Transcript Group B: Ana Rita Domingues, Nicholas Gauthier, Ilka Hoof

Possible HIV Knockout?
Group B: Ana Rita Domingues, Nicholas Gauthier, Ilka Hoof,
Eleonora Kulberkyte, and Nicolas Rapin
Immunological Bioinformatics, June 2006
Introduction
Acquired immunodeficiency syndrome (AIDS)
is one of the worst diseases affecting mankind today.
Since it was discovered in 1981, AIDS has become a
major pandemic, and has killed approximately 16.3
million people. AIDS is caused by HIV (human
immunodeficiency virus) and is spread by sexual
contact, infected blood, and can be passed from
mother to infant. It is estimated that more than 34
million people are infected worldwide, most of which
Figure 1: Diagram of HIV with proteins [3].
are in Sub-Saharan Africa and South/Southeast Asia
[1,2]. HIV is a retrovirus, which means it carries its genetic information in the form
of RNA and replicates via a DNA intermediate. HIV infects the components of the
human immune system such as CD4+ T-cells, dendritic cells and macrophages.
Much scientific research has been done to find a treatment/vaccine for HIV.
The most common therapies used today are antiretroviral drugs that either interfere
with reverse transcription or inhibit the viral protease [2]. Finding a vaccine is a
difficult task as HIV mutates frequently and the infection can remain latent for long
periods before causing AIDS.
This poster aims to present an in silico approach to developing a vaccine for
HIV. We have chosen several epitopes from HIV-1 that are predicted to trigger Bcell and T-cell responses. The proteins we focused on were gp120, a surface protein
that is involved in the binding of virus to the host cell, and gag, a structural protein
from the virus core. The polyprotein gag is cleaved into several sub-proteins of
which p24 and p17 are the largest in size.
The C-terminal cleavage probability is predicted to be high (see Fig. 3 below)
for all MHC class I epitopes, with only one of them containing a strong internal
cleavage site.
Figure 3: Epitope Atlas of the polytope construct. Generated from the four chosen MHC class I and one MHC class II epitopes.
Since proteasomal cleavage is a stochastic process, the epitopes are expected to
be frequently processed correctly, presented by MHC class I molecules, and trigger a
CD8+ T-cell response. The MHC class II epitope was added to the polytope as there
is evidence that even endogenous antigens can be presented by MHC class II
molecules in antigen presenting cells.
Methods
All HIV-1 proteins were processed using NetCTL [4] in order to find putative
MHC class I epitopes. NetCTL combines the prediction of the main stages in
antigen processing: probability for proteasomal cleavage, TAP transport, and MHC
class I binding affinity. Predictions were run on HLA supertypes A1, A2, A3, and
B7. These HLA supertypes were used because they cover over 95% of the human
population [5]. After extensive screening, gag was chosen because it produced the
largest number of high scoring epitopes in NetCTL prediction throughout the
selected supertypes. The best epitope for each supertype was chosen based on its
NetCTL score and the degree of sequence conservation. The sequence conservation
was derived from a multiple sequence alignment of 614 gag sequences provided by
the Los Alamos HIV Sequence Database [6].
To maximize the activation of the immune system, MHC class II presentation
of gag epitopes was also investigated using the EasyGibbs sampler program [7].
The dataset used to train the Gibbs sampler consisted of known HLA-DR4 epitopes
(provided by EasyGibbs).
A polytope containing both MHC class I and II epitopes was generated using
a polytope optimization program in order to minimize internal proteasomal
cleavage and maximize C-terminal cleavage.
B-cell epitopes were predicted using the DiscoTope 1.0 server [8]. A suitable
candidate was chosen according to conservation score and visual evaluation of
accessibility in the 3D-structure of gp120. Both the B-cell epitope and the highest
scoring MHC class II epitope were included into the protein vaccine construct. The
3D-structure of the construct was predicted using the CPHmodels 2.0 server [9].
Results
The final polytope construct has a length of 70 amino acids and contains four
MHC class I epitopes along with the highest scoring MHC class II epitope. All MHC
class I epitopes are positioned in highly conserved -helical structures (see Fig. 2
below) .
Figure 4: 3D-structure of the protein-based vaccine predicted by CPHmodels [9]. The
B-cell epitope is highlighted in red and the MHC class II epitope is highlighted in blue.
As a suitable target to boost B-cell response the surface protein gp120 was
chosen. Most putative epitopes which were predicted by DiscoTope were located in
loop regions, and as these are highly variable in gp120, the conserved helix 1 was
chosen as a qualified B-cell epitope. The peptide vaccine construct consists of the
complete gp120 structure 1RZJ (from the Protein Data Bank) into which the
predicted MHC class II epitope FYKTLRAEQASQEVKNWM was inserted in the
loop region after position 292 (blue in Fig. 4). The 3D-structure appears stable and
the B-cell epitope is still accessible (red in Fig. 4).
Discussion
Throughout the course project we have identified several promising MHC
class I, MHC class II, and B-cell epitopes that can be used as possible HIV vaccines.
Our final vaccine strategy contains both a polytope that can be used for a DNA
vaccine and a peptide that can be used as a protein vaccine. Our strategy is to target
and activate both the humoral and cell-mediated immune response in the same
vaccine. This strategy should make it very difficult for HIV to survive within a
vaccinated individual.
Due to time constraints we were unable to consider all aspects needed for a
trial vaccine. Autoimmunity could be a problem if antigen in our epitopes closely
matches sequestered self antigen. In addition RNA splicing could destroy our
polytope and result in lower epitope synthesis. Furthermore it may be beneficial to
add ubiquitination signals into the polytope to help target it to be degraded by the
proteasome. Placing a signal peptide on the MHC class II epitope which targets it to
lysosomes and endosomes can further enhance MHC class II presentation.
References
Figure 2: 3D-structure of HIV-1 p24 and p17 structural proteins. The four predicted MHC class I
epitopes are highlighted by overall conservation. Note that the two epitopes in p17 overlap.
[1] - Immunobiology. Janeway, C. et al, Garland Publishing, 2001;
[2] - Immunology. Goldsby, R. et al, W.H.Freeman and Company, 2003;
[3] - http://en.wikipedia.org/wiki/HIV
[4] - An integrative approach to CTL epitope prediction. A combined algorithm integrating MHC-I binding, TAP transport
efficiency, and proteasomal cleavage predictions. Larsen M.V., Lundegaard C., Kasper Lamberth, Buus S,. Brunak S., Lund O.,
and Nielsen M. European Journal of Immunology. 35(8): 2295-303. 2005
[5] - Nine major HLA class I supertypes account for the vast preponderance of HLA-A and -B polymorphism. Sette, A., and J.
Sidney. 1999. Immunogenetics 50:201-212.
[6] - http://hiv-web.lanl.gov
[7] - Improved prediction of MHC class I and class II epitopes using a novel Gibbs sampling approach. Nielsen M, Lundegaard C,
Worning P, Hvid CS, Lamberth K, Buus S, Brunak S, Lund O. Bioinformatics. 2004 20:1388-97
[8] - Prediction of discontinuous antibody binding epitopes in proteins. Pernille H. Andersen, Morten Nielsen and Ole Lund
2006, submitted
[9] - CPHmodels 2.0: X3M a Computer Program to Extract 3D Models. O. Lund, M. Nielsen, C. Lundegaard, P. Worning..
Abstract at the CASP5 conferenceA102, 2002.