Novel antagonist antibodies and their Fab fragments against GPVI
and uses thereof
FIELD OF THE INVENTION
The present invention discloses novel antibodies that specifically bind to the human
platelet membrane protein Glycoprotein VI (GPVI) and their monovalent fragments or
derivatives. The antibodies of the invention are antibodies from hybridoma clone 390
and antibodies that bind to an epitope which is similar to the conformational epitope
recognized by the antibody from hybridoma clone 390. These antibodies and Fab
fragments are able to block collagen binding and thus preventing platelet activation
by collagen. The invention also relates to hybridoma clones and expression plasmids
for the production of disclosed antibodies and Fab fragments. The present invention
further refers to the uses of monovalent antibody fragments to manufacture
research, diagnostic and immunotherapeutic agents for the treatment of thrombosis
and other vascular diseases. In another aspect, the invention concerns a method to
prevent receptor activation through patient specific anti-Fab antibodies induced
receptor clustering.
BACKGROUND OF THE INVENTION
Platelet function is of fundamental importance in the development of arterial
thrombosis and cardiovascular diseases. Nowadays it is a matter of course that
patients suffering from cardiovascular diseases are treated with antiplatelet drugs.
Despite the availability of various clinically successful antiplatelet therapies, there is
still a large unmet medical need for new treatments. This deficiency is mainly caused
by the limited efficacy of the currently available drugs, particularly in regard to the
drugs efficacy - safety (bleeding) correlation. Interfering with early events of platelet
activation and adhesion, a mechanism not targeted by drugs currently in use, would
be an attractive approach for the improvement of the efficacy - safety margin. The
collagen receptor glycoprotein (GPVI) is of central importance in these early events
of platelet activation, and therefore a major target for the interference with this
mechanism (Nieswandt B and Watson SP, Blood. 2003 Jul 15;102(2):449-61). The
antiplatelet and antithrombotic effects of GPVI have been described in several in
vitro and in vivo systems, using platelets from mice and men. Platelets deficient in
GPVI are rendered unresponsive to collagen, one of the most important
thrombogenic components of the subendothelial matrix (Lockyer S. et al, Thromb
Res. 2006; 118(3):371-80). Moreover, mouse studies have shown that GPVI
deficiency causes an effective inhibition of arterial thrombus formation at the
damaged vessel wall without increasing the susceptibility to spontaneous bleeding.
All these data suggest that GPVI represents an effective and safe target for the
treatment of arterial thrombosis. The central role of GPVI in the initiation of thrombus
formation indicates that inhibition of this receptor may be beneficial in syndromes of
arterial thrombosis. This makes the use of biotherapeutic proteins such as antibodies
a clinically meaningful strategy for the inhibition of GPVI. All the more since the
interaction of GPVI and its ligand collagen seems to involve an expanded protein
surface, a successful interference with this protein-protein interaction is more likely
with inhibitory GPVI-binding proteins as compared to other strategies. An in vivo
proof of concept for the inhibition of GPVI function by antibodies and Fab fragments
has been shown in several animal models.
There is still a need for clinically effective inhibitors of GPVI activity.
GPVI is a major collagen receptor expressed exclusively on platelets and
megacaryocytes. Binding to collagen induces receptor clustering and subsequent
platelet activation. This is one of the initial events in thrombus formation. Current
anti-platelet drugs interfere with thrombus formation through targeting late events in
this process. A serious side effect of these drugs is prolonged bleeding which limits
their use. There are several lines of evidence that targeting early events in thrombus
formation (such as GPVI) can be highly antithrombotic without a major impact on
bleeding liability. The interaction between collagen and GPVI can be successfully
inhibited by neutralizing monoclonal antibodies (mAb). However as mAb's are
bivalent molecules, they can induce GPVI-receptor clustering and therefore platelet
activation. To circumvent this, monovalent antibody fragments such as Fab
fragments have been developed.
Unfortunately, in depth safety profiling of these monovalent anti-GPVI Fab fragments
reveals a potential to still induce platelet activation in a patient specific manner. To
offer a safe therapeutic for the treatment of ischemic events, this activator/ potential
of the developed Fab molecules had to be abolished. Until now, this problem has not
been resolved.
SUMMARY OF THE INVENTION
In a first aspect, this invention provides a novel monoclonal antibody (from clone
390) which specifically binds to human GPVl, as well as to primate GPVI.
In another aspect, the invention concerns recombinant Fab fragments derived from
the variable domain sequences of anti-GPVI mAb from clone 390. These anti-GPVI
antibody and Fab fragments recognize a specific conformational epitope in D1 and
D2 domains of the human GPVI protein.
Both the anti-GPVI mAb and the Fab fragments are able to inhibit collagen binding to
GPVI. The anti-GPVI Fab fragments have been humanized and engineered with the
aim to improve their affinity to human GPVI. The biophysical characteristics of the
humanized variants are described.
In addition to the inhibition of collagen, the anti-GPVI Fab fragments are able to
inhibit platelet aggregation induced by collagen both in human platelet rich plasma
and in human whole blood. These Fab Fragments are also able to inhibit the
thrombus formation under flow on a collagen coated surface.
Thus, inhibition of GPVI by antibodies in humans appears as an attractive
antithrombotic strategy.
Surprisingly, the anti-GPVI Fab of the invention induces a GPVI depletion
phenotype. Thus, anti-GP VI Fab fragments able to induce GPVI depletion
phenotype constitute another object of the invention.
The invention also provides a method for prevention of recognition of Fab fragments
by pre-existing antibodies consisting in masking the C-terminal extremity of the Fab
by addition of a molecule.The invention also provides a method for prevention of
recognition of Fab fragments by new antibodies directed toward Fab C-terminal
extremity consisting in masking the C-terminal extremity of the Fab by addition of a
molecule.
The invention also provides a method for prevention of platelet activation when an
anti-GPVI Fab is used, where the C-terminal extremity of the Fab is masked by
addition of a molecule.
Another object of the invention is a Fab fragment bearing a molecule at the C-
terminal extremity.
DNA and protein sequences as well as vectors for expression of the anti-GPVI and
Fab fragments of the invention are also provided.
The antithrombotic agents of the invention may be used for the treatment of
thrombotic or vascular diseases.
The invention also provides an antithrombotic composition comprising a
pharmaceutically effective amount of a GPVI specific monoclonal antibody fragment
of the invention with appropriate excipients.
In another aspect of the invention, the antibodies can be used for diagnosis of
patients at risk requiring anti-thrombotic treatment. The invention also encompasses
a kit for diagnosis including anti-GPVI antibodies or fragments thereof.
The invention also provides a method for the preparation of GPVI antibody, antibody
fragments and masked antibody Fab fragments of the invention.
DETAILED DESCRIPTION OF THE INVENTION
In a first aspect, this invention provides a novel monoclonal antibody and fragments
thereof which specifically binds to human GPVI, in particular the anti-GPVI mAb from
hybridoma clone 390. The antibodies of the present invention antagonize, totally or
partially, the activity of GPVI.
In a second aspect of the invention, the inventors have demonstrated that the
addition of a peptide at the C-terminal extremity of the heavy chain (HC) of a Fab
has a masking effect that avoids recognition of Fab molecules by preexisting anti-
Fab antibodies in some patients and therefore prevents platelet activation and
associated side effects.
As used herein, "monoclonal antibody from hybridoma clone 390" or "monoclonal
antibody from clone 390" refers to an antibody defined by SEQ ID NO.6 and SEQ ID
NO.8 as well as its derivatives including humanized antibodies, Fab fragments and
humanized Fab fragments or any improved version of these molecules especially as
described in the present ap
plication.
The term "antibody" is used herein in the broadest sense and specifically covers
monoclonal antibodies (including full length monoclonal antibodies) of any isotype
such as IgG, IgM, igA, IgD and IgE, polyclonal antibodies, multispecific antibodies,
chimeric antibodies, and antigen binding fragments. An antibody reactive with a
specific antigen can be generated by recombinant methods such as selection of
libraries of recombinant antibodies in phage or similar vectors, or by immunizing an
animal with the antigen or an antigen-encoding nucleic acid.
A typical IgG antibody is comprised of two identical heavy chains and two identical
light chains that are joined by disulfide bonds. Each heavy and light chain contains a
constant region and a variable region. Each variable region contains three segments
called "complementarity-determining regions" ("CDRs") or "hypervariable regions",
which are primarily responsible for binding an epitope of an antigen. They are usually
referred to as CDR1, CDR2, and CDR3, numbered sequentially from the N-terminus.
The more highly conserved portions of the variable regions are called the
"framework regions".
In a particular embodiement, the antibodies and fragments thereof of the invention,
including the antibody Fab fragment, comprise the 6 CDR defined by SEQ ID NO:18,
SEQ ID NO:19, SEQ ID NO:20, SEQ ID NO:21, SEQ ID NO:22 and SEQ ID NO:23.
In another embodiement, the antibodies and fragments thereof of the invention,
including the antibody Fab fragment, comprise the 6 CDR defined by SEQ ID NO:18,
SEQ ID NO:19, SEQ ID NO:38, SEQ ID NO:21, SEQ ID NO:39 and SEQ ID
NO:23.ln another particular aspect, the antibody Fab fragment of the invention
comprises:
(a) complementarity determining regions (CDRs) of a heavy chain variable
region (HCVR) having amino acid sequences defined by SEQ ID NO: 18, SEQ ID
NO: 19, and SEQ ID NO: 20 (or SEQ ID NO:38); and
(b) complementarity determining regions (CDRs) of a light chain variable
region (LCVR) having amino acid sequences defined by SEQ ID NO: 21, SEQ ID
NO: 22 (or SEQ ID NO:39), and SEQ ID NO: 23; and
wherein at least 2 amino acid residues of each CDR can be changed to
another amino acid residue without directly disrupting a contact with a GPVI epitope
residue.
As used herein, "directly disrupting a contact with the GPVI epitope residue" refers to
changing an antibody amino acid residue that is in contact with a GPVI epitope
residue and described in the crystal structure of the examples provided herein.
As used herein, "VH" refers to the variable region of an immunoglobulin heavy chain
of an antibody, including the heavy chain of an Fv, scFv, dsFv, Fab, Fab' or F(ab')2
fragment. Reference to "VL" refers to the variable region of the immunoglobulin light
chain of an antibody, including the light chain of an Fv, scFv, dsFv, Fab, Fab' or
F(ab')2 fragment.
A "polyclonal antibody" is an antibody which was produced among or in the presence
of one or more other, non-identical antibodies. In general, polyclonal antibodies are
produced from a B-lymphocyte in the presence of several other B-lymphocytes
producing non-identical antibodies. Usually, polyclonal antibodies are obtained
directly from an immunized animal.
A "monoclonal antibody", as used herein, is an antibody obtained from a population
of substantially homogeneous antibodies, i.e. the antibodies forming this population
are essentially identical except for possible naturally occurring mutations which might
be present in minor amounts. These antibodies are directed against a single epitope
and are therefore highly specific.
As used herein, the terms "antigen-binding portion" of an antibody, "antigen-binding
fragment" of an antibody, and the like, include any naturally occurring, enzymatically
obtainable, synthetic, or genetically engineered polypeptide or glycoprotein that
specifically binds an antigen to form a complex. Antigen-binding fragments of an
antibody may be derived, e.g., from full antibody molecules using any suitable
standard techniques such as proteolytic digestion or recombinant genetic
engineering techniques involving the manipulation and expression of DNA encoding
antibody variable and optionally constant domains. Non-limiting examples of
antigen-binding fragments include: (i) Fab fragments; (ii) F(ab')2 fragments; (iii) Fd
fragments; (iv) Fv fragments; (v) single-chain Fv (scFv) molecules; (vi) dAb
fragments; and (vii) minimal recognition units consisting of the amino acid residues
that mimic the hypervariable region of an antibody (e.g., an isolated complementarity
determining region (CDR)). Other engineered molecules, such as diabodies,
triabodies, tetrabodies and minibodies, are also encompassed within the expression
"antigen-binding fragment," as used herein.
As used herein, the term "Fab fragment" corresponds to the light chain (LC) plus part
of the heavy chain (HC) of the antibody. Fab molecules are produced either by
proteolytic cleavage of IgG molecules or made through expression of recombinant
molecules. Usually parts of the constant domain of the heavy chain (Fc part) are
removed.
As used herein, the term "masked Fab fragment" or "masked Fab" corresponds to a
Fab fragment containing the light chain (LC) plus part of the heavy chain (HC) of the
antibody and where the C-terminus of the HC has been extended by a peptide to
mask recognition of the Fab by preexisting antibodies in a patient.
The antibodies of the invention have been obtained by the generation of hybridoma
through mouse immunization technique using a human recombinant GPVI protein
containing only extracellular D1 and D2 domains. The sequences of the protein used
for immunization correspond to SEQ ID NO.3 and SEQ ID NO.4.
The obtained mAbs have been selected first for their ability to recognize the human
GPVI protein, but also for their capability to block the binding of collagen to the
human GPVI protein and to bind the GPVI protein with a high affinity.
The antibodies show cross-reactivity to primate GPVI (as shown in Figure 2)
whereas no cross-reactivity exists against murine, rat, dog, swine or rabbit GPVI.
The selected mAb is able to bind to the human GPVI protein with a pM affinity (as
shown in Table 1) and to block collagen binding to the human GPVI protein with
IC50 values of less than 10 µg/mL (as shown in Figure 1A). This antibody is also
able to block collagen binding to non-human primate GPVI (as shown in Figure 1B).
In another aspect, the invention concerns recombinant Fab fragments derived from
the variable domain sequences of anti-GPVI mAb from clone 390. Fab fragments
can be prepared by any method know in the art. For example proteolytic Fab
fragments can be obtained by Ficin cleavage or recombinant Fab fragments can be
prepared by cloning the required sequences into an expression vector (for example
as described in Example 3B). Like the anti-GPVI mAb, the Fab fragments are able to
inhibit collagen binding to GPVI.
The Fab fragments have been co-crystallized with GPVI and analyzed by X-ray
crystallography to determine the epitope recognized by the Fab on the GPVI protein.
This analysis allowed defining a conformational epitope involving both D1 and D2
domains in the extracellular domain of the human GPVI. The numbering system
used in the eptiope corresponds to GPVI with the 20 residue signal sequence
removed and as shown in Figure 16 and SEQ ID NO:32.
In one aspect of the invention, an isolated antibody or antigen binding fragment that
specifically binds a conformational epitope of human GPVI and contacts human
GPVI residues comprising Ser 43, Arg 67 and Asp 81. In another aspect of the
invention, the antibody or antigen binding fragment makes additional contacts with
two or more GPVI residues selected from the group consisting of Pro4, Lys7, Glu40,
Lys41, Leu42, Ser44, Ser45, Arg46, Arg65, Trp76, Leu78, Pro79, Ser80, Gln82,
Ser165, and Arg166. In another aspect of the invention, the antibody or antigen
binding fragment makes additional contacts with three or more, four or more, five or
more, six or more, seven or more, eight or more, nine or more, ten or more, eleven
or more, twelve or more, thirteen or more, fourteen or more, fifteen or more GPVI
residues selected from the group consisting of Pro4, Lys7, Glu40, Lys41, Leu42,
Ser44, Ser45, Arg46, Arg65, Trp76, Leu78, Pro79, Ser80, Gln82, Ser165, and
Arg166.
As used herein, "contacting residues" are defined as residues with a distance of less
than 4.5 A to GPVI antigen residues or the reverse and as measured in a crystal
structure between an antibody to GPVI or antigen binding fragment of the antibody
and a GPVI antigen using a software program such as NCONT of the CCP4
Software package or an equivalent software program.
In one embodiement, the antibody Fab fragment binds a conformational epitope of
human GPVI and contacts human GPVI residues comprising Ser 43, Arg 67 and Asp
81.
In another embodiement, the antibody Fab fragment binds a conformational epitope
of human GPVI and contacts human GPVI residues comprising:
Pro4, Lys7, Glu40, Lys41, Leu42, Ser43, Ser44, Ser45, Arg46, Arg65, Arg67, Trp76,
Leu78, Pro79, Ser80, Asp81, Gln82, Ser165, Arg166
The mAb of the invention recognize the same conformational epitope since the Fab
fragments contain the variable domain sequences of anti-GPVl mAb from clone 390.
Thus, in another aspect, the invention provides an anti-GPVl antibody including Fab
fragment, which recognizes a conformational epitope comprised in the extracellular
D1 and D2 domains of the human GPVI protein.
Thus in another aspect, the invention provides a monoclonal anti-GPVl antibody
which recognizes a conformational epitope comprising the following amino acids:
Glu40, Lys41, Leu42, Ser43, Ser44, Ser45, Arg46.
In another aspect, the invention provides an anti-GPVl antibody which recognizes a
conformational epitope comprising the following amino acids:
Trp76, Leu78, Pro79, Ser80, Asp81, Gln82.
In another embodiment, the anti-GPVl antibody recognizes a conformational epitope
comprising the following amino acids:
Glu40, Lys41, Leu42, Ser43, Ser44, Ser45, Arg46, Trp76, Leu78, Pro79, Ser80,
Asp81,Gln82.
In another embodiment, the anti-GPVl antibody recognizes a conformational epitope,
comprising the following amino acids:
Pro4, Lys7, Glu40, Lys41, Leu42, Ser43, Ser44, Ser45, Arg46, Arg65, Arg67, Trp76,
Leu78, Pro79, Ser80, Asp81, Gln82, Ser165, Arg166.
In another embodiment, the anti-GPVl antibody recognizes a conformational epitope
consisting in the following amino acids:
Pro4, Lys7, Glu40, Lys41, Leu42, Ser43, Ser44, Ser45, Arg46, Arg65, Arg67, Trp76,
Leu78, Pro79, Ser80, Asp81, Gln82, Ser165, Arg166.
In a particular embodiement, the antibody Fab fragment of the invention involves
mainly D1 domain of the human GPVI.
The antibody of the present invention can be defined as a monoclonal anti-GPVI
antibody which (i) competitively inhibits the binding of an antibody comprising a HC
of SEQ ID NO.6 and a LC of SEQ ID NO.8 to human GPVI and (ii) binds to an
epitope similar to the conformational epitope recognized by antibody comprising a
HC of SEQ ID NO.6 and a LC of SEQ ID NO.8, or a part of this conformational
epitope, with an affinity of at least 10 nM (KD = 10-8 M) as determined by a
biophysical methods as for example Surface Plasmon Resonance (Biacore as
described in Karlsson R, Larsson A. (2004), J. Mol. Biol. 248, 389-415).
As used herein, the term "KD" refers to the dissociation constant of a particular
antibody/antigen interaction.
The KD reflecting the interaction between the human GPVI protein and the antibody
and Fab of the invention is comprised in the [10-7; 10-10] interval. An antibody specific
for human GPVI present a KD = 10-7 M, preferably a KD = 10-8 M, more preferably a
KD = 10-9 M. The antibody with the highest affinity may have a KD = 10-10 M or below,
for example 10-11M or 10-12M. The term "specifically binds," or the like, means that
an antibody or antigen-binding fragment thereof forms a complex with a GPVI
antigen that is relatively stable under physiologic conditions. Specific binding can be
characterized by a dissociation constant of 1x10-6 M or less. Methods for
determining whether two molecules specifically bind are well known in the art and
include, for example, equilibrium dialysis, surface plasmon resonance as in
Examples 2 and 3, and the like. For example, an antibody that "specifically binds"
human GPVI, as used in the context of the present invention, includes antibodies
that bind human GPVIor portion thereof with a KD of less than about 1000 nM, less
than about 500 nM, less than about 300 nM, less than about 200 nM, less than about
100 nM, less than about 90 nM, less than about 80 nM, less than about 70 nM, less
than about 60 nM, less than about 50 nM, less than about 40 nM, less than about 30
nM, less than about 20 nM, less than about 10 nM, less than about 5 nM, less than
about 4 nM, less than about 3 nM, less than about 2 nM, less than about 1 nM or
less than about 0.5 nM, such as measured in a surface plasmon resonance assay.
An "epitope" is the site on the antigen to which an antibody binds. If the antigen is a
polymer, such as a protein or polysaccharide, the epitope can be formed by
contiguous residues or by non-contiguous residues brought into close proximity by
the folding of an antigenic polymer. In proteins, epitopes formed by contiguous
amino acids are typically retained on exposure to denaturing solvents and are known
as "linear epitopes", whereas epitopes formed by non-contiguous amino acids are
typically lost under said exposure and are known as "conformational epitopes".
As used herein, the term "similar epitope" concern a set of amino acids located in the
same region as the amino acids described by crystallography analysis as being
involved physically in the interaction between antibodies derived from clone 390 and
the extracellular domain of the human GPVI protein. A similar epitope can include
several differences, for example up to five amino acids differences in the vicinity of
the epitope of reference. A similar epitope can also present one or more
modifications in the amino acids identified as forming the epitope, i.e. the amino
acids interacting with the GPVI protein. One or more amino acid identified as part of
the epitope of reference may be absent, and one or more additional amino acid
residues may be present to form the similar epitope. For example, up to five amino
acids may be changed without affecting the characteristics of the antibody compared
to antibody from hybridoma clone 390. Thus, the region wherein the epitope is
located for one specific antibody encompasses some potentially additional
interacting amino acids. In this view, a similar epitope can be defined as a region for
competitive binding with the antibody of reference". When a neutralizing antibody
binds to such region, one or more cellular signalling pathway(s) of the receptor is/are
inhibited leading to a specific impairment of one or more biological function(s) of the
targeted receptor or more generally effective protein.
A "neutralizing" or "blocking" antibody, as used herein, is intended to refer to an
antibody whose binding to GPVI: (i) interferes with the interaction between GPVI and
collagen, (ii) and/or (iv) results in inhibition of at least one biological function of GPVI.
The inhibition caused by a GPVI neutralizing or blocking antibody needs not to be
complete as long as it is detectable using an appropriate assay. Exemplary assays
for detecting GPVI inhibition are described herein
The existence of topographic regions in proteins linked to specific biological activities
is for example illustrated in patent US 6,448,380.
Antibodies recognizing a similar epitope as antibodies derived from clone 390 can be
selected by a competitive ELISA assay using GPVI-Fc fusion protein as capturing
antigen. Plates coated with GPVI-Fc fusion protein are incubated with hybridoma
supernatants or antibody or antigen binding fragments derived from clone 390. A
lack of binding of different anti-GPVI antibodies (labelled by any standard technique)
to epitope blocked GPVI indicates the recognition of a similar epitope. An equivalent
selection strategy can be pursued by competitive Biacore analysis in which GPVI is
immobilized and the antibody (fragments) derived from clone 390 is used to block
the epitope. Other GPVI binding antibodies candidate are now analysed for binding
to epitope blocked GPVI. Thus antibodies that recognize a similar epitope as epitope
recognized by the antibody of the invention can be selected and further
characterized by analysis of co-crystal structures using X-ray analysis.
The invention also concerns a humanized and engineered anti-GPVI antibody and
fragment thereof.
In the context of the invention, the anti-GPVI Fab fragments have been humanized
using a method previously described in WO2009/032661, but any humanization
method known in the art can be used.
Based on the analysis of the crystallographie complex of the Fab with the GPVI,
several mutations have been introduced both for humanization and with the aim to
improve the affinity of the Fab to the human GPVI.
As a result, 3 variants for the LC and 5 variants for the HC have been generated.
These Fab variants are summarized in Table 6 and in Table 7, respectively for LC
(VL)and HC(VH).
Among all possible combinations, the invention concerns in particular the following
combinations:
VL1 with VH1 which correspond to SEQ ID NO.15 and SEQ ID NO.10 respectively
VL1 with VH2 which correspond to SEQ ID NO.15 and SEQ ID NO.11 respectively
VL1 with VH3 which correspond to SEQ ID NO.15 and SEQ ID NO.12 respectively
VL1 with VH4 which correspond to SEQ ID NO.15 and SEQ ID NO.13 respectively
VL1 with VH5 which correspond to SEQ ID NO.15 and SEQ ID NO.14 respectively
VL2 with VH2 which correspond to SEQ ID NO.16 and SEQ ID NO.11 respectively
VL3 with VH2 which correspond to SEQ ID NO.17 and SEQ ID NO.11 respectively
VL3 with VH4 which correspond to SEQ ID NO.17 and SEQ ID NO.13 respectively
In a preferred embodiment, the humanized variant of anti GPVl Fab is association of
VL1 with VH3 corresponding to SEQ ID NO.15 with SEQ ID NO.12.
In one aspect of the invention, the VH of the anti-GPVI antibodies can be further
modified by the addition of tags or amino acid extensions (peptide) to mask the
antibody from recognition by preexisting antibodies. A sequence of six or more
histidine residues, in particular (His)6. or (His)?, or (His)8, or a unit such as
GlyGlyGlyGlySer or (GlyGlyGlyGlySer)2 can be added at the C-terminus of the heavy
chain.
As used herein, "engineered Fab Fragment" refers to Fab fragments that have been
genetically engineered to contain a peptide extension of the heavy chain at the c-
terminus. The extension causes the previous c-terminus to be obscured and
prevents recognition and binding of the Fab fragments to pre-existing anti-Fab
antibodies.
In another embodiment of the invention, an engineered Fab fragment comprising a
combination of a humanized heavy chain (HC) amino acid sequence and a
humanized light chain (LC) amino acid sequence and a c-terminal extension is
provided. The engineered Fab fragment further comprises a combination of a heavy
chain variable region (HCVR) amino acid sequence and a light chain variable region
(LCVR) amino acid sequence, selected from the group consisting of
(a) LCVR (SEQ ID NO.15) and HCVR (SEQ ID NO.10);
(b) LCVR (SEQ ID NO.15) and HCVR (SEQ ID NO.11);
(c) LCVR (SEQ ID NO.15) and HCVR (SEQ ID NO.12);
(d) LCVR (SEQ ID NO.15) and HCVR (SEQ ID NO.13);
(e) LCVR (SEQ ID NO.15) and HCVR (SEQ ID NO.14);
(f) LCVR (SEQ ID N0.16) and HCVR (SEQ ID N0.11);
(g) LCVR (SEQ ID N0.17) and HCVR (SEQ ID N0.11); and
(h) LCVR (SEQ ID N0.17) and HCVR (SEQ ID N0.13).
wherein the c-terminal extension is selected from the group consisting of SEQ ID
NO:35; SEQ ID NO:36; and SEQ ID NO:37.
In a particular embodiment, the invention concerns a monoclonal antibody
comprising the HC of SEQ ID NO.6 and the LC of SEQ ID NO.8 or a sequence
having at least 80%, 85%, 90%, 95% or 99% identity with these sequences but
which retains the same activity as the said monoclonal antibody.
In a more particular embodiement, the antibody Fab fragment of the invention
comprises (a) heavy chain variable region having amino acid sequences defined by
SEQ ID NO:6; and (b) light chain variable region having amino acid sequences
defined by SEQ ID NO: 8.
The invention also concerns nucleic acids encoding anti-GPVI antibodies and Fab of
the invention. In one embodiment, the nucleic acid molecule encodes a HC and/or a
LC of an anti-GPVI antibody. In a preferred embodiment, a single nucleic acid
encodes a HC of an anti-GPVI antibody and another nucleic acid molecule encodes
the LC of an anti-GPVI antibody.
In a particular embodiment, nucleic acids or polynucleotides encoding polypeptides
of the HC and LC from the antibody from hybridoma 390 correspond to SEQ ID NO.5
and SEQ ID NO.7 respectively, or a sequence having at least 80%, 85%, 90%, 95%
or 99% identity with these sequences but which retains the same activity as the said
monoclonal antibody.
The polynucleotide encoding a polypeptide selected from the group consisting in
SEQ ID NO.6, SEQ ID NO.8, SEQ ID NO.9, SEQ ID NO.10, SEQ ID NO.11, SEQ ID
NO.12, SEQ ID NO.13, SEQ ID NO.14, SEQ ID NO.15, SEQ ID NO.16, and SEQ ID
NO.17 or a sequence having at least at least 80%, 85%, 90%, 95% or 99% identity
with these sequences but which retains the same activity as the said monoclonal
antibody are also part of the present invention.
The invention provides vectors comprising the polynucleotides of the invention. In
one embodiment, the vector contains a polynucleotide encoding a HC of an anti-
GPVI antibody. In another embodiment, said polynucleotide encodes the LC of an
anti-GPVI antibody. The invention also provides vectors comprising polynucleotide
molecules encoding, fusion proteins, modified antibodies, antibody fragments, and
probes thereof.
A vector of the invention contains polynucleotides of SEQ ID NO.5 or SEQ ID NO.7
or any polynucleotide encoding a polypeptide selected from the group consisting in
SEQ ID NO.6, SEQ ID NO.8, SEQ ID NO.9, SEQ ID NO.10, SEQ ID NO.11, SEQ ID
NO.12, SEQ ID NO.13, SEQ ID NO.14, SEQ ID NO.15, SEQ ID NO.16, and SEQ ID
NO.17 or a sequence having at least at least 80%, 85%, 90%, 95% or 99% identity
with these sequences but which retains the same activity as the said monoclonal
antibody.
In order to express the HC, fragment of HC and/or LC of the anti-GPVI antibodies or
Fab of the invention, the polynucleotides encoding said HC and/or LC are inserted
into expression vectors such that the genes are operatively linked to transcriptional
and translational sequences. Expression vectors include plasmids, YACs, cosmids,
retrovirus, EBV-derived episomes, and all the other vectors that the skilled man will
know to be convenient for ensuring the expression of said heavy and/or light chains.
The skilled man will realize that the polynucleotides encoding the HC/ or HC
fragment and the LC can be cloned into different vectors or in the same vector. In a
preferred embodiment, said polynucleotides are cloned in the same vector.
Polynucleotides of the invention and vectors comprising these molecules can be
used for the transformation of a suitable mammalian or microbial host cell.
Transformation can be by any known method for introducing polynucleotides into a
cell host. Such methods are well known of the man skilled in the art and include
dextran-mediated transformation, calcium phosphate precipitation, polybrene-
mediated transfection, protoplast fusion, electroporation, encapsulation of the
polynucleotide into liposomes, biolistic injection and direct microinjection of DNA into
nuclei.
The humanized and engineered variants of the invention are fully functional to
antagonize the GPVI pathway activities. In particular, they are able to inhibit the
binding of collagen to GPVI. They are also able to inhibit platelet aggregation
induced by collagen both in human platelet rich plasma and in human whole blood.
Further, they are also able to inhibit thrombus formation under flow on a collagen
coated surface.
The binding of anti-GPVI Fab fragments of the invention to GPVI is characterized by
the development of a GPVI depletion phenotype on the platelet surface. This is a
new and unexpected mechanism of action for anti-GPVI Fab fragments. This unique
characteristic of the Fab fragments described here, may be determined by the new
epitope targeted by these molecules.
As used herein, "GPVI depletion phenotype" refers to the cellular state that results
when an antibody or an antigen binding fragment binds to GPVI receptor molecules
on the surface of a platelet, it prevents , the activation of the GPVI pathway on
platelets. This GPVI depletion phenotype may rely on two different mechanisms: (i)
either GPVI receptors are removed or depleted from the cell surface, or (ii) the
antibody or an antigen binding fragment binds to GPVI receptor molecule in a non
reversible way so that inhibition of the receptor is maintained for the life span of the
cell.. The phenotype will in both situations revert only with the renewal of platelets.
The fact that these anti-GPVI Fab fragments induce a GPVI depletion phenotype
presents two advantages in term of efficacy:
i) If receptor shedding is - at least part of the mechanism, the process is generally
independent of Fab binding to GPVI. This means if occupation of a fraction of cell
surface GPVI by the Fab (e.g. 30%) will induce shedding, the shedding process will
not be restricted to the 30% of GPVI receptors occupied by the Fab but also
extended to free GPVI. This implies that one can achieve 100% of GPVI blockage
(by depletion) with significantly less Fab.
ii) The inhibition effect has a long term effect: It is known that Fab fragments have a
very short plasma half life of approximatively 1-2 hours, which may be problematic
for several indications where a long lasting inhibition of the target is required. With
the described mechanism of depletion or non-reversible occupancy of the receptors,
the duration of the effect (based on pharmacodynamics properties) will be uncoupled
from the pharmacokinetics properties of the Fab. This is because platelets are not
able to replace the affected receptors, once depletion or non-reversible inhibition is
inducedThen the receptors remain absent or unavailable for the life span of the
platelet. That corresponds to an extension of the duration of action to several days,
depending on the half life of platelets (10 days for human platelets).
All these properties demonstrate that Fab of the present invention are suitable
candidates to the treatment of thrombotic and vascular diseases.
Important target classes for antibody based biologies are membrane receptors that
can be blocked with high specificity by monoclonal antibodies. Full length antibodies
in the IgG format bind and also block their target through their Fv part. The Fc part
add further functionality to these molecules that lead to antibody derived cellular
cytotoxicity (ADCC) or complement dependent cytotoxicity (CDC). These activities
are often an important part of the mAb mode of action (MoA) especially for oncology
applications. However, in other application areas ADCC/CDC activities need to be
minimized or are not necessary. Therefore, a number of antibody fragments lacking
the Fc part such as Fab molecules are desireable.
Another reason for developing monovalent antibody fragments derives from the
target biology. Many membrane receptors initiate signaling after receptor clustering
(e.g. GPVI). If such a receptor should be blocked, a monovalent antibody fragment is
the natural format of choice.
Antibody fragments such as Fab fragments are much more suitable formats than
whole antibodies for many clinical applications. This advantage relies on several
unique characteristic of Fab fragments which differentiate them from whole
antibodies, in particular:
- monovalancy: required if cross-liniking of molecules should be avoided
lack of Fc domain: required if Fc-functions are not needed or undesireable
- lower molecular weight: determines pharmacokinetic and pharmacodynamic
behavior as well as tissue distribution
In order to maintain these desired properties of antibody Fab fragments following the
administration to a patient, interaction of Fab fragments with non-target molecules
should be avoided.
As previously mentioned, the Fab fragments are obtained by deletion of the constant
domain of the heavy chain (Fc part) of IgG molecule. Unfortunately, removal of the
Fc exposes a novel C-terminal (or C-terminal) peptide, i.e. a neoepitope. A specific
class of proteins, able to specifically interact with the C-terminal extremity of Fab
fragments has been reported in humans and have been identified as preexisting
antibodies against antibody fragments such as Fab (Kormeier et al. (1968) J.
Immunol. 100(3);612-21; Persselin and Stevens (1985) J. Clin.lnvest. 76; 723-
30).These preexisting antibodies are being formed early in life and are patient
dependant. With the development of monoclonal antibodies and antibody fragments
as novel therapeutics, the existence of preexisting anti-Fab antibodies restricts and
complicates the usage of therapeutic Fab molecules for the reasons mentined
above.
In addition, the C-terminal extremity of the Fab is a preferred epitope for generation
of antibodies as previously described (Christopoulos C, et al 1994). However, even
in cases were the patient has no or low amounts of pre-existing antibodies against
the C-terminus, the generation of new antibodies directed toward this epitope may
appear after therapeutic Fab administration, leading to this same limitations as
described with preexisting antibodies.
Indeed, binding of therapeutic Fab molecule by preexisting or new antibodies against
Fab C-terminal epitope can change the pharmacokinetic and pharmacodynamic
behavior of the molecules (e.g. receptor activation instead of inhibition based on the
change from a monovalent to bivalent molecules), create new complexes and
functions (e.g. adding an antibody Fc-part with all its effector functions) and changes
the size of the complex which may also have consequences for tissue distribution.
Therefore, this phenomenon represents a significant safety and efficacy risk which
needs to be avoided.
In order to avoid such limitations, the neoepitope that is recognized by anti-Fab
antibodies can be masked by addition of a molecule at the C-terminal end of the Fab
as presently described for anti-GPVI Fab molecules. This principle is essential for
development of all therapeutic antibody fragments in which the target biology
prevents the use of bivalent molecules (e.g. in case of receptor activation through
clustering) and where it is therefore strictly required to employ mono-valent
molecules.
Therefore, in order to enable the safest treatment for all patients as well as to avoid
patient specific pharmacokinetic and pharmacodynamic variability, Fab molecules
should be modified to mask Fab specific neoepitopes created at the C-terminal
extremity that could be recognized by preexisting or newly generated antibodies
against Fab fragments.
This invention concerns a Fab fragment where a molecule has been added to the C-
terminus. This molecule can be a peptide or any other kind of molecule, able to mask
the epitope but without interfering with the binding of the Fab to the target.
In a particular embodiement, the said molecule able to mask the neoepitope of the
Fab fragment is a peptide which can comprise 1 to 100 amino acids, 1 to 50 amino
acids, 1 to 20 amino acids, 1 to 15 amino acids or 1 to 10 amino acids.
For example, a His-tag, a G4S or (G4S)2 stretches peptide will constitute an
appropriate molecule.
In another aspect, the invention concerns a Fab bearing a molecule at the C-terminal
extremity of its heavy chain.
In particular embodiment, the Fab fragment bearing a molecule at its C-terminal
extremity specifically recognizes GPVI. In a prefered embodiement, this Fab
fragment is chosen among those previously described.
In a particular embodiement, the sequence of the Fab heavy chain corresponds to a
sequence comprising SEQ ID NO. 29 or SEQ ID NO.30 or SEQ ID NO.31.
In another enbodiement, the sequence of the Fab heavy chain corresponds to a
sequence consisting in SEQ ID NO. 29 or SEQ ID NO.30 or SEQ ID NO.31.
This invention also concerns a method to prevent recognition of Fab fragments by
preexisting antibodies or new antibodies directed toward the C-terminal extremity
consisting in masking the C-terminal end by addition of a molecule. As such, this
molecule allows prevention of unwanted generation of Fab-antibody complexes.
This invention is also directed toward a method of prevention of platelet activation
when an anti-GPVI Fab is used consisting in masking the C-terminal extremity of the
Fab by addition of a molecule.
In another aspect, the invention concerns a method for preparation of a modified Fab
bearing a molecule at its C-terminal extremity comprising the steps of:
a. Addition of a molecule to the C-terminal extremity of the Fab
b. Production of the modified Fab in an appropriate system, including in
bacteria, yeast or mammalian cell lines
c. Purification of the modified Fab
The produced Fab can futher be formulated in an appropriate solution.
In another aspect, the inventions consists in the use of an anti-GPVI antibody, in
particular Fab fragments as described in the present description, to prevent
thrombotic events in the treatment of certain clinical indications, as for example
acute coronary syndrome, percutaneous coronary intervention, ischemic stroke,
carotid artery stenosis or peripheral arterial occlusive disease. Furthermore it could
be used for the prevention of restenosis and atherosclerosis.
The invention concerns also a composition containing an anti-GPVI antibody of the
invention, and in particular Fab fragments with appropriate excipients. This
composition can be used to treat thrombotic and vascular diseases.
In another aspect, the invention concerns a method of manufacture of an antibody
according to the invention
In another aspect of the invention, the antibodies can be used for diagnosis of GPVI
expression changes. It is described that changes in the expression of GPVI on the
platelet surface as well as the occurrence and concentration of soluble GPVI
(cleaved extracellular domain of GPVI) in plasma may well be associated with
pathophysiological conditions such as acute coronary syndromes, transient ischemic
attacks or stroke (Bigalke B, et al., Eur J Neurol., 2009Jul 21 ; Bigalke B. et al.,
Semin Thromb Hemost. 2007 Man33(2):179-84).
Thus, measurement of these parameters could be used to identify patients at risk for
the aforementioned conditions requiring anti-thrombotic treatment and being possibly
particularly susceptible for anti-GPVI treatment. Therefore, antibodies and antibody
fragments described here can be used as diagnostic tool and be part of a diagnostic
kit which determines the presence and quantitative changes of GPVI on the platelet
surface as well as in plasma samples.
The antibodies and fragments thereof can be used to diagnosis patients at risk who
could benefit of an anti-thrombotic treatment.
Such method for diagnosing of GPVI changes in a patient, may comprise (i)
contacting platelets or plasma sample of said patient with an antibody or Fab
fragments thereof according to the invention, (ii) measuring the binding of said
antibody or Fab to the cells present in said sample, and (iii) comparing the binding
measured in step (ii) with that of a normal reference subject or standard.
The invention also encompasses a diagnosis kit for the detection of changes in the
human GPVI expression including an antibody of the invention or a fragment thereof.
In a particular embodiment, a kit according to the invention can be provided as an
ELISA assay kit.
In another aspect of the invention, patients are screened for the presence of anti-Fab
antibodies prior to administration of either a masked Fab or other antibody of the
invention.
The invention also provides a method for the preparation of anti-GPVl antibody or
Fab fragments of the invention comprising the steps of:
a. Culture of a cell line containing DNA sequences encoding one HC or HC
fragment and one LC of the invention
b. Purification of the antibody or Fab expressed in the culture medium
c. Formulation of the antibody in a convenient form
Any expression cell line able to produce immonuglobulin can be used in order to
express the antibodies of the invention. Expression cell lines derived from
mammalian can be used as well as any other expression system as for example
yeast cells (A. H. Horwitz, et al PNAS. 1988 Nov; 85(22): 8678-82) or bacterial cells
(Chiba Y, Jigami Y. Curr Opin Chem Biol. 2007 Dec; 11(6): 670-6).
The purification of the antibody can be realized by any method known by the person
skilled in the art as for example.
The formulation of the antibody depends on the intended use of such antibody.
Depending on the use, for example pharmaceutical use, veterinary or diagnosis use,
it can be lyophilized or not, be solubilized in an appropriate solution.
It should be noticed that this general description as well as the following detailed
description are exemplary and explanatory only and should not be restrictive on the
invention. The drawings included in the description illustrate several embodiments of
the invention intended to explain the principles of the invention.
BRIEF DESCRIPTION OF THE FIGURES
Figure 1: Inhibition of collagen binding by competition ELISA. (A) Dose-dependency
(1C50) against hGPVI-Fc and (B) against Maccaca GPVI-Fc
Figure 2: Cross-reactivity analysis of selected hybridoma mAb against human and
primate GPVI by ELISA
Figure 3: Fab produced by Ficin cleavage of IgG from hybridoma clone 390
Figure 4: Inhibition of collagen induced platelet aggregation by anti-GPVI Fab
fragments (VL1 withVH3).
Figure 5: Crystallographic data showing the interaction between the Fab and the
extracellular domain of the human GPVI protein. (A) Interaction between D2 domain
of GPVI protein and the Fab. (B) Interaction between D1 domain of GPVI protein and
the Fab. (C) Interaction between D1-D2 domain of GPVI protein and the Fab.
Figure 6: Examples of IC50 curves for the inhibitory activity of anti-GPVI Fab-
fragments in collagen induced whole blood aggregation.
Figure 7: Examples of inhibition of thrombus formation under flow by anti GPVI-Fab
fragments.
Figure 8: Anti-GPVI Fab causes a reduction of GPVI surface expression in a
concentration dependent manner.
Figure 9: Anti-GPVI Fab causes a significant GPVI depletion in a time dependent
manner.
Figure 10: Effect of anti-GPVl Fab on ex vivo whole blood platelet aggregation
(agonist: 1 µg/ml). A - after 1 mg/kg IV bolus administration. B - after 3 mg/kg IV
bolus administration.
Figure 11: In vitro activity of anti-GPVl Fab in whole blood platelet aggregation
assay (agonist: 1 µg/ml)
Figure 12: Effect IV bolus administration of anti-GPVl Fab on platelet count. A- After
PBS vehicule at 3ml/kg (control). B- After 0.01 mg/kg of anti-GPVl FAb. C- After 0.1
mg/kg of anti-GPVl FAb. D - After 1 mg/kg of anti-GPVl FAb. E - After 3 mg/kg of
anti-GPVl FAb.
Figure 13: FACS analysis of GPVI receptor expression on platelets after iv
administration of anti-GPVl Fab. The vertical line descriminates between GPVI
negatives platelets on the left side and GPVI positive platelets on the right side. A -
pre-dose. B - at 24h after administration. C - at 48h after administration. D - at 72h
after administration. E - at 150h after administration.
Figure 14: Effect of anti-GPVl Fab on platelet activation. In control panel, addition of
convulxin. In panel A, addition of Fab after IgG depletion. In panel B, addition of
modified Fab.
Figure 15: A - Fab control; B - Fab-His-Tag ; C - FabGly4Ser and Fab-(Gly4Ser)2
Figure 16: GPVI with the 20 residues signal sequence removed (corresponds to
SEQ ID NO: 32). Bolded residues represent the conformational epitope making
contact with antibody CDRs
EXAMPLES
Example 1: Generation of recombinant extracellular D1 and D2 domain of GPVI
A- Construction of hGPVI-hFcfusion expression olasmid (hGPVI-hFc)
Using human cDNA containing plasmids as a PCR template, a DNA fragment
encoding a 237 amino acid residue heavy chain constant region including the hinge
region, CH2 and CH3 domains of human immunoglobulin IgG was amplified.
Using human genomic DNA as PCR template, a DNA fragment encoding a 205
amino acid residue human GPV1 D1 and D2 extracellular domains was amplified.
This human GPVI D1 and D2 fragment includes the signal sequence and
corresponds to amino acids M1 to T205 in the wild type protein (NP_057447 /
Q9HCN6). The resulting amplified, cleaved and purified PCR products encoding
human GPVI D1 and D2 and a human Fc region were combined by ligation PCR and
ligated into baculovirus expression vector pVL1393 by InFusion method using EcoRI
and Notl site. The resulting GPVI-Fc ORF is listed as SEQ ID NO.1 and its
corresponding protein sequence as SEQ ID NO.2.
B - Construction of GPVI-tev-his expression plasmid (GPVI-tev-his)
Using the previously described GPVI-Fc containing plasmid as PCR template,
human GPVI D1 and D2 extracellular domains were amplified, including the signal
sequence corresponding to amino acids M1 to T205 in the wild type protein
(Swissprot: Q9HCN6). The reverse primer contained DNA coding for a tev cleavage
recognition sequence and 7 histidine residues representing the His-tag at the C-
terminus of the construct. The resulting amplified PCR fragment was cleaved with
the restriction endonucleases EcoRI + Notl and was ligated into baculovirus
expression vector pVL1393. The resulting ORF is listed as SEQ ID NO.3 and its
corresponding protein sequence as SEQ ID NO.4.
C - Expression and purification of hGPVI-hFc and GPVI-tev-his protein
SF9 cells growing in SF900 II serum free suspension culture (Invitrogen) were
cotransfected with the expression plasmid and FlashBac baculovirus DNA (Oxford
Expression Technologies). Transfection was performed using Cellfectin transfection
reagent (Invitrogen). After 5h, 5% total bovine fetal serum was added to the
transfected culture. The cells were cultured at 28°C for 5 days. The culture
supernatant containing recombinant virus was used for larger scale virus
amplification in SF900 II suspension culture containing 5% bovine fetal serum.
For protein expression HighFive (Invitrogen) cells growing in ExCell 405 (SAFC)
were transduced with an appropiate amount of virus stock and grown at 27°C for 72
h. After harvest the cell culture supernatant was clarified by filtrated (0.22pm).
For purification the Fc-fusion GPVI protein variants were captured on protein A
matrix (GE Healthcare) and eluted by pH shift. After polishing the protein by SEC
using a Superdex 200 (GE Healthcare) and a final ultrafiltration concentration step
the protein was used for ELISA and futher assays.
For purification the His-tagged GPVI protein variants were directly captured from
supernatant on IMAC matrix (HisTrap, GE Healthcare) and eluted by an imidazol
gradient. After polishing the protein by SEC using a Superdex 75 (GE Healthcare)
and ultrafiltration the protein was used in indicated experiments.
Example 2: Generation and selection of functional anti-GPVI mAbs
A - Generation of anti-GPVI mAbs
GPVI-specific antibodies were generated using the RIMMS method as described by
Kilpatrick et al. (1997. Hybridoma 16: 381389).
6-8 weeks old female BALB/c mice (S082342; Charles River Labs, Bar Harbor, ME)
each received four rounds of immunization with purified soluble his-tagged GPVI
protein (prepared as described in Example 1) over a course of 14 days at intervals of
3-4 days.
For the first immunization on day zero, 5 ug antigen emulsified in Titermax's adjuvant
(TierMax Gold Adjuvant; Sigma #T2684) was administered subcutaneously to six
sites proximal to draining lymph nodes, along the back of the mice. Another 5 pg of
antigen emulsified in RIBl's adjuvant (Sigma Adjuvant system; Sigma #S6322) was
administered to six juxtaposed sites along abdomen. Booster immunizations were
given on days 4, 7 and 11 in a similar fashion.
Four days after the last injection, mice were sacrified. Bilateral popliteal, superficial
inguinal, axillary and branchial lymph nodes were isolated aseptically and washed
with fresh RPMl medium. Lymphocytes were released from the lymph nodes and the
resulting single-cell suspension was washed twice with RPMl medium before being
fused with P3X63-AG8.653 myeloma cells using polyethylene glycol. After fusion,
the cell mixture was incubated in an incubator at 37° C for 16-24 hours. The resulting
cells preparation was transferred into selective semi-solid medium and aseptically
plated out into 100 mm Petri plates and incubated at 37°C.
Ten days after initiation of selection, the plates were examined for hybridoma growth,
and visible colonies were picked-up and placed into 96-well plates containing 200 uL
of growth medium. The 96-well plates were kept in an incubator at 37° C for 2 to 4
days.
B - Screening for mAbs recognizing the human GPVI protein
Primary screening for anti-GPVI IgG production was performed by ELISA using
GPVI-Fc fusion protein (prepared as described in Example 1) as capturing antigen.
Plates were coated with GPVI-Fc fusion protein and hybridoma supematants were
added to the plate and detected by using rabbit anti-mouse IgG conjugated with
horseradish peroxidase (Sigma; #A9044). Antibody binding was visualized by adding
TMB-H202 buffer and read at a wavelength of 450 nm.
Among 367 hybridomas selected from 96-well plates, 129 hybridomas were positive
for anti-human GPVI antibody production and then 111 were confirmed after cell
amplification.
C - Ability of anti-GPVI mAbs to block binding of collagen to human GPVI
A secondary screening was performed to characterize functional properties of all
human GPVI-specific mAbs for their ability to block the binding of collagen to human
GPVI-Fc fusion protein in a competition ELISA binding assay. Custom-made
collagen coated 96- well plates (Pierce) were used. Pre-incubated mixture of human
GPVI-Fc fusion protein and hybridoma supematants were added to the plate and
collagen-human GPVI-Fc complex was detected by using goat anti-human IgG-Fc
conjugated with horseradish peroxidase (Sigma; #A0170). The antibody binding was
visualized by adding TMB-H202 buffer and read at a wavelength of 450 nm. Among
the 100 GPVI-binding hybridomas, 22 hybridomas blocked the binding of GPVI-Fc to
collagen (threshold: 90% inhibition). The blocking properties of pre-selected mAbs to
human GPVI were confirmed to be dose-dependent using a competition ELISA
assay as shown in Figures 1A.
All antagonist mAbs were isotyped as lgG1, kappa, as determined by using a mouse
IgG isotyping kit (SEROTEC; #MMT1) (data not shown).
D - Binding properties of the anti-GPVI mAbs
A last screening was performed by Surface Plasmon Resonance (BIAcore 2000) to
evaluate binding properties of GPVI blocking antibodies. In this analysis, we
evaluated the interaction of the human GPVI protein with the anti-human GPVI mAbs
fixed to anti-Fc antibody covalently linked to CM chips. Binding kinetics of the
individual mAbs were performed using the protocol described by Canziani et at 2004.
Anal. Biochem. 325 : 301-307.
All blocking mAbs displayed affinities in the (sub)nanomolar range to human GPVI
(Table 1). Based on target affinities, the antibody from hybridoma clone 390 was
selected for further development.
Table 1: Affinity and association/dissociation rates of selected anti-hGPVI mAbs

E - Cross-reactivity properties of the anti-GPVI mAbs with the primate GPVI protein
GPVI-specific mAbs listed in table 1 were assessed for their ability to bind primate
GPVI-Fc protein by ELISA. Plates were coated with primate GPVI-Fc fusion protein,
anti-hGPVI mAbs were added to the plate and detected with rabbit anti-mouse IgG
conjugated with horseradish peroxidase (Sigma; #A9044). The antibody binding was
visualized by adding TMB-H202 buffer and read at a wavelength of 450 nm.
The pre-selected hybridomas were cloned by growth in semi-solid medium. Petri
plates were seeded at 125 cell/mL and clones showing significant growth were
screened for GPVI binding, GPVI blocking activity and cross reactivity with primate
GPVI. As shown in Figure 1B and Figure 2, all mAbs were cross-reactive with
primate GPVI.
The sequences used for extracellular domain of primate {Macaca fascicularis or
cynomolgus) GPVI are described as SEQ ID N0.27 and SEQ ID N0.28.
F - Determination of the sequence of the heavy and light chains of the anti-GPVI
mAbs
The cDNA encoding the variable domains of the monoclonal antibodies were
obtained as follows: mRNA was extracted from hybridoma cells with the Oligotex kit
from Qiagen. The corresponding cDNA was amplified by RT-PCR by the RACE
method utilizing the Gene Racer kit (Invitrogen), the transcriptase Superscript III at
55 °C (Invitrogen) and primers described on Table 2 (RACEMOG1 or CKFOR). The
cDNA fragments were amplified by PCR with the polymerase Phusion at 55 °C
(Finnzymes) and primers also described in Table 2.
Table 2:. Primers used for RT-PCR and PCR

The amplified fragments encoding the variable regions of heavy (VH) and light (VL)
chains were cloned into the pGEM-T Easy plasmid from Promega or pCR4-Topo
plasmid from Invitrogen which were amplified in E. coli. Cloned cDNA was then
sequenced on both strands.
Protein sequences were translated from plasmid coding sequences and the masses
of the heavy (HC) and light (LC) chains were calculated (Table 3). The values
obtained were in perfect agreement with mass spectrometry data obtained from
preparations of mAbs purified from cultures of the corresponding hybridoma, see
Table 3. In particular, the occupancy of a N-glycosylation site in the variable region
of the heavy chain (NST) of anti-GPVI antibody 390 was confirmed. Amino acid
sequence of HC and LC are reported in the sequence listing as follows : SEQ ID
NO.5 and SEQ ID NO. 6 corresponds to the HC of anti-GPVI mAb from clone 390
and SEQ ID NO.7 and SEQ ID NO. 8 correspond to the LC of anti-GPVI mAb from
clone 390.
Table 3: Mass spectrometry analysis of Anti-GPVI mAbs from hybridoma

* Compatible with additional high mannose N-glycan. Mass of 49586 Da confirmed
by LC/MS analysis after deglycosylation.
G - Determination of the sequences of the CDR of the anti-GPVI mAbs
The sequences for the CDR regions were deduced from the protein sequence using
the KABAT nomenclature.
For the HC, CDR1 corresponds to SEQ ID NO.18, CDR2 corresponds to SEQ ID
NO.19 CDR3 corresponds to SEQ ID NO.20
For the LC, CDR1 corresponds to SEQ ID NO.21, CDR2 corresponds to SEQ ID
N0.22 CDR3 corresponds to SEQ ID N0.23
Example 3: Preparation and biophysical properties of anti-GPVI mAbs and their Fab
fragments
A - Inhibition of collagen binding to recombinant GPVI by anti-GPVi IqG's or their
proteolytic Fab fragments
Collagen coated 384well plates (Pierce) were blocked with 3% BSA for 2h.
Increasing concentrations of IgG or its corresponding Fab fragment produced by
Ficin cleavage (0,3 - 20 ug/ml) were incubated with recombinant GPVI (Fc-fusion
protein of the extracellular GPVI domain; 3ug/ml) for 30min. The GPVI-IgG mixture
was added to collagen coated plates and incubated for 1h at RT. Next, 384well
plates were washed (DELFIA wash buffer, Perkin Elmer) five times and Eu-labelled
anti-human IgG (100ng/ml Perkin Elmer) was added. Following 1h incubation at
room temperature plates were washed again five times, enhancement solution
(Perkin Elmer) was added and incubated for 10min. Fluorescence was detected at
360/612nrn using a Tecan Ultra reader. Figure 3 shows a typical readout. Table 4
shows calculated IC50 values (ug/ml) for inhibition of collagen binding to GPVI of two
independent experiments.
Table 4: Calculated IC50 values (ug/ml) for inhibition of collagen binding to GPVI of
two independent experiments.

B - Production of recombinant Fab fragments
1. Construction of expression plasmids for recombinant production of anti-GPVI Fab
Amino acid sequences of the variable heavy and light chains of the anti-human GPVI
antibodies were backtranslated into nucleotide sequence and generated respectively
using either a modified protocol of the overlap extension PCR (OE-PCR) described
by Young L. and Dong Q. (Nucl. Acids Res. (2004), 32(7), e59) or by gene synthesis
(Geneart). PCR products were cloned into the pCR Blunt II TOPO vector using the
Invitrogen TOPO cloning kit and sequenced using M13forward and M13 reverse
primers. Each variable heavy chain was fused to the CH1 domain of IGHG1
(Genebank accession number Q569F4) and the variable light chain was fused to the
constant kappa chain (IGKC, Genebank accession number Q502W4) respectively.
These fragments were digested with Nhel and Hindlll and each ligated into the
Nhel/Hindlll sites of the episomal expression vector pXL, an analogon of the pTT
vector described by Durocher et al. (2002), Nucl. Acids Res. 30(2), E9, creating the
plasmids for transient mammalian expression of the chimeric and humanized anti-
GPVI Fabs. The expression plasmids encoding the heavy and light chain of the
antibody were propagated in E.coli NEB 10-beta (DH10B derivative). Plasmids used
for transfection were prepared from E.coli using the Qiagen EndoFree Plasmid Mega
Kit.
2. Transient expression and purification of recombinant anti-GPVI Fab fragments
Hek 293-FS cells growing in Freestyle Medium (Invitrogen) were transfected with
indicated LC and HC plasmids using Fugene (Roche) transfection reagent. After 7
days the cells were removed by centrifugation, 10% Vol/Vol IMTris HCI pH 8,0
was added and the supernatant was passed over a 0.22µm filter to remove particles.
The Fab proteins were captured using KappaSelect matrix (GE Healthcare) and
eluted via pH shift. The protein containing fractions were pooled and desalted using
PD-10 or Sephadex columns. Concentrated and sterile filtered (0.22 urn) protein
solutions were adjusted to 1 mg/ml and kept at 4°C until use.
C - Biophysical characterization of recombinant Fab
Surface plasmon resonance technology on a Biacore 3000 (GE Healthcare) was
used for detailed kinetic characterization of purified antibody fragments. A direct
binding assay was used with the anti-GPVI Fab as the ligand and human GPVI as
analyte. Typically, 500 RU of anti-GPVI Fab were immobilized on a research grade
CM5 chip by amine reactive coupling, resulting in an Rmax of 200 RU for the bound
GPVI molecule. Binding kinetics were measured over a concentration range typically
between 0.8 to 208 nM in HBS-EP (10 mM HEPES pH 7.4, 150 mM NaCI, 3 mM
EDTA, 0.005 % Surfactant P20) at a flow rate of 30 µl/min. Chip surfaces were
regenerated with 10 mM glycine pH 2.0. Kinetic parameters were analyzed and
calculated in the BIAevaluation program package (version 4.1) using a flow cell
without immobilized Fab as reference. A 1:1 binding model with mass transfer was
applied for a global fit of the data for curves.
In order to investigate functional consequences of the additional glycosylation motif
present in the original sequence as derived from glycosylated clone 390 (Fab 390-
G), the conserved glycosylation motif was removed by amino acid replacement (Fab
390-nG). For this purpose, two heavy chain fragments based on clone 390 were
constructed: the first one does not have any modification and corresponds to SEQ ID
NO.9, and the second one has a point mutation n-> Q at position 60 of SEQ ID
NO.9.
It is known that glycosylation can be very heterogeneous and produce glycoprotein
heterogeneity; such source of heterogeneity should be avoided as much as possible.
In the present case, the removal of the additional glycan does not impact the binding
affinity for GPVI as shown in Table 5. Thus, the non-glycosylated Fab 390-nG
construct has been chosen for further experiment.
Table 5: Determination of binding characteristics for recombinant anti-GPVI Fab
molecules by SPR (Biacore)

D - Determination of the epitope recognized bv the anti-GPVI mAb bv co-
crvstallization of anti-GPVI Fab with GPVI and X-rav crystallography
For epitope characterization a co-crystallization strategy has been applied. A
complex between Fab 390-nG (expressed in HEK293 cells) and human glycoprotein
VI (extracellular domain of human GPVI (Met1-Thr205), expressed in Insect cells
High five) was formed by incubation of both proteins at equal molar ratio. The
complex was digested with trypsin, re-purified by gelfiltration, concentrated and
subjected to crystallization screening. Well diffracting crystals were obtained by
mixing 100 nl of protein solution (Fab-antigen complex at 6.5 mg/ml in 20mM Tris pH
8.0,150mM NaCI) with 100 nl of reservoir solution containing 0.1 M BisTris pH 5.5,
25% PEG3350 and 0.2M ammonium acetate and incubating the protein drop against
200 ul reservoir solution at 4°C in a sealed sitting drop vapour diffusion crystallization
plate. One crystal was flash frozen in liquid nitrogen using reservoir solution
supplemented with 25% ethylene glycol for cryoprotection. Crystallographic data was
collected at the European Synchrotron Radiation Facility, Grenoble, France, at
beamline ID14-4. The structure was solved by Molecular Replacement using the Fab
fragment of the pdb-entry 1FDL as model for Fab 390-nG and one monomer of the
crystal structure of human platelet glycoprotein VI, pdb-entry 2GI7, as model for
GPVI. The resolution and final R-factor of the refined structure are 1.72A, R-factor
17.2% and R-free 19.9%. The GPVI-binding epitope is a conformational epitope
characterized by interactions (defined as distance < 4.5 A) of Fab 390-nG to the
following residues of the antigen hGPVI:
Pro4, Lys7, Glu40, Lys41, Leu42, Ser43, Ser44, Ser45, Arg46, Arg65, Arg67,
Trp76, Leu78, Pro79, Ser80, Asp81, Gln82, Ser165, Arg166
These residues belong either to the D1 domain (up to amino acid 84) or constitute
parts of the D2 domain (amino acids 94 to 176) and therefore representing a
conformational epitope. The numbering system for GPVI epitope residues
corresponds to the protein with the signal sequence removed as shown in Figure 16
and SEQ ID NO: 32. The interaction of Fab anti GPVI from clone 390 with the
extracellar D1-D2 domain of GPVI is illustrated on Figure 5.
The antibody residues contacting GPVI epitope residues are described in Tables 6
through Table 11 below.
As used herein, "contacting residues" were defined as antibody residues with a
distance of less than 4.5 A to GPVI antigen residues or the reverse and as measured
in the crystal structure between antibody to GPVI and GPVI antigen. The distances
were determined using the software program NCONT of the CCP4 Software
package.
The X-ray structure highlights residues from the CDRs that can be mutated, that
should not impact binding to GPVI.
In the following descriptions, residues from the CDRs might be modified as indicated,
without disruption of the binding to GPVI. It is also understood that mutations in the
CDRs should not affect the conformation and orientation of the residues from the
CDRs to preserve binding to GPVI. Residue numbering is sequential and does not
follow Kabbat conventions as used in Al-Lazikani ((1997) J. Mol. Biol. 273, 927-948).
"X" represents "any residue", while "-" indicates that no modification should be made
at this position.
CDRH1 sequence as defined in SEQ ID NO: 18 spans ten residues. As shown in
Table 6, six of ten or 60% of CDRH1 residues do not make contact with GPVI. All of
the GPVI residues in contact with antibody residues are serine residues.
Table 6: Antibody CDRH1 residues contacting GPVI antigen residues

CDRH2 sequence is defined in SEQ ID NO: 19 and spans sixteen antibody residues.
As shown in Table 7, nine of sixteen or 56% of CDRH2 residues do not make
contact with GPVI. Tryptophan 52 was found to be in contact with seven different
GPVI antigen residues. Arginine 46 of GPVI was found in contact with five different
CDRH2 residues.
Table 7: Antibody CDRH2 residues contacting GPVI antigen residues

CDRH3 sequence is defined in SEQ ID NO: 20 and spans five antibody residues.
Based on the fact the definition of the CDR sequence can slightly vary depending on
the software used, it has been found that CDRH3 may include one additional
residue. Thus CDRH3 can be as defined in SEQ ID NO:38, spanning six antibody
residues. As shown in Table 8 two of six or 33% of CDRH3 residues do not make
contact with GPVI. Arginine 65 was found to be in contact with three different GPV1
antigen residues. There are three different arginine residues of GPVI contacting
CDRH3 residues.
Table 8: Antibody CDRH3 residues contacting GPVI antigen residues
CDRL1 sequence is defined in SEQ ID NO: 21 and spans eleven antibody residues.
As shown in Table 9, seven of eleven or 64% of CDRL1 residues do not make
contact with GPVI. Tyrosine 32 was found to be in contact with five different GPVI
residues.
CDRL2 sequence is defined in SEQ ID NO: 22 and spans sevenantibody residues.
Based on the fact the definition of the CDR sequence can slightly vary depending on
the software used, it has been found that CDRL2 may include one additional residue.
Thus CDRL2 can be as defined in SEQ ID NO: 39 spanning eight antibody residues.
As shown in Table 10, six of eight or 75% of CDRL2 residues do not make contact
with GPVI. This CDR is in contact with residues in both D1 and D2 domains of
GPVI. Tyrosine 50 is in contact with asp 81 and gin 82 in D1, while proline 56 is in
contact with arg 166 and ser 165 of GPVI D2 domain.
Table 10: Antibody CDRL2 residues contacting GPVI antigen residues

CDRL3 sequence is defined in SEQ ID NO: 23 and spans eight antibody residues.
As shown in Table 11, five of eight or 63% of CDRL3 residues do not make contact
with GPVI.
Table 11: Antibody CDRL3 residues contacting GPVI antigen residues

A summary of the contacts between GPVI residues interacting with antibody
residues revealed that ser 43 has 8 contacts, ser 44 has 7 contacts, arg 46 has 4
contacts, arg 67 has 4 contacts, arg 65 has 3 contacts, asp 81 has 3 contacts, and
Arg 166 has two contacts. Most GPVI residues favor either the fight chain or the
heavy chain, but R67 and R166 have contact with both.
E - Humanization of the Fv domain of anti-GPVI Fab
The humanization protocol described in the patent application WO 2009/032661 was
used to humanize the Fab 390-nG clone. In addition, analysis of the crystallographic
complex between Fab 390-nG and the human GPVI led to several mutations with the
aim of improving or at least retaining the affinity of Fab 390-nG to human GPVI
within the humanization campaign.
The VL & VH sequences of Fab 390-nG were blasted against the August 2007
version of the Protein Data Bank (PDB). The PDB structure 1L7I was used to build
up a homology model of the light chain. The PDB structures 1YY8 and 1ZA6 were
both used to build up a homology model of the heavy chain. The 1ZA6 structure was
specifically used to model the CDR3 subregion, while the 1YY8 structure was used
for the remaining part of the heavy chain. The resulting VL and VH models were
used to build up a homology model of the variable domains which was subsequently
energy minimized using the standard procedure implemented in Molecular Operating
Environment (MOE). MOE is a software for computer assisted drug design
distributed by the Chemical Computing group. The minimized model of Fab 390-nG
was subsequently submitted to Molecular Dynamics simulations in order to identify
the most flexible residues which are more likely to interact with T-cell receptors and
responsible for activation of the immune response. The simulations were run with the
AMBER software distributed by the University of California. 57 amino-acids are
finally identified as flexible in the Fab 390-nG. The motion of the most 60 flexible Fab
390-nG amino-acids (excluding the CDR+5A region), during the 20 ns (10x2ns),
were then compared to the motion of the corresponding flexible amino-acids of 49
human germlines homology models, for each of which were run the 10x2ns MD
simulations. The 49 human germlines models were built by systematically combining
the 7 most common human germline light chains (vk1, vk2, vk3, vk4, vlambdal,
vlambda2, vlambda3) and 7 most common human germline heavy chains (vhla,
vhlb, vh2, vh3, vh4, vh5, vh6). The vk1-vh6 human germline sequences showed a
84% 4D similarity of its flexible amino-acids compared to the flexible amino-acids of
the Fab 390-nG sequences; the vk1-vh6 germline sequences were therefore used to
humanize the Fab 390-nG sequences focusing on the flexible amino-acids. For the
pairwise amino-acid association between Fab 390-nG and vk1-vh6 amino-acids, the
2 sequences were aligned based on the optimal 3D superposition of the alpha
carbons of the 2 corresponding homology models. The following motifs of potentially
problematic sequences were considered in some cases: Asp-Pro (acide labile bond),
Asn-X-Ser/Thr (glycosylation, X=any amino-acid but Pro), Asp-Gly/Ser/Thr
(succinimide/iso-asp formation in flexible areas), Asn-Gly/His/Ser/Ala/Cys (exposed
deamidation sites), Met (oxidation in exposed area). The resulting engineered
sequences were blasted for sequence similarity against UniProtKB/Swiss-Prot
database providing confidence that reasonable assumption has been made. In
addition none of the sequences contains any known B- or T-cell epitope listed in the
Immune Epitope Database and Analysis Resource (IEDB database).
Five versions for the heavy chain (Fab 390-nG VH1, VH2, VH3, VH4, VH5) and
three versions were suggested for the light chain (VL1, VL2, VL3). All versions derive
from the sequences of the ANTI-GPVI Fab 390-nG construct. The starting
sequences for unhumanized Fab 390-nG are provided in SEQ ID N033: for VH and
SEQ ID NO:34 for VL. The L1 version has 4 mutations. The L2 version includes one
additional mutation (N93H) to potentially improve the binding to human GPVI. The L3
version derives from L1 and includes an additional mutation to potentially improve
the binding to human GPVI (N93L).
Version H1 contains 5 humanizing mutations derived from the closest human heavy
chain germline sequence, VH6, as found following the previously described
procedure. Version H2 contains 4 humanizing mutations and derives from H1 without
mutating the amino-acid in position 73 (Asn73), which was found to be important for
a potent affinity as seen in the crystal I ographic complex between Fab 390-nG and
the human GPVI. Version H3 contains 5 humanizing mutations derived from the
human heavy chain germline sequence VH3. VH3 was found to be close to the VH
chain of Fab 390-nG, following the previously described procedure, and to have an
asparagine (Asn) residue in position 73. The H4 version derives from H2 and
includes one additional mutation (G31Y) to potentially improve the binding to human
GPVI. The H5 version derives from H2 and includes one additional mutation (Y103E)
to potentially improve the binding to human GPVI.
Eight combinations of VL and VH variants were recommended for generation of
engineered antibodies: VL1 with VH1, VL1 with VH2, VL1 with VH3, VL2 with VH2,
VL3 with VH2, VL1 with VH4, VL1 with VH5 and VL3 with VH5. As shown in Table
12 and Table 13, the amino acid changes were made in engineered VL and VH
variants of the Fab 390-nG, using the methodology set forth in the detailed
description section of the instant application. The left column indicates the original
amino acids and their positions in the murine Fab 390-nG.
Besides the engineering of the variable domain, the antibodies and their fragments
can be further modified by the addition of tags or amino acid extensions. For
example a sequence of six or more histidine residues, in particular (His)6, or (His)7, or
(His)e, could be added at the C-terminus of the heavy chain or the terminus could be
extended in sequence by addition of amino acids such as GlyGlyGlyGlySer or
(GlyGlyGlyGlySer)2 units. Similarly the framework of the antibodies and their
fragments could be changed from an lgG1 backbone to another IgG backbone like
lgG4.
Table 12: Summary of the mutations introduced intoSEQID NO: 34 for the
humanized light chain of the anti-GPVI Fab 390-nG construct

41

5 F - Biophysical characterization of humanized variants
Surface plasmon resonance technology on a Biacore 3000 (GE Healthcare) was
used for detailed kinetic characterisation of purified antibody fragments. An assay
with immobilised human GPVI was used. Typically, 300 RU of GPVI were
immobilised on a research grade CM5 chip by amine reactive coupling, resulting in
an Rmax of 200 RU for the bound antibody fragment. Binding kinetics were
measured over a concentration range between 3,2 to 112 nM in HBS-EP (10 mM
HEPES pH 7.4, 150 mM NaCI, 3 mM EDTA, 0.005 % Surfactant P20) at a flow rate
of 30 ul/min. Chip surfaces were regenerated with 10 mM glycine pH2.0. Kinetic
parameters were analysed and calculated in the BIAevaluation program package
(version 4.1) using a flow cell without immobilised GPVI as reference. A 1:1 binding
model with mass transfer was applied for a global fit of the data for curves
corresponding to analyte concentrations from 3-56 nM of Fab fragment.
The binding kinetics of anti-GPVI antibody fragments from the humanisation
campaign are shown in Table 14. Note that the Biacore assay used to measure the
data in this example is different in terms of immobilized binding partner (GPVI versus
the mAb/Fab fragment) and this leads to the different Biacore affinities as reported in
Table 5.
Table 14: Determination of binding characteristics of humanized variants of Fab 390-
nG against extracellular domain of human GPVI by SPR (Biacore)

Example 4 : Inhibition of collagen binding and of platelet aggregation bv anti-
GPVI Fab fraqments.A. Inhibition of binding of recombinant GPVI to collagen by
recombinant anti-GPVI Fab fragments
Collagen coated 384well plates (Pierce) were blocked with 3% BSA for 2h.
Increasing concentrations of anti GPVI Fab fragments (0,3 - 20 pg/ml) were
incubated with recombinant GPVI (Fusion protein of the extracellular domain of GPVI
and Fc-Part of human IgG; 3ug/ml). The GPVI-Fab mixture was added to collagen
coated plates and incubated for 1h at RT. 384well plates were washed (DELFIA
wash buffer, Perkin Elmer) five times and Eu-labelled anti-human IgG (100ng/ml
Perkin Elmer) was added. Following 1h incubation at room temperature plates were
washed again five times, enhancement solution (Perkin Elmer) was added and
incubated for 10min. Fluorescence was detected at 360/612nm using a Tecan Ultra
reader. Shown in Table 15 are examples of the inhibitory effect of recombinantly
produced anti-GPVI Fab fragments on GPVI collagen binding.
Table 15: Measured IC50 values (ug/ml) for the inhibition of collagen binding to
GPVI of three independent experiments.

* Note: This variant corresponds to a non-humanized construct based on the original
clone 390 without the His-tag. Therefore this construct is exactly in the same format
as the humanized variants (DKTHT at the C-terminus of HC) mentioned in the same
table.
B. Inhibition of collagen induced platelet aggregation bv recombinant anti-GPVI Fab
fragments (human platelet rich plasmal
For platelet rich plasma (PRP), blood was collected into syringes containing ACD-A
to a final concentration of 10 %. After centrifugation at 200 x g for 20 min at room
temperature without brake, supernatant (PRP) was separated. Platelet poor plasma
(PPP) was separated from the remaining blood by centrifugation for 10 min at 1500 x
g. PRP was set to a platelet count of 3.0 x 108 cells/ml by dilution with PPP. Fab
fragments were used at final concentration of 0.15-20 ug/ml and incubated with
PRP for 5min at 37°C. Thereafter collagen was added at a concentration of 1-
1,5ug/ml and the aggregation response was monitored by the measurement of light
transmission by either using a 96well plate reader (96well plate aggregation) or Born
aggregometer (classical aggregation). The aggregation response was monitored for
20min. As shown in Table 16 all investigated Fab fragments were able to inhibit
collagen induced platelet aggregation in a concentration dependent manner.
Table 16: Calculated IC50 values (ug/ml) for inhibition of collagen induced platelet
aggregation of three independent experiments.

* Note: This variant corresponds to a non-humanized construct based on the original
clone 390 without the His-tag. Therefore this construct is exactly in the same format
as the humanized variants (DKTHT at the C-terminus of HC) mentioned in the same
table.
C. Inhibition of collagen induced platelet aggregation by anti GPVI Fab fragments
(human whole blood)
For the experiments blood was anticoagulated with 20 µg/ml hirudin and used
immediately. Before measurement whole blood was diluted 1:1 with NaCI. Fab
fragments were used at final concentration of 0.15 - 20 µg/ml and incubated with
blood for 5min at 37°C. Thereafter collagen was added at a concentration of 1 pg/ml
and the aggregation response was monitored by the measurement of the impedance
using Multiplate® analyzer. The reaction was monitored for 6min. The aggregation
response was quantified by the area under the aggregation curve (AUC) as specified
by the manufacturer. As shown in Figure 6 investigated Fab fragments were able to
inhibit collagen induced platelet aggregation in a concentration dependent manner.
P. Inhibition of thrombus formation under flow by anti GPVI Fab fragments
For the experiments blood was anticoagulated with 20 pg/ml hirudin and used
immediately. Rectangular capillary glass microslides with an inner diameter of
1x0.1mm (Camlab, Cambridge, UK) were coated overnight with 100 pg/ml Horm
collagen and blocked with heat-inactivated 0.5 % fatty acid- free BSA at room
temperature for 1 h. Blood platelets were fluorescently labeled with 2 pM DiOC6(3)
and treated with Fab fragments for 5 min at 37 °C. Perfusion through the collagen
coated coverslip was performed for 2 minutes at 2000s-1. After blood perfusion,
Tyrodes buffer 1 was perfused through the microslides for 5 min at the same shear
rate. Thrombus formation was quantified by determination of platelet (thrombus)
surface coverage. For this purpose ten final fluorescence images were recorded
from different areas in the middle of the capillary. Additionally, phase contrast and
DIC pictures were recorded. Imaging recording and analysis was performed using
ImagePro plus imaging software (Mediacy, Silver Spring, USA) connected to a black-
and-white CCD camera (CoolSnap cf, Ropers Scientific GmbH/ Photometries,
Ottobrunn). Figure 7A-E show examples of inhibition of thrombus formation under
flow by anti GPVI-Fab fragments.
E. Effect of the anti-GPVI Fab-fraament in a mouse model of arterial thrombosis
For this in vivo investigation, mice humanized for the GPVI receptor have been used.
The Carotid artery occlusion was photochemically induced resulting in vascular
endothelial injury at the inner vessel side without affecting the outer vessel wall. By
this technique the red dye, rose bengal, is systemically administered and the
endothelium of the carotid artery was irradiated by green laser light resulting in an
"inside-out" injury. The anti GPVI Fab-fragment was administered at 10 mg/kg as an
intravenous bolus via the jugular vein catheter. After a 15 min incubation period the
laser light source was placed 12 cm away from the carotid artery distal to the flow
probe and laser irradiation was started. Blood flow was continuously monitored for
an observation period of 90 min. The measured thrombosis parameters were the
length of time to complete arterial occlusion following vascular injury (time to
occlusion, TTO) and the area under the blood flow curve (AUC).
For thrombosis evaluation, two measured parameters were used : a) time to
occlusion (TTO) and b) the area under the blood flow curve. The time from
thrombotic challenge until vessel occlusion (TTO) was 78 min for anti-GPVI Fab
fragment versus 33 min for the control group resulting in a 2.4-fold increase (table
17). The area under the flow curve was 2.3-fold increased for the group treated with
anti-GPVI Fab fragment compared to the control group (table 18). Intravenous bolus
administration of 10 mg/kg anti-GPVI Fab fragment resulted in a significant
antithrombotic effect in the photochemical induced arterial thrombosis model in
humanized GPVI mice.
Table 17: Thrombosis parameter: Time to occlusion (min)

Table 18: Thrombosis parameter: Area under the blood flow curve

F. GPVl depletion from the platelet surface induced bv anti-GPVI Fab fragments
For the experiments, blood was anticoagulated with 20 µg/ml hirudin and used
immediately. Anti GPVl - Fab fragments were used at indicated concentrations.
Blood or platelet rich plasma (PRP) samples were incubated with the Fab fragments
for 5min, 15min, 1h and 2h, respectively. Thereafter samples were fixed using
paraformaldehyde and GPVl receptor expression was determined. GPVl density on
the platelet surface was measured using a different, fluorescently labelled anti-GPVI
antibody and determined in a flow cytometer (BD LSR II). As a control, it was
previously shown that this antibody is able to bind GPVl independently (and in the
presence) of the investigated Fab fragment.
As shown in Figure 8, the anti-GPVI Fab causes a reduction of GPVl surface
expression in a concentration dependent manner. Further experiments show that 5
min of anti GPVl Fab exposure already caused a significant GPVl depletion at
10µg/ml and 2ug/m (Figure 9). Also lower concentrations of the anti-GPVI Fab were
able to decrease the GPVl surface density, although with a delayed time course.
These results support the fact that the anti-GP VI Fab induces the GP VI depletion at
the platelet surface.
G. In vitro and ex vivo effect of an anti GPVl Fab on collagen-induced whole blood
aggregation in cynomolgus monkey (Macaca fascicularis) with concomitant
hematology assessment.
1. Animal details and dose regimen:
Animal studies were conducted in an AAALAC (Association for Assessment and
Accreditation of Laboratory Animal Care)-accredited facility according to the local
animal welfare regulations and registered by the veterinary authorities. The
species/strain used in the study was cynomolgus monkeys (Macaca fascicularis).
Only female monkeys were used for the study. Two animals per dose group were
studied. Doses studied included phosphate buffered saline (PBS) vehicle control,
0.01, 0.1,1, and 3 mg/kg of anti GPVI Fab all administered at 2 ml/kg IV bolus. The
bodyweight of the included monkeys ranged between 3.52 and 7.34 kg. One of the
animals (monkey D) used in the 1 mg/kg dose group did not show any platelet
aggregation response at pre-dose sample. Hence, no calculation of intra-individual
relative change of aggregation response (inhibition of aggregation in % compared to
pre-dose) over time was possible and no data is provided.
2. Blood sampling and processing for hematology and plasma preparation
Whole blood was collected from healthy conscious single-housed cynomolgus
monkeys from the antecubital vein after needle puncture at various time intervals
before and after drug or vehicle administration (pre-dose, 0.5, 1, 2, 4, 6, 8, 24, 48,
72, 149.5 hours) into tubes containing 3.13% sodium citrate (Eifelfango, Bad
Neuenahr-Ahrweiler, Germany) at 1/10 of the total tube volume after blood sampling.
From the PBS vehicle treated group the sampling schedule varied slightly and the
following time points were sampled: pre-dose, 0.5,1, 2, 4, 6, 8, 24, 48,126 hours.
Whole blood cell counts with a particular focus on platelet count were determined to
monitor the physiological state of hematology by using an automated hematology
analyzer Scil Vet abc (Scil animal care Company GmbH, Viernheim, Germany). The
remaining blood sample after whole blood aggregation assays was centrifuged for
plasma preparation from each time point. For plasma preparation blood was
centrifuged at 5000 U/min for 15 min and the supernatant collected in a separate
tube and frozen at -20°C for analysis of plasma levels of anti GPVI Fab at a later
time point.
3. Measurements of whole blood platelet aggregation
Whole blood platelet aggregation assays were performed using the multiplate ®
platelet function analyzer (Dynabyte Informationssysteme GmbH, Munich, Germany).
The agonist used was equine type I collagen (Horm collagen, Nycomed, Munich,
Germany) at a final concentration of 1 ug/ml. The analysis was performed according
to the manufactures instruction and percent inhibition of whole blood aggregation
was calculated relative to each individual whole blood aggregation response at pre-
dose. Briefly described, the cartridge was preloaded with 297 µl CaCI2 and 297 µl of
whole blood was added. After five minutes of equilibration 6 µL of the agonist
collagen was added in a 100-fold concentration and the measurement started. The
recording took place for 7 min and the result was expressed as area under the curve
(AUC, arbitrary unit) over time in minutes. Relative change of AUC over time
compared to the pre-dose value was calculated to percent inhibition of platelet
aggregation and plotted in a graph.
With a separate set of blood samples from other monkeys out of the same colony in
vitro dose-response measurements were performed to determine an in vitro IC50 for
the anti GPVI Fab. Therefore blood samples were handled in the same way as
described above for ex vivo measurement. Briefly, different concentrations of anti
GPVI Fab (30,10, 3,1, 0.3, 0.1, 0.03, 0.01 µg/ml) were added to the CaCI2/whole
blood mixture in the test cell of the multiplate ® analyzer and incubated for 5
minutes. After adding the agonist collagen at 1 µg/ml the measurement was started
and the AUC recorded for 7 min. By plotting the respective dose-response an IC50
was calculated using an in-house statistical software tool (Speed 2.0 LTS).
4. Measurmentof GPVI receptor surface expression
Whole blood was collected from healthy conscious single-housed cynomolgus
monkeys from the antecubital vein after needle puncture at various time intervals
before and after drug or vehicle administration into tubes containing 3.13% sodium
citrate at 1/10 of the total tube volume after blood sampling and used immediately.
Samples were fixed using 5% paraformaldehyde and GPVI receptor expression was
determined. GPVI density on the platelet surface was measured using a different,
fluorescently labelled anti-GPVl antibody and determined in a flow cytometer. As a
control, it was previously shown that this antibody is able to bind GPVI independently
(and in the presence) of the investigated Fab fragment.
5. Results on whole blood platelet aggregation.
To compare the respective dose regimen, percent inhibition of whole blood platelet
aggregation is calculated relative to the pre-dose value of each individual animal.
The PBS vehicle group revealed up to 67 % of inhibition of platelet aggregation at 6
hours after IV bolus administration (data not shown). Therefore, at least 80 %
inhibition of platelet aggregation over at least two consecutive time points was
considered to be physiologically relevant. Both two low doses of anti GPVI Fab
tested (0.01 and 0.1 mg/kg anti GPVI fab) did not show any relevant inhibition of
whole blood aggregation. At 1 mg/kg, 93 % platelet inhibition was reached already at
1 hour after administration and stayed above 80% up to 24 hours (apart from a slight
decrease in effect at 2 Fig 10 A). Inhibition of platelet function subsequently
decreased overtime. At 3 mg/kg the inhibition of platelet aggregation was stable at
greater values than 80 % during the first 24 hours of observation and sustained until
72 hours with values between 69 % and 77 % (Fig. 10 B). In both higher dose
groups tested the platelet function fully recovered at the last time point (149.5 hrs,
Fig. 10 A and B). Based on this data an ED50 was estimated at 0.5 hours post IV
bolus administration and revealed 0.5 mg/kg for the tested anti GPVI Fab.
This experiment demonstrates that anti GPVI Fab inhibit ex vivo whole blood platelet
aggregation in a dose-dependent manner when compared to vehicle using collagen
(1 µg/ml).
In a separate set of experiments the in vitro activity of the anti GPVI Fab was
determined. The calculated IC50 revealed 0.81 µg/ml [0.51; 1.28 µg/ml] CV=21.6%
(Fig. 11).
6. Hematology assessment
To monitor the physiological state of hematology during the time-course of the
experiment, whole blood cell counts with a particular focus on platelet count were
determined. Mean platelet counts at pre-dose varied between 428 x 103 / uL in PBS
vehicle, 369 x 103 / µL at 0.01 mg/kg anti GPVI Fab, 235 x 103 / µL at 0.1 mg/kg anti
GPVI Fab, 357 x 103 / µL at 1 mg/kg anti GPVI Fab, and 312 x 103 / µL at 3 mg/kg
anti GPVI Fab (Figure 12 A-E). During the time-course of the experiment the platelet
count did not change substantially (i.e. values below 100 x 103/µL)and the following
platelet count was determined at 126 hours in the PBS vehicle group: 358 x 103 / µl
(Figure 12 A). In all anti GPVI Fab treated groups the platelet count determined at
149.5 hours revealed the following values: 411 x 103/ µL at 0.01 mg/kg anti GPVI
Fab, 329 x 10s / µL at 0.1 mg/kg anti GPVI Fab, 321 x 103 / µL at 1 mg/kg anti GPVI
Fab, and 320 x 103 / µL at 3 mg/kg anti GPVI Fab (Figure 12 B-E). All other
determined hematology parameters, hematocrit, red blood cell count, and
hemoglobin were not changed substantially during the time-course of the experiment
(data not shown).
These data demonstate that the GPVI Fab do not affect the physiological state of
hematology.
7. GPVI receptor expression
As shown in Figure 13, before iv administration of the anti GPVI Fab all platelets
were positive for GPVI expression (pre-dose). In blood taken after drug
administration no specific signal for GPVI could be observed on the platelet surfaces
(24h). Beginning at 48h after drug administration a new population of platelets
arises, which are GPVI positive. 150h after drug administration all platelet were
again positive for GPVI receptor expression.
This experiment confirms that GPVI Fabs induce GPVI receptor depletion on platelet
and that this effect was reversible.
Example 5: Modification of Fab fragments properties by recognition by auto-
antibodies.
A - Determination of the activatorv component in donor plasma
To investigate the importance of IgG's present in plasma for the activatory effect of
the anti-GPVI Fab, 51 different blood samples were tested. Samples which have
been identified as activatory were depleted of IgG's using protein A.
For the experiments, blood was anticoagulated with 20 µg/ml hirudin and used
immediately for the preparation of plasma (centrifugation of blood samples for 10 min
at 1600 g). Thereafter, plasma was depleted of IgG's for 2h at 4°C using protein A.
Protein A was removed by centrifugation and platelets were added at a final
concentration of 2x10E8/ml. Anti GPVI - Fab fragments were used at 20ug/ml.
Plasma samples were incubated with the Fab fragments or convulxin (or "cvx", a GP
VI specific agonist) for 15min followed by a staining with the FITC labelled Pac-1
antibody (specific for activated GPIlbllla, which is a platelet activation marker) for 30
min. Thereafter, samples were fixed using paraformaldehyde and Pac-1 labelling of
platelet was determined in a flow cytometer (BD LSR II).
As seen in Figure 14, on Control panel, treatment of plasma with protein A had no
effect on platelet activation by the GPVI specific agonist convulxin (cvx). However,
Figure 14, panel A shows that the activatory effect of Fab was greatly reduced after
IgG depletion, suggesting that preformed IgG's are an essential component for this
response.
B - Determination of the expression of platelet activation markers following modified
anti GPVl-Fab fragment exposer
To further investigate the role of plama IgG on activation of platelet activation
through Fab, the different Fab fragments were used which were identical in their
heavy and light chain (HC and LC) sequence but differ in their C-terminal
modification on the heavy chain.
The four different molecules used are described below and illustrated in Figure 15.
Fab: control molecule with no C-terminal additional amino acids on HC (Figure 15 A)
Fab-His-tag: hasHisHisHisHisHis peptide sequence (SEQ ID NO:35) on HC corres-
ponding to the natural occurring sequence plus C-terminal His-tag (Figure 15 B)
Fab-Gly4Ser: has a GlyGlyGlyGlySer peptide sequence (SEQ ID NO:36) on HC
corresponding to the natural occurring sequence plus C-terminal Gly4Ser - tag
(Figure 15 C)
Fab-(Gly4Ser)2: has a GlyGlyGlyGlySerGlyGlyGlyGlySer peptide sequence (SEQ ID
NO:36) on HC corresponding to the natural occurring sequence plus C-terminal
(Gly4Ser)2 - tag (Figure 15 C)
As illustrated in Figure 14, panel B, the modified Fab do not induce any activatory,
even without IgG depletion, suggesting that modifid Fab are not recognized by
preformed IgG's.
For the experiments, blood was anticoagulated with 20 µg/ml hirudin and used
immediately. Anti GPVI - Fab fragments were used at20µg/ml. Blood samples (1-51)
were incubated with the Fab fragments for 15min followed by a staining with the
FITC labelled Pac-1 antibody (specific for activated GPIIbllla, which is platelet
activation marker) for 30 min. Thereafter samples were fixed using
paraformaldehyde and Pac-1 labelling of platelet was determined in a flow cytometer
(BD LSR II).
The activatory potential of the 4 Fab formats was tested on 51 different blood
samples. Fab with no overhang induced a significant increase in Pac-1 binding
(defined as > 5 fold in 23 samples (corresponds to 42%). In sharp contrast, as seen
in Table 19 below, Fabs modified at the C-terminus of the HC were much less active
in this test with Fab-His-tag and Fab-Gly4Ser only showing activity on 1 sample and
Fab-(Gly4Ser)2 (longest C-terminal extension) was not active over the threshold.
This differential activatory pattern is also reflected in the cases of minor activation (2-
5 fold over basal). These results indicated that the activatory potential of Fab
fragments in this assay is not determined by the antigen binding CDR sequences but
appears to reside in the HC C-terminus because the activity greatly reduced by C-
terminal modifications of the HC chain. Thus these results suggest that the C-
terminal extremity of Fab were recognized by IgG preexisting in patients blood,
which induces platelet activation.
Table 19: Fold increase in Pac-1 binding to platelets incubated with Fab fragments
with no overhang, Fab-His-tag, Fab-Gly4Ser and Fab-(Gly4Ser)2. 51 different
samples (donors) were investigated. Grey field represent sample with activation over
5 fold basal values.
1. An antibody Fab fragment that specifically binds to human GPVI and induces
GPVl depletion phenotype.
2. The antibody Fab fragment of claim 1, wherein the antibody Fab fragment binds
a conformational epitope of human GPVI and contacts human GPVI residues
comprising Ser 43, Arg 67 and Asp 81.
3. The antibody Fab fragment of claim 1 or 2, wherein the Fab fragment
comprises
(a) complementarity determining regions (CDRs) of a heavy chain variable
region (HCVR) having amino acid sequences defined by SEQ ID NO: 18, SEQ ID
NO: 19, and SEQ ID NO: 20; and
(b) complementarity determining regions (CDRs) of a light chain variable
region (LCVR) having amino acid sequences defined by SEQ ID NO: 21, SEQ ID
NO: 22, and SEQ ID NO: 23; and
wherein at least 2 amino acid residues of each CDR can be changed to
another amino acid residue without directly disrupting a contact with a GPVI epitope
residue.
4. The antibody Fab fragment of any one of claims 1 to 3, wherein the antibody
Fab fragment comprises a heavy chain of SEQ ID NO.6 and a light chain of
SEQ ID NO.8 or sequences having at least 80% identity with these sequences,
as far as the antibody Fab fragment binding specificity is maintained.
5. The antibody Fab fragment of any preceding claim which is a humanized.
6. The antibody Fab fragment of claim 5, wherein the humanized Fab fragment
comprises a combination of a heavy chain variable region (HCVR) amino acid
sequence and a light chain variable region (LCVR) amino acid sequence,
selected from the group consisting of
(a) LCVR (SEQ ID NO.15) and HCVR (SEQ ID NO.10);
(b) LCVR (SEQ ID NO.15) and HCVR (SEQ ID NO.11);
(c) LCVR (SEQ ID NO.15) and HCVR (SEQ ID NO.12);
(d) LCVR (SEQ ID NO.15) and HCVR (SEQ ID NO.13);
(e) LCVR (SEQ ID N0.15) and HCVR (SEQ ID N0.14);
(f) LCVR (SEQ ID N0.16) and HCVR (SEQ ID N0.11);
(g) LCVR (SEQ ID N0.17) and HCVR (SEQ ID N0.11); and
(h) LCVR (SEQ ID N0.17) and HCVR (SEQ ID N0.13).
7. An engineered Fab fragment comprising a combination of a humanized heavy
chain (HC) amino acid sequence and a humanized light chain (LC) amino acid
sequence, wherein the humanized heavy chain further comprises a c-terminal
extension comprising additional amino acid residues and wherein the c-terminal
extension prevents recognition by anti-Fab antibodies.
8. The engineered Fab fragment of claim 7, wherein the c-terminal extension is
selected from the group consisting of SEQ ID NO:35, SEQ ID NO:36 and SEQ
ID NO:37.
9. The engineered Fab fragment of claim 7 or 8, wherein the Fab fragment is as
defined in any one of claims 1 to 6.
10. A pharmaceutical composition comprising an antibody Fab fragment as defined
in any one of claims 1 to 9 and a pharmaceutically acceptable carrier or
excipients.
11. A polynucleotide encoding a polypeptide selected from the group consisting in
SEQ ID NO.6, SEQ ID NO.8, SEQ ID NO.9, SEQ ID NO.10, SEQ ID NO.11,
SEQ ID NO.12, SEQ ID NO.13, SEQ ID NO.14, SEQ ID NO.15, SEQ ID NO.16,
and SEQ ID NO.17 or a sequence having at least 80% identity with these
sequences, as far as the polypeptide retains its binding specificity.
12. A method for the preparation of an antibody Fab fragment as defined any one
of claims 1 to 9 comprising the steps of:
a. Culture of a cell line expression an antibody Fab fragment as defined
in claims 1 to 9
b. Purification of Fab fragment expressed in the culture medium
c. Formulation of the Fab fragment in a convenient form
13. An antibody Fab fragment as defined in claims 1 to 9 for use to prevent or treat
thrombotic and vascular diseases.
14. A method for prevention of recognition of an antibody Fab fragments by
preexisting antibodies consisting in masking the c-terminal extremity of the
antibody Fab fragment by addition of a molecule.
15. A method for prevention of platelet activation when an anti-GPVI Fab is used
consisting in masking the c-terminal extremity of the antibody Fab fragment by
addition of a molecule.

ABSTRACT

FR2009/113 PCT PATENT APPLICATION TITLE Novel antagonist antibodies and their Fab fragments against
GPVI and uses thereof SANOFI-AVENTIS ABSTRACT The present invention discloses novel antibodies that specifically bind to
the human platelet membrane protein Glycoprotein VI (GPVI) and their monovalent fragments or derivatives. The antibodies of
the invention are antibodies from hybridoma clone 390 and fragment antibodies thereof able to induce a GPVI depletion phenotype. These antibodies and Fab fragments are able to block collagen binding and thus preventing platelet activation by collagen.
The invention also relates to hybridoma clones and expression plasmids for the production of disclosed antibodies and Fab fragments. The present invention further refers to the uses of monovalent antibody fragments to manufacture research, diagnostic and
immunotherapeutic agents for the treatment of thrombosis and other vascular diseases. The invention also concerns a Fab bearing
a molecule at the C-terminal extremity, as well as method for prevention of recognition of Fab by antibodies using such modified
Fab. The invention concerns a method for prevention of platelet activation when an anti-GP VI Fab is used.