See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/326909374

B-cell and T-cell epitope identification with stability analysis of AI-2 import
ATP-binding cassette LsrA from S. typhi - In silico approach

Article  in  Microbial Pathogenesis · August 2018

DOI: 10.1016/j.micpath.2018.08.012

CITATIONS READS

4 467

7 authors, including:

Princy Vijayababu Gopinath Samykannu

Bharathiar University Bharathiar University
16 PUBLICATIONS   55 CITATIONS    17 PUBLICATIONS   84 CITATIONS   

SEE PROFILE SEE PROFILE

Christian Bharathi Syed ibrahim Basheer Ahamed

Bioinformatics Centre Pondicherry University
11 PUBLICATIONS   80 CITATIONS    23 PUBLICATIONS   69 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Outer Membrane Proteins View project

Function Annotation View project

All content following this page was uploaded by Christian Bharathi on 31 August 2018.

The user has requested enhancement of the downloaded file.

 Microbial Pathogenesis 123 (2018) 487–495

Contents lists available at ScienceDirect

Microbial Pathogenesis
journal homepage: www.elsevier.com/locate/micpath

B-cell and T-cell epitope identification with stability analysis of AI-2 import T
ATP-binding cassette LsrA from S. typhi eIn silico approach
Princy Vijayababua,∗, Gopinath Samykannua, Christian Bharathi Antonyrajb,
SundaraBaalaji Narayanana, Syed Ibrahim Basheer Ahamedb, Perumal Perumalc,
Shanmughavel Piramanayagamd
a
Structural Biology Laboratory, Department of Bioinformatics, Bharathiar University, Coimbatore, Tamil Nadu, India
b
Centre for Bioinformatics, Pondicherry University, Pondicherry, India
c
Membrane Protein Biology Group, International Centre for Genetic Engineering and Biotechnology, New Delhi, India
d
Computational Biology Laboratory, Department of Bioinformatics, Bharathiar University, Coimbatore, Tamil Nadu, India

A R T I C LE I N FO A B S T R A C T

Keywords: Typhoid fever is a severe illness in humans, caused by Salmonella typhi, a Gram-negative bacterium. Membrane
Quorum sensing proteins of S. typhi have strong potential for its use in development of subunit vaccine against typhoid. In current
LsrA study, peptide-based subunit vaccine constructed from AI-2 import ATP-binding cassette transporter protein
Protein peptide docking (LsrA) against S. typhi. B-cell and T-cell epitopes were identified at fold level with validated 3-D theoretical
Epitope
modelled structure. T-cell epitope from LsrA (LELPGSRPQ) has binds to maximum number (82.93%) of MHC
MD-Simulation
class I and class II alleles. LsrA epitope was docked with HLA-DR4 and contact map were constructed to analyze
molecular interaction (docking) studies. Simulation search for the binding site for full flexibility of the peptide
from CABS-dock shows the stable interactions. MD simulation analysis reveals that LsrA epitope was binding and
interacting firmly with the HLA-DR4. Hence, we are proposing that LsrA epitope would be a prominent epitope
vaccine for human specific pathogen of S. typhi, which requires further steps to be elevated as a vaccine drug in
near future.

1. Introduction
Salmonella typhi are Gram-negative pathogens that causing the dis-
eases ranging from gastroenteritis and diarrhea to life-threatening
systemic syndrome including typhoid fever to human [1–4]. Typhoid
fever remains a global health problem, resulting in more than 2,00,000
annual deaths, mostly children in developing countries [3,5]. Unlike
other Salmonella enterica serovars such as S. typhimurium or S. enteritidis,
S. typhi is an exclusive human pathogens [6]. Presently, two typhoid
vaccines of established safety and efficacy are available on the inter-
national market: (1) the oral vaccine based on the live, attenuated
mutant strain of S. typhi Ty21a (Ty21a vaccine), is supplied in enteric
coated capsules. (2) The injectable Vi capsular polysaccharide vaccine
(ViCPS vaccine) is given intramuscularly in a single dose (WHO). For-
tunately, with adequate treatment, most patients recover from the acute
phase of typhoid; however, 3–5% of individuals who are infected with
S. typhi develop a chronic infection in the gall bladder [6,7]. Further-
more, S. typhi can most often by colonizing as biofilm to gallbladder and
these abnormalities have risk to developing cancer in gallbladder [6,8].
Cell-to-Cell communication in bacteria facilitates recognition of
their population density in a confined environment and expression of
co-ordinated behaviors. This bacterial talk or cell-to-cell communica-
tion called as Quorum Sensing (QS) system. This is achieved through
synthesis, release and detection of signal molecules called as auto-
inducers (AIs). The signals are called as autoinducers, since they pro-
mote their own expression. Each bacterial cell produces a small amount
of autoinducers, which may diffuse or is pumped out of the cell. At a
low population density, the concentration of AI in the environment is
low. With an increase in bacterial density, more amount of AI is pro-
duced, which accumulates in the environment. When a threshold level,
called as quorate level is achieved, the AIs are able to activated protein
then binds to specific promoters and regulates the expression of

∗
Corresponding author.

https://doi.org/10.1016/j.micpath.2018.08.012
Received 9 May 2018; Received in revised form 7 August 2018; Accepted 7 August 2018
Available online 08 August 2018
0882-4010/ © 2018 Elsevier Ltd. All rights reserved.
P. Vijayababu et al.
downstream genes. The entire process leads to expression of secondary
phenotypes in bacteria, such as bioluminescence, conjugation, anti-
biotic production, virulence, biofilm maturation, DNA competence, etc.
Thus, QS system allows bacteria to detect their number in the sur-
roundings, and co-ordinately expresses certain phenotypes, which do
not provide the same benefits when expressed individually [9].
LsrA, a key molecule and essential for QS system of S. typhi. It is a
member of ATP-binding cassette (ABC) transporters and responsible for
energy coupling in QS system [10]. Generally, components of ABC
transporters that anticipated being most immunogenic in Gram-nega-
tive bacteria are the associated outer membrane proteins [11], since
components that are predicted to be located in the periplasmic space or
inner membrane may also be immunogenic, which are apparently not
surface located yet can induce strong antibody responses [12]. There-
fore, they can used as a potential target in development of vaccine
against S. typhi.
B-cell epitope mapping is the crucial issue in vaccinology. Epitope
vaccine or an epitope-based subunit-vaccine has lesser side effects when
compared to conventional vaccines [13], easier to produce, cheaper to
manufacture, when compared to engineered subunit vaccines, does not
contain any complete component of the pathogens. Limitation on ex-
perimental methods in the biological side, dataset and availability of
computational resources opens a chance on developing prediction
method, which can accelerate epitope identification [14].
In this study, we focused on in silico identification and character-
ization of B-cell and T-cell epitopes from S. typhi LsrA. In addition,
protein-peptide docking and molecular dynamics simulation were car-
ried out to know the epitope interactions and stability at molecular
level.
2. Materials and methods
2.1. Target protein domain architecture and antigenicity prediction
LsrA (Q8Z2X5) amino acid sequence was retrieved from UniProtKB
[15] and subjected to VaxiJen v2.0 [16] to predict antigenic property.
This Bioinformatics tool used to analyze antigenic property, based on
transformation of protein sequences into uniform vectors of principal
amino acid properties. The default parameter of threshold is 0.4 and
Anti-autocross-covariance (ACC) output was used against both full
length proteins [16]. The achievable domain characteristic included in
LsrA was investigated by InterPRO [17]. Subcellular localization were
predicted by PSORTb [18].
2.2. Prediction of B -cell and T-cell epitopes
LsrA amino acid sequence were subjected to BCPred [19]. Antigenic
linear non-overlapping 20-mer B-cell epitopes were predicted from
whole antigen using BCPreds software with two methods. AAP method,
based on amino acid pair antigenicity [20] and BCPreds, using the
subsequence kernel, provided by BCPreds were followed to predict
more reliable B-cell epitope. Only B-cell epitopes having a score > 0.8
were accepted to further analysis. Selected B-cell epitopes were sub-
jected to VaxiJen v2.0 to check antigenicity. B-cell epitope having an-
tigenic property, BCPreds score > 0.8 and VaxiJen score > 0.4 were
selected for T-cell epitope prediction.
Screened B-cell epitopes with antigenic property were subjected to
ProPred I [21] and ProPred [22] analysis. ProPred predicted MHC class
II binding peptides (HTL epitopes) for 51 alleles and ProPred1 to pre-
dict MHC class I binding peptides (CTL epitopes) for 47 alleles.
Common epitope that bind to both the MHC class were selected. The
selected epitope was then analyzed with VaxiJen v2.0. IC50 values of
488
Microbial Pathogenesis 123 (2018) 487–495
corresponding epitopes were inferred from MHCPred server [23].
IC50 < 50 nm indicate a good binder, IC50 = 50–500 nm an inter-
mediary binder, IC50 = 500–5000 nm a week binder and IC50 > 5000
a non-binder [24]. T-Epitope Designer [25] was used to screening the
HLA-peptide binding epitope based on virtual binding pockets using
crystal structures of HLA-peptide complexes followed by calculation of
peptide binding to binding pockets.
2.3. 3-D modeling, refinement and validation
To know the crystal structure similarity of LsrA, BLAST was per-
formed against PDB database with default parameters [26]. I-TASSER, a
hierarchical protein structure modeling approach based on the sec-
ondary structure enhanced of Profile-Profile threading Alignment (PPA)
[27] was used to model LsrA. The best model was selected based on the
computed C-score. Modelled 3D structures were subjected to Mod-
Refiner [28] for refinement of energy minimization. PROCHECK [29]
was used to validate the refined 3-D structures.
2.4. Fold level topology analysis and epitopes characterization
The folding and clusters of selected epitopes in folded LsrA was
analyzed to confirm the topology of epitope using Pepitope server [30],
which use PepSurf and Mapitope algorithms [31] to predict irregular
epitopes based on a set of peptides and affinity-selected against a
monoclonal antibody of interest.
Short sequence of predicted epitopes of LsrA was modelled using
CABS-fold server [32]. Secondary structures were analyzed; resultant
epitopes were subjected to ProGene 1.0 [33] to determine molecular
weight and isoelectric point (pI).
2.5. Peptide docking
Identified epitope LsrA was docked with HLA-DR4 (PDB ID: 1D5M)
using CABS-dock, which performs simulation search for the binding site
permitting for full flexibility of the peptide and small fluctuations of the
receptor backbone [34]. These searches were used for favorable inter-
actions between epitope and HLA molecule. Protein-peptide docking in
CABS-dock is divided into three repeated steps realized by separate
protocols: (I) prediction of the binding site(s) on the receptor structure,
(II) initial modeling of the peptide backbone in the binding site(s), and
finally (III) refinement of the initial protein-peptide complexes to high
resolution. Accuracy of the best obtained models was calculated based
on coordinate root-mean-square deviation (cRMSD) [35] between the
epitope and the MHC molecule.
2.6. Molecular dynamics simulation of receptor-ligand complex
In order to understand dynamic behavior of HLA-DR4 and its epi-
tope was subjected to MD simulation for the time period of 20 ns using
GROMACS 5.0.4 package [36]. Protein topology was obtained by
GROMOS96 54a7 force field for apo and complexes of HLA-DR4. The
docked complex was centered in cubic box while the PBC condition was
kept as 1 Å from each edge of the box. We used SPC water model to fill
the surrounding of the protein or protein-complexes within the box and
the system was neutralized by the addition of exact counter ions by
replacing equal number of water molecules. Steepest descent method
was applied for 50,000 steps with the maximum of 1000 kJ/mol/nm
force to minimize the energy of the system and conjugate gradient was
continued with the same parameter setup for the system. Van der Waals
and long-range electrostatic interactions within the system was treated
with PME (Partical mesh ewald) [37] while all covalent bonds with
P. Vijayababu et al.
LINCS algorithm [38] and geometry of water with SETTLE algorithm
[39]. Moreover, the system was equilibrated by isothermal-isochoric
(or NVT) ensemble method with the temperature of 300 K using V-re-
scale method [40] further proceeded with (NPT) at a constant pressure
of 1 atm using Parrinello-Rahman method [41] for the time period of
500 ps. With this, the whole system was subjected to MD simulation
production for 20ns using leapfrog integrator. The energy and co-
ordinates of protein and protein complexes was stored in their corre-
sponding trajectories at each 2 ps interval and analyzed with xmgrace
tool. Overall, workflow is shown in Fig. 1.
3. Result and discussion
3.1. Domain prediction and antigenicity scoring
LsrA is inner membrane protein belongs to ABC transporter domain
(Inter Pro number IPR003439) and amino acid position ranging from
10-230 and 281–481. LsrA antigenicity was analyzed using VaxiJen
v2.0 assigns a threshold value of 0.4 [16] for antigenic LsrA scored 0.4.
Subcellular location on PSORTb shows that LsrA have 99.9 in cyto-
plasmic membrane region.
3.2. Identification of B-Cell epitope
Amino acid sequence of LsrA (511 aa) was subjected to BCPred for
Fig. 1. Overall workflow of B-and T-cell LsrA epitopes identification from Salmonella typhi. Arrow mark indicate the flow of work start from sequence retrieval and
end with molecular dynamic simulation of HLA-DR4 complex.
489
Microbial Pathogenesis 123 (2018) 487–495
B-cell epitope prediction. Both BCPred and AAP method [42] were used
and predicted B-cell epitopes were listed in Table 1 and Table 2. Epi-
tope length was set as 20 and specificity 75%. Epitopes having BCpreds
and VaxiJen scores > 0.8 and > 0.4 respectively. Finally, 3 out of 8
epitopes from BCPred and 2 out of 8 from AAP method were selected
for further analysis.
3.3. Identification of T-Cell epitope
Selected B-cell epitope from LsrA were subjected to ProPred-1 and
ProPred with default parameters for identification of T-cell epitopes
[43]. The ProPred1 and ProPred implements matrix based prediction
algorithm. The obtained matrices were multiplication matrices, where
the scores are calculated by multiplying and summing the score of
each amino acid position. MHCPred uses Partial least squares (PLS)
based approach for the prediction of binding affinity to MHC mole-
cules, which was used in the present study and these methods also
reported in yellow fever and T- Cell identification [24,44]. By using
MHCPred, malarial merozoite surface protein-1 T-cell epitopes were
identified [45]. MHCPred with the combination of SYFPEITHI,
NetMHC has been used for Epstein-Barr virus latent membrane pro-
tein-2A T-cell epitope prediction [46]. Predicted number of major
histocompatibility complex class eI (47 alleles) with ProPred-I and
MHC class eII (51 alleles) binding regions with their antigenicity
score were shown in Table 3. LsrA epitopes were selected based on
P. Vijayababu et al. Microbial Pathogenesis 123 (2018) 487–495

Table 1
Predicted B-cell epitopes from LsrA using BCPred (BCPred method). Antigenicity of B-cell epitopes of LsrA were calculated using VaxiJen v2.0. Selected B-cell
epitopes were highlighted. (*A-Antigen/NA- Non antigen).
S.No position BCPred epitope sequence of LsrA BCPred scores Vaxijen Score AA/NA*

1. 53 LMKIIAGIVPPDGGTIDIAG 0.991 0.4538 A

2. 249 QKLWLELPGSRPQNERGATV 0.95 1.0707 A
3. 1 MQISHNTASPLICVQNIYKS 0.908 0.1376 NA
4. 167 PTASLTPAETDRLFTRLQEL 0.886 0.0044 NA
5. 470 ADRVYVMHQGELGGPALCGE 0.856 −0.1080 NA
6. 299 GAGRTELAETLYGIRPVNAG 0.825 0.4830 A
7. 365 LTHNQKGFWIKPQRDNATLE 0.818 0.2290 NA
8. 225 HDLSTDEIIQAITPATQGVS 0.759 0.2977 NA

Table 2
Predicted B-cell epitopes from LsrA using BCPred (AAP method). Antigenicity of B-cell epitopes of LsrA were calculated using VaxiJen. Selected B-cell epitopes were
highlighted (*A-Antigen/NA- Non antigen).
S.No Position AAP epitope sequence of LsrA AAP score Vaxijen score A/NA*

1. 304 ELAETLYGIRPVNAGRMLFN 1 −0.1134 NA

2. 166 EPTASLTPAETDRLFTRLQE 1 0.0352 NA
3. 111 LQGRQASTEKMQQLLKAMGC 1 0.5782 A
4. 366 THNQKGFWIKPQRDNATLER 1 0.0483 NA
5. 247 ASQKLWLELPGSRPQNERGA 1 0.9066 A
6. 466 IEQMADRVYVMHQGELGGPA 0.996 0.0523 NA
7. 5 HNTASPLICVQNIYKSYSGV 0.863 0.0156 NA
8. 78 LTPLKAHQYGIYLVPQEPLL 0.809 0.2719 NA

Table 3
T- Cell identification from identified B- Cell through BC-Pred and AAP methds.Based on the Vaxijen scoe and antigenicity T-cell epitope were identified. Epitope
(LELPGSRPQ) shows the antigenic property in both ProPred-I and Propred; with Vaxijen score of 1.2371.
S.No B-cell epitope ProPred-I (MHC-I ProPred (MHC-II Vaxijen score Total Number of binding
binding region) binding region) (Antigenicity) (MHC-I + MHC II)

T-Cell identification from BC- 1. LMKIIAGIVPPDGGTIDIAG IVPPDGGTI LMKIIAGIV −0.0523(NA) 8 + 2 = 10

Preds method LMKIIAGIV MKIIAGIVP 0.2331 (NA) 5 + 14 = 19
PPDGGTIDI IIAGIVPPD −0.4077(NA) 6+2=8
VPPDGGTID IVPPDGGTI 0.0073(NA) 1+2=3
KIIAGIVPP 1+0=1
2. QKLWLELPGSRPQNERGATV PQNERGATV LELPGSRPQ 1.2371(A) 1 + 1=2
KLWLELPGS LWLELPGSR 0.3954(NA) 4+1=5
LWLELPGSR 4+0=4
SRPQNERGAT 3+0=3
RPQNERGAT 4+0=4
LELPGSRPQ 2+0=2
LPGSRPQNE 1+0=1
WLELPGSRP 2+0=2
3. GAGRTELAETLYGIRPVNAG RTELAETLY LYGIRPVNA 0.7557 (NA) 5 + 15 = 20
LAETLYGIR YGIRPVNAG 0.4157 (NA) 4 + 24 = 28
ELAETLYGI 16 + 0 = 16
ETLYGIRPV 6+0=6
GRTELAETL 6+0=6
ELAETLYGI 3+0=3
YGIRPVNAG 1+0=1
LYGIRPVNA 2+0=2
T-Cell identification from 4. LQGRQASTEKMQQLLKAMGC KMQQLLKA MQQLLKAMG −0.9627(NA) 8 + 23 = 31
AAP method ASTEKMQQL LQGRQASTE 2.1276(NA) 14 + 4 = 18
STEKMQQLL 8+0=8
GRQASTEKM 4+0=4
QQLLKAMGC 2+0=2
TEKMQQLLK 1+0=1
5. ASQKLWLELPGSRPQNERGA KLWLELPGS LELPGSRPQ 1.2371(NA) 4+1=5
ASQKLWEL LWLELPGSR 0.3954(NA) 12 + 1 = 13
LWLELPGSR 2+0=2
SRPQNERGA 3+0=3
LELPGSRPQ 2+0=2
WLELPGSRP 2+0=2
LPGSRPQN 1+0=1

490
P. Vijayababu et al. Microbial Pathogenesis 123 (2018) 487–495

VaxiJen and antigenicity. Table 4

The T-epitope Designer was based on a model that defines peptide Epitope (LELPGSRPQ) characterization. B-cell and T-cell mediated LsrA epitope
binding pockets, using x-ray information from crystal structures of HLA- immunity are represented along with their various parameters.
peptide complexes, followed by the estimation of peptide binding to T-Epitopes LELPGSRPQ
binding pockets [25]. The predicted epitope that bind to > 80% alleles
in T-epitope Designer (Supplementary data S1). This epitopes also bind Amino acid position 179–187
VaxiJen score 1.2371
to DRB1*0401 alleles belong to MHC class-II. DRB1 *0401 allele gives
IC50 value of for DRB1*0101 (MHCPred) 58.21 (Good binder)
resistance against typhoid fever [47]. A complex between this allele and IC50 value of for DRB1*0401 (MHCPred) 520.00 (Weak binder)
selected epitope subjected to interaction study. MHC binding regions (ProPred I & ProPred) 5
Confidence prediction (max = 1) 0.89
Percentage of allele binding from T-epitope Designer 82.93%
3.4. Modeling and structure validation
Lowest score 2.02 (C*1202)
Highest score 1503.84 (B*1561)
Blast search against PDB database showed that 4AYT have 29% Molecular weight (MW) 1017.06 kDa
identity with LsrA. Due to the lack of experimental 3D structure, Isoelectric point (PI) 6.00
modeling was done through the automated server I-TASSER [27],
where the modelled LsrA possessed a C-score of −0.52 (Fig. 2). C-score
is a confidence score for estimating the quality of predicted models by I-
TASSER [48–51]. It was calculated based on the significance of
threading template alignments and the convergence parameters of the
structure assembly simulations [52]. Ramachandran plot parameters
were estimated through PROCHECK.

3.4.1. Topology and characterization of T-cell epitopes

Position of predicted epitopes on theoretical model of LsrA was
identified using Pepitope server. Input requirement was epitope se-
quences and modelled structure of the antigen [53]. The epitopes bind
to HLA molecules with 82.93% for LsrA in T-epitope designer. The LsrA
epitope “LELPGSRPQ” from Cluster-I (Score: 14.45, Residue No: 9) is
found antigenic (VaxiJen score: 1.2755) and can bind 5 MHC molecules
of both the MHC class I and II. The confirmed T-cell epitopes and their
properties were listed in Table 4 and position of the predicted epitopes
were shown in Fig. 3.

3.4.2. Theoretical model of epitope

The CABS-fold web server was used to generate 3-D structure of
predicted epitope by de novo modeling [35] and predicted LsrA epitope
(LELPGSRPQ) shown in Fig. 4. Molecular weight and theoretical Iso-
electric point (pI) for LsrA epitope was calculated as 996.13 Da and

Fig. 3. Topology of predicted epitope. Position of predicted epitope shown in

red colour on theoretical model of LsrA. (For interpretation of the references to
colour in this figure legend, the reader is referred to the Web version of this
article.)

Fig. 2. Three dimensional theoretical structure of LsrA. Secondary structure,
coil, sheets and helix represented in green, yellow and red respectively. N and
C-terminus were labeled. (For interpretation of the references to colour in this
figure legend, the reader is referred to the Web version of this article.)
Fig. 4. Theoretical 3-D structure of LsrA epitope. CABS-fold server was used to
predict theoretical epitope and (LELPGSRPQ) shows helical secondary struc-
ture.

491
P. Vijayababu et al. Microbial Pathogenesis 123 (2018) 487–495

6.00 respectively.

3.5. Epitope docking with HLA-DR4

LsrA epitopes was docked with HLA-DR4 CABS-dock web server,

generates 10,000 model structures of the protein-petide complex [35].
Clustering based on the RMSD of the entire protein-epitope complex
and resulting structures are grouped in clusters of similar complexes
and ranked according to cluster size from the largest to the 10th largest
[35]. Ten top ranked CABS-dock models representatives of the 10 most
numerous clusters details shown in supplementary table S2 and top
ranked docked model was shown in Fig. 5.

3.5.1. Contact map analysis

The interactions of highly dynamic complexes from LsrA epitope
with HLA-DR4 can be analyzed using contact maps derived from the
CABS-dock simulations. Intermolecular protein-peptide interactions
were analyzed and the contact map is shown in Fig. 6 besides Fig. 6. LsrA epitopes contact map. The contact map shows LsrA epitopes have
contact residues are shown in Supplementary Table S3. We ex- good interaction with HLA-DR4 and proline 256 has the maximum interaction
amined the contact-forming residues of the peptide to find out with HLA-DR4.
which parts of the LsrA modeling peptide are most important for its
interaction with HLA-DR4. According to interaction study the most
important complex-stabilizing contacts are formed by five LsrA re-
sidues (LEU 181, PRO 182, SER 184, ARG 185 and PRO 186) with
the binding threshold value as 4. In our docking results we observed
the highest contact frequencies for their neighbors as shown in
histogram Fig. 7.

3.6. Molecular dynamic simulation

From MD simulation analysis, it shown that apo- HLA-DR4 shown

more RMSD fluctuations compare to LsrA-epitope except at 12.5–15 ns
time period (Fig. 8A). The difference in RMSD was observed by large
conformational changes in 14–19, 34–40, 58–77, 93–102 and
153–161 AA regions of HLA-DR4 (Fig. 8B). The epitope complex
structure was more stable throughout the simulation and LsrA inter- Fig. 7. Frequencies of contact formation by LsrA peptide residues. Proline256 is
action did not changed the radius of gyration of apo and LsrA epitope most frequent residue contact with HLA-DR4 compared to the other amino acid
bound form of HLA-DR4where the Rg shuttles between 1.8 and in LsrA epitope.
1.95 nm (Fig. 9).
The analysis confirms that LsrA epitope interaction with HLA-DR4

Fig. 5. Three dimensional crystal structure of HLA-DR4 docked with LsrA epitope (LELPGSRPQ). LsrA epitope represented as stick (red colour) and HLA-DR4
represented in blue colour helix. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

492
P. Vijayababu et al.
Fig 8. A) MD simulation analysis- Backbone RMSD profile. The difference in RMSD was observed by large conformational changes in 14–19, 34–40, 58–77, 93–102
and 153–161 AA regions of HLA-DR4. B) RMS fluctuation. Average residual fluctuation where significant fluctuated regions are highlighted as R1 to R4.
can inhibit its function activities. In additional, the increase in the
hydrogen bond pattern with HLA-DR4 by the LsrA epitope interaction
conveys that binding was more favored with four fold increase from the
initial hydrogen bond numbers and it stabilized more than half of the
time period of the MD simulation. Analyzed structural transformation.
Overall MD analysis reveals that LsrA epitope was binding and inter-
acting firmly with the HLA-DR4. In silico epitope analysis were already
reported with other organism [20,54,55], functional studies and char-
acterization were need. Hence, we are proposing that LsrA epitope
would be a prominent epitope for human specific pathogen of S. typhi
that requires further steps to be elevated as a vaccine drug in near fu-
ture.
4. Conclusion
The systematic and successful identification of the epitopes as
well as MHC Class I and II development aided by computational
493
Microbial Pathogenesis 123 (2018) 487–495
based analysis can provide in-depth information on the possible
function in pathogenesis. This present discovery undergoes, the uti-
lity of in silico systematic drug target identification from S. typhi
LsrA, the predicted epitopes, (253 LELPGSRPQ 261) was antigenic
and have much potential to interact with most common human HLA
alleles and HLA -DR4. Specifically, our molecular dynamics simula-
tion studies, HLA-DR4-LsrA epitope complex were found to be flex-
ible and remain stable upto 20 ns. This epitope (253 LELPGSRPQ
261) could be considered as good peptide-based subunit vaccine
candidate. This, in turn will provide new knowledge for the devel-
opment of novel therapeutic strategies for S. typhi causing dieses like
gallbladder cancer and typhoid fever. This identified epitope also
requires proper experimental validation through competitive ELISA
functional studies and characterization for vaccine against human
pathogens.
P. Vijayababu et al.
Fig. 9. MD simulation-Radius of gyration and hydrogen analysis. A) Hydrogen bond was most similar to the apo- and LsrA epitope bound form of HLA-DR4 and B)
Hydrogen bond pattern between HLA-DR4 and LsrA epitope.
Acknowledgement
The authors thank to Gracy Fathima, Department of Virology, King
Institute of Preventive Medicine & Research, Guindy, Chennai - India,
for valuable suggestions in the production of manuscript.
Appendix A. Supplementary data
Supplementary data related to this article can be found at https://
doi.org/10.1016/j.micpath.2018.08.012.
References
[1] G. Samykannu, P. Perumal, S. Narayanan, Computational and experimental char-
acterization of integral membrane proteins at molecular level, in: A.R. Rao (Ed.),
Biotechnology: Bioinformatics and Computational Biology, stadium press LLC, USA,
2014, pp. 225–258.
[2] J. Kaur, S. Jain, Vi antigen of Salmonella enetrica serovar Typhi — biosynthesis,
regulation and its use as vaccine candidate, Open Life Sci. 7 (5) (2012).
[3] J.A. Crump, E.D. Mintz, Global trends in typhoid and paratyphoid fever, Clin.
Infect. Dis.: Offic. Publ. Infect. Dis. Soc. Am. 50 (2) (2010) 241–246.
[4] J.A. Crump, S.P. Luby, E.D. Mintz, The global burden of typhoid fever, Bull. World
Health Organ. 82 (5) (2004) 346–353.
[5] T. Pang, M.M. Levine, B. Ivanoff, J. Wain, B.B. Finlay, Typhoid fever-important
issues still remain, Trends Microbiol. 6 (4) (1998) 131–133.
[6] C.M. Parry, T.T. Hien, G. Dougan, N.J. White, J.J. Farrar, Typhoid fever, N. Engl. J.
E-mail addresses: princymcc@gmail.com, princy.vijayababu@buc.edu.in (P. Vijayababu).
494
Microbial Pathogenesis 123 (2018) 487–495
Med. 347 (22) (2002) 1770–1782.
[7] U. Dutta, P.K. Garg, R. Kumar, R.K. Tandon, Typhoid carriers among patients with
gallstones are at increased risk for carcinoma of the gallbladder, Am. J.
Gastroenterol. 95 (3) (2000) 784–787.
[8] C.W. Lai, R.C. Chan, A.F. Cheng, J.Y. Sung, J.W. Leung, Common bile duct stones: a
cause of chronic salmonellosis, Am. J. Gastroenterol. 87 (9) (1992) 1198–1199.
[9] S. Swift, M.C. Rowe, M. Kamath, Quorum sensing, Bacterial Physiol. (2008)
179–232.
[10] T. Xue, L. Zhao, H. Sun, X. Zhou, B. Sun, LsrR-binding site recognition and reg-
ulatory characteristics in Escherichia coli AI-2 quorum sensing, Cell Res. 19 (11)
(2009) 1258–1268.
[11] H.S. Garmory, R.W. Titball, ATP-binding cassette transporters are targets for the
development of antibacterial vaccines and therapies, Infect. Immun. 72 (12) (2004)
6757–6763.
[12] P.R. Jungblut, D. Bumann, Immunoproteome of Helicobacter pylori, Meth.
Enzymol. (2002) 307–316.
[13] M.S. Sakib, M.R. Islam, A.K.M.M. Hasan, A.H.M.N. Nabi, Prediction of epitope-
based peptides for the utility of vaccine development from fusion and glycoprotein
of nipah virus using in silico approach, Adv. Bioinformat. 2014 (2014) 1–17.
[14] S. Sankar, M. Ramamurthy, B. Nandagopal, G. Sridharan, Short peptide epitope
design from hantaviruses causing HFRS, Bioinformation 13 (7) (2017) 231–236.
[15] R. Apweiler, A. Bairoch, C.H. Wu, Protein sequence databases, Curr. Opin. Chem.
Biol. 8 (1) (2004) 76–80.
[16] I.A. Doytchinova, D.R. Flower, VaxiJen: a server for prediction of protective anti-
gens, tumour antigens and subunit vaccines, BMC Bioinf. 8 (2007) 4.
[17] S. Hunter, R. Apweiler, T.K. Attwood, A. Bairoch, A. Bateman, D. Binns, et al.,
InterPro: the integrative protein signature database, Nucleic Acids Res. 37
(Database issue) (2009) D211–D215.
[18] N.Y. Yu, J.R. Wagner, M.R. Laird, G. Melli, S. Rey, R. Lo, et al., PSORTb 3.0: im-
proved protein subcellular localization prediction with refined localization
P. Vijayababu et al.
subcategories and predictive capabilities for all prokaryotes, Bioinformatics 26 (13)
(2010) 1608–1615.
[19] Y. EL‐Manzalawy, D. Dobbs, V. Honavar, Predicting linear B‐cell epitopes using
string kernels, J. Mol. Recogn. 21 (4) (2008) 243–255.
[20] J. Chen, H. Liu, J. Yang, K.-C. Chou, Prediction of linear B-cell epitopes using amino
acid pair antigenicity scale, Amino Acids 33 (3) (2007) 423–428.
[21] H. Singh, G.P.S. Raghava, ProPred1: prediction of promiscuous MHC Class-I binding
sites, Bioinformatics 19 (8) (2003) 1009–1014.
[22] R.E.M. Toes, A.K. Nussbaum, S. Degermann, M. Schirle, N.P.N. Emmerich, M. Kraft,
et al., Discrete cleavage motifs of constitutive and immunoproteasomes revealed by
quantitative analysis of cleavage products, J. Exp. Med. 194 (1) (2001) 1–12.
[23] P. Guan, MHCPred: a server for quantitative prediction of peptide-MHC binding,
Nucleic Acids Res. 31 (13) (2003) 3621–3624.
[24] M. Oprea, F. Antohe, Reverse-vaccinology strategy for designing T-cell epitope
candidates for Staphylococcus aureus endocarditis vaccine, Biologicals 41 (3)
(2013) 148–153.
[25] P. Kangueane, M.K. Sakharkar, T-Epitope Designer: a HLA-peptide binding pre-
diction server, Bioinformation 1 (1) (2005) 21–24.
[26] S.F. Altschul, W. Gish, W. Miller, E.W. Myers, D.J. Lipman, Basic local alignment
search tool, J. Mol. Biol. 215 (3) (1990) 403–410.
[27] A. Roy, A. Kucukural, Y. Zhang, I-TASSER: a unified platform for automated protein
structure and function prediction, Nat. Protoc. 5 (4) (2010) 725–738.
[28] D. Xu, Y. Zhang, Improving the physical realism and structural accuracy of protein
models by a two-step atomic-level energy minimization, Biophys. J. 101 (10)
(2011) 2525–2534.
[29] R.A. Laskowski, M.W. MacArthur, D.S. Moss, J.M. Thornton, PROCHECK: a pro-
gram to check the stereochemical quality of protein structures, J. Appl. Crystallogr.
26 (2) (1993) 283–291.
[30] I. Mayrose, T. Shlomi, N.D. Rubinstein, J.M. Gershoni, E. Ruppin, R. Sharan,
T. Pupko, Epitope mapping using combinatorial phage-display libraries: a graph-
based algorithm, Nucleic Acids Res. 35 (1) (2006) 69–78.
[31] I. Mayrose, O. Penn, E. Erez, N.D. Rubinstein, T. Shlomi, N.T. Freund, et al.,
Pepitope: epitope mapping from affinity-selected peptides, Bioinformatics 23 (23)
(2007) 3244–3246.
[32] M. Blaszczyk, M. Jamroz, S. Kmiecik, A. Kolinski, CABS-fold: server for the de novo
and consensus-based prediction of protein structure, Nucleic Acids Res. 41 (W1)
(2013) W406–W411.
[33] P. Vijayababu, G. Samykannu, S. Narayanan, ProGene 1.0-an in silico tool for
protein-gene analysis, Int. J. Life Sci. Sci. Res. 3 (3) (2017) 1089–1093.
[34] M. Kurcinski, M. Jamroz, M. Blaszczyk, A. Kolinski, S. Kmiecik, CABS-dock web
server for the flexible docking of peptides to proteins without prior knowledge of
the binding site, Nucleic Acids Res. 43 (W1) (2015) W419–W424.
[35] M. Blaszczyk, M. Kurcinski, M. Kouza, L. Wieteska, A. Debinski, A. Kolinski,
S. Kmiecik, Modeling of protein–peptide interactions using the CABS-dock web
server for binding site search and flexible docking, Methods 93 (2016) 72–83.
[36] D. Van Der Spoel, E. Lindahl, B. Hess, G. Groenhof, A.E. Mark, H.J.C. Berendsen,
GROMACS: fast, flexible, and free, J. Comput. Chem. 26 (16) (2005) 1701–1718.
[37] M. Kawata, M. Mikami, Rapid calculation of two-dimensional Ewald summation,
Chem. Phys. Lett. 340 (1–2) (2001) 157–164.
[38] B. Hess, H. Bekker, H.J. Berendsen, J.G. Fraaije, LINCS: a linear constraint solver for
495
View publication stats
Microbial Pathogenesis 123 (2018) 487–495
molecular simulations, J. Comput. Chem. 18 (12) (1997) 1463–1472.
[39] S. Miyamoto, P.A. Kollman, Settle: an analytical version of the SHAKE and RATTLE
algorithm for rigid water models, J. Comput. Chem. 13 (8) (1992) 952–962.
[40] H.J.C. Berendsen, J.P.M. Postma, W.F. van Gunsteren, A. DiNola, J. Haak,
Molecular dynamics with coupling to an external bath, J. Chem. Phys. 81 (1984)
3684–3690.
[41] M. Parrinello, A. Rahman, Polymorphic transitions in single crystals: a new mole-
cular dynamics method, J. Appl. Phys. 52 (12) (1981) 7182–7190.
[42] C.F. Chin, J.Y. Lai, Y.S. Choong, A.A. Anthony, A. Ismail, T.S. Lim, Delineation of B-
cell epitopes of Salmonella enterica serovar typhi hemolysin E: potential antibody
therapeutic target, Sci. Rep. 7 (1) (2017).
[43] M. Koblischke, M.S. Mackroth, J. Schwaiger, I. Fae, G. Fischer, K. Stiasny, et al.,
Protein structure shapes immunodominance in the CD4 T cell response to yellow
fever vaccination, Sci. Rep. 7 (1) (2017).
[44] N. Khatoon, R.K. Pandey, V.K. Prajapati, Exploring Leishmania secretory proteins to
design B and T cell multi-epitope subunit vaccine using immunoinformatics ap-
proach, Sci. Rep. 7 (1) (2017).
[45] V. Wiwanitkit, Predicted epitopes of Lig A of Leptospira interrogans by bioinfor-
matics method: a clue for further vaccine development, Vaccine 25 (15) (2007)
2768–2770.
[46] B. Wang, K. Yao, G. Liu, F. Xie, F. Zhou, Y. Chen, Computational prediction and
identification of epstein-barr virus latent membrane protein 2A antigen-specific
CD8+ t-cell epitopes, Cell. Mol. Immunol. 6 (2) (2009) 97–103.
[47] S.J. Dunstan, H.A. Stephens, J.M. Blackwell, C.M. Duc, M.N. Lanh, F. Dudbridge,
et al., Genes of the class II and class III major histocompatibility complex are as-
sociated with typhoid fever in vietnam, J. Infect. Dis. 183 (2) (2001) 261–268.
[48] G. Samykannu, P. Vijayababu, C.B. Antonyraj, S. Narayanan, S.I. Basheer Ahamed,
Investigations of binding mode insight in Salmonella typhi type-III secretion system
tip protein (SipD): a molecular docking and MD simulation study, Inf. Med.
Unlocked 9 (2017) 166–172.
[49] G. Samykannu, P. Vijayababu, J. Natarajan, Substrate specificities in Salmonella
typhi outer membrane protease ( PgtE ) from Omptin family – an in silico proteomic
approach, Inf. Med. Unlocked 12 (2018) 6–13.
[50] P. Vijayababu, G. Samykannu, S. Narayanan, Structural insights of cell-cell com-
munication autoinducer-2 kinase (LsrK) model from Salmonella typhi, Int. J. Curr.
Res. 9 (2017) 51169–51172.
[51] P. Vijayababu, G. Samykannu, C.B. Antonyraj, J. Thomas, S. Narayanan,
S.I. Basheer Ahamed, S. Piramanayagam, Patulin interference with ATP binding
cassette transferring auto inducer −2 in Salmonella typhi and biofilm inhibition via
quorum sensing, Inf. Med. Unlocked 11 (2018) 9–14.
[52] J. Yang, R. Yan, A. Roy, D. Xu, J. Poisson, Y. Zhang, The I-TASSER Suite: protein
structure and function prediction, Nat. Meth. 12 (1) (2015) 7–8.
[53] K. Prabhavathy, P. Perumal, N. SundaraBaalaji, In silico identification of B- and T-
cell epitopes on OMPLA and LsrC from Salmonella typhi for peptide-based subunit
vaccine design, Indian J. Biotehnol. 10 (2011) 440–451.
[54] S. Idrees, U.A. Ashfaq, Structural analysis and epitope prediction of HCV E1 protein
isolated in Pakistan: an in-silico approach, Virol. J. 10 (1) (2013) 113.
[55] K. Wang, C. Xie, J. Zhang, W. Zhang, D. Yang, L. Yu, et al., The identification and
characterization of two novel epitopes on the nucleocapsid protein of the porcine
epidemic diarrhea virus, Sci. Rep. 6 (1) (2016).