the work and wrote the manuscript. Gee Jun Tye, Theam Soon Lim and Rahmah Noordin co-wrote the manuscript. Yee Siew Choong designed the overall research plan and co-wrote the manuscript. All authors read and approved the final manuscript.

    Conflicts of Interest

    The authors declare no conflict of interest.

    References

    1. World Health Organization. Global programme to eliminate lymphatic filariasis. Wkly. Epidemiol. Rec. 2011, 86, 518–588.
    2. Noordin, R.; Aziz, R.; Ravindran, B. Homologs of the Brugia malayi diagnostic antigen BmR1 are present in other filarial parasites but induce different humoral immune responses. Filaria J. 2004, 3, 10, doi:10.1186/1475-2883-3-10.
    3. Noordin, R.; Itoh, M.; Kimura, E.; Abdul Rahman, R.; Ravindran, B.; Mahmud, R.; Supali, T.; Weerasooriya, M. Multicentre evaluations of two new rapid IgG4 tests (WB rapid and panLF rapid) for detection of lymphatic filariasis. Filaria J. 2007, 6, 9, doi:10.1186/1475-2883-6-9.
    4. Abdul Rahman, R.; Hwen-Yee, C.; Noordin, R. Pan LF-ELISA using BmR1 and BmSXP recombinant antigens for detection of lymphatic filariasis. Filaria J. 2007, 6, 10, doi:10.1186/1475-2883-6-10.
    5. Buchan, D.W.A.; Ward, S.M.; Lobley, A.E.; Nugent, T.C.O.; Bryson, K.; Jones, D.T. Protein annotation and modelling servers at University College London. Nucleic Acids Res. 2010, 38, W563–W568, doi:10.1093/nar/gkq427.
    6. Cole, C.; Barber, J.D.; Barton, G.J. The Jpred 3 secondary structure prediction server. Nucleic Acids Res. 2008, 36, W197–W201, doi:10.1093/nar/gkn238.
    7. Cheng, J.; Randall, A.Z.; Sweredoski, M.J.; Baldi, P. SCRATCH: A protein structure and structural feature prediction server. Nucleic Acids Res. 2005, 33, W72–W76, doi:10.1093/nar/gki396.
    8. Pollastri, G.; McLysaght, A. Porter: A new, accurate server for protein secondary structure prediction. Bioinformatics 2005, 21, 1719–1720, doi:10.1093/bioinformatics/bti203.
    9. Heinig, M.; Frishman, D. STRIDE: A web server for secondary structure assignment from known atomic coordinates of proteins. Nucleic Acids Res. 2004, 32, W500–W502, doi:10.1093/nar/gkh429.
    10. Martí-Renom, M.A.; Stuart, A.C.; Fiser, A.; Sánchez, R.; Melo, F.; Šali, A. Comparative protein structure modeling of genes and genomes. Annu. Rev. Biophys. Biomol. Struct. 2000, 29, 291–325, doi:10.1146/annurev.biophys.29.1.291.
    11. Xu, D.; Zhang, J.; Roy, A.; Zhang, Y. Automated protein structure modeling in CASP9 by I-TASSER pipeline combined with QUARK-based ab initio folding and FG-MD-based structure refinement. Proteins Struct. Funct. Bioinform. 2011, 79 (Suppl. S10), 147–160, doi:10.1002/prot.23111.
    12. Chivian, D.; Kim, D.E.; Malmström, L.; Bradley, P.; Robertson, T.; Murphy, P.; Strauss, C.E.M.; Bonneau, R.; Rohl, C.A.; Baker, D. Automated prediction of CASP-5 structures using the Robetta server. Proteins Struct. Funct. Bioinform. 2003, 53, 524–533, doi:10.1002/prot.10529.
    13. Chivian, D.; Kim, D.E.; Malmström, L.; Schonbrun, J.; Rohl, C.A.; Baker, D. Prediction of CASP6 structures using automated Robetta protocols. Proteins Struct. Funct. Bioinform. 2005, 61, 157–166, doi:10.1002/prot.20733.
    14. Bonneau, R.; Tsai, J.; Ruczinski, I.; Chivian, D.; Rohl, C.; Strauss, C.E.M.; Baker, D. Rosetta in CASP4: Progress in ab initio protein structure prediction. Proteins Struct. Funct. Bioinform. 2001, 45, 119–126, doi:10.1002/prot.1170.
    15. Zhang, Y. I-TASSER server for protein 3D structure prediction. BMC Bioinform. 2008, 9, 40, doi:10.1186/1471-2105-9-40.
    16. Jayaram, B.; Bhushan, K.; Shenoy, S.R.; Narang, P.; Bose, S.; Agrawal, P.; Sahu, D.; Pandey, V. Bhageerath: An energy based web enabled computer software suite for limiting the search space of tertiary structures of small globular proteins. Nucleic Acids Res. 2006, 34, 6195–6204, doi:10.1093/nar/gkl789.
    17. Laskowski, R.A.; MacArthur, M.W.; Moss, D.S.; Thornton, J.M. PROCHECK: A program to check the stereochemical quality of protein structures. J. Appl. Crystallogr. 1993, 26, 283–291, doi:10.1107/S0021889892009944.
    18. Eisenberg, D.; Lüthy, R.; Bowie, J.U. VERIFY3D: Assessment of protein models with three-dimensional profiles. Methods Enzymol. 1997, 277, 396–404, doi:10.1016/S0076-6879(97)77022-8.
    19. Colovos, C.; Yeates, T.O. Verification of protein structures: Patterns of nonbonded atomic interactions. Protein Sci. 1993, 2, 1511–1519, doi:10.1002/pro.5560020916.
    20. Nielsen, M.; Lundegaard, C.; Lund, O.; Petersen, T.N. CPHmodels-3.0—Remote homology modeling using structure-guided sequence profiles. Nucleic Acids Res. 2010, 38, W576–W581, doi:10.1093/nar/gkq535.
    21. Wiederstein, M.; Sippl, M.J. ProSA-web: Interactive web service for the recognition of errors in three-dimensional structures of proteins. Nucleic Acids Res. 2007, 35, W407–W410, doi:10.1093/nar/gkm290.
    22. Melo, F.; Feytmans, E. Assessing protein structures with a non-local atomic interaction energy. J. Mol. Biol. 1998, 277, 1141–1152, doi:10.1006/jmbi.1998.1665.
    23. Bordoli, L.; Kiefer, F.; Arnold, K.; Benkert, P.; Battey, J.; Schwede, T. Protein structure homology modeling using SWISS-MODEL workspace. Nat. Protoc. 2008, 4, 1–13, doi:10.1038/nprot.2008.197.
    24. Offredi, F.; Dubail, F.; Kischel, P.; Sarinski, K.; Stern, A.S.; van de Weerdt, C.; Hoch, J.C.; Prosperi, C.; François, J.M.; Mayo, S.L.; et al. De novo backbone and sequence design of an idealized α/β-barrel protein: Evidence of stable tertiary structure. J. Mol. Biol. 2003, 325, 163–174, doi:10.1016/S0022-2836(02)01206-8.
    25. Von Mering, C.; Jensen, L.J.; Snel, B.; Hooper, S.D.; Krupp, M.; Foglierini, M.; Jouffre, N.; Huynen, M.A.; Bork, P. STRING: Known and predicted protein–protein associations, integrated and transferred across organisms. Nucleic Acids Res. 2005, 33, D433–D437.
    26. Sigrist, C.J.A.; de Castro, E.; Cerutti, L.; Cuche, B.A.; Hulo, N.; Bridge, A.; Bougueleret, L.; Xenarios, I. New and continuing developments at PROSITE. Nucleic Acids Res. 2013, 41, D344–D347, doi:10.1093/nar/gks1067.
    27. Laskowski, R.A.; Watson, J.D.; Thornton, J.M. ProFunc: A server for predicting protein function from 3D structure. Nucleic Acids Res. 2005, 33, W89–W93, doi:10.1093/nar/gki414.
    28. Faucher, F.; Wallace, S.S.; Doublié, S. The C-terminal lysine of Ogg2 DNA glycosylases is a major molecular determinant for guanine/8-oxoguanine distinction. J. Mol. Biol. 2010, 397, 46–56, doi:10.1016/j.jmb.2010.01.024.
    29. Faucher, F.; Duclos, S.; Bandaru, V.; Wallace, S.S.; Doublié, S. Crystal structures of two archaeal 8-oxoguanine DNA glycosylases provide structural insight into guanine/8-oxoguanine distinction. Structure 2009, 17, 703–712, doi:10.1016/j.str.2009.03.007.
    30. Golovin, A.; Henrick, K. MSDmotif: Exploring protein sites and motifs. BMC Bioinform. 2008, 9, 312, doi:10.1186/1471-2105-9-312.
    31. Gasteiger, E.; Gattiker, A.; Hoogland, C.; Ivanyi, I.; Appel, R.D.; Bairoch, A. ExPASy: The proteomics server for in-depth protein knowledge and analysis. Nucleic Acids Res. 2003, 31, 3784–3788, doi:10.1093/nar/gkg563.
    32. Ponomarenko, J.; Bui, H.-H.; Li, W.; Fusseder, N.; Bourne, P.; Sette, A.; Peters, B. ElliPro: A new structure-based tool for the prediction of antibody epitopes. BMC Bioinform. 2008, 9, 514, doi:10.1186/1471-2105-9-514.
    33. Yasser, E.-M.; Dobbs, D.; Honavar, V. Predicting flexible length linear B-cell epitopes. Comput. Syst. Bioinform. Conf. 2008, 7, 121–132.
    34. Chen, J.; Liu, H.; Yang, J.; Chou, K.C. Prediction of linear B-cell epitopes using amino acid pair antigenicity scale. Amino Acids 2007, 33, 423–428, doi:10.1007/s00726-006-0485-9.
    35. Larsen, J.; Lund, O.; Nielsen, M. Improved method for predicting linear B-cell epitopes. Immunome Res. 2006, 2, 2, doi:10.1186/1745-7580-2-2.
    36. Kringelum, J.V.; Lundegaard, C.; Lund, O.; Nielsen, M. Reliable B cell epitope predictions: Impacts of method development and improved benchmarking. PLoS Comput. Biol. 2012, 8, e1002829, doi:10.1371/journal.pcbi.1002829.
    37. Haste Andersen, P.; Nielsen, M.; Lund, O. Prediction of residues in discontinuous B-cell epitopes using protein 3D structures. Protein Sci. 2006, 15, 2558–2567, doi:10.1110/ps.062405906.
    38. Konc, J.; Janežič, D. ProBiS-2012: Web server and web services for detection of structurally similar binding sites in proteins. Nucleic Acids Res. 2012, 40, W214–W221, doi:10.1093/nar/gks435.
    39. Eswar, N.; Webb, B.; Marti-Renom, M.A.; Madhusudhan, M.S.; Eramian, D.; Shen, M.; Pieper, U.; Sali, A. Comparative protein structure modeling using Modeller. In Current Protocols in Bioinformatics; John Wiley & Sons, Inc.: Hoboken, NJ, USA, 2002.
    40. Choong, Y.S.; Lim, T.S.; Chew, A.L.; Aziah, I.; Ismail, A. Structural and functional studies of a 50 kda antigenic protein from Salmonella enterica serovar Typhi. J. Mol. Graph. Model. 2011, 29, 834–842, doi:10.1016/j.jmgm.2011.01.008.
    41. Errami, M.; Geourjon, C.; Deléage, G. Detection of unrelated proteins in sequences multiple alignments by using predicted secondary structures. Bioinformatics 2003, 19, 506–512, doi:10.1093/bioinformatics/btg016.
    42. Hansen, S.F.; Bettler, E.; Wimmerová, M.; Imberty, A.; Lerouxel, O.; Breton, C. Combination of several bioinformatics approaches for the identification of new putative glycosyltransferases in arabidopsis. J. Proteome Res. 2008, 8, 743–753.
    43. Moult, J.; Fidelis, K.; Kryshtafovych, A.; Tramontano, A. Critical assessment of methods of protein structure prediction (CASP)—Round ix. Proteins Struct. Funct. Bioinform. 2011, 79, 1–5.
    44. Eswar, N.; Eramian, D.; Webb, B.; Shen, M.-Y.; Sali, A. Protein structure modeling with MODELLER. In Structural Proteomics; Kobe, B., Guss, M., Huber, T., Eds.; Humana Press: New York City, NY, USA, 2008; Volume 426, pp. 145–159.
    45. Barh, D.; Barve, N.; Gupta, K.; Chandra, S.; Jain, N.; Tiwari, S.; Leon-Sicairos, N.; Canizalez-Roman, A.; Rodrigues dos Santos, A.; Hassan, S.S.; et al. Exoproteome and secretome derived broad spectrum novel drug and vaccine candidates in Vibrio cholerae targeted by piper betel/derived compounds. PLoS One 2013, 8, e52773, doi:10.1371/journal.pone.0052773.
    46. Arnold, K.; Bordoli, L.; Kopp, J.; Schwede, T. The SWISS-MODEL workspace: A web-based environment for protein structure homology modelling. Bioinformatics 2006, 22, 195–201, doi:10.1093/bioinformatics/bti770.
    47. Bates, P.A.; Kelley, L.A.; MacCallum, R.M.; Sternberg, M.J.E. Enhancement of protein modeling by human intervention in applying the automatic programs 3D-JIGSAW and 3D-PSSM. Proteins Struct. Funct. Bioinform. 2001, 45, 39–46, doi:10.1002/prot.1168.
    48. Lambert, C.; Léonard, N.; de Bolle, X.; Depiereux, E. ESyPred3D: Prediction of proteins 3D structures. Bioinformatics 2002, 18, 1250–1256, doi:10.1093/bioinformatics/18.9.1250.
    49. Combet, C.; Jambon, M.; Deléage, G.; Geourjon, C. Geno3D: Automatic comparative molecular modelling of protein. Bioinformatics 2002, 18, 213–214, doi:10.1093/bioinformatics/18.1.213.
    50. Ahmadi Adl, A.; Nowzari-Dalini, A.; Xue, B.; Uversky, V.N.; Qian, X. Accurate prediction of protein structural classes using functional domains and predicted secondary structure sequences. J. Biomol. Struct. Dyn. 2012, 29, 1127–1137, doi:10.1080/07391102.2011.672626.
    51. Lin, H.-N.; Sung, T.-Y.; Ho, S.-Y.; Hsu, W.-L. Improving protein secondary structure prediction based on short subsequences with local structure similarity. BMC Genomics 2010, 11, S4.
    52. Lee, L.; Leopold, J.L.; Frank, R.L. In Protein secondary structure prediction using blast and exhaustive rt-rico, the search for optimal segment length and threshold. In Proceedings of the 2012 IEEE Symposium on Computational Intelligence in Bioinformatics and Computational Biology (CIBCB), San Diego, CA, USA, 9–12 May 2012; pp. 35–42.
    53. Garofalo, A.; Kläger, S.L.; Rowlinson, M.-C.; Nirmalan, N.; Klion, A.; Allen, J.E.; Kennedy, M.W.; Bradley, J.E. The FAR proteins of filarial nematodes: Secretion, glycosylation and lipid binding characteristics. Mol. Biochem. Parasitol. 2002, 122, 161–170, doi:10.1016/S0166-6851(02)00097-X.
    54. Notredame, C.; Higgins, D.G.; Heringa, J. T-Coffee: A novel method for fast and accurate multiple sequence alignment. J. Mol. Biol. 2000, 302, 205–217, doi:10.1006/jmbi.2000.4042.
    55. Cheng, X.; Xiang, Y.; Xie, H.; Xu, C.-L.; Xie, T.-F.; Zhang, C.; Li, Y. Molecular characterization and functions of fatty acid and retinoid binding protein gene (Ab-far-1) in Aphelenchoides besseyi. PLoS One 2013, 8, e66011, doi:10.1371/journal.pone.0066011.
    56. Jennifer, L.B.; Amber, E.F. The potential role of binding proteins in human parasitic infections: An in-depth look at the novel family of nematode-specific fatty acid and retinol binding proteins. In Binding Protein; Abdelmohsen, K., Ed.; InTech: Rijeka, Croatia, 2012.
    57. Fairfax, K.C.; Vermeire, J.J.; Harrison, L.M.; Bungiro, R.D.; Grant, W.; Husain, S.Z.; Cappello, M. Characterisation of a fatty acid and retinol binding protein orthologue from the hookworm Ancylostoma ceylanicum. Int. J. Parasitol. 2009, 39, 1561–1571, doi:10.1016/j.ijpara.2009.06.005.
    58. Moreno, Y.; Geary, T.G. Stage- and gender-specific proteomic analysis of Brugia malayi excretory-secretory products. PLoS Negl. Trop. Dis. 2008, 2, e326, doi:10.1371/journal.pntd.0000326.
    59. Conesa, A.; Götz, S.; García-Gómez, J.M.; Terol, J.; Talón, M.; Robles, M. Blast2GO: A universal tool for annotation, visualization and analysis in functional genomics research. Bioinformatics 2005, 21, 3674–3676, doi:10.1093/bioinformatics/bti610.
    60. Bennuru, S.; Semnani, R.; Meng, Z.; Ribeiro, J.M.C.; Veenstra, T.D.; Nutman, T.B. Brugia malayi excreted/secreted proteins at the host/parasite interface: Stage- and gender-specific proteomic profiling. PLoS Negl. Trop. Dis. 2009, 3, e410, doi:10.1371/journal.pntd.0000410.
    61. Jordanova, R.; Groves, M.R.; Kostova, E.; Woltersdorf, C.; Liebau, E.; Tucker, P.A. Fatty acid- and retinoid-binding proteins have distinct binding pockets for the two types of cargo. J. Biol. Chem. 2009, 284, 35818–35826, doi:10.1074/jbc.M109.022731.
    62. Kennedy, M.W.; Garside, L.H.; Goodrick, L.E.; McDermott, L.; Brass, A.; Price, N.C.; Kelly, S.M.; Cooper, A.; Bradley, J.E. The Ov20 protein of the parasitic nematode Onchocerca volvulus: A structurally novel class of small helix-rich retinol-binding proteins. J. Biol. Chem. 1997, 272, 29442–29448, doi:10.1074/jbc.272.47.29442.
    63. Novotný, J.; Handschumacher, M.; Haber, E.; Bruccoleri, R.E.; Carlson, W.B.; Fanning, D.W.; Smith, J.A.; Rose, G.D. Antigenic determinants in proteins coincide with surface regions accessible to large probes (antibody domains). Proc. Natl. Acad. Sci. USA 1986, 83, 226–230, doi:10.1073/pnas.83.2.226.
    64. Barlow, D.J.; Edwards, M.S.; Thornton, J.M. Continuous and discontinuous protein antigenic determinants. Nature 1986, 322, 747–748, doi:10.1038/322747a0.
    65. Kulkarni-Kale, U.; Bhosle, S.; Kolaskar, A.S. CEP: A conformational epitope prediction server. Nucleic Acids Res. 2005, 33, W168–W171, doi:10.1093/nar/gki460.
    66. Wang, H.-W.; Lin, Y.-C.; Pai, T.-W.; Chang, H.-T. Prediction of B-cell linear epitopes with a combination of support vector machine classification and amino acid propensity identification. J. Biomed. Biotechnol. 2011, 2011, 12.
    67. Marchler-Bauer, A.; Bryant, S.H. CD-search: Protein domain annotations on the fly. Nucleic Acids Res. 2004, 32, W327–W331, doi:10.1093/nar/gkh454.
    68. Marchler-Bauer, A.; Lu, S.; Anderson, J.B.; Chitsaz, F.; Derbyshire, M.K.; DeWeese-Scott, C.; Fong, J.H.; Geer, L.Y.; Geer, R.C.; Gonzales, N.R.; et al. CDD: A conserved domain database for the functional annotation of proteins. Nucleic Acids Res. 2011, 39, D225–D229, doi:10.1093/nar/gkq1189.
    69. Quevillon, E.; Silventoinen, V.; Pillai, S.; Harte, N.; Mulder, N.; Apweiler, R.; Lopez, R. InterProScan: Protein domains identifier. Nucleic Acids Res. 2005, 33, W116–W120, doi:10.1093/nar/gki442.
    70. Letunic, I.; Doerks, T.; Bork, P. Smart 6: Recent updates and new developments. Nucleic Acids Res. 2009, 37, D229–D232, doi:10.1093/nar/gkn808.
    71. Finn, R.D.; Mistry, J.; Tate, J.; Coggill, P.; Heger, A.; Pollington, J.E.; Gavin, O.L.; Gunasekaran, P.; Ceric, G.; Forslund, K.; et al. The Pfam protein families database. Nucleic Acids Res. 2010, 38, D211–D222, doi:10.1093/nar/gkp985.
    72. Lee, J.H.; Im, Y.J.; Bae, J.; Kim, D.; Kim, M.-K.; Kang, G.B.; Lee, D.-S.; Eom, S.H. Crystal structure of Thermus caldophilus phosphoglycerate kinase in the open conformation. Biochem. Biophys. Res. Commun. 2006, 350, 1044–1049, doi:10.1016/j.bbrc.2006.09.151.
    73. Kim, Y.; Joachimiak, G.; Ye, Z.; Binkowski, T.A.; Zhang, R.; Gornicki, P.; Callahan, S.M.; Hess, W.R.; Haselkorn, R.; Joachimiak, A. Structure of transcription factor HetR required for heterocyst differentiation in cyanobacteria. Proc. Natl. Acad. Sci. USA 2011, 108, 10109–10114, doi:10.1073/pnas.1106840108.
    74. Fu, Z.; Voynova, N.E.; Herdendorf, T.J.; Miziorko, H.M.; Kim, J.-J.P. Biochemical and structural basis for feedback inhibition of mevalonate kinase and isoprenoid metabolism. Biochemistry 2008, 47, 3715–3724, doi:10.1021/bi7024386.
    75. Kim, D.E.; Chivian, D.; Baker, D. Protein structure prediction and analysis using the Robetta server. Nucleic Acids Res. 2004, 32, W526–W531, doi:10.1093/nar/gkh468.
    76. Salomon-Ferrer, R.; Case, D.A.; Walker, R.C. An overview of the Amber biomolecular simulation package. Wiley Interdiscip. Rev. Comput. Mol. Sci. 2013, 3, 198–210, doi:10.1002/wcms.1121.
    77. Berendsen, H.J.C.; Postma, J.P.M.; van Gunsteren, W.F.; DiNola, A.; Haak, J.R. Molecular dynamics with coupling to an external bath. J. Chem. Phys. 1984, 81, 3684–3690, doi:10.1063/1.448118.
    78. Ryckaert, J.-P.; Ciccotti, G.; Berendsen, H. Numerical integration of the cartesian equations of motion of a system with constraints: Molecular dynamics of n-alkanes. J. Comput. Phys. 1977, 23, 327–341, doi:10.1016/0021-9991(77)90098-5.
    79. Yang, J.; Roy, A.; Zhang, Y. Protein-ligand binding site recognition using complementary binding-specific substructure comparison and sequence profile alignment. Bioinformatics 2013, 29, 2588–2595, doi:10.1093/bioinformatics/btt447.
    Int. J. Mol. Sci. EISSN 1422-0067 Published by MDPI AG, Basel, Switzerland RSS E-Mail Table of Contents Alert