CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
TITLE: KIT FOR PREDICTING BINDING OF A SPECIFIC ANTIBODY TO A POTENTIAL
IMMUNOGEN AND METHOD OF SCREENING
FIELD OF INVENTION
The present invention relates to a kit for predicting binding of a specific
antibody to at least one
potential immunogen, as well as a high throughput screening method for testing
the presence
of antibodies specific for at least one structural epitope comprised in at
least one potential
immunogen.
Further the invention relates to a use of the kit and/or the high throughput
screening
to method for predicting binding of specific antibodies, in one or more
samples, to at least one or
more potential immunogen(s).
Still further the invention relates to a vaccine comprising an antigenic
peptide se-
quence corresponding to a structural epitope comprised in a potential
immunogen.
Finally the invention relates to a use of at least one antigenic peptide
sequence corre
sponding to a specific structural epitope in at least one potential immunogen
for the prepara
tion of a vaccine, a method for preparing such a vaccine and use of such a
vaccine.
BACKGROUND OF THE INVENTION
An increasing number of proteins, including enzymes, are being produced
industrially, for use
2o in various industries, housekeeping and medicine. Being proteins they are
likely to stimulate an
immunological response in man and animals, including an allergic response.
As the food market becomes more globalized, the average consumer runs a higher
risk of encountering unexpected allergens. .These foreign allergens add up to
the increased
use of mixtures of proteins as well as additives by a more and more
industrialized food pro-
duction.
Humans or animals may become sensitised to allergens e.g. by inhalation,
direct con-
tact with skin and eyes, or injection. The general mechanism behind an
imunnogenic, and in
particular an allergic response, is divided in a sensitisation phase and a
symptomatic phase.
The sensitisation phase involves a first exposure of a human or animal to an
allergen. This
3o event activates specific T- and B-lymphocytes, and leads to the production
of allergen specific
IgE antibodies (in the present context the antibodies are denoted as usual,
i.e. immunoglobulin
E is IgE etc.). These IgE antibodies eventually facilitate allergen capturing
and presentation to
T-lymphocytes at the onset of the symptomatic phase. This phase is initiated
by a second ex-
posure to the same or a resembling antigen. The specific IgE antibodies bind
to the specific
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
IgE receptors on mast cells and basophils, among others, and capture at the
same time the
allergen. The polyclonal nature of this process results in bridging and
clustering of the IgE re-
ceptors, and subsequently in the activation of mast cells and basophils. This
activation triggers
the release of various chemical mediators involved in early as well as late
phase reactions of
the symptomatic phase of allergy.
For certain forms of IgE-mediated allergies, a therapy exists, which comprises
re-
peated administration of allergen preparations called 'allergen vaccines'
(Int. Arch. Allergy Im-
munol., 1999, vol. 119, pp1-5). This leads to reduction of the allergic
symptoms, possibly due
to a redirection of the immune response away from the allergic (Th2) pathway
and towards the
to immunoprotective (Th1) pathway (Int. Arch. Allergy Immunol., 1999, vol.
119, pp1-5). However,
for most of the allergies avoiding contact with the allergen still is the only
available treatment.
Whatever therapeutic strategy, a proper diagnosis of the allergy, i.e. proper
identifica-
tion of the challenging allergen, is required to optimise either the 'allergen
vaccination' therapy
or the 'abstinence' approach.
The diagnosis of humans or animals with allergic symptoms is not well
developed.
Moreover, there is a gap between the identification of single IgE-binding
allergens and the
quantitative risk assessement.
Numerous tests exist for determination of the biological potency of molecules
or mix
tures. Challenge of human patients are considered as closest to the relevant
biological re
2o sponse, i.e. elicitation of an actual immunogenic, and in particular an
allergic response, albeit
under controlled and safe circumstances. Skin tests obviously involves the
skin mast cells,
which must be sensitised by IgE in order to respond to the offending allergen.
A biological in
vitro system is the sensitised basophil granulocyte. This system mimicks the
sensitized mast
cell present in the relevant target organ of the patient. Moving even further
away from the ac-
a5 tual patient, basophils from, e.g. cord blod, of a non-allergic donor, may
be used as a reagent.
These cells must be sensitised by IgE derived from the actual patient.
Presently, double blind placebo controlled food challenge (DBPCFC) is
considered
valid in diagnosing food allergy, and compared to this gold-standard, there
are many examples
of in vivo and in vitro diagnostic tools which produce misleading results. The
reason for the
30 low specificity of these tests is the extensive cross-reactions between
species, and between
environmental allergens and food allergens.
A pure system can be obtained by immunochemical assays detecting IgE-allergen
binding, directly or indirectly, by inhibition designs. These assays should
preferentially include
single allergen specific IgE epitopes in order to allow direct risk
assessment.
2
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
Several similar techniques for localization of B-cell epitopes are disclosed
by Walshet
et al, J. Immunol. Methods, vol. 121, 1275-280, (1989), and by Schoofs et al.
J. Immunol. vol.
140, 611-616, (1987). All of these documents, relate to identification of
linear epitopes.
Slootstra et al; Molecular Diversity, 2, pp. 156-164, 1996 discloses the
screening of a
semi-random library of synthetic peptides for their binding properties to
three monoclonal anti-
bodies.
WO 99/47680 (ALK-ABELLO) discloses the identification and modification of B-
cell
epitopes by protein engineering.
WO 00/26230 (Novozymes A/S) describes the use of phage-display libraries for
iden-
to tifying linear as well as conformational epitope sequences and patterns on
proteins. This in-
formation is stored in a database, and provides a rational approach for
identifying antigenic
and allergenic areas on proteins.
Conformational/structural epitopes are less likely to be present on different
immuno-
gens and the use of such epitopes in diagnosis or characterization og
immunoglubulins from a
human or animal will therefore give a more precise answer without the problems
of cross reac-
tivity.
Identification of such conformational/structural epitopes can be used in the
context of
the present invention in order to precisely identify interactions between such
conformational
epitopes and specific antibodies and provides a fast method of screening a
large number of
2o different allergenic epitopes at the same time.
SUMMARY OF THE INVENTION
The present invention relates to a kit for predicting binding of a specific
antibody to at
least one potential immunogen, comprising
a) at least one antigenic peptide sequence comprising less than 26 amino acids
wherein
said antigenic peptide sequence corresponds to a structural epitope comprised
in the at
least one potential immunogen and the antigenic peptide sequence is capable of
binding at
least one antibody specific for the structural epitope comprised in the said
potential immu-
ao nogen, and
b) solid support suitable for immobilising the at least one antigenic peptide
sequence.
3
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
A second aspect of the present invention relates to a high throughput
screening method
for testing the presence of antibodies specific for a structural epitope
comprised in at least one
potential immunogen of interest, comprising
a) providing one or more antigenic peptide sequences comprising less than 26
amino acids
to
wherein said one or more antigenic peptide sequences) corresponds) to one or
more
structural epitopes comprised in the at least one potential immunogen and the
antigenic
peptide sequences) is/are capable of binding at least one antibody specific
for the struc-
tural epitope(s) comprised in the said potential immunogen(s),
b) immomilizing the one or more antigenic peptide sequences to a suitable
solid support,
c) adding the specific antibodies from a sample, and
d) detecting binding of specific antibodies to any of the one or more
antigenic peptide se-
quences.
A third aspect of the present invention relates to a use of the kit for
predicting bind-
ing of a specific antibody in a sample to at least one potential immunogen,
wherein binding of
2o antibody to at least one antigenic peptide sequence corresponding to at
least one structural
epitope on the at least one potential immunogen is tested.
A forth aspect of the present invention relates to a use of the high
throughput
screening method for screening antibodies from at least one sample.
A fifth aspect of the invention relates to a vaccine comprising at least one
antigenic
peptide corresponding to a structural epitope comprised in at least one
potential immunogen
and said antigenic peptide sequence being capable of binding at least one
antibody specific for
the structural epitope comprised in the potential immunogen
A sixth aspect of the invention relates to a method of preparing a vaccine
compris-
ing adding to a liquid medium at least one antigenic peptide sequence,
corresponding to a
structural epitope comprised in at least one potential immunogen and said
antigenic peptide
sequence being capable of binding at least one antibody specific for the
structural epitope
comprised in the potential immunogen.
A seventh aspect of the invention relates to a use of at least one antigenic
peptide se-
quence, corresponding to a structural epitope comprised in at least one
potential immunogen
4
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
and said antigenic peptide sequence being capable of binding at least one
antibody specific for
the structural epitope comprised in the potential immunogen, for the
preparation of a vaccine.
An eigth aspect of the invention relate to a use of the vaccine of the
invention for the
treatment of a human or an animal.
BRIEF DESCRIPTION OF DRAWINGS
s Figure 1 shows the antibody binding capacity of selected peptides. The
antibody bind-
ing capacity of the two linear peptide sequences, RRFANDHTR (light gray bars)
and
RRFSNATRA (dark gray bars), were tested in an ELISA assay, by measuring
optical density,
OD. The sequences were tested for binding to antibodies in sera raised against
different prote-
ins. The different proteins are marked by capital letters A through H. A =
Alcalase~, B = Savi-
to nase~, C = Subtilisin Novo~, D = Carezyme~ (cellulase), E = Laccase, F =
Natalase~ (amy-
lase), G = SP722 (amylase), H = Lipolase~ (lipase).
DEFINITIONS
Prior to a discussion of the detailed embodiments of the invention, a
definition of spe-
15 cific terms related to the main aspects of the invention is provided.
The term "epitope" is defined as an antigenic determinant and is a set of
amino acids
on a protein that are involved in an immunological response, such as antibody
binding or T-cell
activation. It is the simplest form or smallest structural area on a complex
antigen molecule
that can combine with an antibody or T lymphocyte receptor. An epitope must be
at least 1 kD
20 (about 10 amino acids) in order to be immunogenic. Epitopes can be linear
or conforma-
tional/structural.
The term "linear epitope" is defined as an epitope composed of amino acid
residues
that are contiguous on the linear sequence of amino acids (primary structure).
The term "epitope sequence" is defined as the amino acid residues which makes
up
25 the epitope.
The term "conformational or structural epitope" is defined as an epitope
composed of
amino acid residues that are not all contiguous. The epitope is thus composed
of separated
parts of one or more linear sequences of amino acids that are brought into
proximity to one
another by folding of the molecule (secondary, tertiary and/or quaternary
structures). A con-
ao formational epitope is dependent on the 3-dimensional structure. The term
'conformational' is
therefore often used interchangeably with 'structural'.
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
The term "antibody binding peptide" is defined as a peptide that binds with
sufficiently
high affinity to antibodies. In particular the antibody binding peptide is
linear, but it may also be
circular. Identification of 'antibody binding peptides' and their sequences
constitute the first
step of the method of this invention.
By the term "epitope pattern" is meant a consensus sequence of antibody
binding
peptides. An example is the epitope pattern A R R > R. The sign ">" or "<" in
this notation indi-
cates that the aligned antibody binding peptides may or may not include one or
more non-
consensus amino acids between the second and the third arginine.
By the term "anchor amino acids" is meant the individual amino acids of an
epitope
to pattern.
The term "immunogen" is a substance that is able to induce a humoral antibody
and/or cell-mediated immune response rather than immunological tolerance. The
term 'immu-
nogen' is sometimes used interchangeably with 'antigen', yet the term
specifies the ability to
stimulate an immune response as well as to react with the products of it, e.g.
antibody. By con-
trast, 'antigen' is reserved by some to mean a substance that reacts with
antibody. The princi-
pal immunogens are proteins and polysaccharides, free or attached to
microorganisms.
The terms "immunogenic/immunogenicity" means the capacity to induce humoral an-
tffbody andlor cell-mediated immune responsiveness.
The term "donor protein" means the protein that was used to raise antibodies
for iden-
ao tification of antibody binding sequences, hence the donor protein provides
the information that
leads to the epitope patterns. The donor protein may e.g. be the parent
protein or a part of it.
The term "acceptor protein" is the protein, whose 3D-structure is used to fit
the identi-
fied epitope patterns and/or to fit the antibody binding sequences. Hence, the
acceptor protein
may e.g. be the parent protein or a part of it.
"Monospecific polyclonal antibodies" are polyclonal antibodies that are
specifically
binding to a certain epitope, and hence are monospecific. The polyclonal
nature of these anti-
bodies is explained by the fact that a number of antibody-producing B-cell
clones may produce
antibodies to similar epitopes (but with the same epitope pattern as the
epitope of interest),
that bind to the epitope of interest, though with lower affinity.
so The term "immunogenic response" including allergic response, used in
connection
with the present invention, is the response of an organism to a compound,
which involves IgE
mediated responses (Type I reaction according to Coombs & Gell). It is to be
understood that
sensibilization (i.e. development of compound-specific IgE antibodies) upon
exposure to the
compound is included in the definition of "immunogenic response".
6
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
An "epitope area" is defined as the amino acids situated close to the epitope
se-
quence amino acids. Particularly, the amino acids of an epitope area are
located <5A from the
epitope sequence. Hence, an epitope area also includes the corresponding
epitope sequence
itself. Modifications of amino acids of the 'epitope area' can possibly affect
the immunogenic
function of the corresponding epitope.
"Environmental allergens" are protein allergens that are present naturally.
They in-
clude pollen, dust mite allergens, pet allergens, food allergens, venoms, etc.
"Commercial allergens" are protein allergens that are being brought to the
market
commercially. They include enzymes, pharmaceutical proteins, antimicrobial
peptides, as well
to as allergens of transgenic plants.
By the term "specific polyclonal antibodies" is meant polyclonal antibodies
isolated ac-
cording to their specificity for a certain antigen, e.g. the protein backbone.
DETAILED DESCRIPTION OF THE INVENTION
Identification of antibody binding peptides and epitope pattern
A first step required to carry out the present invention is to identify
peptide sequences,
which bind specifically to antibodies.
Antibody binding peptide sequences can be found by testing a set of known
peptide
2o sequences for binding to antibodies raised against the donor protein. These
sequences are
typically selected, such that each represents a segment of the donor protein
sequence (Mol.
Immunol., 1992, vol. 29, pp.1383-1389; Am. J. Resp. Cell. Mol. Biol. 2000,
vol. 22, pp. 344
351 ). Also, randomized synthetic peptide libraries can be used to find
antibody binding se
quences (Slootstra et al; Molecular Diversity, 1996, vol. 2, pp. 156-164).
In a particular method, the identification of antibody binding sequences may
be
achieved by screening a display package library, particularly a phage display
library. The prin-
ciple behind phage display is that a heterologous DNA sequence can be inserted
in the gene
coding for a coat protein of the phage (WO 92115679). The phage will make and
display the
hybrid protein on its surface where it can interact with specific target
agents. Such target agent
ao may be antigen-specific antibodies. It is therefore possible to select
specific phages that dis-
play antibody-binding peptide sequences. The displayed peptides can be of
predetermined
lengths, for example 9 amino acids long, with randomized sequences, resulting
in a random
peptide display package library. Thus, by screening for antibody binding, one
can isolate the
peptide sequences that have sufficiently high affinity for the particular
antibody used. The pep-
7
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
tides of the hybrid proteins of the specific phages which bind protein-
specific antibodies char-
acterize epitopes that are recognized by the immune system.
The antibodies used for reacting with the display package are particularly IgE
antibod-
ies to ensure that the epitopes identified are IgE epitopes, i.e. epitopes
inducing and binding
IgE. In a particular embodiment the antibodies are polyclonal antibodies,
optionally monospeci-
fic antibodies.
For the purpose of the present invention particularly polyclonal antibodies
are used in
order to obtain a broader knowledge about the epitopes of a protein.
It is of great importance that the amino acid sequence of the peptides
presented by
to the display packages is long enough to represent a significant part of the
epitope to be identi-
fied. In a particular embodiment of the invention the peptides of the peptide
display package
library are oligopeptides having from 5 to 25 amino acids, particularly at
least 8-12 amino ac-
ids, such as 9 amino acids. For a given length of peptide sequences (n), the
theoretical num-
ber of different possible sequences can be calculated as 20~. The diversity of
the package 1i-
brary used must be large enough to provide a suitable representation of the
theoretical number
of different sequences. In a phage-display library, each phage has one
specific sequence of a
determined length. Hence an average phage display library can express 108 -
10'~ different
random sequences, and is therefore well-suited to represent the theoretical
number of different
sequences.
2o Hence, in one embodiment of the invention, antigenic peptides for use in
the kit of the
invention are obtained by screening a random peptide library with antibodies
raised against
any immunogen of interest and sequencing the amino acid sequence of antibody
binding pep-
tides or the DNA sequence encoding the peptides. Once such sequences have been
estab-
lished the peptides may be prepared/produced.
The antibody binding peptide sequences can be further analysed by consensus
alignment e.g. by the methods described by Feng and Doolittle, Meth. Enzymol.,
1996, vol.
266, pp. 368-382; Feng and Doolittle, J. Mol. Evol., 1987, vol. 25, pp. 351-
360; and Taylo,r,.
Meth. Enzymol., 1996, vol. 266, pp. 343-367.
This leads to identification of epitope patterns, which can assist the
comparison of the
ao information obtained from the antibody binding peptide sequences to the 3-
dimensional struc-
ture of the acceptor protein in order to identify epitope sequences at the
surface of the accep-
for protein.
8
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
Epitope patterns
Given a number of antibody binding peptide sequences and possibly the
correspond-
ing epitope patterns, one need the 3-dimensional structure coordinates of an
acceptor protein
to find the epitope sequences on its surface.
These coordinates can be found in databases (NCBI:
http://www.ncbi.nlm.nih.gov/),
determined experimentally using conventional methods (Ducruix and Giege:
Crystallization of
Nucleic Acids and Proteins, IRL PRess, Oxford, 1992, ISBN 0-19-963245-6), or
they can be
deduced from the coordinates of a homologous protein. Typical actions required
for the con-
struction of a model structure are: alignment of homologous sequences for
which 3-
to dimensional structures exist, definition of Structurally Conserved Regions
(SCRs), assignment
of coordinates to SCRs, search for structural fragments/loops in structure
databases to replace
Variable Regions, assignment of coordinates to these regions, and structural
refinement by
energy minimization. Regions containing large inserts (>3 residues) relative
to the known 3-
dimensional structures are known to be quite difficult to model, and
structural predictions must
be considered with care.
One can match each amino acid residue of the antibody binding peptide to an
identi-
cal or homologous amino acid on the 3-D surface of the acceptor protein, such
that amino ac-
ids that are adjacent in the primary sequence are close on the surface of the
acceptor protein,
with close being <10A, particularly <5k, more particularly <3A between any two
atoms of the
2o two amino acids.
Alternatively, one can define a geometric body (e.g. an ellipsoid, a sphere,
or a box) of
a size that matches a possible binding interface between antibody and antigen
and look for a
positioning of this body where it will contain most of or all the anchor amino
acids.
Also, one can use the epitope patterns to facilitate identification of epitope
sequences.
This can be done, by first matching the anchor amino acids on the 3-D
structure and subse-
quently looking for other elements of the antibody binding peptide sequences,
which provide
additional matches. If there are many residues to be matched, it is only
necessary that a suit-
able number can be found on the 3-D structure. For example if an epitope
pattern comprises 4,
5, 6, or 7 amino acids, it is only necessary that 3 matches surface elements
of the acceptor
3o protein.
In all cases, it is desirable that amino acids of the epitope sequence are
surface ex-
posed (see Example 1 ).
9
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
How to use the epitope information.
When applied on structurally and immunologically related immunogens, the
informa-
tion about epitope patterns and sequences, which can been derived by the above
methods,
can be utilized to assist in the selection of structural epitopes that are
specific for the immuno-
gen of interest.
After having identified the structural epitopes that will react with specific
antibodies
from a sample, which e.g. can be obtained from a human or an animal, these
structural epi-
topes in the form of peptide sequences can be applied in a kit for testing
binding of specific an-
tibodies, particularly IgE antibodies, from the human or an animal, to the
peptide sequences.
to This way it will be possible to predict binding of specific antibodies in a
human or animal to
structural epitopes comprised on potential immunogens.
Hence, in one embodiment the antigenic peptide to be employed in the kit of
the inven-
tion is obtained by
(1 ) screening a random peptide library with antibodies raised against an
immunogen of in-
terest,
(2) determining the amino acid sequence of peptides binding to an antibody or
the DNA se-
quences encoding the peptides,
(3) using the peptides or DNA sequences to identify at least one structural
epitope pattern on
the immunogen and
(4) producing antigenic peptides corresponding to structural epitopes on the
immunogen.
In a particular embodiment an antigenic peptide representing a structural
epitope is a
combination of one part of one antibody binding peptide combined with one or
more parts from
one or more different antibody binding peptides.
In a further embodiment the specificity or the affinity of antigenic peptides
corresponding
to structural epitopes on the immunogen may be increased by adding, deleting
or mutating one
or more amino acids in the sequence of the antigenic peptides or a combination
thereof. Addi-
tion, deletion and mutation of amino acids in a sequence is known to the
skilled person and
may be achieved by conventional biochemical and/or genetic engineering
methods.
3o Once the sequence of a suitable antigenic peptide representing a structural
epitope has
been identified, the peptides may be produced in any convenient way, e.g. by
artificially syn-
thesizing the peptides or expressing nucleic acid sequences encoding the
peptides in a host.
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
Diagnostic kit.
Today, a patient suffering from an imunnogenic disease, such as allergy, may
be sub
jected to allergy vaccine therapy using imunnogens selected on the basis of
testing the speci
ficity of the patient's serum IgE against a bank of immunogen extracts (or
similar specificity
tests of the patient's sensibilization such as skin prick test.
One could improve the quality of characterization by using antibody binding
peptides
corresponding to various epitope sequences on the protein imunnogens of
interest. This would
require a kit comprising reagents for such specificity characterization, e.g.
the antibody binding
peptides of desired specificity. It is particularly useful to use antibody
binding sequences in the
to kit, which correspond to defined epitope sequences known to be specific for
the immunogen
under investigation (i.e. not identified on other immunogens and/or not cross-
reacting with sera
raised against other allergens). This kit would be useful to specifying which
immunogenic de-
cease, such as allergy, the patient is suffering from. This kit will lead to a
more specific answer
than those kits used today, and hence to a better selection of immunogen
vaccine therapy for
the individual patient.
In an extension of this approach, one could also characterize the patient's
serum by
identifying the corresponding antibody binding peptides among a random display
library using
the aforementioned methods. This again may lead to optimisation of the epitope
information,
and thus to a better diagnosis.
2o Further, one could use the individual antibody binding sequences as
(immunogen)
vaccines leading to more specific (immunogen) vaccines. These antibody binding
sequences
could be administered in an isolated form or fused to a membrane protein of
the phage display
system, or to another carrier protein, which may have beneficial effect for
the immunoprotec-
tive effect of the antibody binding peptide (Datum et al., Nature
Biotechnology, 1999, Vol. 17,
z5 pp.666-669).
In a first aspect the present invention relates to a kit for predicting
binding of a specific
antibody to at least one potential immunogen, comprising
a) at least one antigenic peptide sequence comprising less than 26 amino acids
wherein
so said antigenic peptide sequence corresponds to a structural epitope
comprised in the at
least one potential immunogen and the antigenic peptide sequence is capable of
binding at
least one antibody specific for the structural epitope comprised in the said
potential immu-
nogen, and
11
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
b) solid support suitable for immobilising the at least one antigenic peptide
sequence.
The kit of the invention would also be useful for other screening purposes
where it
is desirable to test for antibody binding to peptide sequences, e.g. for the
development of epi-
tope variants as mentioned previously.
Suitable solid support could in the present invention be any chemical support,
in-
cluding micro titer plates, beads, capillary tubing or membranes. Each of
these supports could
be activated, supporting covalent, ionic or hydrophobic binding, chelation or
affinity binding, or
inactivated, promoting ionic or hydrophobic binding.
to Immobillisation could take place by attachment through covalent binding,
ionic or
hydrophobic binding, chelation, affinity binding, or through van der Waal
bonds.
In the present invention, a solid support could also be biological in nature,
such as
phages, bacteria, red blood cells or any related system allowing display of
heterologous pro-
teins or peptides.
The above desribed kit can also be used for screening different antigenic
peptide se-
quences corresponding to structural epitopes at the same time.
Given a number of proteins for which diagnosis optimally has to be performed
simul-
taneously, a kit can be produced containing for each of these proteins a
specific peptide corre-
sponding to a structural epitope sequence comprised in the protein and
immobilised on a solid
2o support. As an example, 3 specific peptides are immobilised on beads, each
peptide having its
specific coloured bead. In an agglutination format were specific antisera is
mixed with this mix-
ture of peptide coated beads, the colour of the agglutinate will identify the
specificity of the an-
tibodies present in the patients serum.
In another embodiment the diagnostic kit comprises ten different antigenic
peptide se
z5 quences and in a further embodiment the diagnostic kit comprises at least
100 different anti
genic peptide sequences.
The kit above can also be used in a high throughput screening method for
screening
many samples, obtained e.g. from humans or animals, at the same time and
thereby predicting
which humans or animals will display an immunogenic response towards
particular immuno-
3o gens. Any practical combination of the number of antigenic peptide
sequences and the number
of humans or animals would be possible.
A second aspect of the invention therefore relates to a high throughput
screening method
for testing the presence of antibodies specific for a structural epitope
comprised in at least one
potential immunogen of interest, comprising
12
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
a) providing at least one antigenic peptide sequence comprising less than 26
amino acids
wherein said antigenic peptide sequences) corresponds) to one or more
structural epi-
topes comprised in the at least one potential immunogen, and wherein the
antigenic pep-
s tide sequences) is/are capable of binding at least one antibody specific for
the structural
epitope comprised in the said at least one potential immunogen,
b) immomilizing the at least one antigenic peptide sequences to a suitable
solid support,
to c) adding the specific antibodies from the human or animal, and
d) detecting binding of specific antibodies to any of the one or more
antigenic peptide se-
q uences.
15 In one embodiment antibodies from at least ten samples are screened and in
another em-
bodiment antibodies from at least 100 samples are screened.
Different assay formats are compatible with a high throughput technology.
One such format is the ELISA format in for example 96, 384 or 1536 well
plates. An
20 other format is the agglutination format, where the relevant peptides are
immobilised on (col
oured) beads or are presented by displaying organism, such as phages or
bacteria. A third
format is the blotting format, which uses membranes, such as nitrocellulose or
polyvinyl-based
membranes, as support. This format includes for example dot blot assays, and
line immunoas
says. A fourth assay format is the dipstick or pin based assays, were the
peptide is immobi
25 lined on for example polystyrene or polyethylene pins.
If required, the solid supports can be activated chemically or biochemically
in order to
optimize binding of the target peptide to the support. This optimation might
involve introduction
of groups promoting covalent linkage, chelation, affinity binding, ionic or
hydrophobic binding.
A chemical activation might for example lead to reactive NH2 groups, or
reactive Ni2+
3o complexes. Biochemical activation might include coating with avidin or
streptavidin for cathing
biotin-peptide complexes, short fatty acids for binding hydrophobic peptides,
antibodies for
binding biotin-labelled peptides.
A third aspect of the present invention relates to a use of the kit according
to the
invention for predicting binding of specific antibodies in a sample, e.g.
obtained from a human
13
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
or animal, to at least one potential immunogen, wherein binding to at least
one antigenic pep-
tide sequence corresponding to a structural epitope is tested. Particularly at
least ten antigenic
peptide sequences are tested, and in a further particular embodiment at least
100 antigenic
peptide sequences are tested.
A fourth aspect of the present invention relates to a use of the high
throughput
screening method for screening antibodies from at least one sample, e.g.
obtained from at
least one human or animal, particularly from at least ten humans or animals,
and in a further
particular embodiment from at least 100 humans or animals.
Conformational/structural epitopes are as discussed previously composed of
amino
to acid residues that are not all contained on the same contiguous amino acid
sequence, but are
brought into the right position to one another by folding of the protein. It
is therefore possible
for the distance between amino acids comprised on a structural epitope, but
located on sepa-
rated parts on the primary structure, to vary due to the dynamic nature of the
conformation of
the folded protein.
In the context of the present invention a structural epitope, comprised in the
poten-
tial immunogen, comprises at least a first contiguous linear amino acid
sequence consisting of
at least one amino acid and a second contiguous linear amino acid sequence
consisting of at
least one amino acid, and wherein a distance between any two amino acids
comprised in the
structural epitope, which amino acids are not part of the same contiguous
linear amino acid
zo sequence, and which two amino acids are most proximal to each other, does
not exceed 5~4.
In one embodiment the said distance should not exceed 3 A.
One way of measuring distances between amino acids on primary structures as
well as 3D-structures of proteins uses Swissprot-PDBViewer (known by the
skilled person in
the art), which can be downloaded, free of charge, from www.expasy.com.
In one embodiment the first contiguous linear sequence and the second
contiguous
linear sequence are part of the same primary sequence of the immunogen. In
case the first
and second part of the epitope are both part of the same rimary sequence the
first contiguous
linear sequence and the second contiguous linear sequence are interrupted by
at least one
amino acid, particularly at least one amino acid which is located more than 10
~ away from at
least one amino acid of the first or second contiguous linear sequence. In a
partcular embodi-
ment the first contiguous linear sequence and the second contiguous linear
sequence are in-
terrupted by at least 10 amino acids.
In another embodiment the immunogen contains two or more subunits of primary
sequences and the first contiguous linear sequence and the second contiguous
linear se-
14
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
quence comprised in the epitope are part of two or more different primary
sequences of the
immunogen.
The epitope may contain more than two separated parts, such as three or four
separated parts. However, in one embodiment the first contiguous linear
sequence and the
s second contiguous linear sequence constitutes the structural epitope.
Cross-reactivity
In order to get a more specific and precise diagnosis, the antigenic peptide
sequence,
representing a structural epitopes on an immunogen, which is selected for the
kit should dis-
play a minimal, e.g little or no cross-reactivity between the antibodies
raised against an immu-
to nogen of interest and antibodies raised against any other 'commercial' and
'environmental'
immunogen. When cross-reactivity is observed typically an antibody that will
bind to one epi-
tope (or antigenic peptide sequence representing the epitope) will also be
able to bind to other
epitopes, e.g. on other immunogens. Cross-reactivity is a common problem when
using linear
epitopes or antigenic peptide sequences representing a linear epitope in
diagnostics. However
15 when using antigenic peptide sequences representing a structural epitope in
diagnostics,
cross-reactivity is minimized.
In one embodiment the kit of the invention employs at least one antigenic
peptide se-
quence, which corresponds to a structural epitope on at least one potential
immunogen,
wherein the at least one specific antibody, when present in excess with
respect to the potential
zo immunogen, will not bind to another antigen unless this antigen is present
at a concentration
which is 1000 fold higher than the potential immunogen.
In a further embodiment the kit of the invention employs at least one
antigenic peptide
sequence, which corresponds to a structural epitope on at least one potential
immunogen,
wherein the at least one antigenic peptide sequence has at least a 10 fold
stronger affinity per
25 microgram antigenic peptide towards at least one specific antibody in full
blood or serum from
an animal or human immunized with the full immunogen, than towards a non-
specific antibody
provided that the concentration of the specific antibody and the non-specific
antibody is the
same.
In a further embodiment the serum may be purified so as to mainly or
completely contain
so antibodies of a selected class. Hence, the kit of the invention employs at
least one antigenic
peptide sequence, which corresponds to a structural epitope on at least one
potential immuno-
gen, wherein the antigenic peptide sequence has at least a 10 fold stronger
affinity per micro-
gram antigenic peptide towards at least one specific antibody in purified
serum from an animal
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
or human immunized with the full immunogen than towards a non-specific
antibody provided
that the concentration of the specific antibody and the non-specific antibody
is the same, and
wherein at least 50% of the specific antibodies present in the purified serum
belongs to the
same class of antibodies.
Particular classes of antibodies includes IgE, IgG, IgA, IgM or IgD. Further,
the serum
may be purified so that at least 75% of the antibodies in the purified serum,
such as at least
90%, e.g. at least 98%, particularly at least 99% or even 100% belongs to the
same class.
In a further embodiment the serum may be purified so as to mainly or
completely contain
antibodies which will bind to the employed antigenic peptides. Hense in this
embodiment the
to kit of the invention employs at least one antigenic peptide sequence, which
corresponds to a
structural epitope on at least one potential immunogen, wherein the at least
one antigenic pep-
tide sequence has at least a 10 fold stronger affinity per microgram antigenic
peptide towards
at least one specific antibody in purified serum from an animal or human
immunized with the
full immunogen, than towards a non-specific antibody provided that the
concentration of the
specific antibody and the non-specific antibody is the same, and wherein at
least 90% of the
specific antibodies present in the purified serum binds to the at least one
antigenic peptide se-
quence. Further, the serum may be purified so that at least 95% of the
antibodies in the puri-
fied serum binds to the antigenic peptide sequence, in particular at least
98%, at least 99% or
even 100%.
2o In the previous four embodiments the affinity of the antigenic peptide
sequence may in
particular be at least 20 fold stronger, more particularly at least 50 fold
stronger affinity, more
particularly at least 100 fold stronger. Further in these embodiments the at
least one specific
antibody towards which the antigenic peptide sequence has affinity is in
particular a collection
of 1-10 different antibodies, more particularly 1-5 different antibodies, more
particularly 1-3 dif-
ferent antibodies. In a still further embodiment the at least one specific
antibody is one specific
antibody.
Cross-reactivities between food allergens of different origin are well-known
(Akker-
daas et al, Allergy 50, pp 215-220, 1995). Similarly, cross-reactivities
between other environ-
mental allergens (like pollen, dust mites etc.) and commercial allergens (like
enzyme proteins)
3o have been established in the literature (J. All. Clin. Immunol., 1998, vol.
102, pp. 679-686 and
by the present inventors. The molecular reason for this cross-reactivity can
be explored using
epitope mapping.
The general principle of the present invention, whereby random peptide
libraries are
screened for any peptides capable of binding to specific antibodies, and these
isolated random
16
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
peptides subsequently are fitted by epitope mapping to 3D-models of known
proteins thereby
identifying first epitope patterns and second structural epitope sequences on
the 3D-struture of
the protein, and finally using a antigenic peptide sequence corresponding to
the identified
structural epitope for predicting binding of specific antibodies in a human or
animal to a poten-
tial immunogen, can be applied for any kind of immunogen or immunogenic
protein. Using
other immunogens than those specifically mentioned will be obvious for the
skilled person and
is to be considered within the scope of the present invention.
In one embodiment the immunogen is an antigen, and particularly an allergen.
Immunogenic protein or immunogen
to ~ The "immunogenic protein" or "immunogen" can in principle be any protein
molecule of
biological origin, non-limiting examples of which are peptides, polypeptides,
proteins, enzymes,
post-translationally modified polypeptides such as lipopeptides or
glycosylated peptides, anti-
microbial peptides or molecules, toxins, marker proteins of bacterial, viral
or mammalian origin
which indicate a specific disease, such as e.g. cancer or a specific
infection, and proteins hav-
ing pharmaceutical properties etc.
Accordingly in one embodiment, the "immunogen" is chosen from the group
consisting
of polypeptides, small peptides, lipopeptides, antimicrobials, toxins, marker
proteins, pharma-
ceutical polypeptides, enzymes, industrial proteins and environmental
allergens. Particularly,
the allergen is an enzymes or an environmental allergen or a pharmaceutical
peptide.
The term "pharmaceutical polypeptides" is defined as polypeptides, including
peptides,
such as peptide hormones, proteins and/or enzymes, being physiologically
active when intro-
duced into the circulatory system of the body of humans and/or animals.
Pharmaceutical polypeptides are potentially immunogenic as they are introduced
into the
circulatory system.
Examples of "pharmaceutical polypeptides" contemplated according to the
invention in-
clude insulin, ACTH, glucagon, somatostatin, somatotropin, thymosin,
parathyroid hormone, pig-
mentary hormones, somatomedin, erythropoietin, luteinizing hormone, chorionic
gonadotropin,
hypothalmic releasing factors, antidiuretic hormones, thyroid stimulating
hormone, relaxin, inter-
feron, thrombopoietin (TPO) and prolactin.
ao However, the proteins are particularly to be used in industry, housekeeping
and/or
medicine, such as proteins used in personal care products (for example
shampoo; soap; skin,
hand and face lotions; skin, hand and face cremes; hair dyes; toothpaste),
food (for example in
the baking industry), detergents and pharmaceuticals.
17
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
Antimicrobial peptides
The antimicrobial peptide (AMP) may be, e.g., a membrane-active antimicrobial
pep-
tide, or an antimicrobial peptide affecting/interacting with intracellular
targets, e.g. binding to
cell DNA. The AMP is generally a relatively short peptide, consisting of less
than 100 amino
acid residues, typically 20-80 residues. The antimicrobial peptide has
bactericidal and/or fungi-
cidal effect, and it may also have antiviral or antitumour effects. It
generally has low cytotoxicity
against normal mammalian cells.
The antimicrobial peptide is generally highly cationic and hydrophobic. It
typically con-
to tains several arginine and lysine residues, and it may not contain a single
glutamate or aspa-
ratate. It usually contains a large proportion of hydrophobic residues. The
peptide generally
has an amphiphilic structure, with one surface being highly positive and the
other hydrophobic.
The antimicrobial peptide may act on cell membranes of target microorganisms,
e.g.
through nonspecific binding to the membrane, usually in a membrane-parallel
orientation, in-
teracting only with one face of the bilayer.
The antimicrobial peptide typically has a structure belonging to one of five
major classes: a
helical, cystine-rich (defensin-like), f3-sheet, peptides with an unusual
composition of regular
amino acids, and peptides containing uncommon modified amino acids.
2o Enzymes
In one embodiment of the invention the protein is an enzyme or enzyme variant.
It is
to be understood that enzyme variants (produced, for example, by recombinant
techniques)
are included within the meaning of the term "enzyme". Examples of such enzyme
variants are
disclosed, e.g., in EP 251,446 (Genencor), WO 91/00345 (Novo Nordisk), EP
525,610 (Solvay)
and WO 94/02618 (Gist-Brocades NV).
Particularly the enzyme is selected from the group consisting of of glycosyl
hydrolases, carbohydrases, peroxidases, proteases, lipolytic enzymes,
phytases,
polysaccharide lyases, oxidoreductases, transglutaminases and
glucoseisomerases.
The enzyme classification employed in the present specification with claims is
in
3 o accordance with Recommendations (7992) of the Nomenclature Committee of
the International
Union of Biochemistry and Molecular Biology, Academic Press, Inc., 1992.
Accordingly the types of enzymes which may appropriately be incorporated in
granules of the invention include oxidoreductases (EC 1.-.-.-), transferases
(EC 2.-.-.-),
hydrolases (EC 3.-.-.-), lyases (EC 4.-.-.-), isomerases (EC 5.-.-.-) and
ligases (EC 6.-.-.-).
18
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
In particular oxidoreductases in the context of the invention are peroxidases
(EC
1.11.1 ), laccases (EC 1.10.3.2) and glucose oxidases (EC 1.1.3.4)].
In particular transferases are transferases in any of the following sub-
classes:
a) Transferases transferring one-carbon groups (EC 2.1 );
b ) transferases transferring aldehyde or ketone residues (EC 2.2);
acyltransferases (EC 2.3);
c ) glycosyltransferases (EC 2.4);
d) transferases transferring alkyl or aryl groups, other that methyl groups
(EC 2.5); and
e) transferases transferring nitrogeneous groups (EC 2.6).
A particular type of transferase in the context of the invention is a
transglutaminase (protein
to glutamine y-glutamyltransferase; EC 2.3.2.13). Further examples of suitable
transglutaminases
are described in WO 96/06931 (Novo Nordisk A/S).
In particular hydrolases in the context of the invention are: Carboxylic ester
hydrolases (EC 3.1.1.-) such as lipases (EC 3.1.1.3); phytases (EC 3.1.3.-),
e.g. 3-phytases
(EC 3.1.3.8) and 6-phytases (EC 3.1.3.26); glycosidases (EC 3.2, which fall
within a group
denoted herein as "carbohydrases"), such as a-amylases (EC 3.2.1.1 );
peptidases (EC 3.4,
also known as proteases); and other carbonyl hydrolases].
In the present context, the term "carbohydrase" is used to denote not only
enzymes
capable of breaking down carbohydrate chains (e.g. starches or cellulose) of
especially five-
and six-membered ring structures (i.e. glycosidases, EC 3.2), but also enzymes
capable of
2o isomerizing carbohydrates, e.g. six-membered ring structures such as D-
glucose to five-
membered ring structures such as D-fructose. Carbohydrases of relevance
include the
following (EC numbers in parentheses): a-amylases (EC 3.2.1.1 ), ~i-amylases
(EC 3.2.1.2),
glucan 1,4-a-glucosidases (EC 3.2.1.3), endo-1,4-beta-glucanase (cellulases,
EC 3.2.1.4),
endo-1,3(4)-[3-glucanases (EC 3.2.1.6), endo-1,4-~-xylanases (EC 3.2.1.8),
dextranases (EC
z5 3.2.1.11), chitinases (EC 3.2.1.14), polygalacturonases (EC 3.2.1.15),
lysozymes (EC
3.2.1.17), ~-glucosidases (EC 3.2.1.21 ), a-galactosidases (EC 3.2.1.22), ~i-
galactosidases
(EC 3.2.1.23), amylo-1,6-glucosidases (EC 3.2.1.33), xylan 1,4-~i-xylosidases
(EC 3.2.1.37),
glucan endo-1,3-~i-D-glucosidases (EC 3.2.1.39), a- A sixth aspect of the
invention relates to a
method dextrin endo-1,6-a-glucosidases (EC3.2.1.41), sucrose a-glucosidases
(EC 3.2.1.48),
3o glucan endo-1,3-a-glucosidases (EC 3.2.1.59), glucan 1,4-~i-glucosidases
(EC 3.2.1.74),
glucan endo-1,6-~i-glucosidases (EC 3.2.1.75), arabinan endo-1,5-a-L-
arabinosidases (EC
3.2.1.99), lactases (EC 3.2.1.108), chitosanases (EC 3.2.1.132) and xylose
isomerases (EC
5.3.1.5).
In particular isomerases in the context of the invention are glycoseisomerases
19
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
In particular lyases in the context of the invention are polysaccharide
lyases.
Environmental immunogens
The environmental immunogens that are of interest include allergens from
pollen, dust
s mites, mammals, venoms, fungi, food items, and other plants.
Pollen, allergens include but are not limited to those of the order Fagales,
Oleales, Pi-
nales, Poales, Asterales, and Urticales; including those from Betula, Alnus,
Corylus, Carpinus,
Olea, Phleum pratense and Artemisia vulgaris, such as Aln g1, Cor a1, Car b1,
Cryj1, Amb a1
and a2, Art v1, Par j1, Ole e1, Ave v1, and Bet v1 (WO 99/47650).
to Mite allergens include but are not limited to those from Derm. farinae and
Derm.
pteronys., such as Der f1 and f2, and Der p1 and p2.
From mammals, relevant environmental allergens include but are not limited to
those
from cat, dog, and horse as well as from dandruff from the hair of those
animals, such as Fel
d1; Can f1; Equ c1; Equ c2; Equ c3.
i5 Venum allergens include but are not limited to PLA2 from bee venom as well
as Apis
m1 and m2, Ves g1, g2 and g5, Ves v5 and to Pol and Sol allergens.
Fungal allergens include those from Alternaria alt. and Cladospo. herb. such
as Alt a1 and Cla
h1.
Food allergens include but are not limited to those from milk (lactoglobulin),
egg
20 (ovalbumin), peanuts, hazelnuts, wheat (alfa-amylase inhibitor),
Other plant allergens include latex (hevea brasiliensis).
The above described kit and high throughput screening method will be of great
impor-
tance in order to more specifically identify the exact cause of an observed
immunogenic re-
sponse in a human or animal since the use of a antigenic peptide sequence
corresponding to a
25 structural epitope on an immunogen will give a much more specific answer
than if a linear epi-
tope was used. Also the identification of antigenic peptide sequences
corresponding to struc-
tural epitopes on potential immunogens will facilitate the use of such
antigenic peptide se-
quences in order to get more specific vaccines.
In a fifth aspect the present invention relates to a vaccine comprising at
least one anti-
genic peptide sequence corresponding to a structural epitope comprised in at
least one poten-
tial immunogen and said antigenic peptide sequence being capable of binding at
least one an-
tibody specific for a structural epitope comprised in a potential immunogen,
and also in a sixth
aspect to a method of preparing a vaccine comprising adding at least one
antigenic peptide
sequence corresponding to a structural epitope comprised in at least one
potential immunogen
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
and said antigenic peptide sequence being capable of binding at least one
antibody specific for
a structural epitope comprised in a potential immunogen to a liquid medium.
In the seventh aspect, the invention relates to the use of at least one
antigenic peptide
sequence, corresponding to a structural epitope comprised in at least one
potential immuno-
gen and said antigenic peptide sequence being capable of binding at least one
antibody spe-
cific for a structural epitope comprised in a potential immunogen, for the
preparation of a vac-
cine. Use of the vaccine of the invention for the treatment of a human or an
animal also falls
withing the scope of the present invention.
MATERIALS AND METHODS
Materials
ELISA reagents:
Horse Radish Peroxidase labelled pig anti-rabbit-Ig (Dako, DK, P217, dilution
1:1000).
Rat anti-mouse IgE (Serotec MCA419; dilution 1:100).
Mouse anti-rat IgE (Serotec MCA193; dilution 1:200).
Biotin-labelled mouse anti-rat IgG1 monoclonal antibody (~ymed 03-9140;
dilution 1:1000)
to Biotin-labelled rat anti-mouse IgG1 monoclonal antibody (Serotec MCA336B;
dilution 1:2000)
Streptavidin-horse radish peroxidase (Kirkegard & Perry 14-30-00; dilution
1:1000).
Buffers and Solutions:
- PBS (pH 7.2 (1 liter))
NaCI 8.00 g
KCI 0.20 g
K2HP04 1.04 g
KH2P04 0.32 g
- Washing buffer PBS, 0.05% (v/v) Tween 20
- Blocking buffer PBS, 2% (wt/v) Skim Milk powder
zo - Dilution buffer PBS, 0.05% (v/v) Tween 20, 0.5% (wt/v) Skim Milk powder
- Citrate buffer 0.1 M, pH 5.0-5.2
- Stop-solution (DMG-buffer)
- Sodium Borate, borax (Sigma)
- 3,3-Dimethyl glutaric acid (Sigma)
- Tween 20: Poly oxyethylene sorbitan mono laurate (Merck cat
21
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
no. 822184)
- PMSF (phenyl methyl sulfonyl flouride) from Sigma
- Succinyl-Alanine-Alanine-Proline-Phenylalanine-paranitro-anilide (Suc-AAPF-
pNP) Sigma no.
S-7388, Mw 624.6 g/mol.
- mPEG (Fluka)
Colouring substrate:
OPD: o-phenylene-diamine, (Kementec cat no. 4260)
Methods
to Automatic epitope map~inq
Implementation:
The implementation consists of 3 pieces of code:
1. The core program (see above), written in C (see Appendix A).
2. A "wrapping" cgi-script run by the web server, written in Python (see
Appendix B).
3. A HTML page defining the input/submission form (see Appendix C).
The wrapper receives the input and calls the core program and several other
utilities. Apart
2o from the standard Unix utility programs (mv, rm , awk, etc..) the following
must be installed:
~ A web server capable of running cgi-scripts, eg. Apache
~ Python 1.5 or later
~ Gnuplot 3.7 or later
~ DSSP, version July 1995
The core program:
Inputs
1. A Brookhaven PDB file with the structure of the protein
2. The output of DSSP called with the above PDB file.
3. Maximum distance between adjacent residues
22
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
4. Minimum solvent accessible surface area for each residue
5. Maximum epitope size (max distance between any two residues in epitope)
6. Maximum number of non-redundant epitopes to include (0 = all)
7. The shortest acceptable epitope (as a fraction of the length of the epitope
consensus
sequence).
8. Epitope consensus sequence describing which residues are possible at the
different
positions. An example is shown below:
KR (Lys og Arg allowed)
to AILV- (Ala, Ile, Leu, Val or missing residue allowed)
* (All residues allowed, but there must be a residue)
? (All or missing residue allowed)
DE (Asp or Glu allowed)
(*, ? or - in first or last position is allowed but obsolete. (- in first
position is ignored.))
Examples of matching epitopes:
KAAKD, KLASD, KLYSD, KLY-D, R-M-D.
a o The epitope searching algorithm:
The "core" of the program is the algorithm that scans the protein surface for
the epitope pat-
terns. The principle is that several "trees" are built, where each of their
branches describes one
epitope:
All residues in the protein are checked according to: a) Does the residue type
match the
z5 first residue of the epitope consensus sequence. b) Is the surface
accessibility greater than or
equal to the given threshold. If both requirements are fulfilled, the protein
residue is considered
as one root in the epitope tree. Remark that there are usually many roots.
1. For each of the residues defined as roots, all residues within the the
given threshold
distance between adjacent residues (e.g. 7 Angstroms) are checked for the same
as
3o above: a) Does the residue type match the second residue of the epitope
consensus
sequence. b) Is the surface accessibility greater than or equal to the given
threshold. If
yes, the protein residue is considered as a "child" of the root. The spatial
position of a
residue is defined as the coordinates of its C-alpha atom.
23
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
2. The procedure from step 2 is repeated for the next residue in the epitope
consensus
sequence, where each of the "childs" found in step 2 are now "roots" of new
childs. If a
gap is defined in the epitope consensus sequence, a "missing" residue is
allowed, and
the coordinates of the root (also called "parent") is used.
3. This procedure is repeated for all residues in the epitope consensus
sequence.
4. In this way a number of trees (corresponding to the number of roots found
in step 1 )
are found. Notice that the same protein residue can be present many places in
the
trees.
5. If no epitopes that matches the length of the epitope consensus sequence
are found,
to the longest shorter epitopes that matches the first n residues of the
epitope consensus
sequence are used, where n is an integer smaller than the length of the
epitope con-
sensus sequence. If n is smaller than the length of the epitope consensus
sequence
multiplied by the fraction value defining the shortest acceptable epitope
length, no epi-
topes are written to the output, and steps 7, 8 and 9 are skipped.
6. The epitopes are extracted from the trees by traversing down from each of
the "childs"
in the last level. The algorithm also finds epitopes which have the same
protein residue
present more than once. This is, of course, an artifact and such epitopes are
discarded.
Every epitope is then checked for its size, that is, the maximum distance
between any
two residues which are members of the epitope. If this exceeds the threshold,
the epi-
2o tope is discarded.
7. Redundant epitopes are removed. Epitopes containing one or more gaps are
redun-
dant if they are subsets of other epitopes without or with fewer gaps. For
example:
A82-gap-F45-G44-K43 is a subset of A82-L46-F45-G44-K43, and is therefore dis-
carded.
8. For every epitope, the total solvent accessible surface area is calculated
(by adding the
contributions from each residue as found by the DSSP program). The epitopes
are
sorted according to this area in descending order. If a maximum number of n
non-
redundant epitopes has been specified, the n epitopes with largest solvent
accessible
surface area are selected.
s o 9. The output consists of a list of the found epitopes, along with
information of the epitope
consensus sequence used and other internal parameters. A separate file
containing the
number of epitopes that each of the protein residues is a member of is also
written.
24
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
The wrapper.
Inputs
1. One PDB file, describing one structure, or one ZIP file, containing a
number of PDB
files, each describing one structure. The ZIP file must not contain
subfolders.
2. An epitope consensus sequence or which part of the current epitope library
to use (full
library or IgE part or IgG part).
3. Maximum distance between adjacent residues
l0 4. Minimum solvent accessible surface area for each residue
5. Maximum epitope size (max distance between any two residues in epitope)
6. Maximum number of non-redundant epitopes to include (0 = all)
7. Whether to use sequential numbering (1,2,3,4,..... etc) or PDB-file
numbering.
Description
The core program accepts only one structure and one epitope consensus
sequence. It
is usually desirable to use a library of epitope consensus sequences and
sometimes several
protein structures. The wrapper reads the user input and calls the utility
programs and the core
2o program the necessary number of times. The output is collected and
presented on the web
page returned to the user.
Depending on the type of input, the wrapper works in different modes:
~ Epitope consensus can be given directly or taken from a library
~ Input type can be a single PDB file or a collection of PDB file given as a
ZIP-file.
Any of the four possible combinations are allowed.
The epitope library consists of a number of text files, each containing one
epitope consensus
sequence as specified above.
The layout of the wrapper is like this:
1. Check if the program is already in use from somewhere else (this is done by
checking
for a lock file when the wrapper starts. If it does not exist, it is created
and removed
again when the program is finished).
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
2. If the epitope consensus sequences are to be read from the library, make an
internal
list of the desired library entries.
3. If the input type is a ZIP file, unzip the file and create one new
directory for each of the
conatined PDB files. Move each PDB file to its corresponding directory.
4. Do a loop over the structures and/or epitope consensus sequences. For each
struc-
ture/epitope consensus sequence pair, DSSP and the core program is called with
the
required parameters. If the input type is a ZIP file, the outputs are put in
the appropriate
directories.
5. If the epitope library is used, a sum file containing the total number of
epitopes each
to residue is a member of. (Such a file is generated by the core program for
each epitope
consensus sequence - here a sum of these files is calculated). If input type
is a ZIP
file, a sum file is generated for each structure and put in the appropriate
directory.
6. If the epitope library is used, a file containing the total number of
epitopes found from
each entry in the epitope library. If the input type is a PDB file, the file
contains only one
line (with a number of data corresponding to the library size). If the input
type is a ZIP
file, there is one line for each structure.
7. Depending on the combination of input type (ZIP or single PDB) and epitope
consen-
sus sequence source (typed-in or epitope library), different information is
returned to
the user:
2o Single PDB + typed in epitope: Graph of numbers of epitopes that each
residue is a
member of. List of found epitopes.
ZIP file + typed in epitope: Graphs (one for each structure) of numbers of
epitopes that
each residue is a member of. Lists (one for each structure) of found epitopes.
Single PDB + epitope library: Graph of numbers of epitopes that each residue
is a
a5 member of (total for the complete library).
ZIP file + epitope library: Graphs (one for each structure) of numbers of
epitopes that
each residue is a member of (total for the complete library).
Data flow sheets for the four different are shown in the figure
8. For all modes except Single PDB + typed in epitope, a ZIP file containing
all output files
3o is created and returned to the user.
26
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
ELISA Procedure for detecting serum levels of IaE and IaG:
Specific IgG and IgE levels were determined using the ELISA specific for
human, mouse or rat
IgG or IgE. Differences between data sets were analysed by using appropriate
statistical meth-
ods. The assays were performed as known to the expert.
Activation of CovaLink plates:
A fresh stock solution of cyanuric chloride in acetone (10 mg/ml) is diluted
into PBS, while stir-
ring, to a final concentration of 1 mg/ml and immediately aliquoted into
CovaLink NH2 plates
(100 microliter per well) and incubated for 5 minutes at room temperature.
After three washes
with PBS, the plates are dryed at 50°C for 30 minutes, sealed with
sealing tape, and stored in
to plastic bags at room temperature for up to 3 weeks.
Protein seauences and alignments:
For purposes of the present invention, the degree of homology may be suitably
determined by
means of computer programs known in the art, such as GAP provided in the GCG
program
package (Program Manual for the Wisconsin Package, Version 8, August 1994,
Genetics
i5 Computer Group, 575 Science Drive, Madison, Wisconsin, USA 53711)
(Needleman, S.B. and
Wunsch, C.D., (1970), Journal of Molecular Biology, 48, 443-45).
Examples of alignments are described in WO 01/83559.
Structures
2o The structure of Savinase~ can be found in Betzel et al., J.MoI. Biol.,
vol. 223, p. 427, 1992
(1 svn.pdb).
Homology modelling
As described earlier one needs the 3-dimentional structure coordinates of an
acceptor protein
25 to find the epitope sequences on its surface. These coordinates if not
already in a database
can be deduced from the coordinates of a homologous protein. Typical actions
required for the
construction of a model structure are: alignment of homologous sequences for
which 3-
dimensional structures exist, definition of Structurally Conserved Regions
(SCRs), assignment
of coordinates to SCRs, search for structural fragments/loops in structure
databases to replace
3o Variable Regions, assignment of coordinates to these regions, and
structural refinement by
energy minimization.
27
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
Examples of 3D-structural models are described in WO 01/83559, where three di-
mensional structural models of the subtilisins properase, relase, ProteaseC,
ProteaseD, Prote-
aseE, and PROTEASE B were constructed based on three dimensional structure of
Savi-
nase~ (Protein Data Bank entry 1 SVN; Betzel, C., Klupsch, S., Papendorf, G.,
Hastrup, S.,
Branner, S., Wilson, K. S.: Crystal structure of the alkaline proteinase
Savinase~ from Bacillus
lentus at 1.4 ~, resolution. J Mol Biol 223 pp. 427 (1992)) using the Modeller
50 (Bali, A.; T.L.
Blundell, "Definition of general topological equivalence in protein
structures: A procedure in-
volving comparison of properties and relationships through simulated annealing
and dynamic
programming," J. Mol. Biol., 212 403-428 (1990)) module of the Insight 2000
molecular model-
to ling package (Biosym inc.). Default parameters were used with the
alignments shown in Figure
1A (WO 01/83559) as input, e.g. alignment between the columns labelled
Savinase~ and
PROTEASE B served as input alignment in construction of a PROTEASE B
structural model.
The Modeller module by default output ten structural models, of these the
model with lowest
'modeller objective function' score was chosen as representing PROTEASE B
structure.
The amylase used in the examples of WO 01/83559 is the alpha-amylase of
Bacillus halma
palus (WO96/23873), which is called amylase SP722 (the wild-type). Its
sequence is shown in
SEQ ID NO 2 (WO 01/83559) and the corresponding protein structure was built
from the BA2
structure, as described in W096/23874. The first four amino acids of the
structural model are not
defined, hence the sequence used for numeration of amino acid residues in the
examples of this
a o invention is four amino acids shorter than the one of the full length
protein SP722.
Several variants of this amylase are available (W096/23873). One particularly
useful variant
has deleted two amino acid residues at D-G at positions 183 and 184 of the SEQ
ID NO 2 (WO
01/83559)(corresponding to residues 179 and 180 of the modelled structure).
This variant is
called JE-1 or Natalase.
Another amylase that is particularly useful is the amylase AA560: This
alkaline a-amylase
may be derived from a strain of Bacillus sp. DSM 12649. The strain was
deposited on 25th Janu-
ary 1999 by the assignee under the terms of the Budapest Treaty on the
International Recognition
of the Deposit of Microorganisms for the Purposes of Patent Procedure at
Deutshe Sammmlung
von Microorganismen and Zellkulturen GmbH (DSMZ), Mascheroder Weg 1 b, D-38124
Braun-
3o schweig DE.
28
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
EXAMPLES.
Example 1.
From a phage display library expressing random hexa-, nona- or dodecc peptides
as part of
their membrane proteins, specific phage clones were isolated capable of
binding specific anti-
bodies. The DNA sequence encoding the displayed peptide of one such clone was
determined
according to standard procedure. The amino acid sequence of the corresponding
oligopeptide
was deduced from the DNA sequence. This analysis revealed the peptide
VQVYGDTSA as a
specific antibody binding peptide.
to Epitope pattern.
By sequence alignment using the "geometric body" approach as described
earlier, the epitope
pattern: Q206 > Y214 > D41 > was localized on the 3D structure of Savinase~.
The identified epitope pattern, Q206 > Y214 > D41 > , was then fitted with the
3D-structure of
Savinase~.
Epitope sequence.
The potential epitope sequence was identified by incorporation of the non-
anchor amino acids
identified by the phage display (settings: 100% homology, and > 20 A2
accessibility for each
amino acid). This identified the potential epitope sequence: Q206 V81 Y214 G80
D41 T208.
ao Detailed description of how to map epitopes and identify potential epitope
sequences
is also disclosed in WO 00/26230 and WO 01/83559 the content of which is
hereby incorpo-
rated by reference.
Example 2.
Epitope mapping was also used to identify epitope patterns specific for
Alcalase~, Savinas~,
and Subtilisin Novo~. These proteases crossreact significantly in ELISA using
specific rabbit
antibody. The specific epitope patterns are shown in Table 1 below and epitope
patterns,
which are specific for each of these proteases, are underlined.
Table 1.
Alcalase~ Savinase~ Subtilisin Novo~
29
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
E~_
Epi#05
Epi#05 Epi#05
Epi#06
Eal#08
Epi#09
Epi#09 Epi#09
Eli#10
Epi#12
Epi#12
Epi#14
Epi#17 Epi#17 Epi#17
Epi#18 Epi#18 Epi#18
Epi#19
Epi#19 Epi#19
Epi#22 . Epi#22
Epi#23 Epi#23
EJai#24
Epi#25
Epi#26
Eli#27
Epi#28 Epi#28
EJ~i#29
E
Epi#31 Epi#31
Epi#32
Epi#33 Epi#33
Epi#34 Epi#34
E~i#35
Epi#36 Epi#36
Epi#37 Epi#37
Epi#40 Epi#40 Epi#40
Epi#41
Ehi#42
Epi#44 Epi#44 Epi#4.4
Epi#45
E~i#4.6
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
Eli#47
Epl#48
Epi#53 Epi#53
Example 3.
From the example above, epitope#10 appears to be specific only for Savinase~,
and this epi-
tope can be translated into the following structural epitope sequences on the
3D-structure of
the protease:
8180 F183 S182 N179 A173 T218 A156
8180 F183 S182 N179 D175 H220 T218
3D imaging of the protease showed that the epitopes were localized on the
surface of the 3D
structure.
The following sequences were synthezised, and immobilized on biotin, through a
linker molecule:
RRFANDHTR,and
RRFSNATRA
Alternatively, these sequences were cloned into the P8 membrane protein of
phage lambda.
2o The biotin-complex was immobilized in ELISA well plates, pre-coated with
strepta-
vidin. If phages were used, these were directly coated into the ELISA plate
wells. ELISA was
performed as described elsewhere, on sera from rabbits, rats, and mice raised
against Alca-
lase~, Savinase~, and Subtilisin Novo~ as well as a number of less relevant
proteins.
In Fig. 1 the reactivity of the selected peptides in terms of antibody binding
capacity is
a5 shown (ELISA assay).
The different proteins are marked by capital letters A through H. A =
Alcalase~, B =
Savinase~, C = Subtilisin Novo~, D = Carezyme~ (cellulase), E = Laccase, F =
Natalase~
(amylase), G = SP722 (amylase), H = Lipolase~ (lipase).
The light gray bar represents the sequence R R F A N D H T R, and the dark
gray bar
ao represents the sequence R R F S N A T R A.
31
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
Reactivity was observed with anti-Savinase~ antibody only, demonstrating the
speci-
ficity of both linear antigenic peptide sequences corresponding to the two
structural epitope
sequences:RRFANDHTR,andRRFSNATRA.
32
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
Appendix A
SOURCE CODE FOR THE CORE C PROGRAM (EPITOPE.C)
/* This is epitope.c */
/* EPF 25-10-2000 */
to
/* _______________________ pEFINES -_________________________ */
#define MAXRESIDUES 1000
#define MAXCONSENSUS 15
#define MAXEPITOPERES 30000
#define MAXEPITOPES 10000
#define AMINOACIDS "ACDEFGHIICLMNPQRSTVWY"
#define AMINOACIDS3 "ALA CYS ASP GLU PHE GLY HIS ILE LYS LEU MET ASN PRO
GLN ARG SER THR VAL TRP TYR "
zo #define REVISIONDATE "12-02-2001"
#define max(A, B) ((A) > (B) ? (A) : (B))
#define min(A, B) ((A) < (B) ? (A) : (B))
/* _______________________ INCLUDES -_________________________ */
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <math.h>
so #include <limits.h>
/* _______________________ STRUCTS -_________________________ */
struct residue
33
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
{
char Itr3[3];
char Itr;
float x, y, z;
int sass, number;
int member of epitopes; /* how many epitopes is this residue part of ? */
struct epitoperesidue
to {
int parent; /* -1 if top level */
int residue; /* -1 if gap */
char level;
struct epitope
{
int sasa, gaps, residues, res[MAXCONSENSUS];
char epi[255];
2o char subset; /* is this epitope a subset of another */
float size;
/* _______________________ GLOBALS -_________________________ */
struct residue res[MAXRESIDUES];
struct epitoperesidue epires[MAXEPITOPERES];
char consensus[MAXCONSENSUS][22];
3o struct epitope epi[MAXEPITOPES];
int numofres = 0, numofepires = 0, consensuslength = 0;
int minsasa = 0, numofepitopes = 0, numofsubsets = 0;
float mindist = 7, sqmindist, maxsize, sqmaxsize, minlength = 0;
int maxepi = 0, minlength_residues, longestepitope;
34
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
/* _______________________ FILE FUNCTIONS -______________________ */
int readconsensus(char *filename)
/* return length of consensus sequence */
Intl=0;
FILE *infile;
char buffer[255], end = 0;
if (infile = fopen(filename, "r"))
/* This code adds linefeeds to the consensus file. This is because there must
be a newline after the last line. Because of permission problems, this has
been moved to
2o the wrapping cgi-script instead
fclose(infile);
infile = fopen(filename, "a");
fprintf(infile,"\n\n");
fclose(infile);
infile = fopen(filename, "r");
*/
while (!feof(infile) && !end)
fgets (buffer, 255, infile);
if (strlen(buffer) > 22)
printf ("Too many residue types in consensus residue %d\n",i+1 );
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
printf ("using all 20 types instead.\n");
strcpy (consensus[i], AMINOACIDS);
}
else if (strchr(buffer,'*')) /* wildcard '*' means any residue, but no gap *l
s strcpy (consensus[i], AMINOACIDS);
else if (strchr(buffer,'?')) /* wildcard '*' means any residue or gap */
f
strcpy (consensus[i], AMINOACIDS);
strcat (consensus[i], "-");
}
else if (!strpbrk(buffer,"ACDEFGHIKLMNPQRSTVWY*?")) /* empty line, end the
loop */
end = 1;
}
else
strncpy (consensus[i], buffer, strlen(buffer)-1 );
i++;
}
}
fclose(infile);
consensuslength = i;
return i;
int readpdbCA(char *filename)
f
ao /* return number of residues */
inti=0;
char *j;
FILE *infile;
36
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
char buffer[255];
char aminoacids[20] = AMINOACIDS;
char aminoacids3[80] = AMINOACIDS3;
if (infile = fopen(filename, "r"))
while (!feof(infile))
f
fgets (buffer, 255, infile);
to if (!strncmp(buffer,"ATOM",4) && !strncmp(buffer+13,"CA",2)) l* get only
the CA atoms */
strncpy(res[i].Itr3,buffer+17,3);
if Q = strstr(aminoacids3,res[i].Itr3))
res[i].Itr = aminoacids[(j-aminoacids3)/4];
else
printf("Unknown residue type: %s\n",res[i].Itr3);
res[i].Itr = 'X';
)
zo res[i].x = atof(buffer+30);
res[i].y = atof(buffer+38);
res[i].z = atof(buffer+46);
res[i].member of epitopes = 0;
res[i].number = atoi(buffer+22);
z 5 i++;
)
numofres = i;
3 o return i;
int readdssp(char *filename)
37
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
/* return number of residues *!
inti=0;
char *j;
FILE *infile;
char buffer[255];
strcpy (buffer," ");
to
if (infile = fopen(filename, "r"))
f
while (!feof(infile) && strncmp(buffer," # RESIDUE AA",15)) /* find where data
begins */
fgets (buffer, 255, infile);
while (!feof(infile))
fgets (buffer, 255, infile);
if (!feof(infile))
if ((buffer[13] -- res[i].Itr && atoi(buffer+5) -- res[i].number
~~(strchr("abcdefghijklmnopqrstuvwxyz",buffer[13]) && res[i].Itr =- 'C' &&
atoi(buffer+5) _-
res[i].number ) )
a5 res[i].sasa = atoi(buffer+35);
i++;
)
else
printf("Inconsistency between pdb and dssp file at residue %c%d\n",res[i].Itr,
3o res[i].number);
)
)
if (i != numofres)
38
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
printf("Inconsistency between pdb and dssp file: wrong # of residues (%d) in
pdb, (%d) in
dssp\n", numofres, i);
return i;
void writedatafile(char *filename)
(
to
int i;
FILE *outfile;
If (outfile = fopen(filename, "w"))
fprintf(outfile,"# seq pdb AA epitopes\n");
fprintf(outfile,"# seq \n");
for (i=0; i<numofres; i++)
zo fprintf(outfile,"%4d %4d %c %4d\n",i+1 , res[i].number, res[i].Itr,
res[i].member of epitopes);
fclose(outfile);
/* ___________________ ANALYSIS FUNCTIONS -______________________ */
ao int addchild(int parent, int residue, char level)
if (numofepires == MAXEPITOPERES)
39
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
printf("Sorry, program constant MAXEPITOPERES exceeded, increase and recompile
pro-
g ram\n");
exit (0);
epires[numofepires].parent = parent; /* should be -1 for the top level */
epires[numofepires].residue = residue; /* should be -1 for a gap */
epires[numofepires].level = level;
to numofepires++;
/*
*/
if (numofepires % 10 == 0)
printf ("Added %d epires\n",numofepires);
return numofepires;
zo float sqdist(int i, int j)
f
/* returns the square of the distance between the coordinates for residues i
and j */
. return (res[i].x-res[j].x)*(res[i].x-res[j].x)+(res[i].y-res[j].y)*(res[i].y-
res[j].y)+(res[i].z-
res[j].z)*(res[i].z-res[j].z);
void findepitopes(void) /* This is the core algorithm */
int i, j, k, nogapanchestor;
/* --- Find parents --- */
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
fior(i=0; i<numofres; i++)
if (res[i].sasa >= minsasa && strchr(consensus[0],res[i].Itr))
addchild(-1,i,0);
/* ---- do 'consensuslength-1' number of child cycles -------- */
to for (i=1; i<consensuslength; i++)
for (j=numofepires-1; j>=0 && epires[j].level == i-1; j--)
f
if (strchr(consensus[i],'-')) /* is a gap allowed at this position in the
consensus ? */
addchild(j,-1,i);
if (epires[j].residue =- -1) l* this a gap, so use distance to parents (or
older anchestor)
instead */
/* the following line is for handling multiple gaps after each other */
for (nogapanchestor = epires[j].parent; epires[nogapanchestor].residue =- -1;
nogapan-
chestor = epires[nogapanchestor].parent);
fork=0; k<numofres; k++)
z5 /* if (res[k].sass >= minsasa && strchr(consensus[i],res[k].Itr) && k !_
epires[epires[j].parent].residue && sqdist(k,epires[epires[j].parent].residue)
<= sqmindist) */
if (res[k].sasa >= minsasa && strchr(consensus[i],res[k].Itr) && k !_
epires[nogapanchestor].residue && sqdist(k,epires[nogapanchestor].residue) <=
sqmindist)
addchildQ,k,i);
else
fork=0; k<numofres; k++)
41
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
if (res[k].sass >= minsasa && strchr(consensus[i],res[k].Itr) && k !=
epires[j].residue &&
sqdist(k,epires[j].residue) <= sqmindist)
addchild(j,k,i);
longestepitope = epires[numofepires-1].level+1;
int cmp(const void *a, const void *b)
struct epitope *aa = (struct epitope *)a;
struct epitope *bb = (struct epitope *)b;
if (aa->sasa < bb->sasa)
return 1;
2 o else if (aa->sasa == bb->sasa)
return 0;
else
return -1;
void processepitopes(void) /* Go through the epitopes, remove copies, nonsense
sequences
etc. */
3o int i, j, k, I, n, thisepinumbers[MAXCONSENSUS], processed=0;
char thisepi[255], tmp[50];
char discarded, toobig, onepresent, allpresent;
float maxsqdist;
42
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
for (i=numofepires-1; i>=0 && epires[i].level == epires[numofepires-1].level;
!--)
f
discarded = 0; toobig = 0;
strcpy(thisepi,"");
j=i;
n=0;
maxsqdist = 0;
do {
to thisepinumbers[n++] = epires[j].residue;
if (epires[j].residue =- -1 ) /* its a gap */
sprintf(tmp "- ")'
else
sprintf(tmp,"%c%d, ", res[epires[j].residue].Itr,
res[epires[j].residue].number);
if (strstr(thisepi,tmp) && epires[j].residue !_ -1 ) /* only gaps can be
present twice! */
discarded = 1;
else
2 o strcat(thisepi,tmp);
j=epires[j].parent;
} while (j !_ -1 );
z5 for (k=0; k <= epires[numofepires-1].level; k++)
for (I=k+1; I <= epires[numofepires-1].level; I++)
if (thisepinumbers[k] !_ -1 && thisepinumbers[I] !_ -1 ) /* if there are no
gaps involved */
maxsqdist = max(maxsqdist, sqdist(thisepinumbers[k],thisepinumbers[I]) );
3o if (maxsqdist > sqmaxsize)
toobig = 1;
if (toobig)
43
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
discarded = 1;
if (!discarded) /* put the found epitopes into the epitope list */
sprintf(epi[numofepitopes].epi,"%s\n",thisepi);
epi[numofepitopes].sasa = 0;
epi[numofepitopes].gaps = 0;
epi[numofepitopes].residues = 0;
to epi[numofepitopes].size = sqrt(maxsqdist);
for Q = 0; j < n; j++) /* loop over the residues in this epitope */
epi[numofepitopes].res[j] = thisepinumbers[j]; /* copy the residue numbers to
the epi-
tope list */
if (thisepinumbers[j] !_ -1) /* if it is not a gap */
epi[numofepitopes].sasa += res[thisepinumbers[j]].sasa;
epi[numofepitopes].residues++;
zo }
else
epi[numofepitopes].gaps++;
z5 numofepitopes++;
if (numofepitopes == MAXEPITOPES)
printf("MEXEPITOPES exceeded. Increase and recompile program.\n");
exit(0);
44
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
/* now indetify epitopes which are a subset of others */
for (i=0; i<numofepitopes; i++) /* initialize array */
epi[i].subset = 0;
for (i=0; i<numofepitopes; i++)
f
for Q=0; j<numofepitopes; j++)
to if (epi[i].residues > epi[j].residues)
allpresent = 0;
for (k=0; k<epi[i].residues; k++)
If (epi[i].res[k] !_ -1 )
onepresent = 0;
for (I=0; I<epi[j].residues; I++)
if (epi[i].res[k] _= epi[j].res[I]) /* if the residues are the same and not
gaps */
a o onepresent = 1;
allpresent ~= onepresent;
)
if (allpresent)
epi[j].subset = 1;
/* numofsubsets++; */
3 0 j'
/* now sort the epitopes according to SASA */
CA 02462651 2004-04-O1
to
WO 03/031981 PCT/DK02/00665
qsort(&(epi[0]),numofepitopes,sizeof(struct epitope), &cmp);
/* counts the ones that are subsets of others */
for (i=0; i<numofepitopes; i++)
if (epi[i].subset == 1 )
numofsubsets++;
/* now count how many epitopes each ressidue is a member of,
considering only non-redundant epitopes, and the number of epitopes wanted */
for (i=0; i < numofepitopes && processed < maxepi; i++)
if (epi[i].subset == 0) /* count only if the epitope is not a subset of
another */
processed++;
for Q=0; j < epi[i].residues; j++)
(res[epi[i].res[j]].member of epitopes)++; /* add the counter for epitopes for
the resi-
2o dues */
void printepitopes(void)
f
int i, processed = 0;
3o for (i=0; i < numofepitopes && processed < maxepi; i++)
if (epi[i].subset == 0)
printf("SAS: %3d, Size %5.2f: %s",epi[i].sasa, epi[i].size, epi[i].epi);
processed++;
46
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
void usage (void)
fprintf(stderr,"USAGE: epitope <epitope template> <filename template> dist acc
maxsize
number minlength\n");
fprintf(stderr,"\n");
to fprintf(stderr,"filenames <filename template>.pdb and <filename
template>.dssp\n");
fprintf(stderr," must be present.\n");
fprintf(stderr,"dist is the maximum distance between adjacent residues in
epitope.\n");
fprintf(stderr,"acc is minimum surface accessible area in square
angstroms.\n");
fprintf(stderr,"maxsize is the maximum distance between any two residues in
the epitope.\n");
fprintf(stderr,"number is the maximum number of non-redundant epitopes to
consider
(0=all)\n");
fprintf(stderr,"minlength is the minimum length of the epitope seqs (in
fractions\n");
fprintf(stderr," of the consensus sequence length).\n");
fprintf(stderr,"A file <filename template>.dat containing the number of
epitopes\n");
2o fprintf(stderr,"each residue participates in is written.\n");
fprintf(stderr,"\n");
exit(0);
int main (int argc, char **arg)
ao int i;
char pdbfile[256], dsspfile[256], datfile[256];
if (argc != 8)
usage();
47
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
readconsensus(arg[1]);
printf ("Epitope consensus sequence read from %s\n",arg[1]);
printf ("-_________________________________________________\n");
for (i = 0; i < consensuslength; i++)
printf("%s\n",consensus[i]);
printf("\n");
to strcpy(pdbfile,arg[2]);
strcat(pdbfile,".pdb");
strcpy(dsspfile,arg[2]);
strcat(dsspfile,".dssp");
strcpy(datfile,arg[2]);
strcat(datfile,".dat");
readpdbCA(pdbfile);
printf ("Sequence read from %s\n",pdbfile);
printf ("-_____________________________\n~~);
for (i = 0; i < numofres; i++)
printf("%c",res[i].Itr);
if (!((i+1)%70))
printf("\n");
3o printf("\n\n");
readdssp(dsspfile);
mindist = atof(arg[3]);
48
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
minsasa = atoi(arg[4]);
maxsize = atof(arg[5]);
maxepi = atoi(arg[6]);
if (maxepi == 0)
maxepi = INT_MAX;
minlength = atof(arg[7]); /* minimum length of epitope sequence (in fractions
of the con-
sensus length) */
sqmindist = mindist*mindist;
to sqmaxsize = maxsize*maxsize;
minlength residues = (float) ceil(minlength*consensuslength);
findepitopes();
if (longestepitope >= minlength_residues)
processepitopes();
printf ("Parameters and internal numbers\n");
2o printf ("_______________________________\n");
printf ("Program revision date : %s\n", REVISIONDATE);
printf ("Consensus sequence length : %d\n", consensuslength);
printf ("Minimum epitope seq length threshold : %.2f (%d residues)\n",
minlength,
minlength residues);
z5 printf ("Longest epitope sequence found : %d\n", longestepitope);
printf ("Number of residues in PDB file : %d\n", numofres);
printf ("Distance threshold value (angstroms) : %.1f\n", mindist);
printf ("Minimum surface accessible area of each res : %d\n", minsasa);
printf ("Maximum epitope size : %.1f\n", maxsize);
s o printf ("Number of nodes in epitope tree : %d\n", numofepires);
printf ("Total number of epitopes.... : %d\n", numofepitopes);
printf ("....of which are subsets of others : %d\n", numofsubsets);
printf ("Max number of non-redundant epitopes : %d\n", maxepi);
printf ("\n");
49
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
printf ("Epitopes found\n");
printf ('~______________\n");
if (longestepitope >= minlength residues)
printepitopes();
writedatafile(datfile);
l o /*
for (i = 0; i < numofepires; i++)
printf("~%4d %4d %4d %4d ",i, epires[i].level, epires[i].residue,
epires[i].parent);
*/
15 return 0;
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
Appendix B
THE WRAPPER (PYTHON) (EPITOPE5.CG1)
#!/z/vaks/bin/python
# Automatic epitope mapping
to
import cgi, os, time, commands, string, sys
FormFile = "epitope.html"
scriptdir = "/z/edhome/epf/public html/epitope/"
i5 epitopepath = "/z/edhome/epf/epitope/epitope3"
dssppath = "/z/vaks/bin/dssp"
gnuplotpath = "/z/edhome/epf/gnuplot-3.7/gnuplot"
zippath = "/usr/freeware/bin/zip"
unzippath = "/usr/freeware/bin/unzip"
timestamp = str(int(time.time()))
liball = range(1,53)
libigg = [3,4,7,11,14,16,17,30,31,32,34,35,38,39,41,42,43,47,48,49,50,51,52]
libige =
[1,2,5,6,8,9,10,12,13,15,18,19,20,21,22,23,24,25,26,27,28,29,33,36,37,40,44,45,
46]
# ------------------ the page startes here
print "Content-type: text/html\n\n" # HTML is following
print '<html>\n'
print '<head>\n'
51
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
print'<title>Automatic epitope mapping</title>\n'
print '</head>\n'
print '\n'
# ------------------- check for lock file
if os.path.isfile("epitope.lock"):
print 'Sorry - lock file exists. This means that automatic epitope mapping is
already in use,'
to print'or that an error has occured.<BR>'
print "If you are absolutely sure that no one are using automatic epitope
mapping, you can"
print "press the button below. <BR>"
print "If you are not sure, just press 'back' in your browser now."
print'<BR><BR>'
print '<form METHOD=GET AC-
TION="http:/leaks.novo.dk/~epf/epitope/epitope removelock.cgi"><input
type="submit"
name="SUBMIT BUTTON" value="Remove lock file"></form>'
sys.exit(0)
25
# ----- create lock file
os.system ("touch epitope.lock")
# ______________ Clean up directory ________________________________
# --- (delete everything but and analysis.cgi and and analysis.html) ---
#commands.getoutput("Is -I ~ awk'$9 !~ /~epitope/ {print \"rm\",$9}'
>cleanup.sh")
#commands.getoutput(". "+scriptdir+"cleanup.sh")
#if os.path.isfile("cleanup.sh"):
52
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
# os.remove ("cleanup.sh")
commands.getoutput ("rm *.png")
commands.getoutput ("rm *.dat.txt")
commands.getoutput ("rm *.out.txt")
# remove any subdirs
commands.getoutput ("find . -type d -name '???*' -exec rm -rf {} \;")
to
# ------- the page continues here
form = cgi.FieIdStorage()
infile = form["pdbfile"].value
namebase = form["pdbfile"].filename
namebasenum = string.rfind(namebase,'\\')
zo if namebasenum < -1:
namebasenum = 0
namelist = string.split(namebase[namebasenum+1:],'.')
z5 pdbname = namelist[0]+'.pdb'
dsspname = namelist[0]+'.dssp'
datname = namelist[0]+'.dat'
dattxtname = namelist[0]+'.dat.txt'
zipname = namelist[0]+'.zip'
ao inzipname ='submitted.zip'
consensusname = namelist[0]+'.cons'
epiname = namelist[0]+'.out.txt'
53
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
minsasa = form["minsasa"].value
mindist = form["mindist"].value
maxsize = form["maxsize"].value
consensus = form["consensus"].value
threshold = form["threshold"].value
number = form["number"].value
minlength = form["minlength"].value
plotmode = form["plot mode"].value
operatemode = form["operate mode"].value
to if (operatemode[0:7] __ "library"):
operatemode = "library"
if (form["operate mode"].value =_ "library_all"):
lib = liball
elif (form["operate mode"].value =_ "library_igg"):
lib = libigg
elif (form["operate mode"].value =_ "library_ige"):
lib = libige
if (operatemode =_ "library"):
zo libsize = len(lib)
if (string.upper(namelist[1]) =='PDB'):
inputtype = 'PDB'
if (string.upper(namelist[1]) =='ZIP'):
z5 inputtype ='ZIP'
# ------ write submitted file
if (inputtype =='PDB'):
3o f=open(pdbname, "w")
if (inputtype =='ZIP'):
f=open(inzipname, "w")
f.write(infile)
f.close()
54
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
# ------ If the submitted file is a zip-file, extract it and make a list of
the entries ------
if (inputtype =='ZIP'):
s pdbfiles = string.split(commands.getoutput(unzippath+" -I "+inzipname+" ~
awk'{ if (NR > 3 &&
NF == 4) print $4}"'))
numofpdbfiles = len(pdbfiles)
commands.getoutput(unzippath+" -j "+inzipname)
to # ----- make directories and move the zipfiles there
for i in pdbfiles:
dirname = i(0:-4]
commands.getoutput("rm -ri "+dirname)
15 os.mkdir(dirname)
os.rename(i,dirname+"l"+i)
else:
pdbfiles = [pdbname]
if (operatemode =_ "single"):
f=open(consensusname, "w")
f.write(consensus)
f.close()
so print'<CENTER>\n'
if form.has key("pagetitle"):
print '<H 1 >'+form["pagetitle"].value+'</H 1 >\n'
print time.ctime(time.time())+'<BR><BR>\n'
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
if (operatemode =_ "single"):
print'<BR><H2>You should print or save this page!</H2>\n'
print 'The results shown on this page are not stored anywhere else.\n\n'
if (operatemode =_ "library"):
if (inputtype =='ZIP'):
print'<H2><A HREF="collected.zip">Download</A> your results!</H2>\n'
if (inputtype =='PDB'):
so print'<H2><A HREF="'+zipname+"'>Download</A> your results!<lH2>\n'
print 'Downloading is strongly recommended! The results are shown on this page
and in-
cluded\n'
print'in this archive. They are not stored anywhere else.<BR><BR>\n'
print 'Filename given by you:<BR>\n'
print '<B>'+form["pdbfile"].filename+'</B>\n'
z o # ----------------- ru n the prog ram
#if (inputtype =='ZIP'):
if (1==1):
z5 for currentpdbname in pdbfiles:
# --------- the naming stuff - identical to that at the top of the file ---
namebase = currentpdbname
so namebasenum = string.rfind(namebase,'\\')
if namebasenum < -1:
namebasenum = 0
namelist = string.split(namebase[namebasenum+1:],'.')
56
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
if (inputtype =='PDB'):
nameroot = namelist[0]
if (inputtype =='ZIP'):
nameroot = namelist[0]
# nameroot = currentpdbname[0:-4]+"/"+namelist[0]
pdbname = nameroot+'.pdb'
dsspname = nameroot+'.dssp'
to datname = nameroot+'.dat'
dattxtname = nameroot+'.dat.txt'
zipname = nameroot+'.zip'
epiname = nameroot+'.out.txt'
# ----- here comes the treatment of the individual structures -----
if (inputtype =='ZIP'):
os.chdir(currentpdbname[0:-4])
if (operatemode =_ "single"):
# add extra newlines to the consensus file
commands.getoutput("echo \\\\n\\\\n » "+consensusname)
commands.getoutput(dssppath+" "+pdbname+" "+dsspname)
3o if (inputtype =='ZIP'):
commands.getoutput(epitopepath+" ../"+consensusname+" "+namelist[0]+"
"+mindist+"
"+minsasa+" "+maxsize+" "+number+" "+minlength+" > "+epiname)
else:
57
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
commands.getoutput(epitopepath+" "+consensusname+" "+namelist[0]+" "+mindist+"
"+minsasa+" "+maxsize+" "+number+" "+minlength+" > "+epiname)
commands.getoutput("mv "+datname+" "+dattxtname)
if (operatemode =_ "library"):
commands.getoutput(dssppath+" "+pdbname+" "+dsspname)
# for i in range(1,libsize+1):
to for i in lib:
if (inputtype =='ZIP'):
commands.getoutput(epitopepath+" ../"+string.zfill(str(i),3)+".epi
"+namelist[0]+"
"+mindist+" "+minsasa+" "+maxsize+" "+number+" "+minlength+" >
"+string.zfill(str(i),3)+".out.txt")
else:
commands.getoutput(epitopepath+" "+string.zfill(str(i),3)+".epi
"+namelist[0]+"
"+mindist+" "+minsasa+" "+maxsize+" "+number+" "+minlength+" >
"+string.zfill(str(i),3)+".out.txt")
commands.getoutput("mv "+datname+" "+string.zfill(str(i),3)+".dat.txt")
2o residues = int(commands.getoutput("grep -v'#'
"+string.zfill(str(lib[0]),3)+".dat.txt ~ we ~ awk
'{print $1}"'))
commands.getoutput("rm sum.dat.txt")
for i in range(1,residues+1):
grepstr = "~"+string.rjust(str(i),4)
a5 commands.getoutput("grep "'+grepstr+"' *.dat.txt ~ awk
'BEGIN{sum=0}{sum+=$5; res=$2;
pdbres=$3; AA=$4} END{print res, pdbres, AA,sum}' » sum.dat.txt")
commands.getoutput("rm "+datname)
# -------------- collect generated files
if (inputtype =='PDB'):
commands.getoutput("rm "+zipname)
commands.getoutput(zippath+" "+zipname+" *.out.txt *.dat.txt")
58
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
# ------------------- if in library mode, create and show the sum graph -----
if (operatemode =_ "library"):
timestamp = str(int(time.time()))
f=open("epitope.gnp", "w")
if (plotmode =_ "sequential"):
to f.write('set xlabel "Residue number (sequential)"\n')
else:
f.write('set xlabel "Residue number (PDB)"\n')
f.write('set ylabel "Epitopes"\n')
f.write('set title "'+currentpdbname[0:-4]+"'\n')
f.write('set size ratio 0.3 1, 0.5\n')
f.write('set term png small color\n')
f.write('set out "epi'+timestamp+'.png"\n')
if (plotmode =_ "sequential"):
f.write('plot "sum.dat.txt" using 1:4 title "Number of epitopes" with steps 1,
'+threshold+'
2o title "Threshold" with lines 3\n')
else:
f.write('plot "sum.dat.txt" using 2:4 title "Number of epitopes" with steps 1,
'+threshold+'
title "Threshold" with lines 3\n')
f.close()
commands.getoutput(gnuplotpath+" epitope.gnp")
print'<H1>Epitope frequency sums for each residue</H1><BR>\n'
s o if (form["operate mode"].value =_ "library_all"):
print'<H2>Library of+str(libsize)+' epitopes (IgG+IgE)</H2>'
elif (form["operate mode"].value =_ "library_igg"):
print '<H2>Library of '+str(libsize)+' epitopes (IgG)</H2>'
elif (form["operate mode"].value =_ "library_ige"):
59
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
print'<H2>Library of+str(libsize)+' epitopes (IgE)</H2>'
if (inputtype =_ 'PDB'):
print'<BR><BR><IMG SRC="epi'+timestamp+'.png"><BR><BR>\n'
print'<A HREF="sum.dat.txt">View the frequency sums table data</A><BR>\n'
print '<A HREF="'+zipname+"'>Download</A> a zip file with all results from the
individual
epitopes.<BR>\n'
print '</CENTER>\n'
to if (inputtype =='ZIP'):
print '<BR><BR><IMG SRC="'+currentpdbname[0:-
4]+'/epi'+timestamp+'.png"><BR><BR>\n'
print '<A HREF="'+currentpdbname[0:-4]+'lsum.dat.txt">View the frequency sums
table
data</A><BR>\n'
# --------- now make gnuplot graphs and data lists for individual epitopes --
# --- so far this goes only for the "single" operating mode
if (operatemode =_ "single"):
timestamp = str(int(time.time()))
z5 # Create gnuplot control file
f=open("epitope.gnp", "w")
if (plotmode =_ "sequential"):
f.write('set xlabel "Residue number (sequential)"\n')
3 o else:
f.write('set xlabel "Residue number (PDB)"\n')
f.write('set ylabel "Epitopes"\n')
f.write('set size ratio 0.3 1, 0.5\n')
f.write('set term png small color\n')
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
f.write('set out "epi'+timestamp+'.png"\n')
if (plotmode =_ "sequential"):
f.write('plot "'+dattxtname+"' using 1:4 title "Number of epitopes" with steps
1,
'+threshold+' title "Threshold" with lines 3\n')
else:
f.write('plot "'+dattxtname+"' using 2:4 title "Number of epitopes" with steps
1,
'+threshold+' title "Threshold" with lines 3\n')
f.closeQ
to commands.getoutput(gnuplotpath+" epitope.gnp")
if (inputtype =='ZIP'):
print '<BR><BR><IMG SRC="'+currentpdbname[0:-
4]+'/epi'+timestamp+'.png"><BR><BR>\n'
print '<A HREF="'+currentpdbname[0:-4]+'/'+dattxtname+"'>View the table da-
ta</A><BR>\n'
else:
print'<BR><BR><IMG SRC="epi'+timestamp+'.png"><BR><BR>\n'
print'<A HREF="'+dattxtname+"'>View the table data</A><BR>\n'
2o print'</CENTER>\n'
# ------------ print the table
print'<PRE>'
f=open(epiname,"r")
line = f.readline()
while line !_ "":
line = string.replace(line,'\n',")
print line
line = f.readlineQ
f.close()
61
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
print'</PRE><BR><BR><BR>'
if (inputtype =='ZIP'):
os.chdir("..")
to # ---------- for ZIP-mode (library only): count number of epitopes found
from each lib consensus
If (inputtype =='ZIP' and operatemode =_ "library"):
numofepitopes = []
f=open("epitopecount.txt", "w")
2o f.write(string.ljust("PDB file",20))
for i in lib:
f.write(string.rjust(str(i),6))
f.write('\n')
z5 forj in range(len(pdbfiles)):
currentpdbname = pdbfiles[j]
f.write(string.Ijust(currentpdbname[0:20],20))
for idx in range(len(lib)):
i = lib[idx]
3o filename = currentpdbname[0:-4.]+"/"+string.zfill(str(i),3)+".out.txt"
numofepitopes.append(0)
tmp = commands.getoutput("grep 'Total number of epitopes' "+filename+" ~ awlc
'{print
~6)~~~)
if (tmp I= ""):
62
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
numofepitopes[j*len(pdbfiles)+idx] = int(tmp)
numofepitopes[j*len(pdbfiles)+idx] - numofepitopes[j*len(pdbfiles)+idx]-
int(commands.getoutput("grep 'of which are subsets' "+filename+" ~ awk '{print
$8}"'))
else:
numofepitopes[j*len(pdbfiles)+idx] = 0
f.write(string.rjust(str(numofepitopes[j*len(pdbfiles)+idx]),6))
f.write('\n')
f.close()
to
# ---------- for ZIP-mode: Collect all dirs and files
if (inputtype =='ZIP'):
commands.getoutput("rm collected.zip")
for currentpdbname in pdbfiles:
commands.getoutput(zippath+" -r -a collected.zip "+currentpdbname[0:-4])
if (operatemode =_ "library"):
commands.getoutput(zippath+" -a collected.zip epitopecount.txt")
# ---- Last lines ----
print '</body>\n'
print '</html>\n'
# ---- remove lock file
os.remove ("epitope.lock")
# ------ remove temporary files ----------
#if (inputtype =='ZIP'):
# for currentpdbname in pdbfiles:
# commands.getoutput("rm -rf "+currentpdbname[0:-4])
63
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
commands.getoutput ("rm "+pdbname)
commands.getoutput ("rm "+dsspname)
commands.getoutput ("rm "+consensusname)
s commands.getoutput ("rm "+epiname)
Appendix C
to
THE HTML INPUT FORM (EPITOPES.HTML)
15 <!doctype html public "-//w3cl/dtd html 4.0 transitional//en">
<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1">
<title>Automatic epitope mapping</title>
20 </head>
<BODY BGCOLOR="#FFF9E6" text="#000000" link="#000040" vlink="#4.04040">
<center>
<TABLE>
25 <TR>
<TD><IMG SRC="epitope design.gif'></TD>
<TD> <H1>Epitope mapping tool </H1></TD>
</TR>
</TABLE>
</center>
<form ENCTYPE="multipart/form-data" action="./epitope5.cgi" method="POST">
64
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
<H2>Title<lH2>
Page title: <INPUT type=text name="pagetitle" size="40" maxlength="80"
value="Automatic Epitope Mapping">
<HR WIDTH=80%>
<H2>Parameters</H2>
<TABLE>
<TR>
<TD>File name (on your local machine)</TD>
to <TD><INPUT type=file name="pdbfile" size="40" maxlength="256"
value="*.pdb"></TD>
</TR>
<TR><TD COLSPAN=2>You may submit either a PDB file containing a single
structure
or a ZIP-archive containing a number of PDB files, each defining a single
structure.
The ZIP-archive must not contain subdirectories.
<TD></TR>
</TABLE>
<BR>
<INPUT TYPE=RADIO NAME="operate mode" VALUE="library_all" CHECKED>
Use epitope library (Full library).<BR>
<INPUT TYPE=RADIO NAME="operate mode" VALUE="library_igg">
Use epitope library (IgG library).<BR>
<INPUT TYPE=RADIO NAME="operate mode" VALUE="library_ige">
Use epitope library (IgE library).<BR>
z5 <INPUT TYPE=RADIO NAME="operate mode" VALUE="single">
Specify epitope consensus sequence here:<BR>
<TABLE>
<TR><TD>
3o Epitope consensus sequence<BR>
<TEXTAREA NAME="consensus" ROWS="12" COLS="21" WRAP="OFF">
</TEXTAREA></TD>
</TD><TD>
<TD>
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
Example of consensus sequence input:<BR>
<BR>
<TABLE BORDER="0" CELLSPACING=0>
<TR><TD>KR </TD><TD></TD><TD> (Lys og Arg allowed)</TD><TR>
<TR><TD>AILV-</TD><TD></TD><TD> (Ala, Ile, Leu, Val or missing residue al-
lowed)</TD><TR>
<TR><TD>* </TD><TD></TD><TD> (All residues allowed, but there must be a resi-
due)</TD><TR>
<TR><TD>? </TD><TD></TD><TD> (All or missing residue allowed)</TD><TR>
to <TR><TD>DE </TD><TD></TD><TD> (Asp or Glu allowed)</TD><TR>
</TABLE>
<BR>
*, ? or - in first or last position is allowed but obsolete.
(- in first position is ignored.)
</TD></TR>
</TABLE>
<BR><HR WIDTH=80%><BR>
<TABLE>
<TR>
<TD>Maximum distance between adjacent residues </TD><TD><INPUT type=text na-
me="mindist" size="5" maxlength="8" value = "10"></TD>
</TR>
2s <TR>
<TD>Minimum solvent accessible surface area for each residue</TD><TD><INPUT
type=text
name="minsasa" size="5" maxlength="8" value = "5"></TD>
</TR>
<TR>
<TD>Maximum epitope size (max distance between any two residues in epi
tope)</TD><TD><INPUT type=text name="maxsize" size="5" maxlength="8" value -
~~~5~~><n-D>
</TR>
<TR>
66
CA 02462651 2004-04-O1
WO 03/031981 PCT/DK02/00665
<TD>Maximum number of non-redundant epitopes to include (0 =
all)</TD><TD><INPUT
type=text name="number" size="5" maxlength="8" value = "0"></TD>
</TR>
<TD>Minimum epitope sequence length (in fractions of consensus
length)</TD><TD><INPUT
type=text name="minlength" size="5" maxlength="8" value = "0.80"></TD>
</TR>
</TABLE>
<BR><HR WIDTH=80%><BR>
<H2>Graph<lH2>
<INPUT TYPE=RADIO NAME="plot mode" VALUE="sequential" CHECKED>
Use sequential numbering of residues.<BR>
<INPUT TYPE=RADIO NAME="plot mode" VALUE="pdb">
Use PDB numbering of residues. (Will sometimes produce funny
re-
suits.)<BR>
Threshold value <INPUT type=text name="threshold" size="5"
max-
length="8" value = "2"><BR>
<BR><HR WIDTH=80%><BR>
<input type="submit" name="SUBMIT BUTTON" width=100 value="Find
epitopes"></form>
<form METHOD=GET ACTION="./epitope.html"><input type="submit"
name="SUBMIT BUTTON" width=100 value="Reset form">
</form>
<HR WIDTH=80%><BR>
<BR>
<CENTER>
Comments and bug reports to <A HREF="mailto:epf@novo.dk">epf</A>.
ao <BR><BR>
<IMG SRC="./epitope_nz.gif'>
</CENTER>
</body>
</html>
67
