Molecular Modeling and Structural Analysis of Five SE Clan (S12 Family) Serine Proteases
Sumit Kumar Dey
SE clan of serine proteases have Lysine, Serine and Tyrosine residues as catalytic triad in their active site which is an exception to all other serine proteases having conserved residues Serine, Aspartic acid and Histidine in their active sites. The aim of this study was to explore the structural distinctiveness of different S12 family (D-Ala-D-Ala carboxypeptidase B family) serine peptidases from diverse groups of species belonging to SE clan, using molecular modeling techniques. Amino acid sequences of those proteases from each group of organisms, i.e., archean, protozoan, fungal, plant and human were taken from the MEROPS database in FASTA format. Homology and threading modeling approach were used to construct all the structures by SWISS MODEL and LOOPP server respectively. Structural quality assessing validation programs (PROCHECK, MODELYN, MOLPROBITY and PROSA) of the final model have demonstrated its reliability for further studies. A full structural database was constructed starting from bacteria up to human, focusing on the catalytic triad of the proteases. MODELYN study showed statistically that the region of the catalytic triad is highly conserved from bacteria to human. Phylogenetic analysis also supports their evolutionary connection. MOLMOL study showed that the catalytic sites of all the SE Clan proteases exhibit acidic regions in the surface electrostatic potential maps but a few of them contain both large patches of positive and neutral potentials. Hence some of the proteases with predominant positive and neutral regions around the catalytic site can be used as potential drug target against bacterial and protozoan pathogens.
Received: March 30, 2011;
Accepted: May 20, 2011;
Published: July 09, 2011
Proteases (proteinases, peptidases or proteolytic enzymes) are enzymes that
can split a protein into peptides and break peptide bonds between amino acids
of proteins. Proteases work by lowering the activation barrier for peptide hydrolysis
(Barrett et al., 2003; Rawlings
and Barrett, 1993). The activated state is usually a tetrahedral intermediate.
The process is called proteolytic cleavage, a common mechanism of activation
or inactivation of enzymes especially involved in blood coagulation or digestion.
Peptidases and their natural, protein inhibitors are of great relevance to biology,
medicine and biotechnology (Perona and Craik, 1995;
A clan contains all the modern-day peptidases (Barrett
et al., 2001) that have arisen from a single evolutionary origin
of peptidases. It represents one or more families that show evidence of their
evolutionary relationship by their similar tertiary structures, or when structures
are not available, by the order of catalytic-site residues in the polypeptide
chain. The MEROPS database (Rawlings et al., 2006)
aims to fulfill the need for an integrated source of information about the proteins.
The organizational principle of the database is a hierarchical classification
in which homologous sets of proteins of interest are grouped into families and
the homologous families are grouped in clans. There are different clans like,
PA/SA, SB, SC and SE clans.
In this study, SE clan is considered that contains peptidases which have specialized
for roles in bacterial cell wall metabolism, for the structure prediction and
analysis using advanced molecular modeling techniques. The homology of the three
families S11 (Englebert et al., 1994), S12 (Kelly
et al., 1985) and S13 (Sauvage et al.,
2005) of SE clan is supported by similarities in their sequence, protein
fold, catalytic mechanism and biological functions. For this study, attention
was focused mainly to the members of the S12 Family/D-Ala-D-Ala carboxypeptidase
B family, contains serine-type D-Ala-D-Ala carboxypeptidases (Silvaggi
et al., 2003). The active site residues Ser and Lys form the catalytic
dyad and are found in the motif Ser-Xaa-Thr-Lys. There is a catalytic Tyr residue
further toward the C-terminus in a conserved Tyr-Xaa-Asn motif. The tyrosine
is replaced by serine in other families of SE clan. The aim of this study was
to explore the structural characteristics of different S12 family serine peptidases
belongs to SE clan from diverse groups of the living kingdom using molecular
modeling techniques. Huge availability of protein sequences and computational
approaches for identifying target proteins/ genes has been a great support for
further experimental work (Shakyawar et al., 2011).
Conserved residues in serine proteases (Rawlings and Barrett,
1994; Krem and Cera, 2001) with a wide evolutionary
span (Rawlings and Barrett, 1993) ranging from bacteria
to mammals on the catalytic sites and their interactions are mainly highlighted
for evolutionary insight and structure based drug designing (Anand
et al., 2003; Turk, 2006; Tambunan
and Parikesit, 2010). The present study aims at exploring the structural
uniqueness of different S12 family (D-Ala-D-Ala carboxypeptidase B family) serine
peptidases from different species belonging to SE clan which have specialized
roles in bacterial cell wall metabolism, using molecular modeling techniques.
MATERIALS AND METHODS
Computational tools: In this study, the computational studies were performed
using the following software packages. The study materials were collected from
MEROPS database which aims to fulfill the need for an integrated source of information
about the peptidases. The database contains a hierarchical classification of
peptidases in which homologous sets of proteins of interest are grouped into
clans and families among kingdoms of organisms, giving concise text annotations
for each peptidase family. Homology (Floudas, 2007)
and threading modeling were carried out using SWISS-MODEL server (Schwede
et al., 2003) and LOOPP server (Meller and Elber,
2001) on PC.
SWISS-MODEL, PROSITE, LOOPP, CLUSTALW, MOLPROBITY and MOLMOL were used from
the respective web-servers available through Internet at sites:
Model construction and refinement: Amino acid sequence of SE clan-serine
protease from each of the groups of archea (Pyrococcus abyssi, No MER026262),
protozoan (Dictyostellium discoideum, MER035017), fungal (Neurospora
crassa, MER033421), plant (Arabidopsis thaliana, No. MER004158) and
human (Homo sapiens, MER017071) were taken from the MEROPS database.
The amino acid sequences of the three proteases were subjected to NCBI PSI BLAST
(Altschul et al., 1997) search in the Protein
Structure Database (PDB) for selection of suitable template for comparative
modeling. The challenge of template based modeling lies in the recognition of
correct templates and generation of accurate sequence-template alignments. When
suitable template was found to be absent in the protein BLAST search result,
the initial models of the proteases were predicted using two web-based molecular
modeling servers e.g. SWISS-MODEL server and LOOPP v4.0 and v3.0 server. Actually,
protein modeling is the only way to obtain structural information if experimental
techniques fail (Prajapat et al., 2010). The
sequences of all the predicted structures both homologous and non homologous
were aligned using the multiple alignment server, CLUSTALW and the structures
were aligned using ABGEN (Mandal et al., 1996)
to select the core structure of the stating scaffold of the target protein for
subsequent refinement. The starting structure was refined using the Insight
II 2005 of Accelrys (San Diego, CA) equipped with DISCOVER as the energy minimization
and molecular dynamics module. Structural optimization involved energy minimization
(100 steps each of steepest descent and conjugate gradient methods) using cff91
force-field followed by dynamics simulations. A typical dynamics run consisted
of 10000 steps of one femto-second (10 picoseconds) after 1000 steps of equilibration
with a conformational sampling of 1 in 10 steps at 300 K. However, dynamics
simulation of 100 picoseconds was also applied to certain bigger loops for proper
regularization. At the end of the dynamics simulation, the conformation with
lowest potential energy was picked for the next cycle of refinement using the
ANALYSIS module of Insight II. This combination of minimization and dynamics
were repeated until satisfactory conformational parameters were obtained. Each
loop was separately regularized applying position constraints to the rest of
the atoms of the protein which were two amino acids away from the desired loop
by energy minimization and molecular dynamics followed by evaluation of the
structural parameters. SCWRL 3.0 (Canutescu et al.,
2003) was used to regenerate the sidechains (Summers
and Karplus, 1991) of the modeled protein. The final structure was energy
minimized 100 steps each with steepest descent and conjugate gradient methods
keeping all the atoms of the protein free.
Structure validation: Structural parameters and quality of the final
3D structures were determined using PROCHECK (Laskowski
et al., 1993; Amir et al., 2011), MOLPROBITY
(Davis et al., 2004), PROSA (Wiederstein
and Sippl, 2007) and MODELYN (Mandal, 1998; Indian
Copyright No. 9/98).
3D structure analysis: The ribbon structure and electrostatic potential
surface of the model were determined by MOLMOL (Koradi et
Multiple sequence alignment and phylogenetic analysis of serine proteases:
Amino acid sequences of serine proteases of archea (Pyrococcus abyssi,
No. MER026262), protozoan (Dictyostellium discoideum, MER035017), fungal
(Neurospora crassa, MER033421), plant (Arabidopsis thaliana, No
MER004158) and human (Homo sapiens, MER017071) were taken from the MEROPS
database in FASTA format and subjected to multiple sequence alignment using
CLUSTALW (Thompson et al., 1994). Phylogenetic
Analysis was also done by PHYLIP 3.5 program package while distance matrix was
computed using PROTDIST with 100 BOOTSTRAP analysis. Phylogeny tree searching
was performed by KITSH program (Chatterjee et al.,
2011; Tambunan et al., 2008).
Model construction: There were X-ray structures of bacterial peptidases for this S12 of SE clan was present in the PDB; one of the X-ray structures, 1ONH.pdb (MER000463) was downloaded for structural analysis. For the rest of the groups i.e. archea, protozoa, fungi, plant and animal, no experimentally determined structures were available. Hence we selected one sequence from each group from the MEROPS database to predict their structures by molecular modeling techniques.
The sequences of the proteases were looked for significant sequence identity (>30%) suitable for homology based structure prediction, finds and compares regions of local similarity between sequences of proteins and calculates the statistical significance of matches. Homology modeling is based on the fact that a structure of a protein can be reliably modeled when its sequence is sufficiently similar to a protein sequence with known 3D structure. As the 3D structure models of target proteins are built upon its alignment to the template, 1KVM.pdb and 1CEG.pdb were considered as templates for homology modeling of archean and plant proteases respectively. Initial models were developed using SWISS-MODEL server which is a fully automated protein structure homology-modeling web server (Fig. 1). The purpose of this server is to make protein modeling accessible to all biochemists and molecular biologists worldwide.
The species of protozoa, fungi and animal as those have <30% sequence identity were considered for threading based modeling. Sequences from protozoan (Dictyostellium discoideum, MER035017), fungal (Neurospora crassa, MER033421) human (Homo sapiens, MER017071) proteases were submitted to the LOOPP server for threading based structure prediction and the search results for secondary structures of the target sequence has the best match with the X-ray structure PDB IDs: 1RGZ, 1SDE and 1CI9, respectively. The percent sequence identities of the templates are 18.96, 25.15 and 18.29%, respectively which shows that they are fit for threading models. The template modeling match scores and the extent of sequence coverage are given in Table 1. Template modeling match score is a measure of similarity between two protein structures, i.e., the modeled structure and the experimentally determined structure (X-ray structure) and they are independent of protein lengths. Here, all the match scores are above 0.5 which interprets that they have roughly assumed the same fold with template proteins. 3D structures of Dictyostellium discoideum, Neurospora crassa and Homo sapiens based on 1RGZ.pdb, 1SDE.pdb and 1CI9.pdb respectively were chosen as initial models.
Refinement of the modeled structures: Quality of the backbone of the
modeled structure was assessed with PROCHECK for reliability. PROCHECK is widely
used to scan a model for unlikely bonds, angles and dihedral values. It was
observed that although most of the φ-Ψ pairs were distributed in the
most favored and additional allowed regions of the Ramachandrans plot,
the backbone conformation of some of the Amino Acids (AA) were in the generously
allowed and disallowed regions as shown in Fig. 2 a and b
and values are given in Table 2.
||The match scores, percentage sequence identity and the extent
of sequence coverage of protozoan (Dictyostellium discoideum), fungal
(Neurospora crassa) and animal (Homo sapiens) proteases using
||Pairwise alignment of the amino acid sequences of the target
with those of known structures/templates: Predicted secondary structures,
s for β-sheets and h for α-helices, are also shown for each sequence.
Columns containing highly conserved AA are marked with '*' and partially
conserved sequences are marked with '.' along the alignments with the color
blue and green
For a model to be validated based on quality, a good quality protein model
should have >80% or more residues in the most favored regions of quadrangle
in the Ramachandrans plot. In the cases of archean and plant protease
models, 81.8 and 88.8% of residues respectively fall in the most favored regions
after refinement. Initially, there were some residues which didnt fall
under the most favored region in Ramachandrans plot. Then using CHARMm
module of Insight II, the dihedral angles (φ-Ψ) of all residues were
fixed so that they fall under the most favored region. The same protocol was
followed for each residue and for each model. The percentage residues of other
models of protozoan, fungal and human proteases are very close to 80% which
fall in the most favored region in the Ramachandrans plot (Table
2). These were grouped into different segments of the structure and refined
by energy minimization and molecular dynamics until all the backbone conformations
fell in the desired regions. The right panels of Fig. 2 show
the Ramachandrans plot of the refined model.
Validation of the structure: The side chains of the modeled structure
were regenerated using SCRWL and the overall structure was energy minimized
to get the refined homology model as well as threading models of the proteases
from SE clan of different species.Overall atom clashscores and rotamer outliers
were measured using MOLPROBITY.
||Results of the backbone refinement of the modeled proteases
from 5 different species of the SE clan as analyzed checking the distribution
of φ-Ψ dihedral angles in the Ramachandran’s plot also in percentage
|Fig. 2 (a-b):
||Ramachandrans plot of the φ-Ψ dihedral angles
of the modeled structure of the protease from Homo sapiens (No.MER017071)
(a) before and (b) after refinement of the backbone. Residues whose φ-Ψ
pairs fell outside the desired most favorable and additional allowed zones
are marked in red. Plot for a single species is shown for brevity
For structure validation and error correction purposes, the clash overlaps
are very much dominant issue. It is considered that a serious clash to occur
where two incompatible atoms overlap by 0.4 Å or more. The overall clashscore
of a structure is the number of serious clashes per 1000 atoms (Gu
and Bourne, 2009). It is seen from the listed values in the Table
3 that the all atom clashscore of predicted models of SE clan proteases
of archean, animal and plant are quite close to the X-ray structures. Again,
for X-ray targets, the target rotamer set consists of all residues for which
a valid rotamer name could be assigned. If that score is = 1%, it assigns the
name of the local rotamer peak and if = 1%, it declares an outlier. In this
work, except for the fungal and animal proteases, all the calculated rotamer
outliers for archean, plant and protozoan proteases are very close to their
X-ray templates. The other structural parameters like deviation of bond lengths
and bond angles from standard values were measured using MODELYN and Z score
of the model were measured using PROSA (Fig. 3). These values
were also compared with the relevant x-ray structures of the modeled structures
used in homology modeling and threading modeling as shown in Table
3. It may be noted that these structural parameters are comparable to those
of the X-ray structure indicating that our modeled structure has good general
Analysis of secondary structure and electrostatic potential surface of the protease of the SE clan: Most of the members of this clan exhibit extensive ordered secondary structures of α-helix and β-sheet except for fungal protease from Neurospora crassa and plant protease from Arabidopsis thaliana. The lack of extensive ordered secondary structures is probably due to less sequence coverage in the modeled structures of these proteases. All the structures of the proteases of this clan, experimental or modeled, contained two catalytic triad residues, Ser and Lys, are present on a single α-helical segment as they are present in a short and highly conserved motif, characteristics of this clan. But, the third catalytic triad residue is either tyrosine or serine and is well separated from the other two showing large variation in sequence and structures around it. The Cα atom RMSD of the triad residues of the modeled protease are within 1 Å which is found among serine proteases.
||Results of structural validity tests in terms of clashscores
(overlaps >0.4 Å) and rotamer outliers (first two χ angles
by >20° from its nearest associated rotamer) calculated using MOLPROBITY
and root mean square deviations from standard bond lengths and bond angles,
measured using MODELYN
|Fig. 3 (a-b):
||(a) Overall model quality (Z score) and (b) local model quality
(residue interaction energy profile) of Homo sapiens (No. MER017071)
after refinement of the backbone are determined using PROSA. Plot for a
single species is shown for brevity
The Electrostatic Surface Potential (ESP) of the X-ray structure (PDB ID: 1ONH) of the bacterial protease of the SE clan showed that the electrostatic potential surface around the catalytic serine residue is mostly neutral along with patches of an intense positive regions.
The ESP of the archean protease from Pyrococcus abyssi of the SE clan (homology modeled structure) showed that the surface electrostatic potentials of this protein around the catalytic triad are mostly positive with a patch of neutral potentials. There is a highly intense negatively charged region near the catalytic triad.
ESP of the threading-modeled structure of the protein from Dictyostellium
discoideum like the first two members of this clan showed the surface electrostatic
potentials around the catalytic site are mostly neutral mixed with a patch of
positive potentials. Thus, it has very different electrostatic environment around
the catalytic site making it useful as a target for drug design (Fig.
|Fig. 4 (a-d):
||Secondary structure (left panel) and surface electrostatic
potentials (right panel) of the X-ray structure of (a) plant protease from
Arabidopsis thaliana of the SE clan (MEROP No. MER004158), (c) protozoan
protease from Dictyostellium discoideum(MEROP No. MER035017) of the
SE clan. In secondary structures β-sheets are shown in light blue with
an arrow pointing to C-terminus, α-helices are shown in red and yellow,
turn/loops are shown in gray and side-chains of catalytic triad residues
are shown in green sticks. In surface electrostatic potentials (right panel)
of Arabidopsis thaliana, (b) and Dictyostellium discoideum and
(d) red, blue and white colors represent negative, positive and neutral
charges respectively. Both the figures are generated in the same orientation
using MOLMOL. Plots for only two species are shown for brevity
In the same way, ESP of the threading model of the fungal protease from Neurospora crassa was constructed. The surface electrostatic potentials around the catalytic site are a combination of negative, positive and neutral regions which is different from the other proteases of this SE clan.
The ESP of the homology model of the plant protease from Arabidopsis thaliana showed that the catalytic triad residues Ser-16 and Tyr-102 are present in the negative electrostatic potentials and Lys-19 is in a positive zone (Fig. 4).
Secondary structure (left panel) and surface electrostatic potential map (right panel) of the threading model of the human (Homo sapiens) protease of the SE clan were also prepared. The surface electrostatic potentials around the catalytic site are a combination of negative, positive and neutral regions. In many human proteases the surface electrostatic potential is highly negative but in this bacterial serine protease is very different and may be useful in drug design.
Analysis of the core structure and active site geometry of the modeled structures
of the proteases of the SE clan in reference to the experimental structures:
In order to analyze the structural environment of the modeled structures and
to compare them with those of the experimental structures, the distances between
the Cα atoms of catalytic triad residues Ser to Lys, Lys to Tyr (/Ser)
and Tyr (/Ser) to Ser in the refined structures were measured using MODELYN.
The mean values of distances between the Cα atoms of Ser to Lys, Lys to
Tyr and Tyr to Ser of the triads in the modeled structures are 5.7±0.07,
11.5±0.06 and 10.6±0.07, respectively. The deviations of values
from mean distances between the Cα atoms of the triad residues are much
smaller than those of the X-ray structures (Table 4). This
is because of the fact that all the triad residues are the same (i.e., the third
residue is Tyr and not Ser).
The 3-D structures of the selected proteases of the SE clan were superpose with the representative x-ray structure (1ONH) of the bacterial protease with respect to a selected set of Cα atoms in order to examine the deviations of the core of the structures of the modeled proteins of the SE clan. It is found that 14 to 37% of the Cα atoms superposed with an RMS deviation below 1 Å. while only 2 to 7% of Cα atoms of the X-ray structures of the bacterial proteases of this clan superposed with an RMS deviation below 2 Å.
Phylogenetic analysis: The consensus phylogenetic tree of all the five S12 family members of our interest is shown in Fig. 5. Although, the core remained relatively conserved some segments of the proteases varied a lot during evolution.
||Analysis of the core structure and active site geometry of
the modeled structures of the proteases of the SE clan in reference to the
experimental structures. All the modeled structures were superposed on the
x-ray structure (1ONH) of the bacterial protease with respect to a selected
set of Cα atoms
||Phylogenetic Analysis of SE clan proteases, One from each
of the groups of bacteria (Enterobacter cloacae No MER000463), archea
(Pyrococcus abyssi, No MER026262), protozoan (Dictyostellium discoideum,
MER035017), fungal (Neurospora crassa, MER033421), plant (Arabidopsis
thaliana, No MER004158) and human (Homo sapiens, MER017071) were
carried out using Phylip web server
Conceptually, there are three basic approaches to protein structure prediction,
e.g., homology modeling, threading and ab initio or template free methods. Comparative
modeling alone or in conjunction with fold recognition analysis provides relatively
accurate structures when the target sequence of a protein is evolutionarily
related to a template which has an experimentally determined structure in the
Protein Data Bank (PDB). (Chatterjee et al., 2011).
In this study, serine proteases from the diverse groups were selected of the
living kingdom based on MEROP classification of the SE clan for 3-D structural
modeling. Five structures of only bacterial proteases of this clan have been
determined by experimental method; we have included them as representative of
structures of the proteases of this clan for comparison with our predicted structures
of proteases from other groups. Most of the studies in this context has predicted
a 3D model of a protein and established it by further refinement and validation.
There are ample numbers of literatures which have predicted 3D structures of
different proteins (Rajesh et al., 2008; Prabhavathi
et al., 2011; Ginalski and Rychlewski, 2003).
Majority of the studies have been reported about the identification of drug
targets by experimental as well as in silico methods as targeting proteases
is going to be very useful in drug design; especially the non-homologous proteins
are very effective (Smith, 2004; Chong
et al., 2006; Suthar et al., 2009;
Shakyawar et al., 2011; Singh
et al., 2006) but till date, no studies have been reported on computer-based
3D modeling of SE clan serine proteases. The prepared models of this clan using
experimental data available in the literature are also submitted to PDB. From
that point of view, the present study is unique and much more virgin in nature
as this has covered a wide arena of model prediction to evolutionary analysis.
In the present study S12 family of serine peptidases of SE clan were chosen for structural study. No 3D structures of the serine proteases of SE clan were available in databases except only the X-ray structures of bacteria, as SE clan is a lower species belonging clan. Therefore, 3D models for the archea, protozoa, fungi, plant and human were built using comparative modeling techniques. The models were evaluated and validated using Discovery Studio visualizer (Accelrys) and PROCHECK. Further refinement and validation were done by Ramachandran plot analysis.
The structural models, both experimental and predicted, are used to analyze various structural properties of the serine proteases of the SE clan mainly focusing on the common core harboring the catalytic site of this very important class of enzymes which have been used and hold high potential in the structure based rational drug design. Examination of the geometry of the triad residues reveal that the modeled enzyme can catalyze the peptide hydrolyzing activity. In general, the catalytic site of all the proteases exhibit acidic regions in the surface electrostatic potential maps but a few of them contain both large patches of positive and neutral potentials. Hence some of the proteases with predominant positive and neutral regions around the catalytic site can be used as drug target against bacterial and protozoan pathogenic organisms as the interactions of inhibitors are strongly influenced by the nature of electrostatic potential surface near the substrate/inhibitor binding sites.
However, the essence of this study has not only prediction of the 3D structures
of the serine proteases; but also the establishment an evolutionary relationship
between them. The residues of the catalytic triad in all the proteases remain
same but they differ in their position in each case. The kinship amongst the
proteases, though belonging in different species inspite of having big phylogenetic
distances have been established by phylogenetic analysis; thats why they
belong to a single "clan".
Most of the studies in this context has predicted a 3D model of a protein and established it by further refinement and validation. From that point of view, the present study is unique and much more virgin in nature as this has covered a wide arena of model prediction to evolutionary analysis.
The authors thank Council for Scientific and Industrial Research (CSIR), New Delhi, India, for providing financial grant for this project. Authors also thank Indian Institute of Chemical Biology (IICB), Kolkata, India, for providing the laboratory support and other infrastructures.
1: Altschul, S.F., T.L. Madden, A.A. Schaffer, J. Zhang, Z. Zhang, W. Miller and D.J. Lipman, 1997. Gapped BLAST and PSI-BLAST: A new generation of protein database search programs. Nucl. Acids Res., 25: 3389-3402.
PubMed | Direct Link |
2: Amir, A., M.A. Siddiqui, N. Kapoor, A. Arya and H. Kumar, 2011. In silico molecular docking of influenza virus (PB2) protein to check the drug efficacy. Trends Bioinform., 4: 47-55.
CrossRef | Direct Link |
3: Anand, K., J. Ziebuhr, P. Wadhwani, J.R. Mesters and R. Hilgenfeld, 2003. Coronavirus main proteinase (3CLpro) structure: Basis for design of anti-SARS drugs. Sci., 300: 1763-1767.
4: Barrett, A.J., N.D. Rawlings and E.A. O'Brien, 2001. The MEROPS database as a protease information system. J. Struct. Biol., 134: 95-102.
5: Barrett, A.J., N.D. Rawlings and J.F. Woessner, 2003. The Handbook of Proteolytic Enzymes. 2nd Edn., Academic Press, USA.
6: Canutescu, A.A., A.A. Shelenkov and R.L. Dunbrack Jr., 2003. A graph theory algorithm for rapid protein side-chain prediction. Protein Sci., 12: 2001-2014.
7: Chatterjee, S., A. Laskar, A. Chatterjee, C. Mandal and S. Chaudhuri, 2011. Insilico structural analysis of an immunotherapeutic glycoprotein T11TS (sheep CD58). Int. J. Biol. Med. Res., 2: 346-359.
Direct Link |
8: Chong, C.E., B.S. Lim, S. Nathan and R. Mohamed, 2006. In silico analysis of Burkholderia pseudomallei genome sequence for potential drug targets. In silico Biol., 6: 341-346.
9: Davis, I.W., L.W. Murray, J.S. Richardson and D.C. Richardson, 2004. MOLPROBITY: structure validation and all-atom contact analysis for nucleic acids and their complexes. Nucleic Acids Res., 32: W615-W619.
10: Englebert, S., P. Charlier, E. Fonze, Y. To'th, M. Vermeire et al., 1994. Crystallization and X-ray diffraction study of the Streptomyces K15 penicillin-binding DD-transpeptidase. J. Mol. Biol., 241: 295-297.
CrossRef | PubMed |
11: Esmon, C.T., 2000. Regulation of blood coagulation. Biochim. Biophys. Acta, 1477: 349-360.
12: Floudas, C.A., 2007. Computational methods in protein structure prediction. Biotechnol. Bioeng, 97: 207-213.
CrossRef | PubMed |
13: Ginalski, K. and L. Rychlewski, 2003. Protein structure prediction of CASP5 comparative modeling and fold recognition targets using consensus alignment approach and 3D assessment. Proteins, 53: 410-417.
14: Gu, J. and P.E. Bourne, 2009. Structural Bioinformatics. 2nd Edn., Wiley-Blackwell Publication, USA.
15: Kelly, J.A., J.R. Knox, P.C. Moews, G.J. Hite and J.B. Bartolone et al., 1985. 2.8-A Structure of penicillin-sensitive D-alanyl carboxypeptidase-transpeptidase from Streptomyces R61 and complexes with beta-lactams. J. Biol. Chem., 260: 6449-6458.
16: Krem, M.M., and D.E. Cera, 2001. Molecular markers of serine protease evolution. EMBO J., 20: 3036-3045.
17: Koradi, R., M. Billeter and K. Wuthrich, 1996. MOLMOL: A program for display and analysis of macromolecular structures. J. Mol. Graph., 14: 51-55.
18: Laskowski, R.A., M.W. MacArthur, D.S. Moss and J.M. Thornton, 1993. PROCHECK: A program to check the stereochemical quality of protein structures. J. Applied Cryst., 26: 283-291.
CrossRef | Direct Link |
19: Mandal, C., Kingery B.D., J.M. Anchin, S. Subramanium and D.S. Linthicum, 1996. ABGEN: A knowledge based automated approach for antibody structure modeling. Nat. Biotechnol., 14: 323-328.
20: Mandal, C., 1998. MODELYN: A molecular modelling programversion, PC-1.0. Indian Copyright No 9/98.
21: Meller, J. and R. Elber, 2001. Linear programming optimization and a double statistical filter for protein threading protocols. Proteins, 45: 241-261.
22: Prajapat, R., R.K. Gaur, R. Raizada and V.K. Gupta, 2010. In silico analysis of genetic diversity of begomovirus using homology modelling. J. Biol. Sci., 10: 217-223.
CrossRef | Direct Link |
23: Prabhavathi, M., K. Ashokkumar, N. Geetha and K.M. Saradha Devi, 2011. Homology modeling and structure prediction of thioredoxin (TRX) protein in wheat (Triticum aestivum L.). Int. J. Biosci., 1: 20-32.
24: Perona, J.J. and C.S. Craik, 1995. Structural basis of substrate specificity in the serine proteases. Protein Sci., 4: 337-360.
25: Rajesh, S., M. Raveendran and A. Manickam, 2008. Comparative modeling and analysis of 3-D structure of EMV2, a late embryogenesis abundant protein of Vigna Radiata (Wilczek). J. Proteomics Bioinform., 1: 401-407.
Direct Link |
26: Rawlings, N.D. and A.J. Barrett, 1993. Evolutionary families of peptidases. Biochem. J., 290: 205-218.
27: Rawlings, N.D. and A.J. Barret, 1994. Families of serine peptidases. Methods Enzymol., 244: 19-61.
CrossRef | Direct Link |
28: Rawlings, N.D., F.R. Morton and A.J. Barrett, 2006. MEROPS: The peptidase database. Nucleic Acids Res., 34: D270-D272.
29: Sauvage, E., R. Herman, S. Petrella, C. Duez, F. Bouillenne, J.M. Frere and P. Charlier, 2005. Crystal structure of the Actinomadura R39 DD-peptidase reveals new domains in penicillin-binding proteins. J. Biol. Chem., 280: 31249-31256.
30: Schwede, T., J. Kopp, N. Guex and M.C. Peitsch, 2003. SWISS-MODEL: An automated protein homology-modeling server. Nucleic Acids Res., 31: 3381-3385.
31: Shakyawar, S.K., A. Goyal and V.K. Dubey, 2011. Database of In silico predicted potential drug target proteins in common bacterial human pathogens. Am. J. Drug Discovery Dev., 1: 70-74.
CrossRef | Direct Link |
32: Silvaggi, N.R., J.W. Anderson, S.R. Brinsmade, R.F. Pratt and J.A. Kelly, 2003. The crystal structure of phosphonate-inhibited D-Ala-D-Ala peptidase reveals an analogue of a tetrahedral transition state. Biochem., 42: 1199-1208.
33: Singh, S., B.K. Malik and D.K. Sharma, 2006. Molecular drug targets and structure based drug design: A holistic approach. Bioinformation, 1: 314-320.
34: Smith, C., 2004. Drug target identification: A question of biology. Nature, 428: 225-231.
35: Summers, N.L. and M. Karplus, 1991. Modeling of side chains, loops and insertions in proteins Methods Enzymol., 202: 156-204.
36: Suthar, N., A. Goyal and V.K. Dubey, 2009. Identification of potential drug targets of Leishmania Infantum by in silico genome analysis. Lett. Drug Des. Discovery, 6: 620-622.
37: Tambunan, U.S.F., O. Hikmawan and T.A. Tockary, 2008. In silico mutation study of haemagglutinin and neuraminidase on banten province strain influenza a H5N1 virus. Trends Bioinform., 1: 18-24.
CrossRef | Direct Link |
38: Tambunan, U.S.F. and A.A. Parikesit, 2010. In silico design of drugs and vaccines for dengue disease. Trends Bioinform., (In Press).
39: Thompson, J.D., D.G. Higgins and T.J. Gibson, 1994. CLUSTAL W: Improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res., 22: 4673-4680.
CrossRef | PubMed | Direct Link |
40: Turk, B., 2006. Targeting proteases: Successes, failures and future prospects. Nat. Rev. Drug Discov., 5: 785-799.
CrossRef | Direct Link |
41: Wiederstein, M. and M.J. Sippl, 2007. ProSA-web: Interactive web service for the recognition of errors in three-dimensional structures of proteins. Nucleic Acids Res., 35: W407-W410.
CrossRef | Direct Link |