J. Biomedical Science and Engineering, 2010, 3, 415-421 JBiSE
doi:10.4236/jbise.2010.34057 Published Online April 2010 (http://www.SciRP.org/journal/jbise/).
Published Online April 2010 in SciRes. http://www.scirp.org/journal/jbise
Insilico structural analysis of parasporin 2 protein sequences
of non-toxic bacillus thuringiensis
Ayyasamy Mahalakshmi, Rajaiah Shenbagarathai
PG Research Department of Zoology and Biotechnology, Lady Doak College, Madurai-2, Tamilnadu, India.
Email: shenbagarathai@rediffmail.com
Received 17 September 2009; revised 5 December 2009; accepted 7 December 2009.
The unusual and remarkable property of parasporin
2 of non-insecticidal Bacillus thuringiensis is specifi-
cally recognizing and selectively targeting human
leukemic cell lines. The 37-kDa inactive nascent
protein is proteolytically cleaved to the 30-kDa ac-
tive form that loses both the N-terminal and the
C-terminal segments. Accumulated cytological and
biochemical observations on parasporin-2 imply
that the protein is a pore-forming toxin. To confirm
the hypothesis, insilico analysis was performed us-
ing homology modeling. The resulting model of
parasporin 2 protein is unusually elongated and
mainly comprises long β-strands aligned with its long
axis. It is similar to aerolysin-type β-pore-forming
toxins, which strongly reinforce the pore-forming
hypothesis. The molecule can be divided into three
domains. Domain 1, comprising a small β-sheet
sandwiched by short α-helices, is probably the tar-
get-binding module. Two other domains are both
β-sandwiches and thought to be involved in oli-
gomerization and pore formation. Domain 2 has a
putative channel-forming β-hairpin characteristic of
aerolysin-type toxins. The surface of the protein has
an extensive track of exposed side chains of serine
and threonine residues. The track might orient the
molecule on the cell membrane when domain 1 binds
to the target until oligomerization and pore forma-
tion are initiated. The β-hairpin has such a tight
structure that it seems unlikely to reform as postu-
lated in a recent model of pore formation developed
for aerolysin-type toxins. Parasporin 2 (Accession no:
BAD35170) protein sequence analysis indicated two
different domains namely, aerolysin toxin and clos-
tridium toxin domain based on different database
searches (CDD and Pfam). It showed a close similar-
ity with the available PDB template (PDB id: 2ZTB)
of parasporin which has cytocidal activity against
MOLT-4, HL60 and Jurkat cell lines. Based on the
PSI Blast analysis, 3D structures of the domains
were predicted by using Swiss model server. Accu-
racy of the prediction of 3D structure of different
domains of parasporin protein was further vali-
dated by Ramachandran plot and PROCHECK
(G-value). The structure is dominated by β-strands
(67%, S1-12), most of which are remarkably ex-
tensive, running all or most of the longer axis of the
molecule. This study helped to elucidate the 3D
structure of parasporin 2 (Acc. No. BAD35170)
which might enable to probe further its specific
mechanism of action. Though the similarity is ob-
served in the domain architecture, there is variation
in the regions of the domains even among the same
group of parasporin 2. Docking of this model struc-
ture and experimental structure with specific recep-
tors of the cancer cells will facilitate to explore
mechanism of parasporin 2 action and also provide
information about its evolutionary relationship with
toxic Cry proteins.
Keywords: Parasporin; Homology Model; Non-Toxic;
Bacillus thuringiensis; Cell Lines
Since the incidence of new cancer patients is increasing
annually due to altered food habits and life styles, efforts
are being made worldwide to identify new molecular
markers and therapeutic agents for the purpose of diagno-
sis and treatment of the same. The existing chemothera-
peutics not only affect tumor cells but also normal cells.
Hence, search for compounds which can specifically target
the cancer cells will overcome the existing problem [1].
At present, four genealogically different parasporins are
identified as parasporin-1 to parasporin-4 that has the
ability to specifically act against cancer cells [2]. This area
of research is still under exploration, since a very few of
the literature is available related to parasporin structure
and mechanism of action. Parasporin 2 is known to inter-
act with GP-I protein and the cell death induced by
parasporin-2 is non-apoptotic, although the apoptotic
process occurs when the cell damage proceeded slowly.
A. Mahalakshmi et al. / J. Biomedical Science and Engineering 3 (2010) 415-421
Copyright © 2010 SciRes. JBiSE
Parasporin-2 increases the plasma membrane permeability
of the target cells as it binds to a detergent-resistan t mem-
brane, the so-called “lipid raft” in a plasma membrane,
and then forms the SDS-resistant oligomer embedded in
the membrane. This toxin binds GPI-proteins in lipid raft,
and then seems to form the oligomer that can permeabi-
lize the plas ma me mbrane. Th is is follo wed by the forma-
tion of oligomers (> 200 kDa) of PS2Aa1 in plasma
membranes, leading to pore formation and cell lysis. The
oligomerization occurs in the presence of membrane pro-
teins, lipid bilaye r and cholesterols [3].
Only two experimentally determined structures (PDB:
2ZTB, 2D42) are available till date as confirmed in the
PDB. Hence, alternative strategies are being applied
to develop theoretical models of protein structure of
parasporin 2 (Accession no: BAD35170) when X-ray
diffraction or NMR structures are not available, aiming
to bridge the structure-knowledge gap. Higher resolution
models, derived from relationships with better than
about 30% sequence identity or refined from lower
resolution starting models, are very helpful in assigning
detailed aspects of molecular function [4].
2.1. Domain identification
The complete sequence of parasporin protein (Bacillus
thuringiensis) available from NCBI till date revealed the
aerolysin toxin and clostridium toxin domain (Ta bl e 1).
Domains present one of the most useful levels at which
to understand protein function and the domain fam-
ily-based analysis has had a profound impact on the
study of individual proteins [5].
2.2. Template Identification by Fold-Recognition
Two approaches are employed to identify the potential
templates by submitting a multiple sequence alignment
(MSA) of all the parasporin sequences and submitting
each of the parasporin sequences individually. Thus, in
order to identify a template structure for modeling of
parasporin protein sequences, we used the comparative
modeling approach (match of secondary structure ele-
ments, compatibility of residue-residue contacts, etc.). In
the former, MSA for the entire sequence from N-to
C-terminus (Figure 1) was submitted to the FUGUE
server; the template 2ZTB was identified [10] with very
high confidence levels (Z-score for the top hit = 31.32;
certain). Even the GeneSilico metaserver identified
2ZTB template with reliable confidence levels (3D-Jury
score for the top hit = 133; reliable). In view of these,
both the servers identified 2ZTB as the top hit (Z-score
= 31.3.2; certain and 3DJury score = 133; reliable).
In the second approach, complete sequence of the
parasporin was used separately as query to search for
homologs in the PDB database using BLAST and
PSI-BLAST. The fold-recognition servers, GeneSilico
metaserver, FUGUE and SAM-T02 identified 2ZTB as
the possible template only for parasporin sequence (Acc.
No. BAD35170) with a high level of confidence (Table 2).
Despite the scores reported by the individual threading
methods were hardly significant, the consensus server
Pcons5 [11] assigned a significant score (1.35) to the
2ZTB structure as a potential modeling template as evi-
dent in its sequence alignment (Figure 2).
2.3. Modeling 3-D Structure and Stereochemical
Evaluation of the Predicted Models
The 3-D structure of parasporin sequence at 2.38 A
resolution (PDB id: 2ZTB, A chain) is the main template
for modeling only one parasporin sequence (Acc. No.
BAD35170) with 88% identity, since, all other p ar as por in
sequences had less than 26% identity with its corre-
Table 1. Domains of parasporin protein sequence as suggested by Pfam-A search and conserved domain database search
(http://pfam.sanger.ac.uk/search; http://www.ncbi.nlm.nih.gov/).
S.no Accession no Significant
Domains identified start end
Bit score E-value
1 BAD35170 ------- ----- ---- ---- -----
2 BAE44986 delta endotoxin, N-terminal domain 114 363 33.1 9.2e-11
3 BAE44985 delta endotoxin, N-terminal domain 95 344 32.9 9.4e-11
4 BAE44984 delta endotoxin, N-terminal doma i n 114 363 36.3 5.8e-11
5 BAE44983 delta endotoxin, N-terminal domai n 114 363 32.5 1e-10
6 BAE79808 delta endotoxin, N-terminal domain 147 396 33.1 9.2e-11
7 BAE79809 delta endotoxin, N-terminal domain 147 396 32.9 9.4e-11
8 BAB11757 delta endotoxin, N-termi n a l d omain 147 396 33.1 9.2e-11
A. Mahalakshmi et al. / J. Biomedical Science and Engineering 3 (2010) 415-421
Copyright © 2010 SciRes. JBiSE
Figure 1. Multiple sequence alignment of parasporin protein residues *-conserved residues.
Ta b le 2 . Summary of the best template sequence profile that was generated at the end of 3rd iteration of PSI Blast
analysis using the parasporin protein sequence of B.thuringiensis as query. Threshold PSI-BLAST E-value = 0.001.
S.no Accession no Query
region Template
sequence id
Region of
template seq
aligned % identityGap (%)E-valueAnnotation of the template
1 BAD35170 28-278 2ZTB 2-252 88 0 2e-114Parasporin 2 crystal structure of
Bacillus thuringiensis
2 BAE44986 119-383 1DLC 18-251 26 15 1e-12
Crystal Structure Of Insecticidal
3 BAE44985 101-364 1DLC 19-251 26 15
Crystal Structure Of Insecticidal
4 BAE44984 109-383 1DLC 5-251 26 16 6e-13
Crystal Structure Of Insecticidal
5 BAE44983 109-383 1DLC 5-251 26 16 1e-12
Crystal Structure Of Insecticidal
6 BAE79808 153-416 1DLC 19-251 25 15 2e-12
Crystal Structure Of Insecticidal
7 BAE79809 153-416 1DLC 19-251 26 15 2e-12
Crystal Structure Of Insecticidal
8 BAB11757 153-416 1DLC 19-251 25 15 2e-12
Crystal Structure Of Insecticidal
A. Mahalakshmi et al. / J. Biomedical Science and Engineering 3 (2010) 415-421
Copyright © 2010 SciRes. JBiSE
Figure 2. Sequence alignment of the parasporin protein (Acc. No. BAD35170) with its tem-
plate (PDB id: 2ZTB-A chain).
sponding template (Figure 3). The final averaged and
optimized model passed all the tests implemented in the
stereochemistry-evaluating WHATCHECK suite [12-15]
and in the VERIFY3D program, which uses contact po-
tentials to assess whether the modeled amino acid resi-
dues occur in the environment typical for globular pro-
teins with h ydrophobic core an d solvent-e xposed surface
(Eisenberg et al., 1997). Moreover, the reasonable ener-
gies are rarely observed for misfolded structures. Thus,
the scores reported for our model by WHATCHECK
(Z-score-4.1) and VERIFY3D (average score 0.3, no
regions scored lower than 0) suggest that both its
three-dimensional fold and the conformation of individ-
ual residues are reasonable.
The selected model, the value of the objective func-
tion, reported as current energy is in the same range as
that if the template is aligned with its own sequence. On
an average, 99.1% of the residues are found in the al-
lowed region of Ramachandran map, PROCHECK con-
siders the model to be very good if it has 90% of the
residues in the most favored region. The inter-atomic
distances are within acceptable range. Verify3D score is
greater than zero for the entire model (1 to 274 residues).
The models were also evaluated using Colorado3D
server, which facilitates the change of amino acid win-
dow size when calculating the overall score. Two win-
dow sizes, 5 and 21, were used to calculate the average
Verify3D and ProsaII score per residue for the model
[12,13]. The scores calculated using these two window
sizes were found to be very similar. The template and
target models were rendered with the residues color-coded
based on ProsaII and Verify3D scores. With ProsaII
score-based coloring, most of the residues are green and
yellow (i.e., average score) in both the target and tem-
plate proteins. Z-score of a model is a measure of com-
patibility between its sequence and structure, the model
Z-score should be comparable to that of the template.
With Verify3D score-based coloring, even the template
proteins has residues in red color (i.e., bad score) al-
though the number of such residues are more in the tar-
2.4. Description of the Model
The structure is dominated by β-strands (67%, S1–12),
most of which are remarkably extensive, running all or
most of the longer axis of the molecule (Figure 3). The
longest strand (S8), comprising 21 residues, runs the
whole length of the molecule; three others (S11, S12,
S14) span two-thirds of the molecule’s length. The
molecule can be divided into three domains: domain 1
(Val29-His85, Gly145-Ser173), domain 2 (Ile36-Pro53,
Ala81-Asp138, Glu160-Ser175, Gly211-Gln236), and do-
main 3 (Val54-Asn80, Phe139-Leu159, Thr237-Ala250).
Domain 1 is composed of a short N-terminal β-strand
Figure 3. Homology model of the parasporin protein sequence
(Acc.No. BAD35170).
A. Mahalakshmi et al. / J. Biomedical Science and Engineering 3 (2010) 415-421
Copyright © 2010 SciRes. JBiSE
(S1: 30-34), a α/β structure (H1 and S2: 46-66), a
β-hairpin (S9 and S10: 152-168), and a α-helix (H2: 79
-83). α-Helices are found only in this domain. The two
α-helices, H1 and H2, are close together and short, oc-
cupying only 8.9% of the molecule. Domains 2 and 3 are
both β-sandwiches: the former is made of a two-stranded
β-hairpin (S6 and S7) and a curled anti-parallel five
stranded β-sheet including S3, S9, S12, S5 and S8; and
the latter is of antiparallel three-stranded and two-
stranded β-sheets (S4, S9 and S12; S5 and S8). These
results (Figure 3, Table 3) are consistent with earlier
reports [14]. This model is expected to yield insight into
the molecular function and mechanism of parasporin
action [15].
The initial step in cytocidal action of PS2Aa1 is the spe-
cific binding of this cytotoxin to a putative receptor lo-
cated in the lipid rafts, followed by its oligomerization
and pore formation in plasma membrane.Secondary
structures observed in the model (Accession No: BAD
35170) are organized virtually in the same way as the
experimentally determined parasporin-2 protein (PDb id:
2ZTB). Even though 2ZTB and 2D42 belong to the
parasporin type 2 protein sequence, BAD35170 bears
greater similarity to 2ZTB in its domain organization.
4.1. Databases
The amino acid sequences of the experimentally chara-
cterized parasporin sequences (Ta ble 4) were retrieved
from the protein sequence database at NCBI http://
www.ncbi.nlm.nih.gov. The 3-D structures of proteins
Table 3. Structural superposition report of
the model generated for BAD35170 with its
corresponding template, 2ZTB.
Table 4. List of parasporin sequences retrieved from NCBI
S.no Accession no Type of protein
1 BAD35170
Parasporin 2 Ab cytotoxic Crystal
protein (Bacillus thuringiensis)
2 BAE44986
Putative uncharacterized protein 1
(Bacillus thuringiensis)
3 BAE44985
Putative uncharacterized protein 2
(Bacillus thuringiensis)
4 BAE44984
Putative uncharacterized protein 3
(Bacillus thuringiensis)
5 BAE44983
Putative uncharacterized protein 4
(Bacillus thuringiensis)
6 BAE79808
81-k da leukemia protein (Bacillus
7 BAE79809
Cry 31-like 82-k da protein (Bacil-
lus thuringiensis)
8 BAB11757
81-k da leukemia toxin (Bacillus
were obtained from the protein data bank [16]. The
fold classification of proteins is from the SCOP data-
base [17].
4.2. Servers
Protein sequence databases we r e searched using PSI-B LAST
[18] servers at NCBI, PHYRE (successo r of 3 D-PSSM) [19],
SAM-T02 [20] and GeneSilico Metaserver [21] were used
for fold-recognition. Multiple sequence alignments were
obtained using the CLUSTALW server [22]. Verify3D
[13] and Colorado3D [23] were used to evaluate the
models. All the servers were used with default values for
the various parameters, except where mentioned otherwise.
4.3. Software and Hardware
SwissPDBviewer [24] was used for visualization and/or
rendering. The stereochemical quality of the generated
model was assessed using PROCHECK [10]. Default
values were used for all the parameters, unless specified
4.4. Te m plate-Target Sequence Al ignmen t
The parasporin sequences were submitted for structure
prediction using comparative modeling technique. The
preliminary models were obtained using unrefined pairwise
alignments reported by PSI-BLAST [25]. Energy
minimization was carried out using GROMOS96 [26]
until all inconsistencies in geometry were rectified and
all the short contacts were relieved. The stereochemical
and energetic properties of modeling intermediates and
of the final model were evaluated using WHATCHECK
[27] and VERIFY3D [28]. Semi-automated and manual
manipulations with protein structures and sequence–structure
alignments were conducted using SWISS-PDB VIEWER
[24]. All the servers provide alignment of the submitted
A. Mahalakshmi et al. / J. Biomedical Science and Engineering 3 (2010) 415-421
Copyright © 2010 SciRes. JBiSE
parasporin sequence (target) with the sequence of the
potential hits (templates).
4.5. Validation of Predicted 3-D Structures
The stereochemical properties of predicted 3-D structures
were assessed by PROCHEC K and t he residue envi ronm ents
by Verify3D and Colorado3D. Regions that are found
by these servers as poorly modeled were improved by
manual adjustment of alignments and re-modeling.
The authors acknowledge N.Lavanya Roselin for acquisition of data.
Authors express their heartfelt thanks for DBT-BIF in providing the
necessary infrastructure facility in the collection, analysis, and inter-
pretation of data; in the writing of the manuscript; and in the decision
to submit the manuscript for publication.
BLAST server: http://www.ncbi.nlm.nih.gov/BLAST/
Colorado3D: http://asia.genesilico.pl/colorado3d/
FUGUE: http://www-cryst.bioc.cam.ac.uk/fugue/
GeneSilico Metaserver: http://genesilico.pl/meta
PDB: http://www.rcsb.org
PHYRE: http://www.sbg.bio.ic.ac.uk/~phyre
http://www.biochem.ucl.ac.uk/~roman / p r o c h ec k / p r o c he c k .
SCOP database: http://scop.mrc-lmb.cam.ac.uk/scop/
SwissPDBviewer: http://ca.expasy.org/spdbv/
Verify_3D: http://nihserver.mbi.ucla.edu/Verify_3D/
[1] Mizuki, E., Ohba, M., Akao, T., Yamashita, S., Saitoh, H.
and Park, Y.S. (1999) Unique activity associated with
non-insecticidal Bacillus thuringiensis parasporal inclu-
sions: In vitro cell-killing action on human cancer cells.
Journal of Applied Microbiology, 86, 477-86.
[2] Katayama, H., Yokota, H., Akao, T., Nakamura, O., Ohba,
M., Mekada, E. and Mizuki, E. (2005) Pa raspo rin-1, a nove l
cytotoxic protein to human cells from non-insecticidal
parasporal inclusions of Bacillus thuringiensis. Biochem-
istry, 137, 17-25.
[3] Abe, Y., Shimada, H. and Kitada, S. (2008) Raft-targeting
and oligomerization of parasporin-2, a Bacillus thuringiensis
crystal protein with Anti-Tumour activity. Biochemistry,
143(2), 269-275.
[4] Murray, D. and Honig, B. (2002) Electrostatic control of
the membrane targeting of C2 domains. Molecular Cell,
9, 145-154.
[5] Copley, R.R., Doerks, T., Letunic, I. and Bork, P. (2002)
Protein domain analysis in the era of complete genomes.
FEBS Letters, 20, 129-134.
[6] Godzik, A. (2003) Fold recognition methods. Methods of
Biochemical Analysis, 44, 525-546.
[7] Kurowski, M.A., Bujnicki, J. M. (2003) GeneSilic o protein
structure prediction meta-server. Nucleic Acids Research,
31, 3305-3307.
[8] Akiba, T., Abe, Y., Kitada, S., Kusaka, Y., Ito, A., Ichi-
matsu, T., Katayama, H., Akao, T., Higuchi, K., Mizuk i, E . ,
Ohba, M., Kanai, R. and Harata, K. (2009) Crystal struc-
ture of the parasporin-2 of Bacillus thuringiensis toxin
that recognizes cancer cells. Journal of Molecular Biol-
ogy, 386, 121-133.
[9] Lundstrom, J., Rychlewski, L., B ujnicki, J.M. an d Elofsson,
A. (2001) Pcons: A neural-network-based consensus predic-
tor that improves fold recognition. Protein Science, 10,
[10] Laskowski, R.A., MacArthur, M.W., Moss, D.S. and
Thornton, J.M. (1993) PROCHECK: A program to check
the stereochemical quality of protein structures. J Appl
Cryst, 26, 283-291.
[11] Hooft, R.W.W., Vriend, G., Sander, C. and Abola, E.E.
(1996) Errors in protein structures. Nature, 381, 272-272.
[12] Sippl, J. (1993) Recognition of error s in three dimensional
structures of proteins. Proteins, 17, 355-362.
[13] Lüthy, R., Bowie, J.U. and Eisenberg, D. (1992) Assessment
of protein models with three-dimensional profiles. Nature,
5, 83-5.
[14] Brenner, S.E. (2001) A tour of structural genomics.
Nature Reviews Genetics, 2, 801-809.
[15] Berman, H.M., Westbrook, J., Feng, Z., Gilliland, G.,
Bhat, T.N., Weissig, H., Shindyalov, I.N. and Bourne, P.E.
(2000) The protein data bank. Nucleic Acids Research,
28, 235-242.
[16] Andreeva, A., Howorth, D., Brenner, S.E., Hubbard, T.J.,
Chothia, C. and Murzin A.G. (2004) SCOP database in
2004: Refinements integrate structure and sequence fam-
ily data. Nucleic Acids Research, 32, 226-229.
[17] Altschul, S.F., Madden, T.L., Schaffer, A.A., Zhang, J.,
Zhang, Z., Miller, W. and Lipman, D.J. (1997) Gapped
BLAST and PSI-BLAST: A new generation of protein
database search programs. Nucleic Acids Research, 25,
[18] Shi, J., Blundell, T.L. and Mizuguchi, K. (2001) FUGUE:
sequence-structure homology recognition using environment-
specific substitution tables and structure-dependent gap
penalties. Journal of Molecular Biology, 29, 243-57.
[19] Kelley, L.A., MacCallum, R.M., Sternberg, M.J.E. (2000)
Enhanced genome annotation using structural profiles in
the program 3D-PSSM. Journal of Molecular Biology,
299, 499-520.
[20] Karplus, K., Karchin, R., Draper, J., Casper, J., Mandel-
Gutfreund, Y., Diekhans, M. and Hughey, R. (2003)
Combining local-structure, fold-recognition, and new
fold methods for protein structure prediction. Proteins,
53, 491-496.
[21] Kurowski, M.A. and Bujnicki, J.M. (2003) GeneSilico
protein structure prediction meta-server. Nucleic Acids
Research, 31, 3305-3307.
[22] Thompson, J.D., Higgins, D.G. and Gibson, T.J. (1994)
CLUSTAL W: Improving the sensitivity of progressive
multiple sequence alignment through sequence weighting,
position-specific gap penalties and weight matrix choice.
Nucleic Acids Research, 22, 4673-4680.
[23] Sasin, M. and Bujnicki, J.M. (2004) COLORADO3D, a
web server for the visual analysis of protein structures.
Nucleic Acids Research, 32, 586-589.
[24] Guex, N. and Peitsch, M.C. (1997) SWISS-MODEL and
the Swiss-PdbViewer: An environment for comparative
protein modeling. Electrophoresis, 18, 2714-2723.
[25] Altschul, S.F, Madden, T.L., Schäffer, A.A., Zhang, J.,
Zhang, Z., Miller, W. and Lipman, D.J. (1997) Gapped
blast and PSI-blast: A new generation of protein database
search programs. Nucleic Acids Research, 25(17),
A. Mahalakshmi et al. / J. Biomedical Science and Engineering 3 (2010) 415-421
Copyright © 2010 SciRes. JBiSE
[26] van Gunsteren, W.F., Billeter, S.R., Eising, A.A., Hünen-
berger, P.H., Krüger, P., Mark, A.E., Scott, W.R.P. and
Tironi, I.G. (1996) Biomolecular simulation. The GRO-
MOS96 Manual and User Guide, Zürich, Groningen.
[27] Hooft, R.W.W., Vriend, G., Sander, C. and Abola, E.E.
(1996) Errors in protein structures. Nature, 381, 272-272.
[28] Eisenberg, D., Lüthy, R. and Bowie, J.U. (1997) VERIFY3D:
Assessment of protein models with three-dimensional
profiles. Methods Enzymol, 277, 396-404.