Note: Descriptions are shown in the official language in which they were submitted.
CA 02399809 2002-08-02
WO 01/57065 l PCT/GBO1/00445
Combinatorial Protein Domains
The present invention concerns the cle nova synthesis of folded protein
domains by the
combinatorial rearrangement of sequence segments. The sequences of the
segments may
correspond directly to those of natural proteins, or be derived from those of
natural
proteins (for example by random or directed mutagenesis), or be derived by
design based
on the known structures of proteins. In particular, the invention makes use of
combinatorial rearrangements of sequence segments which are not single entire
structural
elements of a natural protein and which, in isolation, show no significant
folding.
The clo noun design of proteins is typically based on structure predictions of
predetermined amino acid sequences (Hecht 1994, Sauer 1996. Regan 1998).
Partial
randomisation is often introduced to allow for imperfection in the prediction
al<~orithms.
Resulting repertoires are screened or selected for stably folded structures.
This approach
1 ~ has been successful for the design of helical structures like four helix
bundles with stable
and compact structures exhibiting free energies of unfolding of about 4
kcal/mol
(Kamtekar et a1.1993). Wore problematic has been the design of (3-sheet
proteins. where
even the most recent attempts fell well short of natural (3-sheet proteins
regarding stability
(Quinn et a1.1994, Kortemme et a1.1998, Alba et a1.1999). Problems in the
design of (3-
sheet structures are related to their dependence on backbone hydrogen bonds
between
different secondary structure elements, which are less well understood than
the principles
of helix formation (Hecht 1994). Repertoires of random protein sequences have
also been
screened for the occurrence of folded proteins. About 1% of members in a
random library
of Glu, Leu, Arg rich proteins exhibited some helix formation and cooperative
unfolding
2~ but were unstable (Davidson & Sauer 1994).
Recently, new strategies to select stably folded proteins from repertoires of
phage
displayed proteins based on their resistance to proteolytic degradation have
been used to
improve the stability of natural proteins (Kristensen & Winter 1997, Sieber et
cz1.1998.
Finucane et al. 1999). Proteolvtic degradation is usually restricted to
unfolded proteins or
highly flexible regions of folded proteins. Folded proteins are mostly
resistant to
proteases, because the proteolvtic cleavage requires the polypeptide chain to
adapt to the
specific stereochemistry of the protease active site, and therefore to be
flexible, accessible
CA 02399809 2002-08-02
7
WO 01/57065 PCT/GBO1/00445
and capable of local unfolding (Hubbard et crl.199~1. Fontana et x1.1997).
These methods
have only been described for selection of proteins with point mutations: no
element of
combining sequences from different proteins is involved.
A theoretical approach to protein evolution via combinatorial rearrangement of
defined.
complete structural elements has been described (Bogarad & Deem 1999). The
authors
predict. using statistical algorithms. that rearrangement of a number of
structural elements
(such as helices, strands, loops, turns and others) will result in the
generation of novel
protein functions more rapidly than evolution by point mutation strategies
alone.
However, no allowance for the context dependence of structure is made, nor is
any
reference made to partial structural domains which possess no structural
identity in
isolation. Although some (rare) sequences will form structures in isolation.
others can
adopt a different structure in a different environment as evidenced by
structural
rearrangements following cleavage of some polypeptides by protease or on
ligand
binding. It is therefore not easy to define a structural element except in the
context of the
three dimensional structure of the protein in which it is embedded, and it is
this definition
that we have adopted here. Furthermore this paper does not show that it will
be possible
to undertake this process in vitro. or indicate exactly how to undertake such
experiments.
Summary of the Invention
We have developed a different strategy for the creation and selection of novel
protein
domains which are capable of forming stably folded structures, and thus of
identifying
novel protein structural and functional elements.
The inventors have realised that as the structure of a "structural" element is
dependent on
context, single structural elements taken from one protein and appended to
single
structural elements taken from a second protein will not necessarily retain
their original
structure. Accordingly the inventors have not sought to restrict the segments
to single
complete structural elements. Furthermore the use of parts of structural
elements can
provide new structures that are not simply due to juxtaposition of existing
structural
elements, and the use of segments comprising multiple structural elements (and
making
packing interactions with each other) would be expected to be more stable than
single
CA 02399809 2002-08-02
WO 01/57065 ~ PCT/GBO1/00445
structural elements, and more likely to comprise a significant "nu~~~~et" of
structure in the
chimaeric domain.
Thus. the present invention exploits protein evolution by juxtaposition of
sequence
~ segments. "Sequence segments", as referred to herein, are amino acid
sequences which
are not designed or selected to consist solely of single and complete protein
structural
elements; and are not designed or selected to consist of a complete protein
domain. The
present invention is thus not directed to the juxtaposition of discrete and
single elements
of structure found in naturally-occurring or synthetic proteins, but with the
juxtaposition
of blocks of more than one structural element or with the creation of novel
structural
elements by the juxtaposition of sequences which. in isolation or in their
parent
environments, do not possess a discrete and complete structure.
Therefore, a "sequence segment" is an amino acid sequence which, in its parent
1 ~ environment, does not comprise a complete protein domain and is not
encoded by one or
more complete natural exons. Moreover, a "sequence segment", in its parent
environment,
does not form one or more discrete structural elements, but is either part of
a structural
element or, advantageously, is longer than a structural element. The sequence
segment m
isolation shows no significant folding at the melting temperature of the
chimaeric protein;
in other words, it possesses no independent structure in isolated form.
The "parent environment" of the sequence segment is the protein or polypeptide
from
which that segment is taken, in its folded state. This may be a natural
protein, or an
artificial polypeptide or protein. Preferably, the sequence segment is taken
from an amino
acid sequence which is longer than the sequence segment itself.
According to the present invention in a first configuration, the combinatorial
rearrangement of protein sequence segments permits the selection of novel
folded protein
domains from combinatorial repertoires.
In a first aspect, therefore, the invention provides a chimaeric folded
protein domain when
derived from a repertoire of chimaeric proteins comprising two or more
sequence
segments derived from parent amino acid sequences that are not homologous.
WO 01/57065 ~ PCT/GBO1/00445
Preferably. the parent amino acid sequences are derived from protein domains.
The parent
amino acid sequences may be natural, semi-synthetic or swthetic in origin.
They may be
derived by expression from genes or assembled by chemical synthesis.
Advantageously, the amino acid sequence segments are derived From proteins. In
an
advantageous embodiment. the proteins are selected from the group consisting
of a
naturally occurrin;~ protein, an en~7ineered protein. a protein with a known
bindin~7
activity, a protein with a known binding activity for an organic compound, a
protein with
a known binding activity for a peptide or polypeptide, a protein with a known
binding
activity for a carbohydrate, a protein with a known binding activity for a
nucleic acid, a
known binding activity for a hapten. a protein with a known binding activity
for a steroid,
a protein with a known binding activity for an inorganic compound, and a
protein with an
enzymatic activity.
1~
As used herein. ''amino acid" includes the 20 naturally-occurring amino acids,
as well as
non-naturally occurring amino acids and modified amino acids, such as tagged
or labelled
amino acids. As used herein, the term "protein" refers to a polymer in which
the
monomers are amino acids and are joined together through peptide or disulphide
bonds.
Preferably. "protein" refers to a full-length naturally-occurring amino acid
chain or a
fragment thereof, such as a selected region of the polypeptide that is of
interest in a
binding interaction, or a synthetic amino acid chain, or a combination
thereof.
The sequence segments may be combined, in the chimaeric protein domain, by any
2~ appropriate means. Typically, the segments will be combined by recombinant
DNA
techniques and will thus be joined, in the recombinant protein, by peptide
bonds. In
alternative embodiments, the segments may be synthesised separately and
subsequently
joined. This may be achieved using covalent linkage, for instance peptide
bonds, ester
bonds or disulphide bonds, or non-covalent linkage. Advantageously, sequence
segments
according to the invention comprise one or more reaction groups for covalent
or non-
covalent linkage. For example, linkers capable of associating non-covalently,
such as
biotin/streptavidin, may be incorporated into the sequence segments to effect
non-
covalent linkage.
CA 02399809 2002-08-02
CA 02399809 2002-08-02
WO 01/57065 ~ PCT/GBO1/00445
The repertoire from which the chimaeric protein domain is derived may be or
substantially any size. Preferably, the repertoire comprises at least 10,000
individual
protein domains; advantageously it comprises at least 1,000.000 protein
domains: and
most preferably, at least 100.000,000 protein domains.
The sequence segments may be any appropriate number of amino acids in len<,th
such that
the combined length of the segments represents the length of a complete
domain. which
domains vary from as little as about 3~ residues to several hundred residues
in length.
In an advantageous aspect. the parent amino acid sequences are derived from
the open
reading frames of a genome or part thereof:
(a) wherein said reading frames are the natural reading frame of the genes: or
(b) wherein said reading frames are not the natural reading frame of the
genes.
l~
Sequences may thus be derived from ORFs present in a whole or substantially
whole
genome of an organism, or a part thereof, such as a group or family of genes,
whether
related by structure, function or evolution, or not related. The part of the
genome may also
consist of a single gene.
Sequences may moreover be derived from two or more genomes, from organisms of
related or unrelated species.
The protein domains according to the invention are capable of folding due to
the
combination of two or more polypeptide segments which, in isolation, do not
fold and do
not define a single structural element in the parent protein.
Advantageously, the protein domains according to the invention are selected
according to
their resistance to proteolysis. This provides a useful means to isolate
candidate domains
from libraries; a selection procedure can be configured such that only
proteolysis-resistant
domains are selected from the libraries. Preferably, the proteolysis is
carried out by
exposure to a protease, such as thermolysin.
CA 02399809 2002-08-02
WO 01/57065 6 PCT/GBOi/00445
In a preferred embodiment. the protein domains according to the invention may
be
selected according to their activity. This may for example be a binding
activity. for
example in the case of immunoglobulin-type domains, or an enzymatic activity
in the case
of enzyme domains. Alternatively the protein domain may have the capacity to
bind
antibodies directed against the parent protein. Moreover, a screen for
activity may be
performed in addition to a selection on the basis of folding as determined by
protease
resistance. Such an approach is particularly advantay~eous where an initial
selection on the
basis of activity would be difficult or impossible to perform.
Moreover the invention concerns the juxtaposition of sequence fra<~ments
derived from
non-homologous domains which share a similar polypeptide fold for at least
part of the
structure. We have observed that, in selections of protein domains according
to the
invention, sequence segments derived from parental protein domains havin~~
similar folds
for at least part of their structures are juxtaposed in some of the novel
chimaeric proteins.
1 ~ Accordingly. the present invention provides a chimaeric protein according
to the first
aspect of the invention, wherein the sequence segments originate from parent
domains
with similar polypeptide folds in at least part of the structure.
It has further been observed that, in selections of protein domains according
to the
invention, sequence segments derived from parental protein domains having
entirely
different folds for at least part of their structures are juxtaposed in other
novel structures.
Accordingly, the present invention provides a chimaeric protein domain
comprising two
or more sequence segments derived from parent amino acid sequences, wherein
the
sequence segments originate from parent domains with different polypeptide
folds in at
least part of the structure.
Moreover, in selections of protein domains according to the first
configuration of the
invention, sequence segments derived from the same protein domain may be
observed to
be juxtaposed to form novel structures. In some cases said sequence segments
may
comprise regions in common leading to a duplication of sequence in the
chimaeric
protein. However the common region does not consist of solely of one or more
complete
protein structural elements. Therefore it appears that duplication of amino
acid segments
or parts thereof, without regard to the presence of solely one or more
complete structural
CA 02399809 2002-08-02
WO 01/57065 7 PCT/GBO1/00445
elements, can lead to the formation of stably folded structures. Such
dur'::cations
comprise a second configuration of the invention.
As used herein, "regions in common'' or "common regions'' refers to regions
wh:;.h share
sequence similarity or are of a similar fold. In this context, sequence
similarity preferably
refers to stretches of identical sequence of at least 10 amino acid residu:s:
more
preferably of at least 20 amino acid residues.
According to the second configuration of the invention, the combination of
segments
from homologous proteins. leading to equivalent regions from these homologous
proteins
being brought together in the same chimaeric protein, would also be expected
to lead to
the creation of stably folded structures. Regions, which are equivalent in
homolo~Tous
proteins, are identified by an alignment of their amino acid sequences. Indeed
i~ is even
possible to combine segments from non-homologous proteins which share a common
fold
1 ~ (vide supra), to create stably folded chimaeric proteins from segments
comprising a
common region of the common fold in the parent proteins.
Said stably folded structures based on duplication of amino acid segments hale
been
created as a product of the random shuffling of amino acid segments and were
selected
through proteolytic selection because of their stability. Duplication or
indeed
multimerisation performed in other non random ways have been previously
reported,
including for example by Hardies et al.1979 and Fire & Xu 199. The inventors
envisage
that said methods for duplication and multimerisation may also be used for the
duplication or multimerisation of amino acid segments to create novel and
stabls- folded
domains under the second configuration of the invention. Such stable domains
may be
selected and screened for in ways identical or similar to those in case of
chimaeric
domains derived from combinatorial shuffling.
Protein domains according to both configurations of the present invention may
be created
and selected by any suitable means. Preferred is combinatorial rearrangement
of nucleic
acid segments, for example in phage display libraries. Thus, the invention
provides a
chimaeric protein domain according to any foregoing aspect of the invention.
fused to the
CA 02399809 2002-08-02
WO 01/57065 g PCT/GBO1/00445
coat protein of a filamentous bacteriophage. said bacteriophage encapsidating
a nucleic
acid encoding the protein domain.
Moreover, both configurations of the invention provide a nucleic acid
encodin~~ a protein
domain according to the invention as defined above.
In a further aspect of both configurations of the invention. the amino acid
sequences ot~
any chimaeric proteins may contain sequences designed to display epitopes for
the
vaccination against the parent protein of said amino acid sequences. For
example, a
chosen polypeptide segment from the coat protein of a virus. against which a
vaccine is to
be made, may be incorporated as a constitutive partner in a combinatorial
libran~ of amino
acid sequences generated through the shuffling with one or more segments from
another
genetic source. Resulting chimaeric proteins will then comprise the segment of
the viral
coat protein in a variety of structural environments. By screening or
selection, for
1 ~ example using antibodies from antisera raised against the virus. it is
possible to identify
those folded chimaeric proteins for which the viral sequence is displayed in a
similar three
dimensional configuration to the viral protein. Such stably folded proteins
among these
chimaeric constructs can be used for vaccination and elicit an immune response
against
the chimaeric protein which includes the viral amino acid segment. Vaccination
with such
a protein results in immunisation against the virus. One advantage compared to
vaccination with the viral coat protein is that it is thereby possible to
focus the immune
response against one defined epitope of the virus, such as a neutralisation
epitope.
It is also possible to vaccinate against defined epitopes of human proteins by
the same
2~ strategy by combining a segment from a human protein with that from another
source.
The segment of non-human source should provide T-cell epitopes that will lead
to an
immune response against the human epitope. By way of example, it is possible
to raise a
blocking (IgG) antibody response against the portion of IgE that binds to the
mast cell
receptor. Such response is valuable, for example, in blocking asthma. This is
achieved by
construction of a chimaeric protein as follows. Firstly, segments from IgE are
incorporated into chimaeric proteins by combination with a repertoire of non-
human
segments: secondly the proteins are screened or selected for binding to the
mast cell
receptor or to antibodies known to bind IgE at the critical site; thirdly the
chimaeric
CA 02399809 2002-08-02
WO 01/57065 9 PCT/GB01/00445
proteins with binding activities are used for immunisation. The Iv'E
se'_>ments may be
derived by random fragmentation of the IgE gene. or by using a se~lment
already known to
interact with the receptor. For immunisation it may be necessary to build in
more potent
T-cell epitopes into the non-human part. which can be achieved by making
mutations in
the non-human segment.
Preferably, therefore, the chimeric protein according to the invention
comprises an
epitope of a parent amino acid sequence. Advantageously. the epitope is a
structural
epttope.
Epitopes -comprised in the chimeric proteins according to the invention. in a
preferred
embodiment, cross-react with antibodies raised against a parent amino acid
sequence, or,
advantageously, the folded parent protein.
1 ~ In a further aspect of both configurations of the invention the segments
may be derived
entirely from human proteins. It is expected that these proteins will be less
immunogenic
in humans than foreign proteins as the sequences of the protein will be almost
entirely
human. Although such novel human proteins will be expected to differ in three
dimensional structure from existing human proteins (and therefore to comprise
novel B-
cell epitopes), they will comprise T-cell epitopes derived from other human
proteins (with
the exception of the sequence flanking the join between segments). Such
proteins, that are
not immunogenic, or only weakly so, would be very suitable for therapeutic
purposes or
to avoid sensitisation in humans (for example enzymes in washing powders).
It is unlikely that in every respect that the chimaeric protein will mimic the
three
dimensional surface of the original protein in the region of target segment.
This may be
desirable in that it may allow the protein to adopt a conformation that has
altered bindings
activities. For example, such proteins may be valuable as improved enzyme
inhibitors.
Moreover the invention in either configuration provides for the creation of
small domains
that mimic part of the surface of a larger protein. One advantage of small
domains is that
it may more readily permit the three dimensional structure to be solved by X-
ray
CA 02399809 2002-08-02
WO 01/57065 10 PCT/GBO1/00445
crystallography or NMR, and also at higher resolution. In turn this may
facilitate the
design of non-protein drug's based on the structure.
Further the invention in either configuration allows for the fusion of
individual sequence
segments juxtaposed in the chimaeric protein to additional, stable folded and
complete
protein domains. The function of additional domains may be to provide a means
for
selecting the chimaeric protein domains (see methods below'). Thev may also
serve to
complement the chimaeric protein domain to perform a specific function. for
example
binding, immunogenicity or catalysis.
In the second configuration of the invention, the presence of at least two
regions of the
same sequence or similar (homologous] sequence in the chimaeric protein may
permit the
development of chimaeric proteins that bind to ligand at each of the two
sites. This may
be an advantage by giving improved "avidity" of binding where both heads can
en~~age
1 ~ dimeric ligand (or other multimers), and also in providing two binding
sites with different
affinities, covering a larger dynamic range in binding to a ligand.
A further aspect of the first configuration of the invention relates to a
method for selecting
a protein domain according to the invention as defined above. Accordingly. the
invention
provides a method for preparing a protein domain according to the first aspect
of the
invention, comprising the steps of:
(a) providing a first library of nucleic acids, said library comprising coding
sequences
encoding sequence segments derived from one or more amino acid sequences, said
coding
sequences not being selected or designed such as to solely encode a single and
complete
protein structural element or to encode a complete protein domain;
(b) providing a second library of nucleic acids, said library comprising
coding
sequences encoding sequence segments derived from one or more amino acid
sequences,
said partner coding sequence not being selected or designed such as to solely
encode a
single and complete protein structural element or to encode a complete protein
domain:
(c) combining the coding sequences to form a combinatorial library of nucleic
acids,
said nucleic acids comprising contiguous coding sequences encoding sequence
fragments
derived from the first and second libraries;
CA 02399809 2002-08-02
WO 01/57065 1 1 PCT/GBO1/00445
(d) transcribing andior translating the contiguous coding sequences to prodece
thz
encoded protein domains;
(e) selectin<~ the chimaeric protein domains which are able to adopt a folded
structure
or to fulfil a specific function.
Libraries according to the invention may be constructed such that sequences
homologous
to the partner coding sequence are excluded. For example, the libraries may be
based on
an artificial combination of solved structures, which means that the presence
or absence
of sequences homologous to the partner coding sequence can be controlled.
However. if
genomic libraries are used, it is possible that sequences homologous to the
partner
sequence may be present. In a preferred aspect, therefore, the method
according to the
invention further includes the steps of:
(f) ranalysing the sequence of the selected chimaeric protein domains to
idertih~ the
I ~ origins of the sequence segments; and
(g) comparing the sequences of each of the parent amino acid sequences to
identify
whether the sequences of the parent amino acid sequences are non-homologous.
Similarly, it is possible to construct libraries comprising sequence segments
derived from
defined protein folds. However, if it is required to determine whether the
isolated protein
domain according to the invention is composed of sequence segments derived
from
parental domains having the same fold, the method according to the invention
advantageously includes the step of:
(h) comparing the structures of each of the parent domains to identify whether
they
have same polypeptide folds in whole or in part.
In a further preferred aspect, the first configuration of the invention
relates to the
combination of a library of sequence segments with a unique partner coding
sequence
derived from a protein. The partner sequence is in this aspect provided as a
unique
sequence. Accordingly, steps (b) and (c) in the method according to the first
configuration
of the invention as set forth above may be modified such that:
CA 02399809 2002-08-02
WO 01/57065 12 PCT/GBO1/00445
(b j providing a partner codin'a sequence encoding a sequence se';ment derived
from
one protein. said partner coding sequence not bein~7 selected or designed such
as to solely
encode a sin<~le and complete protein structural element or to encode a
complete protein
domain;
(c) combining the library and partner coding sequences to form a combinatorial
library of nucleic acids, said nucleic acids comprisiny~ contiguous coding
sequences
encoding sequence fragments derived from the first library and the partner
codin~~
sequence.
A further aspect of the second configuration of the invention relates to a
method for
selecting a protein domain, in which the individual sequence segments comprise
common
sequences. Accordin~Tly, the invention provides a method for preparin;~ a
protein domain
according to the first aspect of the invention, comprising the steps of:
I ~ (a) providing a first library of nucleic acids. said library comprising
coding sequences
encoding sequence segments derived from one or more amino acid sequences, said
coding
sequences not being selected or designed to encode a complete protein domain;
(b) providing a second library of nucleic acids. said library comprising
coding
sequences encoding sequence segments derived from one or more amino acid
sequences,
said partner coding sequence not being selected or designed such as to solely
encode a
single and complete protein structural element or to encode a complete protein
domain;
(c) combining the coding sequences to form a combinatorial library of nucleic
acids,
said nucleic acids comprising contiguous coding sequences encoding sequence
fragments
derived from the first and second libraries;
2~ (d) transcribing and/or translating the contiguous coding sequences to
produce the
encoded protein domains;
(e) selecting the chimaeric protein domains, which are able to adopt a folded
structure
or to fulfil a specific function;
and optionally:
(t~ analysing the sequence of the selected chimaeric protein domains to
identify the
origins of the sequence segments; and
(g) comparing the sequences to identify whether they comprise common
sequences.
CA 02399809 2002-08-02
WO 01/57065 1 ~ PCT/GBO1/00445
Similarly. in a further aspect of the second configuration of the invention
relates to a
method for selecting a protein domain, in which the individual sequence
se<~ments
comprise common regions from parent proteins with a common fold. However. if
it is
required to determine whether the isolated protein domain according to the
invention is
composed of sequence segments derived from parental domains havin~.~ the same
fold. tire
method according to the invention advantay~eously does not require step (g)
above. but
includes in its place the steps of:
(g) comparing the structures of the parent amino acid sequences to identify
whether
the parent proteins have a common fold: and
(h) identifying whether the segments comprise a common region of the common
told.
In a further preferred aspect, the second configuration of invention also
relates to the
1 ~ combination of a library of sequence segments with a unique partner coding
sequence
derived from a protein. The partner sequence is in this aspect provided as a
unique
sequence. Accordingly, steps (b) and (c) in the method according to the second
confieuration of the invention as set forth above may be modified such that:
(b) providing a partner coding sequence encoding a sequence segment derived
from
one protein, said partner coding sequence not being selected or designed such
as to solely
encode a single and complete protein structural element or to encode a
complete protein
domain;
(c) combining the library and partner coding sequences to form a combinatorial
library of nucleic acids, said nucleic acids comprising contiguous coding
sequences
encoding sequence fragments derived from the first library and the partner
coding
sequence.
Preferably, in the methods according to the both configurations of the
invention the
domains which are able to adopt a folded structure are selected by one or
several methods
selected from the group consisting of in vivo proteolysis, in vitro
proteolysis, binding
ability. functional activity and expression.
CA 02399809 2002-08-02
WO 01/57065 1 ~ PCT/GBO1/00445
In a further aspect. an amino acid sequence of any chimaeric proteins produced
throu<'h
combinatorial shufflin~~ according to both configurations of the invention may
be mutated
or altered after the ori~~inal juxtaposition of the parent amino acid
sequences. Such
changes may be introduced by any of the following methods:
(a) designing and introducing specific or random mutations at predefined
positions
within the gene of the chimaeric protein:
(b) deleting nucleotides within the gene of the chimaeric protein so as to
delete amino
acid residues;
(c) inserting nucleotides within the gene of the chimaeric protein so as to
insert ammo
acid residues
(d) appending nucleotides to the gene of the chimaeric protein so as to append
amino
acid residues;
(e) randomly introducing mutations in all or part of the gene encoding the
chimaeric
protein through recombinant DNA technology;
(f) randomly introducing mutations in the gene of the chimaeric protein
through
propagation in mutating cells;
(g) introducing derivatives of natural amino acid during chemical synthesis;
(h) chemically derivatising amino acid groups after synthesis;
(i) multimerising the chimaeric proteins through concatenation of two or more
copies
of the gene in a single open reading frame;
(j) multimerising the chimaeric proteins through covalent linkage of two or
more
copies of the chimaeric protein domain after translation;
(k) multimerising the chimaeric proteins through fusion to a multimeric
partner.
Any said changes may improve the stability or the function of the chimaeric
protein. For
example, said changes may be aimed to meet predicted structural requirements
within the
combined segments advantageous for the formation of specific polypeptide folds
or to
introduce specific amino acid sequences to fulfil a desired function. An
example of such
improvements is given in Example 14 in the Experimental section.
The invention moreover encompasses further optimisation of the regions of N-
and C-
termini of recombined amino acid segments. Both their joining and end regions
as part of
WO 01/57065 1 ~ PCT/GBO1/00445
a chimaeric protein are conceivably not optimised as far as stability and,'or
functic', of the
chimaeric protein are concerned. Natural proteins, which may have been created
through
a recombinatorial event. are subsequently optimised through (pointy mutationa'
events
and Darwinian selection. This process may be mimicked in vitro for chimaeric
protein as
defined herein, for example using the above listed methods (including
mutation. deletion
and/or addition of amino acid residues).
Chimaeric proteins containing such improvements may be identified by one .~r
more
methods used for the selection and screening of the original combinatorial
libraw. It may
further be advantageous to produce any selected chimaeric protein domai:~a in
a
multimerised form, for example to increase stability throwgh interdomain
intera;:tion or
improve binding to a ligand through avidity effects.
Brief Description of the Figures
Figure 1. Proteolysis of selected phages and chimaeric proteins. (a) ELISA for
barstar
binding of phages 1c2 (squares), lbl l (circles), 1g6 (diamonds) and espi2
itriangles)
before and after trypsin/thermolysin treatment at different temperatures. (b)
SDS-PAGE
of proteins His-1c2, His-lbll and His-lg6 before and after treatment with
trypsin,
thermolysin and chymotrypsin at 25°C.
Figure 2. Circular dichroism and thermodenaturation of chimaeric proteins. (a~
Circular
dichroism spectra of His-lc2 (upper trace) and His-2f3 (lower trace) at 20''C.
(b)
Ellipticity of His-let (at 20~ nm; upper trace) and His-2f3 (at 223 nm; lower
trace) at
2~ different temperatures.
Figure 3. Nuclear magnetic resonance analysis of chimaeric proteins. 1 D- H-
NMR
spectra of His-2f~ recorded (a) at 25°C in HBO and (b) after incubation
for 2s hours at
?~°C in D~O. 1D-'H-NIVTR spectra of His-lc2 recorded at 30°C (c)
in HBO and ~d) after
incubation for 24 hours at 25°C in DSO. 2D-'H-NOESY spectrum of His-lc2
re:,orded at
30°C (e) in HBO.
CA 02399809 2002-08-02
CA 02399809 2002-08-02
WO 01/57065 16 PCT/GBOI/00445
Figure -t. Biotin-CspA ELISA. A rabbit anti-CspA antiserum was incubated with
varied
amounts of soluble His-CspA. His-1c2. His-2f~, His-lbll or Ivsozvme gas a
ne~Tative
control) before binding to biotinylated CspA immobilised on Streptavidin
coated ELISA
well. Bound rabbit antisera were detected with a HRP-conjugated goat anti-
rabbit I~1G
antiserum.
Detailed Description of the Invention
The present invention relates to chimaeric, folded protein domains. In the
contew of the
present invention. "folded" means that the protein domains concerned are
capable of
adoptin~a, or have adopted, a stable tertiary structure. Stability in this
contest may be
defined as the conformational stability of the protein. which is the
difference in free
ener~v between the folded and unfolded conformations under physiological
conditions:
the higher this value, the greater the energy required to unfold the protein.
and thus the
1 ~ greater the stability of the folded structure. A quantitative measure of
this conformational
stability of proteins. the Gibbs free energy of folding, can be determined
from reversible
thermodynamics. Proteins undergo order-disorder transitions. which are
detectable in
differential scanning calorimetry (DSC) profiles of specific heat vs.
temperature.
Preferably, the free energy of folding possessed by a protein domain according
to the
invention is 1.6 kcal/mol or higher; advantageously, it is 3 kcal/mol or
higher; and most
preferably it is ~ kcal/mol or higher.
Folded proteins which form stable structures are known to be resistant to
proteolysis.
2~ Thus, the invention provides for the selection of folded protein domains in
accordance
with the present invention using protease enzymes, which cleave and preferably
eliminate
unstable or unfolded domains. ''Folded" may therefore be defined in terms of
resistance to
proteolysis under assay conditions. Exemplary conditions are set forth in the
examples
below.
Sequence segments according to the invention are segments of natural protein
sequence.
which occurs in naturally-occurring proteins, or artificial segments of
sequence modelled
on the sequence or structure of naturally-occurring proteins. The sequence
segments may
CA 02399809 2002-08-02
WO 01/57065 17 PCT/GBOI/00445
be between 10 and 100 amino acids in length. or loner: preferably bemveen I ~
and ~0
amino acids in length: and advantageously between ?0 and 4~ amino acids in
length: or.
where nucleic acids are concerned, the necessay length to encode such amino
acid
sequences.
Sequence segments according to the invention are derived from parental protein
domains
which are not homologous.
The term "parent amino acid sequences" (or "parental amino acid sequences")
refers to
any amino acid sequences encoded by open reading frames within DN.~ sequences.
which
form the source of the cloned DNA segments as part of the combinatorial
libraries as
outlined in the claims. Said reading frames may be part of the original
reading frame of
genes, of shifted reading frames or of the reverse strand of genes. They may
also form part
of intragenic regions, which are not known to encode a protein. Originating
<,genes may be
1 ~ natural or synthetic.
As outlined in the introduction, the term homology between two or more
proteins or
proteins domains can refer to a similarity or identity of both their amino
acid sequences
and their structural fold. For the present purposes, the term homology shall
solely refer to
the degree of identity between two parent amino acid sequences.
Homologous amino acid sequences have 3~°io or greater identity (e.g..
at least -10°,-'~
identity. 50% identity, 60% identity, 70% identity, or at least 80% identity.
such as at
least 90% identity, or even at least 9~% identity, for instance at least 97%
identiri ).
2~ Homologous nucleic acid sequences are nucleic acid sequences which encode
homologous polypeptides, as defined. Actual nucleic acid sequence
homology/identit~~
values can be determined using the ''Align" program of Myers & Miller 1988,
("Optimal
Alignments in Linear Space") and available at NCBI. Alternatively or
additionally, the
term "homology", for instance, with respect to a nucleotide or amino acid
sequence, can
indicate a quantitative measure of homology between two sequences. The percent
sequence homology can be calculated as (N,ef - Ndf)* 100/N,ef wherein Ndj is
the total
number of non-identical residues in the two sequences when aligned and wherein
N,e; is
the number of residues in one of the sequences. Hence, the DNA sequence AGTC
AGTC
CA 02399809 2002-08-02
WO 01/57065 I 8 PCT/GBO1/00445
will have a sequence similarity of 7~% with the sequence AATC.~~ATC (N,.~= 8:
N~", _
2). Alternatively or additionally, "homolo~~y" with respect to sequences can
refer to the
number of positions with identical nucleotides or amino acids divided by the
number of
nucleotides or amino acids in the shorter of the two sequences wherein
alignment of the
two sequences can be determined in accordance with the Wilbur and Lipman
algorithm
(V'ilbur & Lipman 1983), for instance, usin~,~ a window size of ?0
nucleotides, a word
length of 4 nucleotides, and a yap penalty of 4, and computer-assisted
analysis and
interpretation of the sequence data including alignment can be conveniently
performed
using commercially available programs (e.g., IntelligeneticsT" Suite.
Intelligenetics Inc.
CAI. When RNA sequences are said to be similar, or have a degree of sequence
identity
or homology with DNA sequences. thymidine (T) in the DNA sequence is
considered
equal to Uracil (U) in the RN A sequence.
RNA sequences within the scope of the invention can be derived from DNA
sequences,
1 ~ by thymidine (T) in the DNA sequence being considered equal to Uracil (U)
in RNA
sequences.
Additionally or alternatively. amino acid sequence similarity or identity or
homology can
be determined using the BlastP program (Altschul et a1.1997) and available at
NCBI.
?0 The following references (each incorporated herein by reference) provide
algorithms for
comparing the relative identity or homology of amino acid residues of two
proteins, and
additionally or alternatively with respect to the foregoing, the teachings in
these
references can be used for determining percent homology or identity: Needleman
&
Wunsch (1970); Smith & Waterman (1981); Smith et a1.(1983); Feng & Dolittle
(1987);
2~ Higgins & Sharp ( 1989); Thompson et al. ( 1994); and Devereux et al. (
1984).
The invention contemplates the recombination of sequence segments which are
derived
from parental proteins with similar folds. In this contest, "similar" is not
eqmvalent to
''homologous''. Indeed. similar folds have been shown to arise independently
during
30 evolution. Such folds are similar but not homologous.
A ''protein structural element'' is an amino acid sequence which may be
recognised as a
structural element of a protein domain. Preferably, the structural element is
selected from
CA 02399809 2002-08-02
WO 01/57065 I9 PCT/GBO1/00445
the group consisting of an a-helm, a ~3-strand. a (3-barrel. a parallel or
antiparalle; ~3-sheet.
other helical structures (such as the 3 "~ helix and the pi helix), and
sequences representing
tight turns or loops. Advantageously, the structural element is an cc-helix or
a ~3-strand.
sheet or barrel.
In a preferred embodiment. the folded protein domains according to the present
invention
are constructed from sequence segments which do not comprise only a single
structural
element; rather, they comprise less than a single structural element, or more
than a single
structural elements or parts thereof.
In accordance with the present invention, the sequence segments used are not
desi~~ned or
selected to comprise only such single elements; in other words, they may
comprise more
than a single structural element, or less than a single structural element.
This may be
achieved through the use of substantially random sequence segments in
constructing a
1 ~ library according to the invention. For example, sonicated genomic or cDNA
or segments
produced by random PCR of DNA may be used. Advantageously, the DNA fragments
are
between 100 and 500 nucleotides in length.
The sequence segments used in accordance with the present invention are unable
to fold
significantly in isolation; that is, they do not contain sufficient structural
information to
form a folded protein domain unless they are combined with another sequence
segment in
accordance with the present invention. The inability to fold significantly may
be measured
by susceptibility to protease digestion, for example under the conditions
given in the
examples below, or by measurement of the free energy of folding .
Proteolysis may be carried out using protease enzymes. Suitable proteases
include trypsin
(cleaves at Lys, Arg), chymotrypsin (Phe, Trp, Tyr, Leu), thermolysin (small
aliphatic
residues), subtilisin (small aliphatic residues), Glu-C (Glu), Factor Xa
(Ile/Leu-Glu-Gly-
Arg), Arg-C (Arg) and thrombin. Advantageously, since the combination of
random
polypeptide sequence segments cannot be guaranteed to generate a precise
cleavage site
for a particular protease. a broad-spectrum protease capable of cleaving at a
variety of
sites is used. Trypsin, chymotrypsin and thermolysin are broad-spectrum
proteases useful
in the present invention.
CA 02399809 2002-08-02
WO 01/57065 ~'0 PCT/GBO1/00445
The ability of a protein domain to fold is also associated with its tunction.
Accordin'~lv.
the invention provides for the selection of folded protein domains by
functional assays.
In the case of immunoglobulins or other polypeptides capable of binding. such
assays may
be performed for binding activity accordin~~ to established protocols:
however. where
bindings is only transitory, the selection may be performed on the basis of
function alone.
Suitable methodology is set forth, for example. in International patent
applications
PCT/GB00/00030 and PCT,~GB98/01889. Such techniques are useful for the
selection of
novel or improved enzymes produced by combinatorial rearrangement accordiny~
to the
present mventton.
The invention also provides for screening for activity after selection
according to protease
resistance. This allows protein domains which hare been selected accordin~~ to
their
ability to fold to be screened for any desired activity. Since the repertoire
sizes are more
limited. as a result of the selection by proteolysis, the screenings step can
be conducted
more easily (for example. in a multiwell plate).
The libraries of the present invention may be created by any suitable means in
any form.
As used herein, the term "library" refers to a mixture of heterogeneous
polypeptides or
nucleic acids. The library is composed of members, each of which has a unique
polypeptide or nucleic acid sequence. To this extent, library is synonymous
with
repertoire. Sequence differences between library members are responsible for
the
diversity present in the library. The library may take the form of a simple
mixture of
2~ polypeptides or nucleic acids, or may be in the form organisms or cells,
for example
bacteria, viruses, animal or plant cells and the like, transformed with a
library of nucleic
acids. Typically, each individual organism or cell contains only one member of
the
library. In certain applications, each individual organism or cell may contain
nvo or more
members of the library. Advantageously. the nucleic acids are incorporated
into
expression vectors. in order to allow expression of the polypeptides encoded
by the
nucleic acids. In a preferred aspect, therefore, a library may take the form
of a population
of host organisms, each organism containing one or more copies of an
expression vector
containing a single member of the library in nucleic acid form which can be
expressed to
CA 02399809 2002-08-02
WO 01/57065 21 PCT/GBOI/00445
produce its corresponding polvpeptide member. Thus, the population of host
or'Tamsms
has the potential to encode a large repertoire of genetically diverse
polypeptide variants.
A number of vector systems useful for library production and selection are
known in the
art. For example, bacteriophage lambda expression systems may be screened
directly as
bacteriophage plaques or as colonies of lysogens, both as previously described
(Huse et
crl.(1989); Caton & Koprowski (1990); Mullinax et a1.(1990); Persson et
a1.(1991) and are
of use in the invention. Whilst such expression systems can be used to
screening up to 10~
different members of a library, they are not really suited to screening of
larger numbers
(greater than 106 members). Other screening systems rely, for example, on
direct chemical
synthesis of library members. One early method involves the synthesis of
peptides on a se.t
of pins or rods. such as described in W084/03~6-1. A similar method involving
peptide
synthesis on beads, which forms a peptide library in which each bead is an
individual
librar~~ member, is described in U.S. Patent No. 4.631,211 and a related
method is
1 ~ described in W09?/00091. A significant improvement of the bead-based
methods
involves tagging each bead with a unique identifier tag, such as an
oligonucleotide. so as
to facilitate identification of the amino acid sequence of each library
member. These
improved bead-based methods are described in W093/06121.
Another chemical synthesis method involves the synthesis of arrays of peptides
(or
peptidomimetics) on a surface in a manner that places each distinct library
member (e.g..
unique peptide sequence) at a discrete, predefined location in the array, or
the spotting of
pre-formed polypeptides on such an array. The identity of each library member
is
determined by its spatial location in the array. The locations in the array
where binding
?S interactions between a predetermined molecule (e.g., a receptor) and
reactive library
members occur is determined, thereby identifying the sequences of the reactive
library
members on the basis of spatial location. These methods are described in U.S.
Patent No.
~,143,8~4; W090/15070 and W092/10092; Fodor et a1.(1991); and Dower & Fodor
(1991).
Of particular use in the construction of libraries of the invention are
selection display
systems. which enable a nucleic acid to be linked to the polypeptide it
expresses. As used
CA 02399809 2002-08-02
WO 01/57065 ~? PCT/GBO1/00445
herein, a selection display system is a system that permits the selection. by
suitable
display means. of the individual members of the library.
Any selection display system may be used in conjunction with a library
according to the
invention. Selection protocols for isolating desired members of lar;~e
libraries are know w
in the art. as typified by phage display techniques. Such systems, in which
diverse peptide
sequences are displayed on the surface of filamentous bacteriopha~Te (Scott &
Smith
(1990), have proven useful for creating libraries of antibody fragments (and
the nucleotide
sequences that encoding them) for the in vitro selection and amplification of
specific
antibody fragments that bind a target antigen. The nucleotide sequences
encoding the V,i
and V~ regions are linked to gene fragments which encode leader signals that
direct them
to the periplasmic space of E. coli and as a result the resultant antibody
fragments are
displayed on the surface of the bacteriophage, typically as fusions to
bacteriophage coat
proteins (e.g., pIII or pVIII). Alternatively, antibody fragments are
displayed externally on
1 ~ lambda phage capsids (phagebodies). An advantage of phage-based display
systems is
that, because they are biological systems, selected library members can be
amplified
simply by growing the phage containing the selected library member in
bacterial cells.
Furthermore, since the nucleotide sequence that encode the polypeptide library
member is
contained on a phage or phagemid vector, sequencing, expression and subsequent
genetic
manipulation is relatively straightforward.
Methods for the construction of bacteriophage antibody display libraries and
lambda
phage expression libraries are well known in the art (McCafferty et a1.(1990);
Kang et
al. ( 1991 ); Clackson et al. ( 1991 ); Lowtnan et al. ( 1991 ); Burton et al.
( 1991 ); Hoogenboom
et al. ( 1991 ); Chang et al. ( 1991 ); Breitling et al. ( 1991 ); Marks et
al. ( 1991 ); Barbas et
al. ( 1992); Hawkins & Winter ( 1992); Marks et al. ( 1992); Lerner et al. (
1992),
incorporated herein by reference).
Other systems for Generating libraries of polypeptides or nucleotides involve
the use of
cell-free enzymatic machinery for the in vitro synthesis of the library
members. For
example, in vitro translation can be used to synthesise polypeptides as a
method for
generating large libraries. These methods which generally comprise stabilised
polysome
complexes, are described further in W088/08453, W090/0~78~, W090/07003.
CA 02399809 2002-08-02
WO 01/57065 23 PCT/GBO1/00445
W091!02076. W091%0~0~8. and W092/02~36. Alternative display systems w:::;:h
are
not phage-based, such as those disclosed in W09~/2262~ and W'09~/11922
(_~:yma~)
use the polysomes to display polypeptides for selection. These and all the
forgoing
documents are incorporated herein by reference.
In order to produce libraries of sequence segments in accordance with the
present
invention, PCR amplification is advantageously employed. Where a defined
partner
sequence is used. one PCR .primer may be deigned to anneal specifically with
the partner
sequence; for random libraries, general random PCR primers may be used. The
resulting
fragments are joined by restriction and ligation and cloned into suitable
vectors. Although
the ligation of two sequence segments is described below, the invention
encompasses the
li;ation of three or more sequence segments, any of which may be the same or
~:~rfierent.
such as to mirror a multiple cross-over event.
1 ~ The invention is further described, for the purpose of illustration. in
the following
experimental section.
Example 1
Preparation of a repertoire of chimaeric proteins comprising t<vo sequence
segments
A repertoire of genes encoding chimaeric proteins, which comprise the N-
terminal 36
residues of the E. toll cold shock protein (CspA) and a C-terminal polypeptide
sequence
encoded by randomly created fragments of the E. toll genome, was prepared.
CspA
2~ comprises 70 residues and forms a stable ~i-barrel (Schindelin et a1.1994).
Its N-terminal
36 residues comprise the first three strands of its six stranded (3-barrel and
are unable to
fold when expressed alone as they are degraded in the E. toll cytoplasm.
The gene fragment encoding the first 36 residues of CspA was complemented with
fragmented DNA from The E. toll genome around 140 base pairs in size. DNA
fragments
were created by random PCR amplification using genomic E. toll DNA as a
template.
Resulting chimaeric genes were inserted between the coding regions for the
infection
protein p3 and an N-terminal tag, a stable but catalytically inactive mutant
of the RNase
CA 02399809 2002-08-02
WO 01/57065 ~~ PCT/GBO1/00445
barnase, as a single continuous gene on a phagemid vector For protein display
on
filamentous phage.
In the resulting genomic library (l.OxlO~ members) an opal (TGA) stop codon
was
incorporated at the 3' end of the chimaeric gene in 60% of clones with the
remainder
containing the Glv-encoding GGA codon in this position. The partial
incorporation of the
TGA codon at the 3' end of the chimaeric genes was achieved through the use of
two
different PCR primers (XTND and NOARG) in the PCR amplifications of the E.
coli
gene fragments. The transfer-RNAT'~ can decode TGA with an efficiency of up to
3°'0
(Eggertsson & Soll 1988) leading to sufficient display of the barnase-chimera-
p3 fusion
on the phaye but avoiding folding, related, toxic effects. Phages displaying
this repertoire
were prepared using the helper phage KM13, which contains a modified fd gene
encoding a trypsin-sensitive p3 due to a modified sequence (Kristensen ~,
Winter 1997.
to reduce infectivity due to helper phage encoded p3 molecules.
1~
Example 2
Preparation of a repertoire of chimaeric proteins comprising two sequence
segments
with common sequences
In a second "plasmid-derived" library the N-terminal CspA gene fragment was
complemented with DNA fragments of around 140 base pairs created by random PCR
amplification using as the PCR template a 3.6 kb plasmid containing the wild
type Csp_
gene. Resulting chimaeric genes were again inserted as a fusion between the
coding
regions for the infection protein p3 and an N-terminal tag, a stable but
catalytically
inactive mutant of the RNase barnase, on a phagemid vector for protein display
on
filamentous phage.
In the plasmid-derived library (1.7x108 members) an opal (TGA) stop codon was
constitutivelv introduced at the 3' end of the chimaeric gene in all clones.
Phages
displaying this repertoire were prepared using the helper phage K1VI13, which
contains a
modified fd gene 3 encoding a trypsin-sensitive p3 due to a modified sequence
CA 02399809 2002-08-02
WO 01/57065 25 PCT/GBO1/00445
(Kristensen & Winter 1997). to reduce infectivity due to helper phage encoded
p3
molecules.
Example 3
Proteolvtic selection of combinatorial libraries
To select stably folded chimaeric proteins from the repertoires of barnase-
chimaera-p3
fusions described in Examples 1 and 2, the phage-displayed libraries were
select for
proteolytic stability in three rounds through treatment at 10°C with
the proteases trypsin
(specific for peptide bonds containing Arg or Lys in the P 1 position) and
thermolysin
(specific for bonds containing a amino acid with an aliphatic side chain in
the Pl
position) followed by capture on barstar, elution, infection and regrowth.
After the first round of selection. 2x10; and 6x10- of 10'°
proteolytically treated phages
were eluted from a single barstar coated microtitre plate well in case of the
plasmid-
derived library and the genomic library, respectively. When protease treatment
is omitted
5x106 phages can be eluted indicating that the vast majority of unselected
phages did not
display a stably folded chimera protein fused between barnase and p3. The
number of
phages rescued after two and three rounds of selection increased to 2x10' for
the
plasmid-derived library and to 2x103 and 4x10 for the genomic library.
Selected phages were grown up individually, bound to immobilised barstar,
treated in situ
with trypsin and thermolysin at 10°C and resistance was measured
through detection of
bound (and therefore resistant) phage in ELISA. For the plasmid-derived
library 27 of 64
phages (42%) retained 80% or more of their barstar binding activity after
protease
treatment. For the genomic library, after two rounds, 6 of 192 (3%) phages
retained at
least 80% of their barstar binding activity. After three rounds, 31 of 86
(36%) phages
retained 80% or more of their barstar binding activity. Selection therefore
clearly enriched
phages displaying protease-resistant p3 fusions.
CA 02399809 2002-08-02
WO 01/57065 26 PCT/GBO1/00445
Example -1
Sequence analysis of selected chimaeric proteins
As an initial characterisation of the selected chimaeric fusion proteins, the
sequences of
the selected clones from Example 3 were deterniined. The chimaeric genes of
all the 2-1
most stable phage clones selected from the plasmid-derived library had an open
reading
frame from the genes for barnase, through the one for chimaeric protein and to
the end of
the p3 gene. They also contained no stop codons (in addition to the opal stop
codon at the
3' end). Twenty of these contained inserts originating from the CspA gene in
the correct
reading frame. These 20 comprised three different clones (A1 was found 12-
times. D6 6-
times, G4 twice). Phage A1 contains a deleted version (residues 1 to ~?) of
the CspA wild
type gene, which must have been created through a deletion within a phagemid
clone
originally harbouring a larder insert (Table 1 ). Phage D6 contains in
addition to the N-
1 ~ terminal half of CspA (residues 1 to 36 as part of the cloning vector) the
core of CspA
(residues 17 to 53) (Table 1). Phage G4 contains as an insert a partial
duplication of the
N-terminal half of CspA (residues 2 to 19). Thus from the plasmid-derived
library phages
with p3-fusion chimeras, in which the N-terminal half of CspA was complemented
with
another fragment from CspA, were strongly enriched by proteolytic selection.
The sequences of 25 protease resistant phage clones selected from the genomic
library
revealed 11 different clones l2 clones were found five times, 1 clone four
times, 3 clones
twice). All inserts kept the reading frame from barnase into p3. They all
contained the
opal stop codon at their 3'end but no additional stop codons. The inserts of
all phages
sequenced could be traced back to the E. coli genome showing an error-rate of
about 1%
presumably due to their generation by PCR. 64% of the sequenced phages
contained
inserts, whose reading frame was identical to that of the originating E. coli
protein. This
suggests an enrichment for DNA fragments in their natural reading frame, as
from a
random distribution based on three possible reading frames and two possible
orientations
of any DNA only 16% of inserts would be expected to retain the natural reading
frame.
However, the selection of clones which originated from open reading frames
(ORFs) that
do not correspond to the natural reading frame of the originating gene in
36°ro of the
CA 02399809 2002-08-02
WO 01/57065 ~~ PCT/GBO1/00445
sequenced inserts indicates that these may also lead to the formation of
stahi:. folded
chimaeras.
As outlined in Example 1. 60% of all clones in the unselected genomic library
contained
an opal (TGA) stop codon at the 3' end of the chimaeric gene while the
remainder
contained the Gly-encoding GGA codon in this position. However, only clones
containing
opal stop codons at this position were found after proteolytic selection from
the genomic
library. In the absence of a constitutive stop codon almost exclusively
chimae;ic gene
fusions leading to a frameshift between the barnase and p3 genes were selected
( data not
shown). These results show that the efficiency (up to 3% according to
Eg~ertsson &: Soll,
1988), with which transfer-RNAT'~ can decode TGA as a tryptophan, leads to
sufficient
display of the barnase-chimera-p3 fusion on the phage but appears to reduce
folding
related, toxic effects. The use of a opal stop codon in the genes encoding the
displayed
fusion proteins was therefore advantageous for selection in the presented
examples.
Example 5
Proteolytic stability of selected chimaera-phages in solution
To show that the sequenced fusion proteins were not only proteolytically
stable after
immobilisation of the displaying phage on a barstar coated surface (as shown
in Example
3) but also in solution, they were tested for proteolytic stability through
exposure to
trypsin and thermolysin in solution (prior to immobilisation) at different
temperatures
(Fig. 1 a). Phages retaining the barnase tag (as a consequence of a
proteolytically stable
fusion protein) were captured on barstar and the percentage of retained
barstar binding
activity was quantitated by ELISA.
Among the phages from the plasmid-derived library two clones (A1 and D6)
retained at
least 80% of their binding activity after treatment at 20°C. From the
genomic library 8 of
the 11 clones (1C2, 1G6, 1A7, 2F3, 1B11, 2F1, 2H2, 3A12) retained at least
80°,0 of their
activity after trypsin/thermolysin treatment at 24°C. The remaining
phages were less well
protected from proteolytic attack in solution than when bound to the barstar
coated
surface (compare Example 3).
CA 02399809 2002-08-02
WO 01/57065 ~ 8 PCT/GBOI /00445
Example 6
Soluble expression of selected chimaeric proteins
J
To characterise the selected chimaeric proteins outside the context of the
barnase-p3
fusion protein, the genes of the ten most stable chimaeras of the selected
clones in
Example ~ were expressed without the fusion partners. For this, their genes
were
subcloned for cyrtoplasmic expression into a His-tag vector. Five of these
proteins (His-al.
His-d6 from the plasmid-derived library; His-1c2, His-2f3 and His-Ibll from
the
Qenomic library) could be purified after expression directly from the soluble
fraction of
the cyrtoplasm via their His-tag. The remaining proteins formed inclusion
bodies in the
expressing cells. One of these, His-1 g6 containing an insert expressed in a
reading frame
different from that of its originating gene (Table II), was refolded via
solubilisation in 8iVl
1 s urea. The remaining clones were not further studied.
Example 7
Biophysical characterisations of chimaeric proteins
The first biochemical analysis of the purified chimaeric proteins described in
Example 6
concerned their multimerisation status. The chimaeric proteins His-al, His-d6,
His-1c2,
His-2f3, His-lg6 formed only monomers according to their elution volume in gel
filtration. while His-lbl 1 formed 30% monomers with the remainder forming
dimers.
To analyse the type of secondary structure formed by these chimaeras, the
purified
proteins were studied by CD and NMR. The CD spectra (Fig. 2a) of the monomeric
proteins and the monomeric fraction of His-1 b 11 were all characteristic of
(3-structure
containing proteins with minima between 21 ~ nm and 22~ nm (Greenfield &
Fasman
1969, Johnson 1990). All proteins exhibited cooperative folding
characteristics with
sigmoidal melting curves (Fig. 2b) and midpoints of unfolding transition
between 46°C
and 62°C (Table I). The cooperative folding behaviour is a strong
indication that each of
CA 02399809 2002-08-02
WO 01/57065 ?9 PCT/GBO1/00445
the analysed chimaeras forms a domain with a sin~~le fold. in contrast to a
mixture oi-
folded or partially folded structures as in a molten globule.
The NMR spectra of His-2f3 and His-le? further suggested the presence of well
folded
protein domains, as can be inferred from the chemical shift dispersion of many
amide
protons to values downfield of 9 ppm (Fig. 3a, c) and of methyl group protons
to values
around 0 ppm in their NMR spectra (Wiithrich 1986j. Finally, downfield
chemical shifts
of C" protons to values between ~ and 6 ppm, as seen in the N1VIR spectrum of
His-lc~'
(Fig. 3e), are also frequently observed in (3-sheet containing polypeptides
like the
immunoglobulin domains (Riechmann & Davies 199~j.
To determine the thermodynamic stability of the selected chimaeras, the energy
of
unfolding (~Gj of the six proteins was inferred from their thermodenaturation
curves as
measured by CD (Fig. 2b). The folding energies of His-al. His-d6, His-lbl 1,
His-2f3 and
1 ~ His-1 g6 are between 1.6 and 2.4 kcal/mol (Table I). These values are
lower than those of
typical natural proteins and similar to the so far most stable of the de novo
designed ~3
structure proteins, betadoublet (2.5 kcal/mol; Quinn et a1.1994). However, the
His-lc?
protein selected from the genomic library had a considerably higher folding
energy of ~.3
kcal/mol, which falls within the normal range of natural proteins (5 to 15
kcal/mol; Pace
1990). His-lc2 is indeed 1.7 kcal/mol more stable than His-CspA.
The relative folding stabilities of His-2f3 and His-lc2 were confirmed through
the rate of
exchange of their amide protons in D'O as observed in NMR experiments. For His-
2~ a
1D-'H NMR spectrum recorded after incubation for 24 hours in DSO buffer at
2~°C
2~ revealed the complete exchange of its amide protons (Fig. 3a, b). In
contrast. amide
exchange in His-1 c2 was slow allowing the observation of many amide protons
in a 1 D-
'H NMR spectrum after 24 hours at 2~°C in DSO (Fig. 3c,dj. A group of
amide signals
between 8.7 and 10 ppm was even detectable three weeks later at about 40% of
their
original intensity.
CA 02399809 2002-08-02
WO 01/57065 ~0 PCT/GBO1/00445
Example 8
Proteolytic stability of chimaeras as soluble proteins
Apart from the spectroscopic zvidence for folding stability (see Example 7),
stability was
also confirmed by the exposure of the isolated chimaeric proteins to proteases
in solution.
The stability data described in Example 7 of the soluble chimaeric proteins
from Example
6 largely correspond to the degree of their protection from proteolysis by
trypsin.
thermolysin (both used during the selection) and chymotrypsin (Fig. l b).
Tryptic
degradation of the N-terminal His-tag through cleavage after Argl 1 was
observed for all
six proteins. This arginine was introduced as part of the expression vector
immediately C-
terminal of the N-terminal His-tag. His-lc2 (with a folding energy of ~.3
kcal/mol) is no
further degraded by any of the proteases confirming its high conformational
stability, but
the other proteins are partially proteolysed within the main body of the
polypeptides. This
1 ~ is consistent with a partial unfolding expected from a folding energy of
about 2 kcal/mol.
Thus although all the proteins are resistant to proteolysis (for example
compared with the
facile cleavage of the His-tag at Arg), the resistance varies between the
proteins and upon
the conditions.
Example 9
Sequence duplications in selected chimaeric proteins
As outlined in Example 4 above, in 20 of 24 sequenced chimaeric proteins,
which were
selected from the plasmid-derived library, the N-terminal half of CspA was
complemented with another fragment from CspA. Indeed, the chimaeric proteins
D6 and
G4 both comprise a partial duplication of their N-terminal half. Phage D6
contains in
addition to the N-terminal half of CspA (residues 1 to 36 as part of the
cloning vector) the
core of CspA (residues 17 to ~3) (Table 1). Phage G4 contains as an insert a
partial
duplication of the N-terminal half of CspA (residues 2 to 19). This result
indicates that
(partial) duplication of amino acid segments can lead to the formation of
stably folded
protein domains.
CA 02399809 2002-08-02
WO 01/57065 31 PCT/GBO1/00445
Example 10
Duplication of homologous elements in stably folded chimaeric proteins
No direct structural information is available for the seven DNA fragments.
which were
found after selection of the genomic library (Example 1 ) and which were
expressed in
their natural reading frame. One however has a high level of sequence identity
with a
sequence neighbour of known three dimensional structure (as identified by
BLAST
analysis of the E. coli genome). The insert of phage 1B11 spans residues 36-1
to 398 in the
E. coli 30S ribosomal subunit protein S 1 (gene identifier 1787140). of which
residues 369
to 397 have a 52°'o identity with residues 11 to 39 of S 1 1Z'sA-
binding domain from the E.
coli polynucleotide phosphorylase. These comprise a stretch of four ~3-strands
in the 3D
structure of the S 1 domain, which like CspA forms a (3-barrel albeit with an
inserted helix
(Bycroft et a1.1997).
The two S 1 domains (of the 30S ribosomal protein and of the phosphorylase)
are
according to their sequence similarity and identity homologous to Csp.-~. The
juxtaposition of the segments in the chimaeric protein 1B11 represents
therefore a
juxtaposition of corresponding regions from homologous polypeptide domains
(which
also forming the same structural fold). This result indicates that a (partial)
duplication of
homologous amino acid segments can lead to the formation of stably folded
protein
domains.
Example 11
Evidence for complementation with elements of similar structure from proteomic
analysis in a chimaeric protein
20 of the 24 most stable phage clones selected from the plasmid-derived
library (Example
2) contained inserts originating from the CspA gene in the correct reading
frame (see
Example 4). These 20 comprised three different clones (A1, D6, G4). Al
contains a
deleted version (residues 1 to 52) of the CspA wild type gene, which must have
been
created through a deletion within a phagemid clone originally harbouring a
larger insert
CA 02399809 2002-08-02
WO 01/57065 ~~ PCT/GBO1/00445
(Table 1 ). Phage D6 contains in addition to the N-terminal half of CspA i
residues 1 to 3C~
as part of the clonin'; vector) the core of Csp.A (residues 17 to ~3 > (Table
1 ). Phage G-1
contains as an insert a partial duplication of the N-terminal half of Csp_~
(residues ? to
19). The complementing sequences in all three clones comprise regions of
Csp:~. which
in the CspA structure form ~3-strand regions. Thus sequences forming the same
type of
secondary structure are juxtaposed in the chimaeric proteins A1. D6 and G-1.
No direct structural information is available for the seven DNA fragments,
which were
found after selection of the genomic library (Example 1) and which were
expressed in
their natural reading frame. One however has a high level of sequence identity
with a
sequence neighbour of known three dimensional structure (as identified by
BLAST
analysis of the E. coli genome). The insert of phage 1 B 11 spans residues 364
to 398 in the
E. coli 30S ribosomal subunit protein S l (gene identifier 1787140), of which
residues 369
to 397 have a ~?% identity with residues 11 to 39 of S1 RNA-binding domain
from the E.
1 ~ coli polynucleotide phosphorylase. These comprise a stretch of four ~3-
strands in the 3D
structure of the S 1 domain, which like CspA forms a (3-barrel albeit with an
inserted helix
(Bvcroft et crl.1997). Thus sequences forming the same type of secondan-
structure are
juxtaposed in the chimaeric protein 1 B 11.
Thus in case of the His-al, His-d6 and His-lbl 1 proteins the juxtaposition of
sequences.
which form the same type of secondary structure, have lead to the formation of
stably
folded chimaeric protein. Overall, gene fragments selected from both libraries
appear to
be enriched for sequences forming primarily [3-structure in their parent
protein. Such
sequences may be more frequently able to from a stable domain with another
gene
2~ fragment, that originally encodes part of a (3-barrel, than sequences of a
helical origin.
Example 12
Evidence for complementation with elements of different structure from
proteomic
analysis in selected chimaeric proteins
No direct structural information is available for the seven DNA fragments.
which were
found after selection of the genomic library (Example 1 ) and which were
expressed in
CA 02399809 2002-08-02
WO 01/57065 » PCT/GBO1/00445
their natural reading frame. One however has a high level of sequence identity
with a
sequence neighbour of known three dimensional structure (as identified by
BLAST
analysis of the E. coli genome). The insert of 3A12 spans residues ~? to 80 in
the putative
transport periplasmic protein (gene identifier 178790) sharing a 48% sequence
identity
with residues 30 to 58 of the Salmonella oligopeptide-binding protein. In its
~D structure
(Tame et al.1994) these residues form a helm and two short antiparallel (3-
strands. Thc:
oligopeptide-binding protein as a mixed a/~3 protein has no structural
homology with
CspA and its residues ~2 to 80 do not form part of a (3-barrel. Thus sequences
from
different folds are juxtaposed in the chimaeric protein 3A12. Thus while gene
fragments
selected from both libraries appear to be enriched for sequences formin~~
primarily (3-
structure in their parent protein, polypeptide sequences originating from
different folds are
also represented.
Example 13
l~
Effects of modified selection conditions
Proteolytic selection seemingly favoured phaQes displaying chimaeric proteins
with
higher folding stahilities than those displaying chimeras with high melting
points. From
the plasmid-derived library the phage clone displaying the more stable protein
A 1 was
selected twice as frequently as the less stable D6, which however has the
higher meltin~l
point (Table I). In case of the genomic library the phages displaying the two
most stable
proteins (1C2, 1G6) were found four and five times, while the phages of the
two less
stable proteins (1B11, 2F3) were only found twice each after selection. Again
the His
2~ lbll and His-2f3 proteins have the higher melting points (Table I). This
suggests that
escape from proteolysis depends more on stability than on the melting point as
long as
proteolysis is performed at temperatures well below melting points. Higher
proteolysis
temperatures than used here may therefore allow more frequent selection for
proteins with
higher melting points. while energetically more stable proteins would probably
be
enriched if phages are proteolysed for longer.
Such modified conditions may increase the frequency, with which polypeptides
exhibiting
stabilities of natural proteins are selected from random combinatorial
libraries. Further
CA 02399809 2002-08-02
WO 01/57065 ~=1 PCT/GBO1/00445
improvements may be expected by use of much larger repertoires, for Trample
created by
scale up. by improvements in the transfection efficiency of plasmid, pha'~emid
or phase
replicons into cells. or by other techniques such as in vivo recombination
usin~~ the cre-lox
system (Sternberg & Hamilton 1981). Alternatively or in addition repertoires
could be
further diversified by mutagenesis before or after selection. Effective
repertoire sizes can
further be increased, when recombination partners are enriched prior to
recombination for
in frame, no stop codon containing DNA fragments.
The presented methodology allows the selection of new' chimaeric proteins.
which have
been created through recombination of natural genes and which can combine
properties
from different molecules. Using suitable combinatorial partners polypeptides
may be
created. which inherit desirable functions (such as a target binding sites or
an anti~~enic
epitope) from parent proteins. while removing undesirable properties (such as
such as
unwanted receptor binding sites or unwanted epitopes). For this purpose,
proteolyic
1 ~ treatment may be combined with selection for binding.
In the case of selection for binding of chimaeric proteins to a ligand, it may
be
advantageous to increase the copy number of phage displayed fusion proteins.
An
increased copy number of displayed p3-fusion proteins, of which there can be
up to five
on each phage particle, would result in multiple binding events for a single
clone, which
may allow selection even in the case of chimaeric proteins with a low affinity
to the
ligand. Copy number of fusion proteins in phage display can for example be
increased,
when phagemid-encoded fusion p3-fusion protein are rescued for phage
preparation with
a helper phage lacking the gene for p3 (Rakonjac et a1.1997)
Example 14
Secondary modifications of selected chimaeras
~U The binding activity of chimaeric proteins created through the random
recombination of
polypeptide segments for a given ligand may be low, even if the parent
proteins of these
segments have a high affinity for such a ligand. Thus any newly juxtaposed
polypeptide
segment is expected to have some effect on the structure of the other w hen
compared with
CA 02399809 2002-08-02
WO 01/57065 ~~ PCT/GBO1/00445
its structure in the parent protein. As a consequence most binding sites will
no lon~__=er lit a
ligand with the same precision and result in a reduced affinity. It is
therefore envisaged
that it may be necessary to improve such binding sites, once a new chimaeric
protein has
been created as part of a combinatorial library.
Improvements of selected chimaeric proteins can be achieved by secondary
modification
or mutation. Such modifications can be made to improve binding, they may also
be made
to increase stability and/or to introduce new binding or enzymatic functions.
The ype of
modification and its location in the chimaeric protein (i.e. which old amino
acid is
replaced with which new one) may be based on rational design principles or
partially or
entirely random. vlodifications can be introduced by a site-directed
mutagenesis
(Hutchison III et a1.1978) or by a site-directed random mutagenesis (Riechmann
& Weill
1993) followed by selection or screening for activity or stability in the
resulting mutant
chimaeras. Alternatively an entirely random mutagenesis (through for example
error-
1 ~ prone PCR amplification, Hawkins et a1.1992) of either one or both
segments (or indeed
their linking sequence) of the chimaeric protein or through passage of the
phagemid
through an E. coli mutator strain (Low et a1.1996) followed by selection
and/or screening
for binding, enzymatic activity or stability.
Modifications can further comprise the deletion of residues or introduction of
additional
residues. In particular the joining and end regions of the recombined
polypeptide
segments may be expected to be not optimised. The joining regions may strain
interactions between the juxtaposed segments, which may be relieved by
introducing
additional residues within the joining region. Regions close to the end of the
chimaeric
2~ protein may comprise terminal residues not participating in the fold of the
domain. and
their deletion may improve the overall integrity of the protein.
We demonstrate for one of the chimaeric protein how its stability was improved
based on
rational design. His-2f3 was created through the combinatorial shuffling of
the N-terminal
half of the E. coli protein CspA with random amino acid segments encoded by
fragments
of the E. coli genome (Example 1). The sequence and genetic origin of the
random
fragment are given in Table II. The spectroscopic analysis of His-?f3 (Example
7)
CA 02399809 2002-08-02
WO 01/57065 ~6 PCT/GBO1/00445
indicates a fold rich in ~3-structure. If His-?~ folds (like CspA) into a ~-
barrel certain
sequence requirements may have to be met to improve the stability of the
barrel.
In CspA the hydrophobic side chain of residue Leu-I~ closes one end of its (3-
barrel and
GIv~l8 and G1n49 form a turn between two (3-strands in the polypeptide fold to
allow the
formation of backbone hydrogen bonds of the following (3-strand with the N-
terminal (3-
strand of CspA. Within this strand the side chains of three hydrophobic
residues (Vale 1.
Phe53 and Ile») point to the inside of the barrel. His-2h does not meet those
requirements exactly but has a similar motifs within its genomic segments. as
the residues
Pro~8, G1y61, A1a62, Met64. Phe66 and A1a68 (in its genomic segment] exhibit
the same
spacing as the motif described for CspA (compare Table III).
We therefore mutated the genomic segment in 2f; at positions ~8 (P to L) . 62
(A to Qj
and 68 (A to L) to match the amino acid types described for the motif in CspA.
while the
1 ~ residues at position 61, 64 and 66 in 2f; were already judged to be
identical or similar
enough. As summarised in Table III. the combined P58L and the A62Q mutations
increased the stability of ?t~ to 6 kcal/mol, which lies within the range of
typical natural
protein domains and is 1.6 kcal/mol higher than that of CspA itself. The A68L
had no
positive effect in 2f;.
In addition, the two C-terminal residues (PW) in 2h (compare Tables II and
III) were
removed, which were partially degraded in the originally expressed, soluble
?f; protein.
The removal of these residues had no significant effect on the overall
stability of 2f;, but
resulted in a more homogeneous protein preparation after expression, which for
example
2~ is advantageous for structural studies like NMR.
This result shows that new chimaeric proteins can be improved upon after
selection
through further modifications, in this case based on rational design.
CA 02399809 2002-08-02
WO 01/57065 ~7 PCT/GBO1/00445
Example 1,
Crossreactivity of anti-CspA antisera with chimaeric proteins
A possible application of chimaeric proteins is their use as vaccines against
the parent
polypeptide of one or more of the recombined amino acid sequences. For this
purpose
antisera against the chimaeric protein will be cross-reactive with the parent
polypeptide
(and indeed vice versa).
A rabbit was immunised with CspA using Freunds adjuvant (see Methods). The
resultin~~
antiserum recognised immobilised, biotinvlated CspA. Binding of the rabbit
antiserum to
the immobilised Biotin-Csp A could be competed with soluble Csp A and to van-
in;~
degrees with the chimaeric protein His-1c2, His-2t3 and His-lbll (Fig. ~).
This result
shows that an immunisation with CspA results in an immune response which
contains
1 ~ antibodies that crossreact with all three of the analysed chimaeras.
Conversely, it should
therefore also be possible to achieve an immune response against CspA when any
one of
the chimaeras is used for vaccination. The immune response can be expected to
be
directed against both linear and against conformational determinants of CspA.
Example 16
Selection of chimaeric proteins for binding
2~ In the earlier examples, stable folded chimaeric domains were selected by
proteolysis
through the combinatorial juxtaposition of the N-terminal half of the E. coli
protein CspA
with amino acid segments encoded by fragments of the E. coli genome (Examples
1 and
3). A number of these chimaeric proteins are expected to form a polypeptide
fold
resembling that of CspA as the secondary structure prediction and
spectroscopic analyses
of the four chimaeras described (Example 7) indicates a fold rich in ~i-
structure.
It is possible that the RNA binding function (Jiang et a1.1997) of CspA is
retained in
some of the selected chimaeras. The nucleic acid binding site in CspA has been
proposed
CA 02399809 2002-08-02
WO 01/57065 38 PCT/GBO1/00445
to be located on a surface formed around Trpl 1. PhelB, Phe?0. Phi ~ l and
Lys60
(Newkirk et a1.1994; Schroder et a1.199~). While the four aromatic residues
are part of
the N-terminal half of Csp~-~. and are therefore present in all members of the
~~enomic
repertoire (Example 1). residue Lys60 is not. It seems likely that in some of
the chimaeric
proteins the nucleic acid binding activity will be retained; such proteins
could be selected
for example by binding of phage displaying the protein to nucleic acid
immobilised on
solid phase. (However as the phage display system used in the experiments
above would
be unsuitable as the barnase tag retains nucleic acid binding activity).
Furthermore. for functional selection it may be necessary to use a phage-
display system
which allows the multiple display of the fusion protein thereby facilitating
selection of
chimaeric proteins with low- affinities for the li'and (in this case nucleic
acid) through the
resulting avidity effect. This may be achieved in the case of chimaeras fused
to the phaye
coat protein p3 for example through the use of a phage vector like phase td
(Zacher et
1 ~ al.1980), through the use of a phagemid in combination with a helper phase
devoid of the
phase p3 gene (Rakonjac et a1.1997) or through an increased expression of
functional
chimaera-p3-fusion protein. Alternatively, multiple display may be achieved
tlu-ou~~h
fusion to a different phase coat protein, like p8.
Of particular importance is the binding of the chimaeric domains to
antibodies. If
antiserum against the parent protein were used for selections, this would be
expected to
direct the selection to any of the epitopes of the chimaeric protein that are
similar to those
in the parent protein and are represented in the anti-serum. Alternatively
monoclonal
antibodies could be used which would select for those clones binding a single
epitope that
?5 is similar to that of the parent protein. A number of these chimaeric
proteins are expected
to form a polypeptide fold resembling that of CspA, as the secondary structure
prediction
and spectroscopic analyses of the four chimaeras described in Example 7
indicates a fold
rich in (3-structure. If any of the recombined chimaeric proteins within the
repertoire
resemble in fold that of CspA, it should therefore be possible to enrich for
such proteins
through binding to antibodies which specifically recognise CspA.
Example 1 ~ already describes that an anti-CspA antiserum crossreacts with
three of the
chimaeric proteins selected through proteolysis (and barstar binding) alone.
The anti- . ,
CA 02399809 2002-08-02
WO 01/57065 ~9 PCT/GBO1/00445
CspA antiserum may therefore serve as a reagent to enrich the combinatorial
libray from
Example 1 specifically for phages displaying chimaeric proteins which resemble
CspA
most closely.
A rabbit anti-CspA serum was fractionated through binding to Biotin-CspA
immobilised
to Streptavidin-agarose to enrich for that against conformational determinants
or CspA.
Purified CspA-specific (anti-CspA) rabbit antibodies (IgG) were tested for
anti-CspA
binding activity as described in Example 1 ~. For use in phage selection the
anti-Csp.~
rabbit antibodies were immobilised on a Streptavidin-coated ELISA-well plate
through a
commercial biotinvlated goat anti-rabbit IgG antiserum. Phages (7x109 cfu)
from the
genomic library of chimaeric proteins (Example 1 ). which had undergone one
round of
proteolytic selection (followed by barstar binding, see Example 3), were
treated with
trypsin and thermolvsin (see Example 3) followed by binding to the CspA-
specific rabbit
antibodies in 2% BSA in PBS. After washing with PBS and 40 mM DTT =l.3xlOJ
bound
phage were eluted at pH2, neutralised and used for infection of bacterial
cells.
96 of the resulting clones were grown up in a multiwell plate and infected
with helper
phage KM13 for phage production. Phage from the culture supernatants from the
infected
bacterial clones were bound to the anti-CspA antibodies. which again had been
?0 immobilised to a Streptavidin-coated plated via a biotinylated goat anti-
rabbit IaG
antiserum. Bound phage was washed with PBS, exposed to trypsin and thermolysin
after
immobilisation as before, washed with PBS and remaining phage was detected
with an
anti-M13-HRP conjugate. Sequences of the nine clones with the strongest signal
remaining after proteolysis were determined. Seven of these clones were
identical (two)
or almost identical (five clones had one residue less at the N-terminal end of
the genomic
insert and two different residues at the C-terminal end) to the clone 2t~,
which had been
previously selected - albeit not at the same high frequency - after
proteolytic
selection/barstar binding (Examples 3 and 4). The two remaining sequences had
not been
previously observed. Purified phage of the 2f3 and 2f3-like clones was
confirmed to be
strongly reactive with the purified rabbit anti-CspA antibodies, also after
exposure to
trypsin in solution, confirming that it is protease-resistant folded sequences
that are
binding to antibody. Together with the fact that the anti-serum had been
fractionated for
binding to the folded CspA, this indicates strongly that the selection has
been towards a
CA 02399809 2002-08-02
WO 01/57065 'l~ PCT/GBO1/00445
conformational determinant. The ELISA in Fig. 4 (see Example I ~) proves that
the
correspondin~~ chimaeric protein also interacts in its soluble version with an
anti-Csp.1
antiserum.
This experiment suggest how it is possible to identify "isosteric" peptides
(same
conformation in parent protein and chimaeric domain). It also indicates that
the method
can be used for vaccination towards a conformational segment of the protein;
thus it
should equally be possible to use 2f~ for vaccination and to produce anti-
serum that
recognises the conformation of the N-terminal portion of CspA.
Methods
Lector constrczctiozzs
The gene for the H10?A mutant of barnase (Meiering et x1.199?) was fused to
the N-
terminus of the gene 3 protein (p3) of phage fd (Zacher et cr1.1980) in a
modified
1 ~ phagemid pHEN 1 (Hoogenboom et al. l 991 ) between the DNA encoding the
pelB leader
peptide and the mature p3 after PCR amplification with suitable
oligonucleotides using
Ncol and PstI restriction sites to create the vector p22-1?. Into p22-12
suitably amplified
parts of the E. coli gene CspA (Goldstein et x1.1990) were cloned between the
barnase
and the p3 genes using PstI and NotI restriction sites. In the resulting
phagemid vector
pC~-7 the barnase gene is followed by the N-terminal 36 residues of CspA (the
N-
terminal Met being mutated to Leu to accommodate the PstI site) and the DNA
sequence
GGG AGC TCA GGC GGC CGC AGA A (SacI and NotI restriction sites in italics)
before the GAA codon for the first residue (Glu) of p3. In pC~-7, the barnase-
Csp
cassette is out of frame with the p3 gene. In the control vector pCsp/2 the
barnase-Csp
2~ cassette is in frame with the p3 gene, but the first codon of the linking
DNA constitutes an
opal stop codon.
Vectors for the cytoplasmic expression of soluble proteins were constructed by
subcloning of genes from the phagemids into the BamHI and HindIII sites of a
modified
QE30 vector (Qiagen). This vector is identical to QE30 except for a tetra-His
tag. During
PCR-aided subcloning using the primers CYTOFOR (~'-CAA CAG TTT A .4G CTT CCG
CCT GAG CCC AGG-3') and CYTOBAK (~'-CCT TTA CAG GAT CCA GAC TGC
AG-3') opal stop-codons were converted into the Trp-encoding TGG triplet.
CA 02399809 2002-08-02
WO 01/57065 '11 PCT/GBO1/00445
Library construction
As templates for random amplifications 100 n~~ of a pBCSK (Stratagene) based
plasmid
containing the entire CspA coding region or genomic DNA (2 u8 di;Tested with
Sacl)
from the E. coli .strain TG1 (Gibson 198-1) prepared as described (Ausubel et
ul.199~) was
used in 2~ PCR cycles with an annealing temperature of 38°C using the
oligonucleotide
SN6NEW (~'-GAG CCT GCA G:-IG CTC AGG _NNN NNN-3' at 40 pmole/ml) for the
plasmid or in 30 PCR cycles with an annealin<, temperature of 38°C
usin<~ the
oligonucleotide SN6IV1IX (~'-GAG CCT GCA G.AG CTC CGG NWT N'NN-3' at ~10
pmole/ml) for the genomic DNA. PCR products were extended in a further 30
cycles
with an annealing temperature of ~2°C using the oligonucleotide NOARG
(~'-CGT GCG
AGC CTG CAG AGC TC:A GG-3' at x.000 pmole/ml) for the plasmid and the
oligonucleotide XTND (~'-CGT GCG AGC CTG CAG .~GC TCC GG-3' at 1.000
pmole/ml) for the genomic DNA. PCR products of around I~10 by were purified
from an
agarose gel and reamplified in 30 PCR cycles using the oligonucleotide NOARG
at an
annealing temperature of ~0°C.
Resulting fragments were digested with SacI, purified and ligated into the
phosphatased
and SacI-digested vector pC~-7. Ligated DNA was electroporated into TG1
creating a
plasmid-derived repertoire of 1.7x108 clones and a genomic repertoire of
l.OxlOs clones.
In both libraries about 60% of the recombinants contained monomeric inserts,
while the
remainder contained oligomeric inserts. Ligation background was less than 1 %
for both
ligations. Due to differences in the 3' end of the PCR primers XTND and NO.ARG
40% of
clones with in-frame inserts in the genomic library contained a GGA-encoded
Gly residue
as part of the 3'-SacI site, while the remaining clones contained the TGA-
encoded opal
2~ stop-codon at the same position. All members of the plasmid-derived library
with in-
frame inserts contained the TGA-encoded opal stop-codon at this position.
Selections
For selections about 10'° colony forming units (cfu) of phage were
treated with 200 WI
trypsin (Sigma T8802) and 384 nM thernlolysin (Sigma P1~12) in TBS-Ca buffer
(2~
mM Tris. 137 mM NaCI, 1mM CaCl2, pH 7.4) for 10 minutes at 10°C.
Proteolysed
phaae was captured for 1 hour with biotinylated C=l0A,C82A double mutant
barnase
inhibitor barstar (Hartlev 1993. Lubienski et x1.1993) immobilised on a
streptavidin
CA 02399809 2002-08-02
WO 01/57065 4~ PCT/GBO1/00445
coated microtitre plate (Boehringer) wells in 3°~o Marvel in PBS. Wells
were washed
twenty times with PBS and once with ~0 mM dithiothreitol (DTT) in PBS for ~
minutes
to elute phage containing proteolysed p3-fusions held together solely by
disulphide
bridges. Bound phage w-as eluted at pH 2, neutralised to pH 7 and propay~ated
after
reinfection.
Phabe ELISA
Proteolysis and binding of purified phage (about 10'° cfu per well) to
immobilised barstar
was performed as above. Phage remaining bound after washes with PBS and DTT
was
detected in ELISA with an anti-M 13 phage antibody - horse radish peroxidase
(HRP)
conjugate (Pharmacia) in 3% IVtarvel in PBS. Non-purified phage from culture
supernatants was bound to the biotinvlated barstar and then proteolysed in
.situ. Purified
phage was proteolvsed in solution and proteases were inactivated with Pefabloc
(Boehringer) and EDTA before capture.
1J
Anti-CpsA antisera
A first anti-CspA serum (as used for Fib. 4) was obtained from an immunised
rabbit. The
rabbit was injected once with refolded (see above) His-CspA (0.~ ml at
1.7~mg/ml PBS)
mixed with 1:1 with Freud's complete adjuvant, followed by two injections with
refolded
His-CspA (0.5 ml at 1.7~mg/ml PBS) mixed 1:1 with Freud's incomplete adjuvant
in 4
week intervals to boost the immune response. The antisera used was obtained
from blood
taken ten days after the second boost.
A second anti-CspA serum (as used for purification of anti-CspA specific
antibodies in
2~ Example 16) was obtained from a different immunised rabbit. The rabbit was
injected
once with refolded (see above) His-CspA (0.~ ml at 1.7~mg/ml PBS) mixed with
1:1 with
Freud's complete adjuvant, followed by three injections with refolded His-CspA
(0.~ ml at
1.75mg/ml PBS) alone in ~l week intervals to boost the immune response. The
antisera
used was obtained from blood taken ten days after the third boost. One ml of
this
antiserum was purified on 0.2 ml of Streptavidin-agarose (Pierce No. X3117).
to which
about 0.1 mg Biotin-CspA (see below) was bound, after washing with PBS.
elution at pH
2 followed by neutralisation and buffer exchange into 3.~ ml PBS (i.e. at 3.~
fold dilution
compared to the original antiserum). When used for phage selection, the
purified anti-
CA 02399809 2002-08-02
WO 01/57065 '13 PCT/GBO1/00445
CspA antibodies were X00-fold diluted in PBS for binding to a biotinylated
goat anti-
rabbit antiserum (Sigma B-7389) immobilised in Streptavidin-coated ELISA
wells.
His-CspA, as used for immunisation and data in Table III and Fig. =l, was
purified from
the unfractionated E. coli cell pellet using NTA agarose after solubilisation
with 8M urea
in TBS. Before elution with 200 mM imidazole in PBS, agarose bound His-CspA
was
renatured with an 81~I to OM urea gradient TBS. Eluted protein was dialvsed
against PBS.
For biotinylation, CspA was modified through addition of cysteine-glutamine-
alanine
residues as a C-terminal tag, introduced on the gene level using suitable PCR
primers.
The corresponding His-CspA-Cys protein was expressed. purified and refolded as
His-
CspA except for the addition of 0.~ mM DTT to all solutions. The NTA agarose
with the
bound His-CspA-Cvs was washed with ~ volumes of PBS (all solutions without DTT
from this step onwards) and mixed with the biotinylation reagent EZ-Linl<TV~
Biotin-
1 ~ HPDP (Pierce) for biotinylation according to the manufacturer's
instructions. After 1 hour
the agarose with the bound and biotinylated protein was washed with 10 volumes
of PBS,
eluted with 200 mM imidazole in PBS and buffer-exchanged into PBS.
Biotinylation of
the now His-Biotin-CspA was verified by MALDI mass spectrometry using a SELDI
(Ciphergen systems).
Binding of rabbit anti-CspA antisera to CspA was analysed after immobilisation
of
biotinylated His-Csp-Gys (at 0.2~ ~g/ml in PBS) onto streptavidin-conjugated
ELISA
plates (Boehringer Mannheim). The rabbit anti-CspA serum was diluted 1/30,000
in 2%
bovine serum albumin in PBS and preincubated with varied amounts of purified
2~ competitors (see Fig. 4) before binding to the ELISA well. Bound rabbit
antibodies from
the serum were detected with a HRP-conjugated goat anti-rabbit IaG antiserum
(Sigma).
2f3 mutants
The gene for the 6H-2f3 protein (compare Table III) was prepared by PCR with
the
primers QEBACK (~'-CGG ATA ACA ATT TCA CAC AG-3') and 2F3FOR (~'-GGC
CGC CTG AAG CTT TTA AGG CGG ATG GTT GAA-3') using the 2f3 gene in QE30
(compare Table II) as a template. Mutant genes for the 6H-2f3 protein were
prepared
through PCR amplification of the partial 2f; gene using accordingly designed
primers and
CA 02399809 2002-08-02
WO 01/57065 44 PCT/GBO1/00445
the same template. For each mutant two PCR products (coverin' the N and C-
terminal
portion of the 2f3 gene respectively) were purified, denatured, annealed and
extended.
Full-length mutant genes were specifically reamplified using the two outside
primers
BACKTWO (~'-CCT TTA CAG GAT CC-3') and 2F3FOR. Complete genes were
digested with HindIII and BamHI and cloned into the unmodified QE30 vector
(Qiagen:
encoding a 6 histidine containing N-terminal tag).
For the mutant 6H-2t~-P>8L the primers 2F3F2 (~'-GGT AAA A~~.G CAT GAT TGC
GCC AAT TTC TAG CTC GCC TGC-3'), CYTOBAK (for the N-terminal halfj, 2F3B0
(~'-GGT AAA AAG CAT GAT TGC G-3') and QEFOR (~'- GTT CTG AGG TCA TTA
CTG G-3~) (for the C-terminal half were used). For the mutant 6H-2f~-P~BL_A62Q
the
primers 2F3F1 (~'-GGT AAA AAG CAT GAT TTG GCC AAT TTC TAG CTC GCC
TGC-3' ), CYTOBAK (for the N-terminal hal f , 2F3B0 and QEFOR (for the C-
terminal
half were used). For the mutant 6H-2f;-P~8L,A62Q,A68L the primers 2F3F1,
1 ~ CYTOBAK (for the N-terminal hal f . 2F3B 1 (~'-AAT CAT GCT TTT TAC CCT AAT
GGA TGG C-3') and QEFOR (for the C-terminal half were used).
Protein expression, purification and analysis
Proteins were expressed by induction of exponential bacterial cultures at
30°C and
purified from the soluble fraction of the cytoplasm using NTA agarose
according to the
Qiagen protocol. His-lg6 was purified after solubilisation with 8 M urea in
TBS and
refolded by dialysis from 8 M, 4 M, 2 M, 1 M, 0.~ M to 0 M urea in TBS.
Proteins were
further purified by gel filtration on a Superdex-7~ column (Pharmacia). The
molecular
weight of proteolytic fragments was determined using the surface enhanced
laser
2~ desorption/ionisation (SELDI) technique (Hutchens & Yip 1993).
Proteolysis of soluble proteins (about 40 pM) was carried out using 40 nM of
trypsin,
thermolysin or a-chymotrypsin (Sigma C3142) in TBS-Ca at 20°C for 10
minutes.
Circular dichroism spectra and thermodenaturation were recorded as described
(Davies &
Riechmann 1995). Thermodenaturation of 10 ~M protein (His-lc2 at 2 ~M) in PBS
was
followed at a wavelength between 220 nm and 22~ nm (His-lc2 in 2.~ mM
phosphate
buffer, pH 7, at 20~ run). Nuclear magnetic resonance experiments were
performed on a
Bruker D~tY-600 spectrometer as described (Riechmann & Holliger 1997) using a
CA 02399809 2002-08-02
WO 01/57065 4~ PCT/GBO1/00445
Watergate sequence (Piotto et al. l99?) for water suppression with protein at
I mvI in ?0
mM phosphate buffer at pH 6.? containing 100 mM NaCI in 93% H~O / 7% DSO or
99.9% DSO.
All publications mentioned in the above specification are herein incorporated
by
reference. Various modifications and variations of the described methods and
system of
the invention will be apparent to those skilled in the art without departing
from the scope
and spirit of the invention. Although the invention has been described in
connection with
specific preferred embodiments, it should be understood that the invention as
claimed
should not be unduly limited to such specific embodiments. Indeed, various
modifications
of the described modes for carrying out the invention which are obvious to
those skilled
in molecular biology or related fields are intended to be within the scope of
the following
claims.
CA 02399809 2002-08-02
WO 01/57065 46 PCT/GBO1/00445
Table I. Amino acid sequences and biophysical parameters of de novo proteins.
Protein Tm,oC ~1G' MW, Da C-terminal sequence'
His-Csp X9.8 3.6 8.566 IQNDGYKSLDEGQK'JSFTIESGAKGPAAGu
VTSLEA
His-Csp/2 no expr.' x,854 WAQAEA
plasmid library:
His-al 46.0 ?.1 7,729 IQNDGYKSLDEGQKVSFTWAQAEA
His-d6 47.6 1.6 10,32 GSSGFGFITPDDGSKDVFVHF SAIQNDGYK
SLDEGQKVSFTWAQAEA
genomic library:
His-lbll X7.1 2.0 10.722 GSSGP.AVRGNPQQGDRVEGKIT~~SI' DFGI
IGLDGGIDGLVHLSDISWAQAEA
His-2f3 61.4 1.8 10,82 GSSGAGEPEIGAIMLFTAMDGSEL~TPGVIRE
INGDSITVDFNHPPPWAQAEA
His-lc2 X4.8 5.3 10,972 GSSGRVISLTNENGSHSVF SYDALDRLVQQ
GGFDGRTQRYHYDLTWAQAEA
His-lg6 48.4 2.4 10,48 GSSGKSGVKTDYRASASIACAYAGAGSSDS
RRSFLCITRSESDGPWAQAEA
1: The conformational stability 0G (kcal/mol) at a given temperature T was
calculated
using the Gibbs-Helmholtz equation 4G(T) = OHm (1-T/Tm) - OCp [(Tm-T)
ln(T/Tm)], while inferring the midpoint of thermal unfolding (Tm) and the
enthalpy
change for unfolding (OHm) at the Tm from the denaturation curve (Agashe &
Udgaonkar
1995) and assuming for ~Cp (the difference in heat capacity between unfolded
and folded
2~ conformation) a value of 12 cal per residue (Edelhoch & Osborne 1976).
2: Sequences shown are those following the N-terminal half of CspA, which is
MRGSHHHHGSRLQSGKMTGIVKWFNADKGFGFITPDDGSKDVFVHFSA.
3: The His-(Csp/2) protein was found neither in the soluble nor insoluble
fraction of the
cytoplasm presumably due to degradation within the cell.
CA 02399809 2002-08-02
WO 01/57065 47 PCT/GBO1/00445
Table II. Sequences and origin of genomic segments
Segments Sequenced Genetic origine Protein originf
la7b GIATSAICDA QVIGEEPGQP 8931 to 9041 in minus strand
TSTTCRFRSK FSAIAFPW ECAE298, gatC
lbl lb~cGAAVRGNPQQ GDRVEGKIKS 6382 to 6514 in 364 to 398 in
ITDFGIFIGL DGGIDGLVHL ECAE193, rpsA RS1 ECOLI
SDISCn~
lc2be GRVISLTNEN GSHSVFSYDA 2178 to 2303 in 645 to 686 in
LDRLVQQGGF DGRTQRYHYD ECAE156, rhsD RHSD ECOLI
L TW
lg6b>c GKSGVKTDYR ASASIACAYA 2694 to 2569 in frameshift
GAGSSDSRRS FLCITRSESD ECAE116, rluA
GPW
2flb GAGTMAEEST DFPGVSRPQD 8558 to 8422 in 452 to 494 in
MGGLGFWYRW NLGWMHDTLD ECAE419, glgB GLGB ECOLI
YMKPHSW
2~b,c GAGEPEIGAI MLFTAMDGSE 5431 to 5551 in 89 to 127 in
MPGVIREING DSITVDFNHP ECAE113, slpA FKBX ECOLI
PPW
2h2b GSAYNTNGLV QGDKYQIIGF 7955 to 7854 in minus strand
PRFNQLTVYF HNLPW ECAE475, yjbC
3a12b GKAVGLPEIQ VIRDLFEGLV 1479 to 1568 in 52 to 80 in
NQNEKGEIVP W ECAE231, b1329 MPPA_ECOLI
CA 02399809 2002-08-02
WO 01/57065 ~8 PCT/GBO1/00445
1 g7 Grr-iLKRKLNLK FNLAS IAGCD 7290 to 7213 in frameshift
._LLNAAW ECAE217. b1 191
1h12 GC VPYTNFSL IYEGKCGMSG 1203) to 11927 in 33~ to 36% in
GRVEGKVIYE TQSTHK~:SW ECAE48~, cadA DCLY-ECOLI
?e2 GMWPLDMVNA IESGIGGTLG 7398 to 714 in 4~ to 83 in
i ~AAVIGPGT ILGKIM L VSW ECAE324, dsdX DSD~ ECOLI
Segments retaining 80% barstar binding activim after proteolysis of phage in
.sitto and in
solutionb and those purified as chimaeric proteins. The sequence of the
genomic
segmentd as a C-terminal appendage to the N-terminal region of CspA
(LQSGKMTGIV
KWFNADKGFG FITPDDGSKD VFVHFSAGSS) is listed; sequences expressed in-
1 ~ frame with the originating gene are shown in italics. The location of each
se;ment within
the E. coli genome is indicated by nucleotide numbers in the EMBL database
entry and
name of the originating genee, and for those expressed in the same frame of
the
originating gene, the residue numbers of the corresponding protein and its ID
in the Swiss
protein database are givenf. A single base pair deletion after the first 29
base pairs in the
DNA insert of lbl 1 renders the first 10 residues out of frame with the rspA
gene=.
CA 02399809 2002-08-02
WO 01/57065 '~9 PCT/GBO1/00445
Table III.
(a) Amino acid sequences of CspA and His-2f3
2C 3C 40
CspA MSGKMTGIVK WFNADKGFGF ~TPDDGSKDV FVHFSAIQND G'yKSLDEGQ=~:
* .~
2f3 ...SGKMTGIVK WFNADKGF GF I TPDDGSKDV FVHFSAGSSG :_~F.-PEIG_'
24 j4 44 54
10 60 7G
CspA VSFT IESGAK GPAAG~'I. SL
2f3 MLFTAMDGSE MPGVI?.J-_viG DSTTVDFNHP P
73 83 93
1J
(b) Folding energy of 2f3 mutants and CspA
Protein 4G at 298K (kcal/mol)
CspA
6H-2f3 1.9
6H-2f3-P58L 2.8
6H-2f3-P58L, A62Q 6.0
2~ 6H-2f3-P58L, A62Q, A68L 3.2
(a) The amino acid sequence of CspA is that from the native gene as in the
E;VIPL
database. The numbering of the 2f3 sequence takes into account the N-terminal
His-tag
(MRGSHHHHHHGSRLQ). The C-terminal residues PWAQAEA (compare 2f; in Table
I) were deleted in the constructs used for the data here, as they were
partially cleaved in the
expressed protein of the original His-2f3 construct indicating that they did
not participate
to the fold of the chimaeric domain. Their deletion had no significant effect
on the overall
folding stability of the domain (1.8 vs. 1.9 kcalimol in the 2f3 constructs
used for data in
CA 02399809 2002-08-02
WO 01/57065 50 PCT/GBO1/00445
Table I and III respectively). The residues important for the ~3 barrel fold
in CspA as
discussed in Example 14 are indicated by an asterisks.
(b) Folding energies were determined as described in Table I. Mutation for 2t~
denote the
original amino acid, followed by the residue number and the new amino acid.
CA 02399809 2002-08-02
WO 01/57065 ~ 1 PCT/GBO1/00445
References
:Ausubel. F.l~~f.. Brent. R., Kin~~ston. R.E.. Moore. D.D.. Seidman. .1.G..
Smith. .1..-~. ~:
Struhl, K. (1990 Current protcols in molecular biology. Chapter 2.4.1. Wiley &
Sons.
Agashe, V.R. & Udgaonkar, J.B. (1990 Thermodynamics of denaturation of
barstar:
evidence for cold denaturation and evaluation of the interaction with
guanidine
hydrochloride. Biochemistry 34, 3286-3299.
Alba, E. de, Santoro, J., Rico, M. & Jimenez, M.A. (1999) De novo design of a
monomeric three-stranded anti-parallel (3-sheet. Protein Sci. 8, 8~4-865.
Altschul, S.F., Madden, T.L., Schaffer, A.A., Zhang, J.H., Zhang, Z.,
ivliller. V'. ~:
Lipman, D.J. (1997). Gapped BLAST and PSI-BLAST: a new Generation of protein
database search programs. Nucleic Acids Res. 2~, 3389-3402.
1~
Barbas, C.F., Crowe, J.E., Cababa, D.. Jones, T.M., Zebedee, S.L., Murphy,
B.R..
Chanock, R.M. & Burton D.R. (1992) Human monoclonal fab fragments derived from
a
combinatorial library bind to respiratory syncytial virus-f glycoprotein and
neutralize
infectivity. Proc. Natl. Acad. Sci. USA 89, 10164-10168.
Bogarad, L., & Deem, M. (1999) A hierarchical approach to protein molecular
evolution.
Proc. Natl. Acad. Sci. USA 96, 2591-2~9~.
Breitling, F., Dubel, S., Seehaus, T., Klewinghaus, I. & Little M. (1991) A
surface
expression vector for antibody screening. Gene 104, 147-153.
Burton, D.R., Barbas, C.F., Persson, M.A.A., Koenig, S., Chanock, R.M. &
Lerner. R. A.
(1991) A large array of human monoclonal-antibodies to type-1 human
immunodeficiency-virus from combinatorial libraries of asymptomatic
seropositive
individuals. Proc. Natl. Acad. Sci. USA 88, 10134-10137.
CA 02399809 2002-08-02
WO 01/57065 ~~ PCT/GBO1/00445
Byeroft. i~M.. Hubbard. T.J.. Proctor. VI.. Freund. S.l~i. & Murzin. .-~.G. (
1997) The
solution structure of the s1 RNA binding domain: a member of an ancient
nucleic acid-
binding fold. Cell 88, ?3~-242.
Caton, A.J. & Koprowski. H. ( 1990) Influenza-virus hemagglutinin-specific
antibodies
isolated from a combinatorial expression library are closely related to the
immune-
response of the donor. Proc. Natl. Acad. Sci. USA 87, 640-644.
Chang, C.N., Landolti, N.F. & Queen, C. (1991) Expression of antibody fab
domains on
bacteriophage surfaces - potential use for antibody selection. J. Immunol.
147. 3610-3614.
Clackson, T., Hoogenboom. H.R.. Griffiths. A.D. & Winter, G. (1991) Making
antibody
fragments usin~~ phage display libraries. Nature 3~2. 624-628.
1 ~ Davies, J. & Riechmann, L. ( 1990 An antibody VH domain with a lox-Cre
site integrated
into its coding region: bacterial recombination within a single polypeptide
chain. FEBS
Lett. 377, 92-96.
Davidson, A.R. & Sauer, R.T. (1994) Folded proteins occur frequently in
libraries of
random amino-acid-sequences. Proc. Natl. Acad. Sci. USA 91, 2146-210.
Devereux, J., Haeberlie, P. & Smithies O. (1984) A comprehensive set or
sequence
analysis program for the VAX. Nucl. Acids Res. 12. 387-39~.
Dower, W.J. & Fodor, S.P.A. ( 1991 ) The search for molecular diversiy .2.
recombinant
and synthetic randomized peptide libraries. Annu Rep Med Chem 26, 271-280.
Edelhoch, H. & Osborne, J.C., Jr. (1976) The thermodynamic basis of the
stability of
proteins, nucleic acids, and membranes. Adv. Prot. Chem. 30, 183-250.
Eggertsson, G. R. Soll, D. (1988) Transfer ribonucleic acid-mediated
suppression of
termination codons in Escherichia Coli. Microbiol. Rev. ~2. 3~4-374.
CA 02399809 2002-08-02
WO 01/57065 ~3 PCT/GBO1/00445
Feng, D.F. & Dolittle. R.F. ( 1987) Progressive sequence alignment as a
prerequisite to
correct phvlo~~enetic trees. J. of lrolec. Evol. 2~. 3~ 1-360.
Fincuane: M.D., Tuna, M., Lees. J.H. ~. Woolfson, D.N. ( 1999) Core-directed
protein
design. I. An e;cperimental method for selecting stable proteins from
combinatorial
libraries. Biochemistry 38, 11604-11612.
Fire, A. & Xu. S.Q. (1990 Rolling replication of short dna circles. Proc.
Natl. Acad. Sci.
USA 92, 4641-464.
Fodor. S.P.A., Read, J.L., Pirrung. M.C., Stryer, L., Lu, A.T. & Solas. D. (
1991 ) Light-
directed, spatially addressable parallel chemical synthesis. Science 2~ 1. 767-
773.
Fontana, A., Laureto, P. de, Filipis, V. de, Scaramella, E. & Zambonin. VI. (
1997) Probing
1~ the partly folded states of proteins by limited proteolysis. Fold. Des. 2.
R17-R26.
Gibson, T.J. (1984) Ph.D. Thesis, University of Cambridge, UK.
Goldstein, J., Pollitt, N.S. & Inouye, M. (1990) Major cold shock protein of
Escherichia
coli. Proc. Natl. Acad. Sci. USA 87, 283-287.
Greenfield, N & Fasman, G.D. (1969) Computed circular dichroism spectra for
the
evaluation of protein conformation. Biochemistry 8, 4108-4116.
2~ Hardies, S.C., Hillen, W., Goodman, T.C. & Wells, R.D. (1979) High
resolution thermal
denaturation analyses of small sequenced DNA restriction fragments containing
Escherichia coli lactose genetic control loci. J. Biol. Chem. 2~4, 527-s~34.
Hartley, R.W. (1993) Directed mutagenesis and barnase-barstar recognition.
Biochemistry
32, 5978-X984.
Hawkins, R.E., Russell, S.J. & Winter, G. (1992) Selection of phage antibodies
by
binding-affinity - mimicking affinity maturation. J. Mol. Biol. 226, 889-896.
CA 02399809 2002-08-02
WO 01/57065 54 PCT/GBO1/00445
Hawkins, R.E. & Winter. G. (1992) Cell selection-strategies For making
antibodies from
variable gene libraries - trapping the memory pool. Eur. J. Immunol. 22. 867-
870.
Hecht, M. (1994) De novo design of (3-sheet proteins. Proc. Natl. Acad. Sci.
USA 91,
8729-8730.
Higgins, D.G. & Sharp. P.M. (1989) Fast and sensitive multiple sequence
alignment on a
microcomputer. CABIOS 5, 151-153.
Hoogenboom; H.R., Griffiths, A.D., Johnson, K.S., Chiswell, D.J., Hudson. P. &
Winter.
G. (1991) Multi-subunit proteins on the surface of filamentous phage:
methodologies for
displaying antibody (Fab) heavy and light chains. Nucleic Acids Res. 19, 4133-
4137.
Hubbard, S.J., Eisenmenger, F. & Thornton, J.M. (1994) Modeling studies of the
change
in conformation required for cleavage of limited proteolvtic sites. Protein
Science 3, 757-
768.
Huse; W.D., Sastry. L., Iverson, S.A., Kang, A.S., Altingmees, M., Burton,
D.R.,
Benkovic, S.J. & Lerner, R.A. (1989) Generation of a large combinatorial
library of the
immunoglobulin repertoire in phage-lambda. Science 246 1275-1281.
Hutchens, T. W., and Yip, T-T. (1993) New desorption strategies for the mass
spectrometric analysis of macromolecules. Rapid Commun. Mass Spectrom. 7, 576-
580.
Hutchison III, C.A., Phillips, S., Edgell, M.H., Gillam, S., Jahnke, P. &
Smith, M. (1978)
Mutagenesis at a specific position in a DNA sequence. J. Biol. Chem. 253, 6551-
6560.
Jiang, W.N., Hou, Y. & Inouye, M. (1997) CspA, the major cold-shock protein of
Escherichia coli, is an RNA chaperone. J. Biol. Chem. 272, 196-202.
Johnson, W.C. Jr (1990) Protein secondary structure and circular-dichroism - a
practical
guide. Proteins 7, 205-214.
CA 02399809 2002-08-02
WO 01/57065 ~J PCT/GBO1/00445
Kamtekar. S.. Schiffer. J.M., Xiong, H., Babik. J.M. & Hecht. M. ( 199= )
Protein desi~Tn
by binary patternin~~ of polar and nonpolar amino acids. Science 262, 1680-
168.
Kang, A.S., Jones, T.M. & Burton, D.R. (1991) Antibody redesign by chain
shuftZin~J
from random combinatorial immunoglobulin libraries. Proc. Natl. Acad. Sci. USA
88.
11120-11123.
Kristensen, P. & Winter, G. (1997) Proteolvtic selection for protein folding
using
filamentous bacteriophages. Folding Des. 3, 321-328.
Kortemme, T., Ramirez-Alvardo, M. & Serrano, L. (1998) Design of a 20-amino
acid.
three-starnded (3-sheet protein. Science 281, 2~3-2~6.
1~ Lerner, R.A., Kang, A.S., Bain, J.D., Burton, D.R. & Barbas, C.F. (1992)
Antibodies
without immunization Science 2~8, 1313-1314.
Low, N.M., Holliger, P. & Winter, G. ( 1996) Mimicking somatic hypermutation:
Affiniy
maturation of antibodies displayed on bacteriophage using a bacterial. J. Mol.
Biol. 260.
359-368.
Lowman HB, Bass SH, Simpson N. & Wells, J.A. (1991) Selecting high-affinity
binding-
proteins by monovalent phage display. Biochemistry 30. 10832-10838.
2~ Lubienski, M.J., Bycroft, M., Jones. D.N.M. & Fersht, A.R. (1993)
Assignment of the
backbone H-1 and N-l~ nmr resonances and secondary structure characterisation
of
barstar. FEBS Lett. 332, 81-87.
Marks, J.D., Hoogenboom, H.R., Bonnert, T.P., McCafferty, J., Griffiths, A.D.
& Winter.
G. (1991) By-passing immunization - human-antibodies from v-gene libraries
displayed
on phage. J. Mol. Biol. 222, 581-597.
CA 02399809 2002-08-02
WO 01/57065 ~6 PCT/GBO1/00445
Marks. J.D.. Hoo~enboom HR. Griffiths AD. Winter G (199?j Molecular evolution
of
proteins on filamentous phage - mimickin<, the strate~v of the immune-system.
J. Biol.
Chem. 267. 16007-16010.
NIcCaffertv. J., Griffiths. A.D., V-inter. G. & Chiswell, D.J. (1990) Phage
antibodies
filamentous phage displayin~~ antibody variable domains. Nature 348. ~~~-~~4.
Meiering, E.M., Serrano, L. & Fersht, A.R. (1992) Effect of active site
residues in barnase
on activity and stability. J. Mol. Biol. 22~, ~8~-589.
Mullinax. R.L., Gross, E.A., Amberg, J.R.. Hay. B.N., Hogrefe. H.H.. Kubitz.
VL~I..
Greener, A., Altingmees, M.. Ardourel. D., Short. J.M.. Sorge. J.A. & Shopes.
B. (1990)
Identification of human-antibody fragment clones specific for tetanus toxoid
in a
bacteriopha~e-lambda immunoeYpression library. Proc. Natl. Acad. Sci. USA 8~.
809~-
1 J 8099.
Myers, E.W. & Miller, W. (1988) "Optimal Alignments in Linear Space", CABIOS
4. 11-
17.
Needleman, S.B. & Wunsch, C.D. (1970) A general method applicable to the
search for
similarities in the amino acid sequences of two proteins. J. Mol. Biol. 48. -
144--1> .
Newkirk, K.. Feng, W.Q., Jiang, W.N., et a1.(1994) Solution nmr structure of
the major
cold shock protein (cspa) from escherichia-coli - identification of a binding
epitope for
2~ DNA. Proc. Natl. Acad. Sci. USA 91, ~ 114-~ 118.
Pace; C.N. (1990) Conformational stability of globular proteins. Trends
Biochem. Sci. 1 ~.
14-17.
Persson. _l~t.A. A., Caothien, R.H. & Burton, D.R. ( 1991 ) Generation of
diverse hi<Th-
affinih~ human monoclonal-antibodies by repertoire cloning. Proc. Natl. Acad.
Sci. US A
88. 2432_2436.
CA 02399809 2002-08-02
WO 01/57065 ~ 7 PCT/GBO1/00445
Piotto: M, Saudek, V. & Sklenar. V. ( 199?) Gradient-tailored excitation for
sin=le-
quantum nmr-spectroscopy of aqueous-solutions. J. Biomolecular NMR ?. 661-66~.
Quinn, T.P.. Tweedy, N.B., Williams, R.W., Richardson, J.S. & Richardson. D.C.
( 1994]
~ Betadoublet: De novo design, synthesis, and characterisation of a (3-
sandwich protein.
Proc. Natl. Acad. Sci. USA 91. 8747-87~ 1.
Rakonjac, J., Jovanovic, G. R. 'Model. P. (1997) Filamentous phage infection-
mediated
gene expression: construction and propagation of the gIII deletion mutant
helper phage
R408d3. Gene 198. 99-103.
Regan, L. (1998) Proteins to order'' Structure 6, 1-4.
Riechmann, L. & Davies, J. ( 1990 Backbone assignment, secondary structure and
Protein
l~ A binding of an isolated, human antibody VH domain. J. Biomol. NMR 6. 141-
152.
Riechmann, L. & Holliger, P. (1997) The C-terminal doamin of ToIA is the
coreceptor for
filamentous phage infection of E. coli. Cell 90, 351-360.
Riechmarm, L., & Weill, M. (1993) Phage display and selection of a site-
directed
randomized single-chain antibody Fv fragment for its affinity improvement.
Biochemistry
32, 8848-88».
Sauer, R.T. (1996) Protein folding from a combinatorial perspective. Folding
Des. 1, R27-
2~ R30.
Schindelin, H.. Maraheil, M.A. & Heinemann, U. (1994) Crystal structure of
CspA. the
major cold shock protein of Escherichia coli. Proc. Natl. Acad. Sci. USA 91, ~
119-~ 123.
Schroder. K.. Graumann, P.. Schnuchel, A., Holak, T.A. & Marahiel, M.A. (1990
Mutational analysis of the putative nucleic acid-binding surface of the cold-
shock domain.
cspb, revealed an essential role of aromatic and basic residues in binding of
single-
stranded-dna containing the y-box motif. Mol. IVIicrobiol. 16, 699-708.
CA 02399809 2002-08-02
WO 01/57065 ~8 PCT/GBO1/00445
Scott, J.K. & Smith. G.P. (1990) Searchinv~ for peptide ligands with an
epitope library.
Science 249. 386-390.
Sieber, V.. Plueckthun, A. & Schmid, F.X. (1998) Selecting proteins with
improved
stability by a phage-based method. Nat. Biotechnol. 16. 9~~-960.
Smith, T.F. & Waterman, M.S. (1981) Comparison of Bio-sequences. Advances in
Applied Mathematics 2, 482-489.
Smith, T.F., Waterman, M.S. & Sadler, J.R. (1983) Statistical characterisation
of nucleic
acid sequence functional domains. Nucleic Acids Res. 11. 2205-2220.
Sternberg, N. & Hamilton, D. (1981) Bacteriophage P1 site-specific
recombination. I.
Recombination between IoxP sites. J. Mol. Biol., 150, 467-486.
Tame J.R., Murshudov, G.N., Dodson, E.J., Neil, T.K., Dodson, G.G., Higgins,
C.F. &
Wilkinson, A.J. (1994) The structural basis of sequence-independent peptide
binding by
OppA protein. Science 264, 1578-1581.
Thompson, J.D., Higgins, D.G. & Gibson, T.J. (1994) ClusterW: improving the
sensitivity of progressive multiple sequence alignment through sequence
weighing,
positions-specific gap penalties and weight matrix choice. Nucleic Acid Res.
22, 4673.
2~ Wilbur, W.J. & Lipman, D.J. (1983) Rapid similarity searches of nucleic-
acid and protein
data banks. Proc. Natl. Acad. Sci. USA 80, 726-730.
Wiithrich, K. (1986) NMR of proteins and nucleic acids. Chapter 3. Wiley &
Sons.
Zacher, A.N., Stock, C.A., Golden, J.W. & Smith, G.P. (1980) A new filamentous
phage
cloning vector: fd-tet. Gene 9, 127-140.
Cited patents and patent applications:
CA 02399809 2002-08-02
WO 01/57065 59 PCT/GBO1/00445
PCT/GB00/00030
PCT/GB98/01889
W084/03564
~ W088/084~3
W091/OSOSB
W090/0578~
W090/07003
W090/1 X070
W091 /02076
W092/00091
W092/02~36
W092/10092
W093/06121
W095/11922
W095/2262~
U.S. Patent No. 4,631,211
U.S. Patent No. 5,143,84
