Note: Descriptions are shown in the official language in which they were submitted.
CA 02355215 2001-06-08
WO 00/39585 PCTNS99/30960
IDENTIFYING SMALL ORGANIC MOLECULE LIGANDS FOR BINDING
FIELD OF THE INVENTION
The present invention is directed to novel avidity-based methods and
compositions for quickly and unambiguously identifying small organic molecule
ligands
that bind to biological target molecules. Small organic molecule ligands
identified
according to the methods of the present invention find use, for example, as
novel
therapeutic drug lead compounds, enzyme inhibitors, labeling compounds,
diagnostic
reagents, affinity reagents for protein purification, and the like.
BACKGROUND OF THE INVENTION
The primary task in the initial phase of generating novel biological effector
molecules is to identify and characterize one or more tightly binding
ligand(s) for a given
biological target molecule. In this regard, many molecular techniques have
been
developed and are currently being employed for identifying novel ligands that
bind to
specific sites on biomolecular targets such as proteins, nucleic acids,
carbohydrates,
nucleoproteins, giycoproteins, glycolipids and lipoproteins. Many of these
techniques
exploit the inherent advantages of molecular diversity by employing
combinatorial
libraries of potential binding compounds in an effort to speed up the
identification of
functional ligands. For example, combinatorial synthesis of both oligomeric
and
non-oligomeric libraries of diverse compounds combined with high-throughput
screening assays has already provided a useful format for the identification
of new lead
compounds for binding to chosen molecular targets.
While combinatorial approaches for identifying biological effector molecules
have
CA 02355215 2001-06-08
WO 00/39585 PCT/US99/30960
proven useful in certain applications, these approaches also have some
significant
disadvantages. For example, often there does not exist an appropriate
screening assay
which allows one to detect binding of a library member to the target molecule
of interest
when the library member may bind only weakly to the target. Moreover, even
when
such screening assays are available, in many cases techniques which allow
rapid
identification of the actual library members) which bind most effectively to
the target
are not available or provide ambiguous results, making the actual s
identification and
characterization of functional ligand molecules difficult, time-consuming or
impossible.
Furthermore, many approaches currently employed to identify novel ligands are
dependent upon only a single specific chemical reaction type, thereby limiting
the
usefulness of such approaches to only a narrow range of applications. Finally,
many of
the approaches currently employed are expensive and extremely time-consuming.
Thus, there is a significant interest in developing new methods which allow
rapid,
efficient and unambiguous identification of small organic molecule ligands for
selected
biomolecular targets.
For the most part, combinatorial libraries that find use in methods for s
screening
against a target biomolecule comprise molecules that are larger than most
small
organic compounds. Techniques for assaying for binding of such "larger"
molecules to a
target biomolecule are known in the art and may often be employed to identify
the
specific library members) that bind to the target. However, when libraries of
relatively
small organic molecules are screened against a biological target molecule,
binding of a
library member to the target is often difficult to detect because even the
"best" small
molecules may bind only weakly. Moreover, even if one is able to detect the
binding of
a library member to the target, actual identification of the bound compound
may be
impossible unless methods for deconvolution are available. As such, we herein
propose
novel methods based upon principles of avidity which significantly enhance the
ability to
screen libraries of relatively small organic molecules for binding to a
biological target.
An interaction between two molecules that are capable of binding to one
another
is often characterized in terms. of the strength with which those molecules
attach or
interact, i.e., the "affinity" that the molecules have for one another.
Generally, the
2
CA 02355215 2001-06-08
WO 00/39585 PCT/US99/30960
affinity that one molecule has for another is a measurement of the strength of
attachment between the molecules assuming that each of the binding molecules
may
interact only through a single specific site. However, in reality, binding
molecules often
have multiple sites through which an interaction with another molecule may
occur. In
such situations, although the affinity at any one binding site may be
unchanged, the
overall strength of the attachment must take into account binding at all of
the available
sites on one or both of the binding partners. This overall strength of
attachment is
known as the "avidity" and will appear as a stronger apparent affinity at any
given
binding site. Mathematically, the strength of the avidity increases for each
occupied site
on a binding molecule. The phenomenon of avidity, therefore, might be
exploited in
ways so as to not only enhance the ability to screen libraries of small
organic molecules
for binding to a biological target molecule but also to identify organic
molecules that
actually bind to the target.
It is, therefore, an object of the present invention to provide novel methods
and
compositions which exploit the avidity phenomenon and allow the quick and
unambiguous identification of organic compounds that are capable of binding to
a
biological target molecule. Such methods are herein described and are quick,
easy to
perform and relatively inexpensive as compared to other currently employed
methods.
SUMMARY OF THE INVENTION
Applicants herein describe an avidity-based molecular approach for rapidly and
efficiently identifying small organic molecule iigands that are capable of
interacting with
and binding to specific sites on biological target molecules, wherein organic
compounds
identified by the subject methods as being capable of binding to the
biological target
may find use, for example, as new small molecule drug leads, enzyme
inhibitors,
labeling compounds, diagnostic reagents, affinity reagents for protein
purification, and
the like. The herein described approaches allow one to quickly screen a
library of small
organic compounds to unambiguously identify those that are capable of binding
to a
site of interest on a biomolecular target. The small organic molecule ligands
identified
by the methods described herein may thernseives be employed for numerous
3
CA 02355215 2001-06-08
WO 00/39585 PCT/US99/30960
applications, or may be coupled together in a variety of different
combinations using
one or more linker elements to provide novel binding molecules.
With specific regard to the above, one embodiment of the present invention is
directed to a method for identifying an organic molecule ligand that binds to
a site of
interest on a biological target molecule, wherein the method comprises the
steps of:
(a) identifying or selecting a biological target molecule that comprises a
site of
interest to which the organic molecule ligand may potentially bind;
(b) obtaining a multimeric form of the biological target molecule, wherein the
multimeric form comprises at least two linked biological target molecules and
a plurality
of the sites of interest;
(c) contacting or combining the multimeric form with at least first and second
members of a library of organic compounds that are potentially capable of
binding to
the site of interest, wherein at least two of the first members of said
library bind to tie
sites of interest of the multimeric form; and
(d) identifying the first member of the library of organic compounds that
bound to
the site of interest.
As is evident from the above, because an embodiment of the described method
employs a multimeric form of a biological target molecule that comprises a
plurality of or
at least two sites of interest that are capable of binding to a member of an
organic
molecule library, the strength with which the multimeric form binds to members
of a
library of small organic compounds is greatly enhanced due to well known
avidity
principles, thereby enhancing the ability to detect that binding has occurred
and to
determine 5 which members of the library are capable of binding to the target
molecule.
In other words, because of the multiple available binding sites on the
multimeric form of
the biological target molecule (i.e., the plurality of sites of interest), the
overall strength
of binding between the multime;ric form of the target and an organic compound
that
binds thereto will be greatly enhanced (i.e., avidity) as compared to "single-
site" binding
(i.e., affinity).
In certain particular embodiments, the biological target molecule is a
polypeptide,
a nucleic acid, a carbohydrate, a nucleoprotein, a glycopeptide, a glycolipid,
or a
4
CA 02355215 2001-06-08
WO 00/39585 PCT/US99/30960
lipoprotein, preferably a polypeptide, which may be, for example, an enzyme, a
hormone, a transcription factor, a receptor, a ligand for a receptor, a growth
factor, an
immunoglobulin, a steroid receptor, a nuclear protein, a signal transduction
component,
an aliosteric enzyme regulatar, and the like. The muitimeric form of the
biological target
molecule comprises either covalently or non-covalently linked biological
target
molecules and may be obtained in a variety of ways including, for example, by
chemically linking two or more pre-existing biological target molecules, by
directly
synthesizing a multimeric form of a biological target molecule or, in the case
of a
polypeptide, by recombinantly Expressing a multimeric form of the polypeptide
target
molecule.
Other embodiments of the herein described methods which employ libraries of
organic compounds which comprise aldehydes, ketones, oximes, hydrazones,
semicarbazones, carbazides, primary amines, secondary amines, tertiary amines,
N-substituted hydrazines, hydrazides, alcohols, ethers, thiols, thioethers,
thioesters,
disulfides, carboxylic acids, esters, amides, ureas, carbamates, carbonates,
ketals,
thioketals, acetals, thioacetals, aryl halides, aryl sulfonates, alkyl
halides, alkyl
sulfonates, aromatic compounds, heterocyclic compounds, anilines, alkenes,
alkynes,
diols, amino alcohols, oxazolidines, oxazolines, thiazolidines, thiazolines,
enamines,
sulfonamides, epoxides, aziridines, isocyanates, sulfonyl chlorides, diazo
compounds
and/or acid chlorides, preferably aldehydes, ketones, primary amines,
secondary
amines, alcohols, thioesters, disulfides, carboxylic acids, acetals, anilines,
diols, amino
alcohols and/or epoxides, most preferably aldehydes, ketones, primary amines,
secondary amines and/or disulfides.
With regard to the above described method, at least the (a) multimeric form of
the biological target molecule or (b) at least first and second organic
compound library
members may optionally be covalently attached to a solid matrix material prior
to
combining the multimeric form vvith the library members to assess binding
therebetween.
Another embodiment of the present invention is directed to a method for
identifying an organic molecule ligand that binds to a site of interest on a
biological
CA 02355215 2001-06-08
WO 00/39585 PCT/US99/30960
target molecule, wherein the method comprises:
(a) identifying or selecting a biological target molecule that comprises a
site of
interest to which the organic molecule ligand may potentially bind;
(b) combining the biological target molecule with at least first and second
members of a library of multivalent organic compounds that are potentially
capable of
binding to the site of interest, wherein the site of interest of at least two
of the biological
target molecules binds to the first member of the library of multivalent
organic
compounds; and
(c) identifying the first member of the library of multivalent organic
compounds
that bound to the biological target molecule.
Similar to the above described embodiment that employs a multimeric form of
the biological target molecule, bE:cause organic compound libraries may be
constructed
that contain members which are "multivalent" (i.e., members which have two or
more
small organic molecules linked together either covalently or non-covalently),
the
members of the organic compound 30 library may also have a plurality of sites
available
for bonding to the site of interest on a biological target molecule.
Therefore, the strength
with which the members of the organic compound library bind to the biological
target
molecule is greatly enhanced due to well known avidity principles, thereby
enhancing
both.the ability to detect the existence of binding as well as to determine
which
members of the library are capable of binding to the target molecule. In other
words,
because of the multiple available: binding sites present on the organic
compound library
members, the overall strength of binding between the biological target and an
organic
compound that binds thereto wilU be greatly enhanced (i.e., avidity) as
compared to
"single-site" binding (i.e., affinity;l.
Another'embodiment of the present invention is directed to methods for
identifying an organic molecule ligand that binds to a site of interest on a
biological
target molecule, wherein the method comprises:
(a) selecting a biological target molecule that comprises a site of interest
to
which the organic molecule ligand may potentially bind;
(b) contacting (i) a multimeric form of the biological target molecule which
6
CA 02355215 2001-06-08
WO 00/39585 PCT/US99/30960
comprises at least two linked biological target molecules and at least two
sites of
interest with (ii) at least first and second members of a library of
multivalent organic
compounds that are potentially capable of binding to the sites of interest,
wherein the
multimeric form binds to the first member of the library of multivalent
organic
compounds; and
(c) identifying the first member of the library of multivalent organic
compounds
that bound to the multimeric form.
Other embodiments of the present invention are directed to solid matrix
materials
comprising materials having at least first and second members of a library of
organic
compounds covalently bound thereto. Such materials may, among other things,
find
use in the presently described methods.
Additional embodiments of the present invention will become evident to the
ordinarily skilled artisan upon review of the present specification.
BRIEF DESCRIPTION OF THE DRAWINGS
Figure 1 shows the binding of biotin-dextran conjugates to anti-biotin
monoclonal
antibodies.
Figure 2 shows the plots for recognition of either biotin, desthiobiotin,
unreacted
wells and wells blocked with propionic acid.
Figure 3 shows soluble biotin titrated against a constant concentration of a-
Biotin
antibody and a-Fc-HRP detection antibody to evaluate the amount of free biotin
required to compete with either immobilized biotin or immobilized
desthiobiotin.
DETAILED DESCRIPTION OF THE INVENTION
The present invention provides rapid and efficient methods for identifying
small
organic molecule ligands that are capable of binding to selected sites on
biological
target molecules of interest. 'The organic molecule ligands themselves
identified by the
subject methods find use, for example, as lead compounds for the development
of
novel therapeutic drugs, enzyme inhibitors, labeling compounds, diagnostic
reagents,
affinity reagents for protein purification, and the like, or two or more of
the identified
7
CA 02355215 2001-06-08
WO 00/39585 PCT/US99/30960
organic molecule ligands may be' coupled together using routine and well known
techniques through one or more linker elements to provide novel biomolecule-
binding
conjugate molecules.
One embodiment of the subject invention is directed to a method for
identifying
an organic molecule ligand that binds to a site of interest on a biological
target
molecule. As an initial step in they herein described methods, a biological
target
molecule is identified or obtained as a target for binding to the small
organic molecule
compounds screened during the process. Biological target molecules that find
use in
the present invention include all biological molecules to which a small
organic molecule
may bind and preferably include, for example, polypeptides, nucleic acids,
including
both DNA and RNA, carbohydrates, nucleoproteins, glycoproteins, glycolipids,
lipoproteins, and the like. The bic>logical target molecules that find use
herein may be
obtained in a variety of ways, including but not limited to commercially,
synthetically,
recombinantly, from purification from a natural source of the biological
target molecule,
and the like.
In a particularly preferred embodiment, the biological target molecule is a
polypeptide. Polypeptides that find use herein as targets for binding to
organic molecule
ligands include virtually any peptide or protein that comprises two or more
amino acids
and which possesses or is capable of being modified to possess a site of
interest that is
potentially capable of binding to a small organic molecule. Polypeptides of
interest
finding use herein may be obtained commercially, chemically, recombinantly,
synthetically, by purification from a natural source, or otherwise and, for
the most part
are proteins, particularly proteins associated with a specific human disease
condition,
such as cell surface and soluble receptor proteins, such as lymphocyte cell
surface
receptors, enzymes, such as proi:eases and thymidylate synthetase, steroid
receptors,
nuclear proteins, aliosteric enzyme inhibitors, clotting factors,
serine/threonine kinases
and dephosphorylases, threonine kinases and dephosphorylases, bacterial
enzymes,
fungal enzymes and viral enzymes, signal transduction molecules, transcription
factors,
proteins associated with DNA and/or RNA synthesis or degradation,
immunoglobulins,
hormones, receptors for various c:ytokines including, for example,
erythropoietin/EPO,
8
CA 02355215 2001-06-08
WO 00/39585 PC'T/US99/30960
granulocyte colony stimulating receptor, granulocyte macrophage colony
stimulating
receptor, thrombopoietin (TPO), I iL-2, IL-3, IL-4, IL-5, IL-6, IL-10, lL-11,
IL-12, growth
hormone, prolactin, human placental lactogen (LPL), CNTF, octostatin, various
chemokines and their receptors such as fZANTES, MIP1-a, IL-8, various ligands
and
receptors for tyrosine kinase such as insulin, insulin-like growth factor 1
(IGF-1 ),
epidermal growth factor (EGF), heregulin-a, and heregulin-Vii, vascular
endothelial
growth factor (VEGF), placental growth factor (PLGF), tissue growth factors
(TGF-a
and TGF-Vii), other hormones and receptors such as bone morphogenic factors,
folical
stimulating hormone (FSH), and' ieutinizing hormone (LH), tissue necrosis
factor (TNF),
apoptosis factor-1 and -2 (AP-1 and AP-2), mdm2, proteins and receptors that
share
20% or more sequence identity to these, and the like.
A "site of interest" on a biological target molecule may be any site on a
target
molecule where it is desired that a small organic molecule binds. Sites of
interest raay
be naturally existing on the target or may be artifrcially introduced into the
target using
techniques that are routinely employed and well known in the art. In preferred
embodiments, the site of interest is an active site of an enzymatic protein or
a binding
site on a target that binds to another biological target molecule.
In one specific embodiment of the present invention, once a biological target
molecule that comprises a site of interest is identified, a multimeric form ~
of that
biological target molecule is then obtained. By "multimeric form" is meant
that at least
two of the biological target molecules of choice are linked, either covalentiy
or
non-covalently, preferably covalently, such that the resulting "multimeric"
structure
comprises multiple copies of the biological target molecule and a plurality of
or at least
two of the site of interest. The multimeric form of the target molecule may be
prepared
or obtained in a number of ways including, for example, by chemically
crosslinking two
or more biological target molecules using any of a number of known chemical
crosslinkers, biotin/streptavidin, immunogiobulin-mediated linkage, and the
like, by
artificially synthesizing a multimeric form of a biological target molecule
or, in the case
of polypeptide targets, by recombinantly expressing a multimeric form of the
target
polypeptide. Multimeric forms may also be obtained by covalent bond formation
9
CA 02355215 2001-06-08
WO 00/39585 PCT/US99/30960
between two or more chemically reactive groups present on the target
molecules.
Techniques for obtaining a multimeric form of a biological target molecule are
well
known in the art and are found in general textbooks such as, for example,
Sambrook et
al., Molecular Cloning: A Laboratory Manual (New York: Cold Spring Harbor
Laboratory
Press, 1989) and Ausubel et al., Current Protocols in Molecular Biology,
Greene
Publishing Associates and Wiley-Interscience 1991.
Since a biological target molecule will comprise at least one of the site of
interest, the multimeric form of that biological target molecule will possess
a plurality of
or at least two of the site of interest against which a library of small
organic compounds
may be screened for binding. Techniques for preparing a multimeric form of a
chosen
biological target molecule are known in the art and may be employed herein in
a routine
manner.
A multimeric form of a biological target molecule of interest will comprise at
Fast
about 2 linked biological target molecules, often from about 2 to about 200
linked
biological target molecules, more' often from about 2 to about 100 linked
biological
target molecules, usually from about 2 to about 50 linked biological target
molecules,
more usually from about 2 to about 20 linked biological target molecules,
preferably
from about 2 to about 10 linked biological molecules and more preferably from
about 2
to about 6 linked biological molecules. The biological target molecules will
be linked
either covalently or non-covaiently to provide the multimeric form, preferably
covalently
using well known molecular linking techniques.
In a particular embodiment of the present invention, the multimeric form of
the
biological target molecule obtained as described above will be combined with
or
contacted with at feast first and second members of a library of organic
compounds that
are potentially capable of binding to the sites of interest on the multimeric
form of the
target molecule. Organic compounds will be "potentially capable of binding to
a site of
interest" if they possess a size or structure which is compatible with the
site of interest
on the biological target, thereby allowing binding therebetween. The step of
combining
the multimeric form and the at least first and second organic compound library
members will be performed under conditions which are capable of allowing
binding to
CA 02355215 2001-06-08
WO 00/39585 PCT/US99/30960
occur therebetween, wherein those conditions will depend upon the nature of
the
components of the system and may be determined routinely and empirically.
The first and second organic compounds that find use preferably are of the
same
chemical class (e.g., are all aldehydes, are all amines, etc.) but may also be
of different
chemical classes. The library of organic compounds to be screened against the
biological target molecule or multimeric form thereof may be obtained in a
variety of
ways including, for example, through commercial and non-commercial sources, by
synthesizing such compounds using standard chemical synthesis technology or
combinatorial synthesis technology (see Gallop et al., J. Med Chem. 37:1233-
1251
(1994), Gordon et al., J. Med. Chem. 37:1385-1401 (1994), Czarnik and Ellman,
Acc.
Chem. Res. 29:112-170 (1996), Thompson and Ellman, Chem. Rev. 96:555-600
(1996), and Balkenhohl et al., Arngew. Chem. Int Ed. 35:2288-2337 (1996)), by
obtaining such compounds as de:gradation products from larger precursor
compounds,
e.g. known therapeutic drugs, large chemical molecules, and the like.
The monovalent "organic compounds" employed in the methods of the present
invention will be, for the most past, small chemical molecules that will
generally be less
than about 2000 daltons in size, usually less than about 1500 daltons in size,
more
usually less than about 750 daltons in size, preferably less than about 500
daltons in
size, often less than about 250 daltons in size and more often less than about
200
daltons in size, although organic molecules larger than 2000 daltons in size
may also
find use herein. Organic molecules that find use may be employed in the herein
described methods as originally obtained from a commercial or non-commercial
source
(for example, a large number of small organic chemical compounds are readily
obtainable from commercial suppliers such as Aldrich Chemical Co., Milwaukee,
WI and
Sigma Chemical Co., St. Louis, MO) or may be obtained by chemical synthesis.
Organic molecule compounds that find use in the present invention include, for
example, aldehydes, ketones, oximes, such as O-alkyl oximes, preferably O-
methyl
oximes, hydrazones, semicarbazones, carbazides, primary amines, secondary
amines,
such as N-methylamines, tertiary amines, such as N,N-dimethylamines, N-
substituted
hydrazines, hydrazides, alcohols, ethers, thiols, thioethers, thioesters,
disulfides,
11
CA 02355215 2001-06-08
WO 00/39585 PCTNS99/30960
carboxylic acids, esters, amide:>, areas, carbamates, carbonates, ketals,
thioketals,
acetals, thioacetals, aryl halides, aryl sulfonates, alkyl halides, alkyl
sulfonates,
aromatic compounds, heterocyc;lic compounds, anilines, alkenes, alkynes,
diols, amino
alcohols, oxazoiidines, oxazolines, thiazolidines, thiazolines, enamines,
sulfonamides,
epoxides, aziridines, isocyanates, sulfonyl chlorides, diazo compounds, acid
chlorides,
and the like. In fact, virtually any small organic molecule that is
potentially capable of
binding to a site of interest on a biological target molecule may find use in
the present
invention with the proviso that it is sufficiently soluble and stable in
aqueous solutions to
be tested for its ability to bind to the biological target molecule.
Libraries of organic compaunds which find use herein will generally comprise
at
least 2 organic compounds, often at least about 25 different organic
compounds, more
often at least about 100 different organic compounds, usually at least about
300
different organic compounds, more usually at least about 500 different organic
compounds, preferably at least about 1000 different organic compounds, more
preferably at least about 2500 different organic compounds and most preferably
at least
about 5000 or mare different orc,~anic compounds. Populations may be selected
or
constructed such that each individual molecule of the population may be
spatially
separated from the other molecules of the population (e.g., each member of the
library
is a separate microtiter well) or two or more members of the population may be
combined if methods for deconvolution are readily available. The methods by
which the
populations of organic compounds are prepared will not be critical to the
invention.
Usually, each member of the orc,~anic molecule library will be of the same
chemical
class (i.e., all library members are aldehydes, all library members are
primary amines,
etc.), however, libraries of organic compounds may also contain molecules from
two or
more different chemical classes.
Reaction conditions for screening a library of organic compounds against a
site
of interest-containing biological target molecule will be dependent upon the
nature of
the site of interest and the chemical nature of the chosen library of organic
compounds
and- can be determined by the skilled artisan in an empirical manner. For the
step of
screening a population of organic molecules to identify those that bind to a
target
12
CA 02355215 2001-06-08
WO 00/395$5 PCT/US99/30960
polypeptide, it will be well within the skill level in the art to determine
the concentration
of the organic molecules to be employed in the binding assay.
Because the presently described methods take advantage of the principle of
binding avidity, covalent bond formation between chemically reactive groups on
the
biological target and members) of an organic compound library is not desired
when
those components are combined. Thus, in a particularly preferred embodiment of
the
present invention, because the members of the organic molecule library and the
biological target molecules may both possess chemically reactive groups which
potentially would allow undesired covalent bond formation therebetween, the
chemically reactive groups on either or both of the biological target molecule
or the
organic compound library members, preferably the biological target molecule,
may be
"capped" by treatment with a "capping agent'" prior to combining the target
and the
organic compound library members. By "capping" chemically reactive groups on
a,,
target molecule or organic library members, preferably a target molecule, is
meant that
one or more of the available chemically reactive groups are altered such that
they no
longer are capable of forming a covalent bond with another chemically reactive
group.
"Capping agents" that find use for altering chemically reactive groups so as
to prevent
those groups from participating in covalent bond formation are well known in
the art and
may be routinely employed herein. Preferably, the biological target and/or the
organic
compound library members will be chosen such that covalent bond formation
therebetween is not possible.
The methods of the present invention are based upon principles of avidity iri
that
at least one of the (a) biological target molecule or multimeric form thereof
and/or (b)
organic compound library members comprises at least two available sites for
binding to
its counterpart~molecule. By exploiting the phenomenon of avidity, the
strength of
attachment between the target molecule and the organic compound library
members is
increased to such an extent so as to enhance or facilitate the ability to not
only detect
binding between the target and library member(s), but also to allow easy
identification
of the actual library member that bound to the target biomolecule. In other
words,
avidity allows for the identification of organic compounds that bind to the
target when
13
CA 02355215 2001-06-08
WO 00/39585
PCT/US99/30960
they otherwise bind too weakly to be detected by facile detection. As such,
one
embodiment of the present invention as described above employs a multimeric
form of
a biological target molecule that comprises at least two sites of interest
that are
available for binding to an organic compound library member, thereby
exploiting the
phenomenon of avidity. Yet other embodiments of the present invention,
however, may
employ a single biological target molecule or multimeric form thereof and a
library of
"multivalent" organic compounds that are potentially capable of binding to
sites of
interest on the target, thereby again exploiting the phenomenon of avidity. By
"multivalent" when used to describe an organic compound is meant that the
compound
possesses at least two chemical structures, preferably at least two of the
same
chemical structures, that are capable of binding to a site of interest on a
biological
target molecule. Ttiis, in one embodiment, at least two organic compounds as
described above. are linked together, either covalently or non-covalently, so
as to
produce a "multivalent" molecule that comprises at least two of the organic
compound
of interest, each of which are available for binding to a site of interest on
a target
molecule. Preferably, a multivalent organic compound possesses two or more of
the
same linked organic compound, however, multivalent organic compounds having
two or
more different linked organic co~~mpounds may also find use herein.
Multivalent organic
compounds for use in the present invention may be obtained in a variety of
ways
including, for example, commercially, synthetically using well known and
routine
chemical synthesis techniques, or as degradation products of a larger chemical
molecule or other novel techniques.
Specific examples of'multivalent" organic compounds that find use herein may
inctude, for example, dextran whose sugar moieties are oxidized to form
aldehyde
functionalities, brganic compounds linked through a polymeric backbone,
dendromers,
poly-amino acids such as polylysine which possesses a plurality of amine
groups, and
the like.
Once a biological target molecule or multimeric form thereof has been combined
with members of an organic cornpound library, multimeric or not, and binding
therebetween has been allowed to occur, one must identify the organic
compound{s)
14
CA 02355215 2001-06-08
WO 00/39585 PC'T/US99/30960
that bound to the sites) of interest on the target molecule. In this regard,
many
techniques exist for identifying the bound organic compound. For example, the
well
known technique of mass spectrometry may preferably be employed either alone
or in
combination with other means for detection for identifying the organic
compound ligand
that bound to the target of interest. Prior to employing mass spectrometry,
one may
wish to chemically crosslink the bound organic compound to the target
molecule.
Techniques employing mass spectrometry are well known in the art and have been
employed for a variety of applications (see, e.g., Fitzgerald and Siuzdak,
Chemistry &
Biology 3:707-715 (1996), Chu et al., J. Am. Chem. Soc. 118:7827-7835 (1996),
Siuzdak, Proc. NatL Acad. Sci USA 91:11290-11297 (1994), Burfingame et al.,
Anal.
Chem. 68:5998-651 R (1996), Wu et al., Chemistry & Biology 4:653-657 (1997)
and Loo
et al., Am. Reports Med. Chem. 31:319-325 (1996)).
in other embodiments, subsequent to the binding of the library member to the
target molecule and covalent crossiinking of the bound components, the target
molecule/organic compound conjugate may be directly subjected to mass
spectrometry
or may be fragmented and the fragments then subjected to mass spectrometry for
identification of the organic compound that bound to the target molecule. The
success
of mass spectrometry analysis of the intact target proteinlorganic compound
conjugate
or fragments thereof will depend upon the nature of the target molecule and
can be
determined on an empirical basis.
In addition to the use of mass spectrometry, one may employ a variety of other
techniques to identify the organic compound that bound to the biological
target
molecule of interest. For example, one may employ various chromatographic
techniques such as liquid chromatography, thin layer chromatography, and the
like, for
separation of the components of the reaction mixture so as to enhance the
ability to
identify the bound organic molecule. Such chromatographic techniques may be
employed in combination with mass spectrometry or separate from mass
spectrometry.
One may optionally couple a labeled probe (fluorescently, radioactively, or
otherwise) to
the organic compound library members or biological target molecule so as to
facilitate
its identification using any of the above techniques. Other techniques that
may find use
CA 02355215 2001-06-08
WO 00/39585 PCT/US99/30960
for identifying the organic compound that bound to the target biomolecule
include, for
example, nuclear magnetic resonance (NMR), capillary electrophoresis, X-ray
crystallography, and the like, all of which will be well known by those
skilled in the art.
In a particularly preferred embodiment, the identification of the bound
organic
compound is facilitated by attaching either or both of (a) the biological
target molecule
or multimeric form thereof or (b) the organic compound library members or
multivalent
forms thereof to a solid matrix material prior to the step of combining those
components. When the biological target molecule or multimeric form thereof is
attached
to a solid matrix material during the step of combining that solid phase-
linked target with
members of an organic compound library, the library members which bind to the
site of
interest on the target necessarily become themselves linked to the solid
matrix material
which may then be washed to remove any unbound reaction components, thereby
facilitating the subsequent identification of the bound members. Along the
same vein,
when the organic compound library members or multivalent forms thereof are
attached
to a solid matrix material during the step of combining those members with the
biological target molecule or multimeric form thereof, the target molecules
that bind to a
solid phase-linked library member necessarily become linked to the solid
matrix material
which may then be washed to remove unbound contaminants, thereby facilitating
the
identification of the bound members. In another preferred embodiment,
different organic
compound library members may be covalently attached to spatially distinct
regions of
the solid matrix material, wherein binding of the target molecule to any
specific region of
the solid matrix will then necessarily identify the organic compound being
bound.
"Solid matrix materials" that find use in the present invention include all of
those
solid matrix materials that are known in the art and to which biological
target molecules
and/or organic compounds may be covalently immobilized including, for example,
those
materials that are routinely employed in the various types of chromatography,
affinity
purification, or any other techniques that requires that a ligand or potential
ligand
molecule be covalently immobilized on a solid substrate. The solid matrix
materials
employed herein may be organic or inorganic in nature and may be, for example,
formed from any resin material which will support the attachment of a
biological target
16
CA 02355215 2001-06-08
WO 00/39585 PCT/US99/30960
molecule or organic compound as described above. For example, synthetic
polymer
resins such as poly(phenol-formaldehyde), polyacrylic, or polymethacrylic acid
or nitrite,
amine-epichlorohydrin copolymers, graft polymers of styrene on polyethylene or
polypropylene, poly(2-chloromethyl-1,3-butadiene), poly(vinylaromatic) resins
such as
those de rived from styrene, alpha-methylstyrene , chlorostyrene,
chloromethylstyrene,
vinyltoluene, vinyinaphthalene o~r vinylpyridine, corresponding esters of
methacrylic
acid, styrene, vinyltoluene, vinyinaphthalene, and similar unsaturated
monomers,
monovinylidene monomers including the monovinylidine ring-containing nitrogen
heterocyclic compounds and copolymers of the above monomers are all suitable.
Techniques for the preparation of such solid matrix materials may be found,
for
example, in Ikada et al., Journal of Polymer Science 12:1829-1839 (1974) or as
described in U.S. Patent No. 4,382,124 to Meitzner et al. Other techniques for
the
synthesis of such solid matrix materials can be found in U.S. Patent Nos.
3,915,64,
3,918,906, 3,920,398, 3,925,019 and the monograph "Dowex: Ion Exchange" 3rd.
edition, (1964) published by the Dow Chemical Company, Midland, Michigan.
Additional solid matrix materials that find use in the present invention
include, for
example, biacore, gold-plated carboxymethylated dextran and other gold films,
glass or
glass containing matrices, and the like. Preferably, the molecule being
immobilized on
the solid matrix material and the solid matrix material possess or are
modified to
possess compatible chemical functionalities such that covalent bonding
therebetween
may be easily accomplished. In this regard, techniques for immobilizing
ligands on a
solid matrix material are well known in the art and will depend upon the
chemical
nature of the components being linked. Detailed conditions for immobilizing
ligands
onto solid matrix materials may be determined in an empirical manner without
undue
experimentation. Examples of linkages and chemistries that may be employed for
covalently linking a target molecule or organic compound library members to a
solid
matrix material include, for example, linkage through sulfhydryl groups,
linkage through
NHS-ester groups, reductive aminations between aldehydes and ketones and
amines
(March, Advanced Organic Chernistry, John Wiley & Sons, New York, 4th edition,
1992,
pp.898-900), alternative methods for preparing amines (March et al., supra, p.
1276),
17
CA 02355215 2001-06-08
WO 00/39585 PCT/US99/30960
reactions between aldehydes and ketones and hydrazine derivatives to give
hydrazones and hydrazone derivatives such as semicarbazones (March et al.,
supra,
pp.904-906), amide bond formation (March et ai., supra, p.1275), formation of
ureas
(March et al., supra, p.1299), formation of thiocarbamates (March et al.,
supra, p.892),
formation of carbamates (March et al., supra, p.1280), formation of
sulfonamides
(March et al., supra, p.1296), formation of thioethers (March et al., supra,
p.1297),
formation of disulfides (March et al., supra, p.1284), formation of ethers
(March et al.,
supra, p.1285), formation of esters (March et al., supra, p.1281 ), additions
to epoxides
(March et al., supra, p.368), additions to aziridines (March et al., supra,
p.368),
formation of acetals and ketals (March et al., supra, p.1269), formation of
carbonates
(March et al., supra, p.392), formation of enamines (March et al., supra,
p.1284),
metathesis of alkenes (March eat al., supra, pp.1146-1148 and Grubbs et al.,
Acc.
Chem. Res. 28:446-452 (1995)), transition metal-catalyzed couplings of aryl
halide
and sulfonates with alkenes and acetylenes (e.g., Heck reactions) (March et
al., supra,
pp.717-178), the reaction of aryl halides and sulfonates with organometallic
reagents
(March et al., supra, p.662), such as organoboron (Miyaura et al., Chem. Rev.,
95:2457
(1995)), organotin, and organozinc reagents, formation of oxazolidines (Ede et
al.,
Tetrahedron Letts. 38:7119-7122 (1997)), formation of thiazolidines (Patek et
al.,
Tetrahedron Letts. 36:2227-22:30 (1995)), amines linked through amidine'groups
by
coupling amines through imidoesters (Davies et al., Canadian J. Biochem.
50:416-422
(1972)), and the like. In fact, covalent immobilization of a target molecule
or organic
molecule library members to a ;solid matrix material may be accomplished if
any
compatible chemically reactive groups exist therebetween.
An additional aspect that facilitates the ability to detect binding between
the
biological target and organic compound library members is to increase the
density or
local concentration of the solid matrix bound component, preferably the
density of the
organic molecule library components as bound to the solid matrix. For example,
in
embodiments where organic compound library members are covaiently attached to
a
solid matrix material and a multimeric form of a biological target molecule is
passed
over the solid matrix to allow binding therebetween, one may facilitate the
ability to
18
CA 02355215 2001-06-08
WO 00/39585 PCT/US99I30960
detect binding and to identify the bound components by employing a relatively
high
local concentration of library members in the matrix. Such concentrations may
be
empirically determined.
Additional embodiments of the present invention are directed to solid matrix
materials having at least first and second members of a library of organic
compounds or
multivalent organic compounds covalently bound thereto. Such solid matrix
materials
will find use, for example, in the methods of the present invention.
EXPERIMENT 1
To determine whether activated dextrans would be a suitable soluble scaffold
for
polyvalent display of small molecules, biotin was conjugated to aldehyde-
activated
dextrans and tested for avidity in binding to immobilized anti-biotin
monoclonal
antibodies. To examine the independence of avidity on valency of display,
dextrans
were prepared having variable numbers of conjugated biotins per molecule, and
these
were tested in an anti-biotin binding assay.
Preparation of biotin-dextran conjugates
Aldehyde-activated dextran (Pierce Biochemicals) was reacted with mixtures (10
total equivalents per aldehyde group) of varying ratios of 1-
aminopropoxylamine
(serving as a bifunctional linker) and methoxylamine (serving as a capping
agent), in
0.3 M sodium acetate, pH 4.7, for 16 h at room temperature. Reactions were
desalted
through NAP-5 columns (Pharmacia) into PBS buffer (10 mM sodium phosphate, 150
mM NaCI, pH 7.4), and amine contents determined by Fluoraldehyde assay
(Pierce).
The dextran solutions were trE:ated with 5 equivalents (per maximal
concentration of
amine linker + ~0 mM) of the sulfo-N-hydroxysuccinnamide ester of biotin
(which was
prepared by reaction of biotin with 1 equivalent each of EDC and sulfo-NHS in
DMSO
for 1 hour at room temperature) overnight at 4°C. The biotin content of
each conjugate
was determined by titration with HABA-avidin complex, following the
manufacturer's
instructions (Pierce).
19
CA 02355215 2001-06-08
WO 00/39585 PCT/US99/30960
Anti-biotin MAb binding assay
Nunc Maxisorp 96-well plates were precoated with 5 Ng/ml goat anti-mouse Fc
polyclonal antibody (Boehringer Mannheim; 100 pllwell in 50 mM sodium
carbonate
buffer, pH 9.6) overnight at 4°C. 4'llells were blocked with superblock
in PBS (Pierce) for
1 hour at room temperature, washed (PBS plus 0.05% Tween 20), coated with
mouse
anti-biotin ascites fluid (Sigma, 3..8 Ng/ml in total IgG) in binding buffer
(superblock plus
0.05% Tween 20) for 1 hour at room temperature, and washed again. Serial
dilutions of
biotindextran conjugates were added, followed by biotin-horseradish peroxidase
(HRP)
conjugate at a concentration predetermined to give yield subsaturating
binding. After 2
hours at room temperature, wells were washed and assayed for HRP activity.
ICso
values were determined from 4-parameter fits of displacement plots using
Kaleidagraph
(Synergy Software). The amine content of the dextrans prior to reaction with
sulfo-NHS-biotin were calculated assuming 100% recovery of dextran from the
desalting column, and a dextran rnolecular weight of 40,000.
Results
Plots of the IC5° concentrations, on a per biotin-basis (biotin
concentrations
determined by direct assay), for anti-biotin binding versus the number of
amine linkers
per dextran precursor to the conjugates are shown in Figure 1. The optimal
avidity
effect was seen from the conjugate that was derived from the .dextran
containing 60
amine linkers per molecule, and amounted to an 80-fold decrease in the
apparent biotin
ICso relative to that of free biotin. The inhibition seen in the dextran
sample having no
available amine groups is due to the presence of free biotin that arose from
incomplete
desalting of the biotin conjugates reactions. The per-biotin ICso of this
material (64 nM)
corresponds well with the ICso determined for free biotin (61 nM).
EXPERIMENT 2
Purpose: To evaluate whether binding of weak affinity ligands can be detected
using a protein possessing multiple binding sites. A commercially available
anti-Biotin
antibody (Sigma, clone BN-34) binds biotin and also a biotin analog known as
CA 02355215 2001-06-08
WO 00/39585 PCTIUS99/30960
desthiobiotin with affinities of 100 nM and 130 NM respectively. The small
molecule
ligands biotin and desthiobiotin will be immobilized on a solid support
through a
covalent linkage. The affinity for desthiobiotin (130 NM) is substantially
weaker than
what is expected to be readily detected using a standard ELISA based assay
with a 1:1
protein to ligand binding interaction.
Immobilization and multivalent detection of small molecule ligands:
Method A. To a Covalink 96 well plate (Nunc) was added 100 NL of either
biotin-NHS, desthiobiotin-NHS, or NHS-propionate at a concentration of 10 uM
in
phosphate buffered saline pH 7.2 (PBS) containing 1 % DMSO. After 1 hour
incubation
at room temperature, the plate was washed on a plate washer and blocked with
0.05%
Tween 20 in Superblock (Pierc;e) for 2 hours. An appropriate concentration of
a-biotin
IgG (Sigma) together with a-Fc-HRP conjugate IgG (Boerhinger) as previously
determined to be subsaturating by titration was added for 1 hour. The plate
was
washed, and TMB substrate (Fierce) was added and signal allowed to develop
following the manufacturers instructions (Pierce).
Method B. Amino PEGA resin (0.25 mmol) (Calbiochem) was swelled in
dimethylformamide and treated with a premixed solution of biotin (1 mmoi),
HBTU (1
mmol) and diisopropylethylamine (1.5 mmol) in DMF. After mixing for 1 hour the
resin
was drained and washed with DMF and dichloromethane. The same method was used
to covalently link desthiobiotin and acetic acid. Resin was aliquoted into
separate wells
of a 96 well polystyrene filter plate and an appropriate concentration of a-
biotin IgG
(Sigma) together with a-Fc-HRP conjugate IgG (Boerhinger) in PBS with 0.05%
Tween
20 was added. After 1 hour the resin was drained, washed with PBS containing
0.05%
Tween 20, and TMB substrate was added according to the manufacturers
instructions
(Pierce).
Method C. Glass microscope slides (VWR) were cleaned using a mixture of
sulfuric acid and hydrogen peroxide. The slide were treated with a 5% solution
of
2'i
CA 02355215 2001-06-08
WO 00/39585 PCT/US99/30960
aminopropyl triethoxy silane in 95% ethanol for 1 hour. After treatment, the
slides were
washed with ethanol and annealed at 120°C for 2 hours. The slides were
derivatized at
specific sites by the addition of a 10 NM solution of either biotin-NHS,
desthiobiotin-NHS, or NHS-acetate in PBS containing 1 % DMSO for 1 hour. The
slides
were then washed with water and methanol. Sections of the glass slide were cut
into
pieces according to where the different small molecules were coupled, these
pieces
were distributed into a 96 well polystyrene filter plate and an appropriate
concentration
of a-biotin IgG (Sigma) together with a-Fc-HRP conjugate IgG (Boerhinger) in
PBS with
0.05% Tween 20 was added. After 1 hour the resin was drained, washed with PBS
containing 0.05% Tween 20, and TMB substrate was added according to the
manufacturers instructions (Pierce).
Results
Covalink plates derivatized as described in Method A were first titrated
against
the a-Biotin antibody. Figure 2 below shows the plots for recognition of
either biotin,
desthiobiotin, unreacted well~~ and wells blocked with propionic acid. The
antibody
readily detects both the immobilized biotin and the immobilized desthiobiotin
at sub nM
concentrations. The relatively similar sensitivity for detection of both
biotin and
desthiobiotin is indicative of the avidity effect. Even though the affinity of
the antibody
for each of these two ligands differs by 1000 fold, they are both readily
recognized. The
detection system is potentialh/ tetravalent as the a-Fc-HRP detection antibody
can
dimerize the a-Biotin antibody which already possesses two binding sites. The
potential
affinity of a tetravalent antibody is well below the sensitivity of this assay
which is most
likely titrating the affinity of the a-Fc-HRP detection antibody for the a-
Biotin antibody.
These results indicate multivalent binding systems will give very high
observed binding
sensitivity and may require specialized assays to accurately quantitate the
results.
In Figure 3, soluble biotin is titrated against a constant concentration of a-
Biotin
antibody and a-Fc-HRP detection antibody to evaluate the amount of free biotin
required to compete with either immobilized biotin or immobilized
desthiobiotin. As
expected, the concentration of soluble biotin (approximately 1 NM) required to
compete
22
CA 02355215 2001-06-08
WO 00/39585 PCT/US99/30960
with the immobilized biotin is much higher than the measured affinity of the
antibody for
biotin (100 nM). The second curve shows the titration of soluble biotin in a
well
containing immobilized desthiobiotin. Much lower concentrations of soluble
biotin
(approximately 50 nM) are required to compete with this interaction due to the
substantially lawer affinity of the antibody for desthiobiotin.
Similar experiments canducted with supports constructed using Method B and
Method C gave similar results.
EXPERIMENT 3
Techniques to prapare Multimeric Proteins
Method A. To a solution of IL-4 (50 NM, MES pH 6.0) was treated with 2
equivalents of biotin-NHS (Pierce) at 4°C for 12 hours. Analysis by
mass spectrometry
revealed a mixture of approximately equal parts of mono-biotinylated and non-
mod~ed
IL-4. The protein was separated from unreacted biotin by purification on a NAP-
5
column (Pharmacia). The mixture of protein was further purified by incubation
with 0.2
equivalents of neutravidin (Pierce) to form the tetravalent complex (4:1
IL-4:neutravidin). The desired tetravalent complex was purified by size
exclusion
chromatography using a Bio-Silect SEC 125-5 column (Bio-Rad). Additionally,
site
specific biotinylation was achieved by expression of a cysteine mutant A104C
(other
sites could easily be used for the same purpose) and reaction with Biotin HPDP
(Pierce). This strategy is potentially important to prevent blocking of the
target protein
binding site through modification.
Method B. Aldehyde-activated dextran (Pierce) was treated with 3 equivalents
of
pyridyidithiopropionyl hydrazide (PDPH, Pierce), in 50 mM MES pH 6.0 for 12
hours.
The reaction was purified using a NAP-5 column (Pharmacia) and the number of
thiopyridyl groups incorporated could be quantified by reduction of the
disulfcde with
DTT and monitoring the UV absorbance at 340 nM (thiopyridone). The number of
aldehyde sites that were derivatized with the thiopyridyl functionality could
be adjusted
by adding a capping hydrazide reagent, such as semicarbazide, to react with a
certain
23
CA 02355215 2001-06-08
WO 00/39585 PCT/US99/30960
percentage of the available aldehydes. The thiopyridyl-dextran was then
reacted with
IL-4 A104C to generate IL-4 derivatized dextrans. These constructs were
purified by
size exclusion chromatography using a Bio-Silect SEC 125-5 column (Bio-Rad).
Method C. To prepare a lipid that could be readily derivatized, 10 mg of
phosphatidy! ethanolamine (Sigma, P7943) was dissolved in chloroform (2 mL)
with 3
equivalents of diisopropylethylamine. To incorporate a thiol specific reactive
functionality, 5 mg of a bifunctional maleidimide-NHS ester crosslinker (BMPS,
Pierce)
was added. After 2 hours at room temperature, the reaction was concentrated
and
purified on a silica get column. Liposomes were prepared using extrusion
techniques
with a mixture of the maieimide derivatized lipid at levels or 1 % to 5% with
phosphatidylcholine making up the remainder. Liposomes were purified by size
exclusion chromatography using a Bio-Silect SEC 125-5 column (Bio-Rad).
Proteins
containing a free cysteine, such as tL-4 A104C or the anti-biotin antibody
derivatized
with a free thiol using the reagent SATA (Pierce), were covalently linked to
the extewor
of the liposome through reaction between the thiol and the maleimide.
The foregoing description details specific methods which can be employed to
practice the present invention. Having detailed such specific methods, those
skilled in
the art will well enough know how to devise alternative reliable methods at
arriving at
the same information in using the fruits of the present invention. Thus,
however,
detailed the foregoing may appear in text, it should not be construed as
limiting the
overall scope thereof; rather, the ambit of the present invention is to be
determined only
by the lawful construction of the appended claims. All documents cited herein
are
expressly incorporated by reference.
24
