Transcript Mycobacterium tuberculosis - CBS

Stine R. Lund, Charlotte Kvennefors, Hengameh Mirsepasi, Rikke V. Benjaminsen
27685 Immunological Bioinformatics CBS, DTU, 2006
Introduction
Vaccines
The bacterial pathogen Mycobacterium tuberculosis (Mtb) is the causative
agent of Tuberculosis (TB) in humans. This disease can be easily spread and has
infected an estimated 2 billion people worldwide [1], causing millions of deaths
each year. Mtb is a facultative intracellular parasite that usually infects the lung
alveoli and alveolar ducts by entering inactivated macrophages in this area. The
lymphocytes, especially the T-cells, are the major defense against tuberculosis.
CD4+ T cells especially, play a major role in restricting infection, where Th1
response mediates protection [2]. Although T-cell immune response has been
mentioned as the major protective response in TB immunity, a recent study has
also suggested B-cells and antibodies to mediate protection against TB [3]. Here
we present MHC class I and II epitopes, B-cell epitopes as well as a polytope
from the Ag85A and B (swissprot ID: A85A MYCTU A85B MYCTU respectively)
There is an existing vaccine against TB, the so called BCG vaccine. This
was constructed using the attenuated strain of Mycobacterium bovis. However, the
vaccine has several limitations such as varying efficiency, waning protection in
adolescence and no protection against pulmonary TB in adults [2]. There is also an
increasing need for safer vaccines in immuno-comprimised individuals. Many HIV
infected patients often develop TB, therefore therapeutically or protective new
vaccines not involving live vaccines are required.
Several molecules from Mtb have been identified as potential antigens in inducing
immune responses towards Mtb. These are often secreted molecules that may aid in
walling off the bacteria. The AG85 complex belongs to this category, is an early
stage protein and thought to facilitate tubercule formation [4]. It includes the
molecules Ag85A, B and C. These molecules has been shown to stimulate cellular as
well as humoral immunity [5]. Hence, the molecules in this complex are
suitable for prediction of epitopes to be used in vaccine design.
B-cell epitope
MHC class II epitope
Two of the epitopes (10 and 11) are potential
major immuno-dominants [8]; however,
these areas have variable amino acid regions
and are not conserved. Epitopes 6 and 9 are
minor immuno-dominants, but both contain
conserved amino acids, with epitope 6 being
the most conserved. Hence, although epitope
10 and 11 might appear as superior epitopes
for vaccine design, the pathogen may induce
mutations in this area for immune escape.
Therefore, epitopes 6 (res 157-163) and 9
(res 222-233) are preferred (as in
order) in this prediction.
Figure 2: Structural model of Ag85A showing
predicted epitopes as colored sticks. Epitope 6 (res
157-163), magenta; epitope 9 (res 222-233), yellow;
epitope 10 (res 256-267), red; epitope 11 (res 286308), blue.
MHC class I epitopes
The secreted Mtb protein Ag85A has been shown
to induce antibody and cell-mediated immune
response [6] and the crystal structure of this
protein has been solved [7]. This molecule
therefore provides a solid basis for B-cell epitope
predictions. In addition, the structure of the
protein has not been shown to induce any
structural changes when binding to targets of
the protein [7].
The Ag85A protein structure was modeled on the
CPHmodels 2.0 Server (E-value e-160). Potential
B-cell epitopes were predicted using the
BepiPred 1.0 server with score threshold for
epitope assignment set at 0.85, resulting in
several
potential
epitopes.
Non-protruding
epitopes were deselected and loops and turns for
the protein were viewed on the swissprot
directory. Immuno-dominant epitopes as found
from the related Ag85B (highly homologous to
Ag85A) [8] together with structurally favorable
epitopes were further selected. This resulted in 4
potential epitopes as viewed in figure 1.
When predicting MHC class I epitopes it is
important to consider MHC polymorphism. Each
MHC molecule has a different specificity and
epitopes need to be selected carefully to cover the
entier, or at least most of the population. To solve
this, epitopes are selected from HLA supertypes,
which is groups of HLA molecules with similar
specificity. So far HLA class I molecules are divided
into 9 supertypes and previous work has show that
6 of them (A1, A2, A3, A24, B7 and B44) cover
98.1 to 100 % of the population [9].
Ag85A was used for prediction of MHC-class I
binding epitopes using the NetCTL network. This
method includes prediction of the antigenprocessing steps MHC class I binding (NetMHC),
proteasomal cleavage (NetChop C-term 3.0) and
transporter associated with antigen processing
(TAP) transport efficiency. The MHC class I binding
was based on the six supertypes mentioned above.
The two epitopes with highest score, for each
supertype were selected.
The identified epitopes were:
DSGTHSWEY
SSALTLAIY
GLLDPSQAM
AIYHPQQFV
AMGDAGGYK
ALYLLDGLR
IYHPQQFVY
GWDINTPAF
RVRGAVTGM
GPTLIGLAM
WETFLTSEL
FEWYDQSGL
Polytope
For this method the proteins Ag85A and B were
used. Ag85B is also included since it has been shown
that a combination of Ag85A and B results in a higher
CTL response than Ag85A alone [10]. The epitopes were
selected as described above in the MHC-class I epitopes
section. The two epitopes with highest score, from each
supertype from both proteins were selected. A total
number of 24 epitopes were linked and the resulting
polytope was optimized in Polytope Optimizer. This
program optimizes the position of the epitopes and the
linker region for better expression of the epitopes,
stronger C-terminal cleavage and less internal cleavage
(see figure 3 for Polytope construct before and after
optimisation). For the final polytope the MHC class II
epitope was attaced at the end resulting in the final
sequence.
Linker
MHC class I epitopes
MHC class II epitope
msGPSLIGLAMviRAWGRRLMISSALTLAIYyywGWDINTPAFaadALLDPSQGMtnsyaNTPAFE
WYYeaGLLDPSQAMyWETFLTSELyydRVRGAVTGMalyyQSSFYSDWYALYLLDGLRslDSGTH
SWEYaqllFEWYDQSGLraAMGDAGGYKkfAYHPQQFIYycrlyFEWYYQSGLAVYLLDGLRafwrv
gGPTLIGLAMAIYHPQQFVstwrwWETFLTSELnIYAGSLSALalAMGDAGGYKyiakwLMIGTAAA
VaIYHPQQFVYnivLPGWLQANRHVKPTG
Discussion/Conclusion
When designing a vaccine several factors must be taken into account. B-cell
response often require partial or whole proteins, whilst T-cell response can be
induced by epitopes. The use of epitopes also avoid potential toxic properties of
whole proteins. DNA vaccines may induce both humoral and cellular responses
which can be modulated via specific cytokine co-expression [11]. DNA vaccines
may not be entirely risk free (integration into genome) or as of yet too efficient,
but the method is under development. A polytope DNA vaccine as seen in this
study might be preferred as T-cell response is most important in TB immunity.
For the MHC class II epitope a 15 mer
peptide starting at position 148 of the protein
Antigen 85A is chosen as the best target for a
vaccine. The peptide sequence is:
LPGWLQANRHVKPTG
With the EasyGibbs web-server a matrix was
obtained by training with a set of peptides
known to bind to HLA-DRB1*0401. When
comparing the logo obtained and the anchor
positions found with SYFPEITHI the matrix was
improved and the Pearson coefficients and Aroc
values of the method can be seen in table 1.
These values are also shown for the TEPITOPE
method. In figure 2 the logo for the improved
training method is shown.
Table 1: The Pearson coefficient and the Aroc
value of the three methods examined. Method
1 is the TEPITOPE method. Method 2 is the
training of a prediction method done with a set
of peptides known to bind to HLA-DRB1*0401.
Method 3 is the 2. method improved.
Method
Perason coefficient
Aroc value
Method 1
0.358
0.469
Method 2
0.578
0.875
Method 3
0.741
0.885
Figure 2: The logo
obtained by training a
method with a set of
peptides known to
bind
to
HLADRB1*0401. From the
knowledge
of
the
anchor positions and
the amino acids these
are restricted to, the
training method was
set put more weight
on position 1, as this
position
and
the
amino acid in this
position
is
very
important
for
the
binding.
Figure 3: Top figure shows the
polytope construct where the linkers
and the line of epitopes are random.
The bottom figure shows the polytope
construct after optimisation.
The epitopes are sorted in the optimal
order And the most optimal sequence
and length of each linker is selected.
For final sequence see the Polytope
section.
References
1.
Frieden T. R., Sterling., T., R., Munsiff, S., S. Watt, C., J. Dye, C. (2003) Tuberculosis, Lancet 362, 887-99.
2.
Girard M., P., Fruth, U., Kieny, M-P. (2005) A review of vaccine research and development: Tuberculosis. Vaccine 23, 2725-2731.
3.
Bosio C. M., Gardner D., Elkins, K. L. (2000) Infection of B-cell deficient mice with CDC1551, a clinical isolate of Myobacterium tuberculosis: delay in
dissemination and development of lung pathology. J Immunol 164, 6417-25
4.
Kenneth Todar University of Wisconsin-Madison Department of Biology. Tuberculosis. 2005 http://textbookofbacteriology.net/tuberculosis.html
5.
De Groot AS., McMurry J., Marcon L., Franco J., Rivera D., Kutzler M., Weiner D., Martin B. (2005) Developing an epitope-driven tuberculosis (TB)
vaccine. Vaccine 23, 2121-2131.
6.
Montgomery D. L., Huygen K., Yawman A. M., Deck R. R., Dewitt C. M., Content J., Liu M. A. , Ulmer J. B. (1997) Induction of humoral and cellular
immune responses by vaccination with M. tuberculosis antigen 85 DNA. Cell Mol Biol 43(3),285-92.
7.
Ronning, D. R., Vissa V, Besra G. S., Belisle J. T., Sacchettini J. C. (2004). Mycobacterium tuberculosis Antigen 85A and 85C Structures Confirm Binding
Orientation and Conserved Substrate Specificity. The Journal of Biological Chemistry 279 (35), 36771-36777.
8.
Naito M., Ohara, O., Matsumoto, S., Yamada, T. (1998) Immunological Characterization of α antigen of Myobacterium kansaii: B-Cell epitope mapping.
Scand J. Immunol. 48, 73-78.
9.
Lund C., Nielsen M., Lundegaard C., Keşmir C., Brunak S. (2005) Immunological bioinformatics. Massachusettes Institute of Technology.
10. Sable SB., Kaur S., Verma I., Khuller GK. (2005) Immunodominance of low molecular weight secretory polupeptides of Mycobacterium tuberculosis to
induce cytotoxic T-lymphocyte response. Vaccine 23, 4947-4954.
11. Hansen J., Center for Biologisk Sekvensanalyse, Bioteknologisk Institut, DTU. http://www.biokemi.org/biozoom/1999_3/bz_0399f.htm