Note : Les descriptions sont présentées dans la langue officielle dans laquelle elles ont été soumises.
CA 02446921 2003-10-16
WO 02/083708 PCT/IL02/00307
SELECTIVE COVALENT-BINDING COMPOUNDS HAVING THERAPEUTIC
DIAGNOSTIC AND ANALYTICAL APPLICATIONS
FIELD OF THE INVENTION
The present invention pertains to compounds, herein designated COBALTs, or
Covalent-Binding Antibody-Like Trapping or -Trap, characterized by specific
binding to a
target with antibody or antibody-like affinity for a desired, preselected,
target substance (a
small molecule; a macromolecule such as a protein, a carbohydrate, a nucleic
acid, etc.; a cell;
a viral particle; etc.) and which contain chemical groups that allow these
COBALTs to form
strong, specific bonds, such as irreversible covalent bonds, with the target
substance for
which the COBALT was specifically designed.
BACKGROUND OF THE INVENTION
The superior selectivity of antibodies, especially monoclonal antibodies, has
led to
their widespread use in many biomedical and other applications such as immuno-
assay
diagnostics and therapeutic drugs. However, antibodies have a number of
drawbacks
compared to synthetic compounds including: the difficulty of optimization or
modification via
chemical modification; the necessity of their being maintained at low
temperatures; their short
shelf life (subject to thermal and microbial degradation); bio-contamination;
high production
costs. In addition, since antibodies are generally produced using the immune
system, the
genetic limitations of the animals used may restrict the binding site
variability. This last
restriction may be overcome by using phage-display and other genetic
engineered methods for
antibody production but here, too, it has not always been possible to raise
effective antibodies
for every desired target material. While some antibodies have very high
affinity constants
others do not have the degree of binding needed for a particular, desired
application. As a
result of these and other considerations, significant efforts have been
devoted to develop
1
CA 02446921 2003-10-16
WO 02/083708 PCT/IL02/00307
antibody mimics. However, the antibody mimics produced thus far have had
limited practical
application. In particular, these antibody mimics suffer from one or more of
the following:
lack of general applicability; insufficient selectivity; low affinity
constants. Thus, there is a
clear need for improved antibody mimics and this invention addresses those
needs by
providing antibody mimics, here termed COBALTs, as described above, which have
broad
application to biomedical and other areas.
Biological recognition, i.e., the specific, attractive interactions between
the myriad
substances of nature which involve selective interactions and which are
essential for life, is
universally based on reversible, non-covalent interactions. Even the
"essentially irreversible",
extraordinarily tight binding interaction between biotin and avidin, Ka ~
1015, involves no
covalent bonds. However, for many diagnostic, therapeutic and other
applications,
advantages would accrue to synthetic substances which exhibit antibody-like
specific
recognition for a given biological substance but which would bind covalently,
i.e., essentially
irreversibly under normal circumstances, to that selected substance.
The background art does not teach or suggest a general approach to enable the
design
of covalently bound 'traps', 'tags' or 'labels' for particular biological
substances, (beyond the
restricted, narrowly defined enzyme inhibitors, which form covalent bonds
within active sites,
or activated ligands which form covalent bonds within antibody or receptor
binding sites).
SUMMARY OF THE INVENTION
The present invention overcomes the deficiencies of the background art by
providing a
variety of substances which have an affinity for a desired, preselected,
target substance (a
small molecule; a macromolecule such as a protein, a carbohydrate, a nucleic
acid, etc.; a cell;
a viral particle; etc.) and which contain chemical groups that allow these
substances to form
strong bonds, such as irreversible covalent bonds, with the desired target
substance. These
substances are termed COBALTs, or Covalent-Binding Antibody-Like Trapnin~ or -
Trap,
2
CA 02446921 2003-10-16
WO 02/083708 PCT/IL02/00307
as they are characterized by tight, specific binding to a target, with
antibody or antibody-like
affinity. Since these substances according to the present invention preferably
feature a
mechanism in which a covalent bond forms specifically between the desired
target and the
COBALT substance, the letter "T" of COBALT may also stand for "tagging" or
"tag", as the
covalent bond may cause binding between the desired target and the COBALT
substance to
be irreversible.
The present invention optionally and more preferably includes a method wherein
a
target species (as above) is chosen and then, by synthetic chemical procedures
and
modifications, novel substances (COBALTs) are obtained or selected from
combinatorial
libraries that exhibit selective and covalent binding to the preselected
target species. The
COBALT substances themselves may optionally be categorized according to
different types
of structures, including but not limited to, molecularly imprinted polymers,
cyclodextrins,
triazines and peptides, which may be selected from these combinatorial
libraries of
chemically reactive substances that can covalently react with a target
substance. A
molecularly imprinted polymer (MIP), as discussed in greater detail below, is
typically
synthesized in the presence of the target molecule, and hence is designed to
bind specifically
to that target molecule. The peptide derivative may include at least one of
cyclic, linear and
modified peptides and derivatives thereof.
The COBALT substances of the present invention are preferably designed to be
highly
specific, and to bind to the target substance with a high degree of
specificity, regardless of the
category or type of COBALT substance which is used. The uses of the COBALTs
include
diagnostic, analytical, therapeutic and industrial applications.
Without wishing to be limited to a single hypothesis, the binding mechanism
between
the COBALT and the target substance may occur as follows. An initial non-
covalent
complex may be pictured as forming rapidly between the COBALT and the target;
other
substances, different from the target substance, may be expected to either not
form such
3
CA 02446921 2003-10-16
WO 02/083708 PCT/IL02/00307
complexes or alternatively to form complexes with a much shorter lifetime (in
other words,
their stability will be much lower) than that of the COBALT-target substance
complex. The
orientation of the two components, the COBALT and the target substance, in the
complex is
such that a chemical reaction can rapidly ensue to produce a new substance
having a strong,
covalent bond (or bonds) between the two initial components. Complexes formed
by the
COBALT and substances other than the target substance, even if they exhibit
some stability,
may either not have the orientation or lifetime required for covalent bond
formation or the
reaction rate would be expected to be far less than that between the COBALT
and the desired
target substance. Use of the term complex, does not rule out the possibility
of more than one
complex between the COBALT and the desired target substance. Indeed, for some
of the
applications described herein, there may be many different complexes formed,
but the overall
result is that the COBALTs are expected to display a preferred affinity for
the desired target
substance, relative to other materials, and to preferentially react chemically
with the desired
target substance.
Examples of situations where the chemically reactive, COBALT, approach may
have
important therapeutic and other applications include increased binding of a
target molecule
compared to conventional, noncovalently binding agents. The irreversible
chemical reaction
can eventually tag or trap selectively essentially all targets that initially
bind, whereas with
conventional, noncovalent binding, once the equilibrium constant is reached,
no additional
target molecules may be trapped. Illustrative potential applications include
assay detection at
low concentrations of the target analyte or the more effective therapeutic
action of a
COBALT allowing lower dosages of more effective drugs.
According to the present invention, there is provided a compound for
specifically
binding a molecular structure, comprising a selective and chemically reactive
compound with
an enhanced apparent affinity constant. Preferably, the compound is an
antibody mimic.
Optionally and more preferably, the compound is a molecularly imprinted
polymer (MIP), the
4
CA 02446921 2003-10-16
WO 02/083708 PCT/IL02/00307
MIP being modified through chemical activation in order to react covalently
with the
molecular structure.
Alternatively, the target substance is a chemically reactive substance and the
COBALT is designed or selected so as to selectively react with the target. An
important
implementation of the present invention involves MIPs designed to react
covalently with
reactive organophosphate toxins (OP-agents) such as DFP, soman, sarin, VX,
etc.
According to one implementation of the present invention, this COBALT, which
is
preferably an MIP, is chemically modified by including an isocyanate, or
isothiocyanate,
functional group. Alternatively and preferably, the MIP is chemically modified
by including
an isocyanate, or isothiocyanate functional group and the molecular structure
is a steroid.
More preferably, the steroid is cholesterol or a bile acid. Also alternatively
and preferably, the
MIP contains a nucleophile such as at least one of an oxime, a hydroxylamine,
a hydrazine, a
phenol and a 2-iodosobenzoic acid, and so forth, for specific and tight
binding to
organophosphates. Also alternatively and preferably, the MIP contains two or
more boronic
acids or two or more aldehyde functions for specific and tight binding to
carbohydrates.
Other optional functional groups which the COBALT may contain, preferably as
part
of the MIP implementation, include but are not limited to chloromethylphenyl
and 2,5-diketo-
N-phenyltriazoline or any other triazine related functional group; functional
group includes at
least one of an alpha-halomethyl ether, wherein a halogen moiety may be
fluoro, chloro,
bromo or iodo; a beta-haloethyl ether, wherein a halogen moiety may be chloro,
bromo or
iodo; and a halomethylaryl, wherein a halogen moiety may be fluoro, chloro,
bromo, or iodo;
or carboxylic acid chloride or an activated carboxylic acid (e.g., 4-
nitrophenyl ester, N-
hydroxysuccinimide ester, pentafluorophenyl ester, etc.). This list of
functional groups is not
intended to be inclusive but illustrates the scope of appropriate functional
groups.
According to another embodiment of the present invention, the COBALT compound
is a cyclodextrin derivative which has been chemically modified to react
covalently with the
5
CA 02446921 2003-10-16
WO 02/083708 PCT/IL02/00307
target molecular structure. According to one implementation of the present
invention, this
COBALT, which is preferably a cyclodextrin, and more preferably is an alpha-,
beta-, or
gamma-cyclodextrin, includes at least one or more amino groups (replacing one
or more of
the hydroxyl groups). More preferably, one or more of the hydroxyl and amino
groups are
linked directly to arylcarboxylic acid groups, arylmethylcarboxylic acid (or
other
arylalkylcarboxylic acid) groups via amide or ester bonds. Also more
preferably, one or more
of the hydroxyl and amino groups are linked directly to aryl or arylmethyl
groups, where aryl
refers to, but is not limited to, phenyl and substituted phenyl, pyridyl and
substituted pyridyl,
naphthyl and substituted naphthyl groups.
According to yet another embodiment of the present invention, the COBALT
compound is a triazine derivative being chemically modified to react
covalently with the
target molecular structure. According to one implementation of the present
invention, this
COBALT, is preferably a triazine, and more preferably is a derivative of 2,4,6-
trichloro-1,3,5-
triazine (cyanuric chloride) wherein one or more of the chloro groups are
replaced by
alcohols, phenols or preferably by amine-containing derivatives, as described,
for example, by
R-X Li, V. Dowd, D.J. Stewart, S.J. Burton and C.R. Lowe (1998) Nature
Biotechnology 16,
190-195, and references therein.
According to still another embodiment of the present invention, the COBALT
compound is a peptide or derivative thereof being chemically modified to react
covalently
with the target molecular structure. Preferably, the peptide derivative is at
least one of cyclic,
linear and modified peptides and derivatives thereof.
According to yet another embodiment of the present invention, the COBALT
compound is an antibody or antibody fragment or derivative thereof, being
chemically
modified to react covalently with the target molecular structure.
Preferably, the chemical modification to the cyclodextrin, triazine, peptide
or antibody
includes introduction of one or more isothiocyanate groups. Other optional
chemical
6
CA 02446921 2003-10-16
WO 02/083708 PCT/IL02/00307
modifications to the cyclodextrin, triazine, peptide or antibody include but
are not limited to
chloromethylphenyl and 2,5-diketo-N-phenyltriazoline or any other triazine
related functional
group; functional group includes at least one of an alpha-halomethyl ether,
wherein a halogen
moiety may be fluoro, chloro, bromo or iodo; a beta-haloethyl ether, wherein a
halogen
moiety may be chloro, bromo or iodo; and a halomethylaryl, wherein a halogen
moiety may
be fluoro, chloro, bromo, or iodo; or carboxylic acid chloride or an activated
carboxylic acid
(e.g., 4-nitrophenyl ester, N-hydroxysuccinimide ester, pentafluorophenyl
ester, etc.). This
list of modifications is not intended to be inclusive but illustrates the
scope of appropriate
functional groups.
Alternatively and preferably, the chemical modification to the cyclodextrins
includes
two or more boronic acids or two or more aldehyde functions for specific and
tight binding to
carbohydrates.
According to another embodiment of the present invention, there is provided a
combinatorial library of compounds containing chemically reactive groups
screened for
selectivity and chemical reaction with the molecular structure as a target,
for creating the
compound of the present invention.
According to still another embodiment of the present invention, there is
provided a
compound for specifically reacting at any site on a target molecule, such that
these sites are
not limited to the conventional active site or ligand binding site of the
target.
Hereinafter, the term 'antibody mimic' includes but is not limited to, any
synthetic
substance such as an MIP or a triazine derivative, or a derivative of a
natural product such as
a cyclodextrin or a peptide, which has been designed or selected so as to
display selective
affinity for a given target structure.
CA 02446921 2003-10-16
WO 02/083708 PCT/IL02/00307
BRIEF DESCRIPTION OF THE DRAWINGS
The invention is herein described, by way of example only, with reference to
the
accompanying drawings, wherein:
FIG. 1 is a schematic flow chart, showing illustrative preparation of
"chemically
reactive" molecularly imprinted polymers (MIP-based COBALT) for the selective
covalent
binding of hydroxyl-containing target substances, ROH;
FIG. 2 shows a schematic depiction of the preparation of a conventional
cholesterol-
binding, amino-containing MIP (MS50);
FIG. 3 shows a schematic view of the two MIPs, one binding non-covalently and
one
binding covalently, the latter being a COBALT;
FIG. 4 shows the calibration curve for cholesterol;
FIGS. 5 and 6 show the Scatchard and binding isotherm plots respectively for
the third
Example;
FIG. 7 is related to the preparation of crMIP - MS71;
FIG. 8 shows the IR Spectrum of MS71 (when the maximum
conversion into NCO is reached);
FIG. 9 describes the overall approach for developing MIP-based COBALTS
for the binding of toxic organophosphates, in which functional monomers, A,
are
polymerized with a large excess of crosslinker, porogen, etc, to create an
MIP, B
(MIP-B), that is hydrolyzed to remove the phosphonate, leaving behind
complementary cavities, containing a nucleophile, X, in MIP-C, that
selectively reacts
covalently with DFP to form D (MIP-D). It should be noted that the
nucleophile, X, in
MIP-C differs from the functional group Z in the monomer as well as Y
following
reaction with DFP. Inclusion of other functional monomers, such as one
containing a
basic group to bind the HF released, was also carned out. Note, too, that the
reaction
of MIP-C with DFP is initially an equilibrium binding reaction, having a given
Ka,
8
CA 02446921 2003-10-16
WO 02/083708 PCT/IL02/00307
which is followed by an irreversible covalent binding reaction to give MIP-D.
Also,
the nature of MIP-D is such that the phosphate is readily hydrolyzed to
dialkyl
phosphoric acid and MIP-C, which now represents a catalytic cycle;
FIG. 10 depicts representative structures of synthesized functional monomers
used for the DFP-binding MIPs that were prepared,
FIG. 11 shows the synthetic scheme used for the preparation of the 4-
vinylbenzaldehyde oxime phosphate and phosphonate functional monomers 3, 9,
and 10;
FIG. 12 shows a calibration curve for converting percent BChe inhibition to
DPFP
concentration;
FIG. 13 shows a calibration curve for DCP concentration;
and
FIG. 14 illustrates the approach taken for preparing widely varying COBALTs
using
combinatorial libraries based on cyclodextrins having isothiocyanates for
covalent reaction
with the target substances, in which R1, R2 = phenyl, substituted phenyl,
naphthyl,
substituted naphthyl, etc; there is almost no limit to the variability of the
structures that can be
prepared by using this scheme, by adding additional R groups, using different
cyclodextrin
derivatives, (eg, diamines, etc) and so forth.
DESCRIPTION OF THE PREFERRED EMBODIMENTS
The present invention is of a variety of substances which have an affinity for
a desired,
preselected, target substance (a small molecule; a macromolecule such as a
protein, a
carbohydrate, a nucleic acid, etc.; a cell; a viral particle; etc.) and which
contain chemical
groups that allow these substances to form strong bonds, such as irreversible
covalent bonds,
with the desired target substance, which as previously described may be termed
COBALTs,
Covalent-Binding Antibody-Like Trapping or -Trap. The present invention
includes a
mechanism wherein a target species (as above) is chosen and then, by synthetic
chemical
9
CA 02446921 2003-10-16
WO 02/083708 PCT/IL02/00307
procedures and modifications, novel substances (COBALTs) are obtained that
exhibit
selective and covalent binding to the preselected target species. The uses of
the COBALTs
include diagnostic, analytical, therapeutic and industrial applications.
Without wishing to be limited to a single hypothesis, the binding mechanism
between
the COBALT and the target substance may occur as follows. An initial non-
covalent
complex may be pictured as forming rapidly between the COBALT and the target;
other
substances, different from the target substance, may be expected to either not
form such
complexes or alternatively to form complexes with a much shorter lifetime (in
other words,
their stability will be much lower) than that of the COBALT-target substance
complex. The
orientation of the two components, the COBALT and the target substance, in the
complex is
such that a chemical reaction can rapidly ensue to produce a new substance
having a strong,
covalent bond (or bonds) between the two initial components. Complexes formed
by the
COBALT and substances other than the target substance, even if they exhibit
some stability,
may either not have the orientation or lifetime required for covalent bond
formation or the
reaction rate would be expected to be far less than that between the COBALT
and the desired
target substance. Use of the term complex, does not rule out the possibility
of more than one
complex between the COBALT and the desired target substance. Indeed, for some
of the
applications described herein, there may be many different complexes formed,
but the overall
result is that the COBALTs are expected to display a preferred affinity for
the desired target
substance, relative to other materials, and to preferentially react chemically
with the desired
target substance.
Examples of situations where the chemically reactive, COBALT, approach may
have
important therapeutic and other applications include increased binding of a
target molecule
compared to conventional, noncovalently binding agents. The irreversible
chemical reaction
will eventually trap all targets that initially bind, whereas with
conventional, noncovalent
binding, once the equilibrium constant is reached, no additional target
molecules may be
CA 02446921 2003-10-16
WO 02/083708 PCT/IL02/00307
trapped. Illustrative potential applications include assay detection at low
concentrations of
the target analyte or the more effective therapeutic action of a COBALT
allowing lower
dosages of more effective drugs.
To illustrate one advantage of the notion of a chemically reactive binding
site, where
the reversible binding of a ligand is followed by a covalent bond-forming
chemical reaction
between the receptor and the ligand, an example may be considered with a
molecularly
imprinted polymer (MIP). The subject of molecularly imprinted polymers has
been
extensively reviewed (e.g., G. Wulff, Angew. Chem., Int. Ed. Engl. 1995, 34,
1812-1832;
A.G. Mayyes and K. Mosbach, Trends Anal. Chem. 1997, 16, 321-332; E.N.
Vulfson, C.
Alexander, and M.J. Whitcombe Chem. Brit. 1997, 33, 23-26; K. Haupt and K.
Mosbach,
Trends Biotechnol. 1998, 16, 468-475; Molecular and Ionic Recognition with
Imprinted
Polymers, ACS Symp. Ser. 703; R.A. Bartsch and M. Maeda, Eds.; American
Chemical
Society, Washington, DC, 1998) and a number of patents on this topic have been
issued [e.g.,
US Patent 4127730 (Wulff, G., Sarhan A.); US Patent 5110833 (Mosbach. K.); US
Patent
5630978 (Domb, A.,); US Patent 5587273 (Yan, M. et al.); US Patent 5872198
(Mosbach, K.
et al.)]. All of these background art references are hereby incorporated by
reference as if fully
set forth herein.
For the purposes of the present example, assume that the MIP has been
synthesized in
the presence of the template molecule X; if the MIP binds a ligand X with an
affinity (given
as KD) of 10-5, then at X concentrations much lower than 10-5 M, say, 10-7 M,
essentially no
detectable amounts of X binding can be measured; a sensor based on this MIP
would not be
expected to detect such low levels of analyte, X.
If the reversible (relatively weak) binding of X by the MIP to give a complex,
MIP.X,
were followed by a covalent-bond forming reaction between X and MIP, this MIP
(a
chemically activated MIP - a COBALT) would now be able to detect X, depending
on the
relative rate constants involved, because the concentration of irreversibly
formed MIP-X
11
CA 02446921 2003-10-16
WO 02/083708 PCT/IL02/00307
gradually builds up to much higher values than the equilibrium concentration
of complex
MIP.X.
kon kreact
MIP + X ~ MIP.X ---~ MIP-X
k°ff non-covalent covalently bonded
complex entities
In the case of the MIPs, the system is somewhat analogous to the irreversible
enzyme
inhibitors and antibody affinity labeling reagents. Note that substances
differing from X to an
appreciable extent will tend not to react effectively within the MIP cavity
because, even if
they enter the MIP cavity, they will not have the orientation necessary for
reaction (the
stereochemical or the "spatiotemporal" [Khanjin NA, Snyder JP, Menger FM, J.
Amer. Chem.
Soc. 121 (50): 11831-11846 (1999)] demands are not met). The exact nature of
the
chemically reactive group on the "antibody mimics", the chemically reactive
"receptor", or
COBALT, whether based upon MIPs, peptides, triazines, cyclodextrins, etc.,
will depend on
the requirements and nature of the substance to be bound. The chemical
reactions presented
are illustrative; other reactions are obvious to those skilled in the art and
may be selected
from, but are not limited to, the abundant literature on other reactions where
covalent bonds
are formed. Various illustrative examples of the kinds of chemical reactions
that may be
carried out can be found in the literature on: affinity labeling, bioconjugate
chemistry, and
enzyme-active site and receptor-binding site labelling reactions.
The term "enhanced apparent affinity constant" is used herein to describe the
ability of
the COBALT to bind more of a target substance than a conventional, non-
covalently binding
substance. Since affinity or binding constants are restricted to equilibrium
systems, and since
the use of irreversible inhibitors technically does not allow equilibrium to
be reached, this
definition is a generalization from a classical equilibrium system. However,
the definition
does permit a quantitative estimate of the improved sequestering of a given
target substance to
be made. Thus, if a conventional MIP, after 24 hours, or 48 hr, is determined
to have bound a
12
CA 02446921 2003-10-16
WO 02/083708 PCT/IL02/00307
given fraction of a target substance, the equilibrium binding constant can be
determined for
this equilibrium system. If a chemically reactive MIP, or a COBALT, is
determined to have
bound an increased fraction of a target substance after a similar period of
time has elapsed,
this increased fraction may be termed "enhanced apparent affinity", even
though the
COBALT is a dynamic system and the "enhanced apparent affinity constant" may
vary with
time.
The COBALTs comprise many different classes of laboratory-synthesized
substances
and also include chemically modified monoclonal antibodies, as well. In the
latter case, an
antibody is elicited to an appropriate hapten and then the antibody is
chemically "activated" in
order to convert it to a COBALT, e.g., amine to isothiocyanate; tyrosine to an
o-quinone; thiol
to chloromethylthioether; etc. The antibody-based COBALT will now complex with
the
desired target substance and subsequently bind covalently and irreversibly.
The many substances that are suitable for producing COBALTs include, but are
not
restricted to: (i) molecularly imprinted polymers (MIPs), which have been
prepared in one of
the conventional manners reported in the literature, and then chemically
converted to an
"activated MIP", i.e., a COBALT, which specifically binds to and then reacts,
forming a
covalent bond with the target substance; (ii) peptides, e.g., cyclic peptides,
which have been
modified to contain a reactive functional group, such as an isothiocyanate
group, so that the
peptide will not only bind to a specific target substance but covalently react
with that
substance; (iii) peptide-derivatives of "platform" molecules, such as
cyclodextrins, where the
peptide or cyclodextrin has been chemically activated to specifically bind to,
and covalently
react with, a desired target substance; (iv) non-peptide substances, such as
triazine
derivatives, which, again, have affinity for and chemically react with a given
target substance;
(v) non-peptide-derivatives, such as triazine derivatives, of molecules, such
as cyclodextrins,
where either the triazine derivative or the cyclodextrin has been chemically
activated (or
where both have been activated) to specifically bind to and covalently react
with, a desired
13
CA 02446921 2003-10-16
WO 02/083708 PCT/IL02/00307
target substance. The target substances are not restricted to any class of
material and include
but are not limited to, small molecules such as steroids, sugars, lipids;
macromolecules such
as proteins, carbohydrates and nucleic acids; cell surface substances and
receptors; and other
molecules of biomedical interest. The invention particularly relates to the
use of COBALTs
as drugs and for detection as well as separation applications.
Production of the COBALT may include, but is not limited to, the following
three
approaches, which are given only as non-limiting illustrative examples:
i. Creating a compound having selectivity for the target substance (e.g., an
MIP) and
then 'activating' the compound by introducing a chemically reactive functional
group at an
appropriate, specific locus on the compound in order for covalent bond
reaction to take place
when the target molecule is in contact with the compound. Alternatively, if
the target
substance has a reactive functional group, e.g., an ester, epoxide, disulfide,
etc. or
fluorophosphates or fluorophosphonate, as in the case of chemical warfare
agents, the MIP or
other COBALT can optionally be prepared such that an appropriate functional
group is
present to react with the target.
ii. Chemically modifying a substance, known to be selective for the target
material, so
that it reacts covalently with the target; this substance may also react at a
number of different
sites on the target substance.
iii. Creating a combinatorial library of related or different COBALTs,
designed for
both selectivity and chemical reactivity; these are then screened for optimum
performance
with the target substance.
In addition, in a manner that is reminiscent of the evolution of improved
binding in
antibodies produced by the immune system, the COBALTs may be further
"evolved", i.e.,
chemically changed and further selected in order to obtain improved binding
plus-reacting
substances for the given, desired target material.
14
CA 02446921 2003-10-16
WO 02/083708 PCT/IL02/00307
In the previous description, the COBALT has been described as having one
reactive
group for covalent bond formation with the target substance, but, as
illustrated below, two or
more activated functional groups may be present on each COBALT.
A large variety of suitable functional groupings that can be incorporated so
as to
convert a substance (such as an MIP, a cyclodextrin derivative, etc.) to a
COBALT. The
examples given below are illustrative and not meant to limit the classes,
number or examples.
Functional groups) on Functional groupFunctional groups on
the the target
potential COBALT having on the COBALT substance that enter
a into
degree of affinity for covalent bond-forming
the target reactions
substance with the COBALT
-NHZ -NCO -NCS -NHZ -OH -SH
-OH -O-CHZ-Cl -NHZ -OH -SH -COZH
-OH -O-CHZ CHZ-Cl -NHZ -OH -SH -COzH
methylphenyl (tolyl) chloromethylphenylNHZ -OH -SH -COZH
conjugated dime (Diels-Alder
2,5-diketo-N-phenyltriazoline 2,5-diketo-N- reaction)
phenyltriazoline
-C02H -COCI or active-NHZ -OH -SH
ester
Although boronic acids were used in one of the first examples of an MIP,
tighter
binding of carbohydrates, sugars, glycolipids, etc. may be achieved using two
or more boronic
acid derivatives. Similarly, acetals and ketals have been used in MIPs for
binding to diols,
but the use of two or more aldehydes has not been reported as for the present
invention.
There are numerous examples of previously reported covalent-bond forming
reactions
but these examples all differ from the present invention. Chemically activated
ligands have
been used as enzyme inhibitors (some irreversible inhibitors form covalent
bonds with the
enzyme active site residues; "suicide" inhibitors; etc.), and specific
labelling agents for
receptors and for antibodies; the latter two are often called affinity
labelling approaches.) The
CA 02446921 2003-10-16
WO 02/083708 PCT/IL02/00307
goals in these studies have invariably been to obtain information regarding
the labelled amino
acids, the binding site structure and topology, etc., but some of the
irreversible enzyme
inhibitors are important drugs, e.g., Orlistat.
The present invention may be differentiated from the above examples, in that,
in an
important and major embodiment of the invention, any site on the target
substance is available
for targeting by the chemically reactive enhancement is designed. This is
reminiscent of the
immune system's approach: antibodies elicited to, say, an enzyme, may bind at
any point on
the surface of the enzyme, and, when characterized, are found to be specific
for a given site
on the enzyme, including but not restricted to the enzyme active site. So,
too, the chemically
reactive, covalently binding compounds of the present invention have a priori
random site-
selectivity. Thus, the present invention is not dependent on the availability
of a known or
defined ligand binding site, or any ligand binding site at all, which is an
essential requirement
in all of the above approaches.
The COBALT approach enables feasible structures having chemically reactive
groups,
such as an isothiocyanate, to be designed or discovered, for which the desired
binding plus
covalent bond-forming reaction (one that may react at any site on the target
substance) occurs.
Although the design or discovery of such compounds according to the present
invention may
exploit any knowledge available regarding the target structure to improve
probability of
discovering effective binders, the approach of the present invention is
primarily and
preferably a discovery process which is dependent, as with the antibodies of
the immune
system, on using a large number and variety of potential binding structures.
In addition, the systems associated with the present invention (COBALTs based
upon
molecularly imprinted polymers (MIPs), cyclodextrins, peptides, triazines,
etc.) have not been
previously disclosed in the literature. There are references to peptide
derivatives that form
covalent bonds with enzymes but these are enzyme inhibitors of the type
mentioned above.
Triazine derivatives have been developed for selective binding but these have
been based
16
CA 02446921 2003-10-16
WO 02/083708 PCT/IL02/00307
upon equilibrium, non-covalent binding [e.g., "Design, Synthesis and
Application of a Protein
A mimetic." R-X Li, V. Dowd, D.J. Stewart, S.J. Burton and C.R. Lowe (1998)
Nature
Biotechnologyl6,190-195].
Example 1
Cholesterol-binding MIP and cholesterol-binding MIP-based COBALT
Bile acid- binding and bile acid-binding MIP-based COBALTs
This Example relates to the design and implementation of COBALTs for binding
cholesterol and bile acids.
Cholesterol, whose efficient, specific elimination from the body represents an
important therapeutic advance in atherosclerosis and deoxycholic acid (DCA), a
toxic bile
acid produced by bacteria present in intestinal flora, represent important
applications. Figure
1 shows a schematic outline of the approach for the creation of specific
COBALTs.
As shown in Figure 1, the hydroxyl-containing cholesterol and DCA targets are
represented by R-OH. An ester R-O-CO-R' or carbamate R-O-C(=O)-NH-R'
derivative of the
target alcohol ROH is prepared where the R' group has a vinyl, polymerizable
function such
as styrene, methacrylyl, etc. The ester or carbamate is then polymerized with
a large excess
of crosslinker in the conventional fashion [G. Wulff and A. Sarhan Angew.
Chem. Int. Ed.
Engl. 11, 341-(1972); G. Wulff. Angew. Chem. Int. Ed. Engl. 34, 1812-1832
(1995); G.
Wulff, et al. Angew. Chem. Int. Ed. Eng. 36, 1962-1964 (1997)] to give a
highly crosslinked
polymer containing many "buried" copies of the R-O-C(=O)- function. The
polymer is then
treated chemically (MeOH/KOH/H20 solution, for example) to hydrolyze
essentially all of
the ester or carbamate bonds, releasing the R-OH imprint molecule and creating
selective
cavities containing a carboxylate (for the ester derivative) or an amino group
(for the
17
CA 02446921 2003-10-16
WO 02/083708 PCT/IL02/00307
carbamate derivative). Typically, approximately 80-95% of the "template"
groups R-OH are
thus removed.
More specifically, as shown, the top portion of Figure 1 utilizes
polymerization of an
ester derivative of ROH to give, after hydrolytic removal of the print or
template molecule
ROH, carboxylic acid-containing cavities complementary to ROH which are
activated to acid
chlorides in order to form the covalent product of step e. The bottom portion
exemplifies
polymerization of a carbamate derivative of ROH to afford, following removal
of ROH,
complementary amine-containing cavities, which are activated to isocyanate (or
isothiocyanate, not illustrated here) to form the covalent product, step i.
All of the steps are
routinely used in MIP technology: step a, synthesis of the methacylic ester of
ROH (synthesis
of the carbamate from 4-vinylbenzeneisocyanate is not illustrated); steps b
and f, the ester or
carbamate is polymerized with a large excess of crosslinker and a porogen
(pore-forming
solvent) using an initiates such as AIBN and heat (or irradiation); steps c
and g, the solid
polymer is ground, sieved and treated with reagents to hydrolzye all ester and
carbamate
bonds and remove ROH; step d, the acid chloride MIP-based COBALT may be
prepared by
treating with SOC12 or COC12; step h, the isocyanate (or isothiocyanate) MIP-
based
COBALT can be prepared by reaction with phosgene, COC12. (or thiophosgene,
CSC12).
The resulting MIPs may be used directly as conventional binding agents or they
may
be converted into COBALTs by chemical modification, or "activation", using
specific
chemical reactions. The carboxylate polymers (derived from the ester
derivative) are
chemically converted to acid chloride, Cl-C(=O)-polymer, while the amino
polymers (derived
from the carbamate derivative) are converted to isocyanate, O=C=N-polymer, or
isothiocyanate, S=C=N-polymer. Since the different reactive functional groups
will have
varying reactivities and orientations and distances from the target molecules,
a family of
many MIP-based COBALTs may be prepared and tested to find the materials having
optimal
performance.
18
CA 02446921 2003-10-16
WO 02/083708 PCT/IL02/00307
Experiments carried out on the resulting MIP-based COBALTs show that the
apparent
equilibrium binding constants of R-OH to the COBALTs are much higher than with
the
parent MIPs; this difference is especially pronounced as the concentration of
R-OH is
decreased and incubation time is increased.
When the two MIPs, the covalently bound (COBALTs) and non-covalently bound are
treated under conditions which dissociate R-OH from the latter, the former
remain bound to
R-OH.
Example 2
' Steroid MIPs
The COBALT approach is illustrated using chemically reactive molecular imprint
polymers (MIPs). Cholesteryl 4-vinylphenyl carbamate was used as a template
monomer;
cholesteryl methacrylate was an added functional monomer to create hydrophobic
binding
and recognition. Cholesterol was cleaved from the polymer (MS40, MS41)
hydrolytically
with the concomitant loss of CO2, resulting in the formation of a conventional
MIP (MS50)
having a non-covalent (or non-reactive) recognition site, bearing an
aminophenyl group,
capable of interacting with cholesterol through hydrogen bonding. By chemical
modification
of the polymer with phosgene and thiophosgene, respectively, the amino group
was
transformed into reactive isocyanate (MS71) and isothiocyanate (MS80) groups,
respectively,
both of which bound the cholesterol in a covalent fashion.
Figure 2 shows a schematic depiction of the preparation of a conventional
cholesterol-
binding, amino-containing MIP (MS50). In separate experiments it was shown
that a polymer
of cholesteryl methacrylate is completely resistant towards hydrolysis (no
release of
cholesterol under more drastic hydrolysis conditions than used to obtain
MS50); cholesteryl
methacrylate was therefore used as a functional monomer to obtain a lipophilic
recognition
cavity.
19
CA 02446921 2003-10-16
WO 02/083708 PCT/IL02/00307
Figure 3 shows a schematic view of the two MIPs, one binding non-covalently
and
one binding covalently, the latter being a COBALT. The isocyanate MIP (MS71)
is
illustrated; the analogous reaction using thiophosgene ~ afforded the
isothiocyanate MIP
(MS80).
Cholesteryl methacrylate
0
II Ci . py O O
HO
CHpCIz, RT
Cholesterol
To a mixture of cholesterol (1.0 g, 2.60 mmol) and pyridine (0.25 ml, 3.12
mmol) in
dichloromethane (10 ml) containing ca. 0.01 g hydroquinone, was added
methacryloyl
chloride (0.3 ml, 3.12 mmol) drop wise at 0 °C. The mixture was stirred
at RT for 4h
transferred to a separatory funnel, washed with saturated aq. sodium
bicarbonate solution
(2x10 ml) followed by brine (2x10 ml). The organic layer was dried over anhyd.
sodium
sulphate, and the solvent evaporated in vacuo. The residue (1.15 g, 97%) was
recrystallized
from hot ethylacetate using ca. 0.05 g of hydroquinone, which afforded long
needles of
cholesteryl methacrylate (1) (1.0 g, 84%), m.p. 98-100 °C.
~H NMR (300 MHz, CDC13): 8 (delta) 0.63 (s, 3H, 18-CH3), 0.83 (d, 6H, 27-CH3,
J=7.7 Hz), 0.86 (d, 3H, 21-CH3, J=6.9 Hz), 1.02 (s, 3H, 19-CH3), 0.88-2.05
(br, steroid
nucleus), 1.89 (s, 3H, allylic CH3), 2.38 (d, 2H, J=8.5 Hz, C7-H2) 4.66 (m,
1H, 3a-H), 5.40
(m, 1 H, C6-H), 5.48 (s, 1 H, =CH anti to allylic CH3), 6.30 (s, 1 H, =CH syn
to allylic CH3).
4-Vinylbenzazide
CA 02446921 2003-10-16
WO 02/083708 PCT/IL02/00307
/ /
/ 25% NaN3 solution _ /
\ I Acetone \
COCI CONS
Freshly distilled 4-vinylbenzoylchloride (1.2 g, 7.2 mmol) was dissolved in
dry
acetone (25 ml) containing 0.1 g of hydroquinone and cooled to 0 °C. A
cold solution of
sodium azide (4.68 g, 72 mmol) in water (20 ml) was slowly added to the acid
chloride
solution. The reaction mixture was stirred at RT for 2h, extracted with
dichloromethane (3 x
20 ml) and the solvent evaporated. The crude product was recrystallized with
ethyl acetate by
adding a pinch of hydroquinone (yield, 1.07 g, 86%), m.p. 167-169 °C.
'H NMR (300 MHz, CDC13): 8 (delta) 5.40 (d, 1H, J=11.7 Hz, vinyl C2-H, cis),
5.84
(d, 1H, J=19.5 Hz, vinyl C2-H, trans), 6.67 (dd, 1H, J~ and Z=19.5 and 11.7
Hz, vinyl C1-H),
7.44 (d, 2H, J=9.7 Hz, aromatic, ortho to vinyl group), 8.00 (d, 2H, J=9.7 Hz,
aromatic, ortho
to acylazide group).
Cholesteryl(4-vinyl)phenylcarbamate
CONS
HO ~ C~O
Benzene, 80 °C, 6h HN
Cholesterol
Solid 4-vinylbenzoylazide (0.495 g, 2.86 mmol) was added in five small
portions with
a spatula to a clear solution of cholesterol (1.0 g, 2.60 mmol) in 25 ml of
benzene (dried
overnight over CaClz) containing ca. 0.01 g of hydroquinone, preheated in an
oil bath to 80
°C. Slow evolution of NZ gas was observed which subsided after 6h. The
clear reaction
mixture was cooled to RT (a solid started precipitating out at this stage) and
the solvent was
21
CA 02446921 2003-10-16
WO 02/083708 PCT/IL02/00307
evaporated (rotary evaporator). The crude solid was dissolved in 50 ml of
dichloromethane,
washed with 5% aq. sodium bicarbonate solution (3 x 50 ml). The solvent was
evaporated and
the crude was recrystallized from hot ethyl acetate using ca. 0.01 g of
hydroquinone (yield,
1.33 g, 92%).m.p. 178-180 °C.
'H NMR (300 MHz, CDC13): ~delta)0.63 (s, 3H, 18-CH3), 0.83 (d, 6H, 27-CH3,
J=7.8
Hz), 0.86 (d, 3H, 21-CH3, J=7.4 Hz), 1.02 (s, 3H, 19-CH3), 0.80-2.04 (br,
steroid nucleus),
2.40 (m, 2H, C7-HZ) 4.61 (m, 1H, 3a-H), 5.18 (d, 1H, J=11.7 Hz, vinyl C2-H,
cis), 5.40 (m,
1H, C6-H), 5.64 (d, 1H, J=19.5 Hz, vinyl C2-H, trans), 6.58 (s, 1H, NH), 6.63
(dd, 1H, J1 a"a
Z=19.5 and 11.7 Hz, vinyl C1-H), 7.38 (s, 4H, aromatic).
22
CA 02446921 2003-10-16
WO 02/083708 PCT/IL02/00307
Illustrative Procedure for Bulk Polymerization:
A clear solution containing cholesteryl (4-vinyl)phenyl carbamate (0.1 g)
ethyleneglycol dimethacrylate (EGDM) (0.736 g), cholesteryl methacrylate
(0.171 g), EVB
(0.132 g) (the mole ratio of these components is 1:20:2:4), porogen
(EtOH/acetonitrile, 1:1
v/v, 3 ml), and initiator, AIBN, 0.012 g, was transferred to a thick-walled
glass
polymerization tube. The tube was subjected to three freeze-evacuate-thaw
cycles followed
by flame sealing. The sealed tube was kept in an oven at 60 °C for 3
days. Following
polymerization, the neck of the tube was broken and the polymer monolith was
broken into
small pieces using a spatula and the broken pieces ground in a mortar. The
polymer was
Sohxlet-extracted using 60 ml MeOH and dried in vacuo at 70 °C and then
weighed. The
polymer (MS41, 1.0g) was suspended in 1M sodium hydroxide in methanol (50 ml)
and
heated to reflux for 24h. periods. The cooled suspension was neutralized with
1N
hydrochloric acid, filtered on a sintered glass funnel and washed with water
until the
washings were neutral, followed by several methanol washings. The mixture of
the
hydrolysate and all washings were extracted with dichloromethane (3x50 ml),
dried over
anhyd. sodium sulphate and evaporated in a pre-weighed flask. The purity of
the resulting
cholesterol was confirmed by TLC and its mass was used to calculate the degree
of
hydrolysis. The cholesterol removed constituted 97% of the calculated amount
due to
hydrolysis of the carbamate.
Control polymer was prepared in a similar way but the carbamate template was
excluded from the polymerization mixture.
EXAMPLE 3
Tests for Cholesterol Sequestrants of
Example 1
23
CA 02446921 2003-10-16
WO 02/083708 PCT/IL02/00307
Analysis of Cholesterol
Analysis of cholesterol was performed based on its color reaction with a
modified and
stable Liebermann-Buchard reagent developed by Huang and co-workers.
Cholesterol (FW: 386.67)
~ Preparation of Liebermann-Buchard reagent: Glacial acetic acid (90 ml)
was slowly added to well stirred acetic anhydride (180 ml).
~ Concentrated sulfuric acid (30 ml) was added to the above solution
carefully at RT, which causes a slight increase in the solution temperature,
followed by 6 g of anhydrous sodium sulfate. The solution was stirred
vigorously
for 5 min. and the clear solution was transferred to a bottle and sealed with
parafilm. The bottle was stored in a refrigerator at < 4 °C. According
to Sommers
and co-workers the reagent is stable for more than a month if stored at 4
°C and
they recommend not storing the reagent at RT (room temperature) longer than
one
day.
~ Preparation of stock solution of cholesterol in cyclohexane: Dissolved
0.4833 g of cholesterol in cyclohexane in a standard flask (25 ml) to obtain a
50
mM stock solution.
~ Standard solutions of cholesterol: From the above stock solution, 0.2, 0.4,
0.6, 0.8, 1.0, 1.2, 1.4 and 1.6 ml were withdrawn using glass syringe (1 ml)
and
diluted to 10 ml in standard flasks to obtain respectively solutions with
concentrations of 1, 2, 3, 4, 5, 6, 7 and 8 mM. All the solutions were
prepared in
duplicates.
~ Developing cholesterol with the LB reagent: Using a glass pipette (5 ml),
the LB-reagent (2.0 ml) was taken in thirteen conicle flasks (10 ml) fitted
with
ground glass stoppers (14'). Above standard solutions (100 ~1) were added to
the
24
CA 02446921 2003-10-16
WO 02/083708 PCT/IL02/00307
twelve conicle flasks containing the reagent using a micro liter syringe at
RT.
Solution temperature remains unchanged. Cyclohexane (200 p1) was added to the
thirteenth flask containing the reagent (4 ml) which served as the reference.
All the
flasks were heated on a water bath at 36 °C for 15 min. After 5 min.
the solutions
turned to blue-green color. The flasks were removed from the water bath and
allowed to cool to RT for 15 min. The OD's of the soutions were measured using
UV-visible spectrophotometer at 618 nm.
Figure 4 shows the calibration curve for cholesterol.
Cholesterol Binding to MS50
Stock solutions of cholesterol in cyclohexane were made in 10 ml standard
flasks. 1,
1.5, 2, 2.5, 3, 3.5, 4, 4.5, 5, 6, 7, 8 mM concentrations were studied in
duplicates for both
MS50 and its control polymer. Since only 12 conical flasks were used for the
study, only
three concentrations were studied at a time (six flasks each for MS50 and
control polymer for
three concentrations in duplicates). MS50 and control polymer were each (20
mg) weighed in
6 conical flasks fitted with ground glass joints separately. The above
standard solution (2 ml)
was added in the above 12 conical flask containing polymers. The flask
stoppers held in place
with parafilm (which also prevented introduction of moisture) and stood
undisturbed
overnight. After 8-10 hours of standing, samples (100 p,1) were taken from the
supernatant
using a microlitre syringe and added directly to the LB reagent (2m1) and
proceeded with
heating as mentioned in 'Analysis of Cholesterol'. The resulting solutions
were analyzed by
UV-visible spectrophotometer at 618 nm. The concentrations obtained were free
cholesterol
concentrations which upon subtraction from initial solution concentrations
gave bound
cholesterol concentrations. The absorption (non-specific binding) due to
control polymer is
subtracted from those of MS50. Using this information, Scatchard plots were
constructed. The
software'Graphpad Prism' was used to construct Scatchard plots.
CA 02446921 2003-10-16
WO 02/083708 PCT/IL02/00307
Table 1: MS50 Binding Studies with Cholesterol
Sample Initial OD (618 nm) Free Bound
Solution Conc Cholesterol (M) Cholesterol
(m~ (n'n
MS50 1.0 0.0251 0.000314 0.000686
MS50 1.5 0.0336 0.000421 0.001079
MS50 2.0 0.0452 0.000566 0.001434
MS50 2.5 0.0689 0.000863 0.001637
MS50 3.0 0.0805 0.001008 0.001992
MS50 3.5 0.0866 0.001085 0.002415
MS50 4.0 0.1235 0.001547 0.002453
MS50 4.5 0.1545 0.001935 0.002565
MS50 5.0 0.1809 0.002266 0.002734
MS50 6.0 0.2262 0.002833 0.003167
MS50 7.0 0.3001 0.003759 0.003241
MS50 8.0 0.3668 0.004594 0.003406
MS50 1.0 0.0256 0.000321 0.000679
MS50 1.5 0.0343 0.00043 0.00107
MS50 2.0 0.0457 0.000572 0.001428
MS50 2.5 0.0693 0.000868 0.001632
MS50 3.0 0.0811 0.001016 0.001984
MS50 3.5 0.0866 0.001085 0.002415
MS50 4.0 0.1231 0.001542 0.002458
MS50 4.5 0.1549 0.00194 0.00256
MS50 5.0 0.1802 0.002257 0.002743
MS50 6.0 0.2268 0.002841 0.003159
MSSO 7.0 0.3001 0.003759 0.003241
MSSO 8.0 0.3666 0.004591 0.003409
26
CA 02446921 2003-10-16
WO 02/083708 PCT/IL02/00307
Table 2: Control Polymer Binding Studies with Cholesterol
Sample Initial OD (618 Free Bound
nm)
Solution Conc Cholesterol Cholesterol
(M)
(m~
Control 1.0 0.0713 0.000893 0.000107
Control 1.5 0.1008 0.001262 0.0002376
Control 2.0 0.1273 0.001594 0.0004057
Control 2.5 0.1624 0.002034 0.0004661
Control 3.0 0.1929 0.002416 0.0005841
Control 3.5 0.2223 0.002784 0.0007159
Control 4.0 0.2565 0.003212 0.0007875
Control 4.5 0.2879 0.003606 0.0008943
Control 5.0 0.322 0.004033 0.0009672
Control 6.0 0.3831 0.004798 0.001202
Control 7.0 0.4612 0.005776 0.0012238
Control 8.0 0.5293 0.006629 0.0013709
Control 1.0 0.0732 0.000917 0.0000832
Control 1.5 0.1038 0.0013 0.0002
Control 2.0 0.1244 0.001558 0.000442
Control 2.5 0.1721 0.002155 0.0003446
Control 3.0 0.1937 0.002426 0.000574
Control 3.5 0.2206 0.002763 0.0007371
Control 4.0 0.2505 0.003137 0.0008627
Control 4.5 0.2957 0.003703 0.0007966
Control 5.0 0.3196 0.004003 0.0009972
Control 6.0 0.3873 0.004851 0.0011494
Control 7.0 0.4599 0.00576 0.0012401
Control 8.0 0.5309 0.006649 0.0013509
27
CA 02446921 2003-10-16
WO 02/083708 PCT/IL02/00307
Table 3: Correction of Non-specific Binding
Initial MS50 bound Control Net CholesterolNet Cholesterol
SolutionCholesterolPolymer bound by MS50 bound by
Conc Conc (M) bound (M550 bound- MS50/Free
(mM) CholesterolControl PolymerCholesterol
Conc (1V1)bound) (Bound/Free)
(non-specificConc (M)
binding)
1.0 0.000686 0.000107 0.000579 1.840701
1.5 0.001079 0.0002376 0.000841 1.99894
2.0 0.001434 0.0004057 0.001028 1.815773
2.5 0.001637 0.0004661 0.001171 1.356992
3.0 0.001992 0.0005841 0.001408 1.396342
3.5 0.002415 0.0007159 0.001699 1.566842
4.0 0.002453 0.0007875 0.001665 1.076617
4.5 0.002565 0.0008943 0.001671 0.863412
5.0 0.002734 0.0009672 0.001767 0.780071
6.0 0.003167 0.001202 0.001965 0.693618
7.0 0.003241 0.0012238 0.002017 0.53677
8.0 0.003406 0.0013709 0.002035 0.443
1.0 0.000679 0.0000832 0.000596 1.859449
1.5 0.00107 0.0002 0.00087 2.026195
2.0 0.001428 0.000442 0.000986 1.722068
2.5 0.001632 0.0003446 0.001287 1.483375
3.0 0.001984 0.000574 0.00141 1.388458
3.5 0.002415 0.0007371 0.001678 1.547387
4.0 0.002458 0.0008627 0.001596 1.034912
4.5 0.00256 0.0007966 0.001763 0.90896
5.0 0.002743 0.0009972 0.001746 0.773605
6.0 0.003159 0.0011494 0.00201 0.707655
7.0 0.003241 0.0012401 0.002001 0.532487
8.0 0.003409 0.0013509 0.002058 0.448165
Figures 5 and 6 show the Scatchard and binding isotherm plots respectively for
this
Example.
Figure 7 is related to the preparation of crMIP - MS71.
To a suspension of MS50 (0.5 g ~ 0.1 mmol of NH2 groups) in dry MeCN (50 ml)
was
added 4.947 g of triphosgene (16.66 mmol) at once. The resulting mixture was
stirred for 30
min. and cooled to 0 °C using ice-salt mixture. Pyridine (4 ml, SO
mmol) was added to the
suspension slowly using a syringe, over a period of one hour. The contents
were stirred at
this temperature for 4h. the reaction was monitored by IR spectroscopy
following the
28
CA 02446921 2003-10-16
WO 02/083708 PCT/IL02/00307
developing NCO stretching at v = 2265 cm-~ (nujol). The suspension was
filtered and the
polymer was washed with dry acetonitrile (2 x 10 ml) and dried in the oven at
80 °C for 30
min. The resulting polymer MS71 was further dried under vacuum for 1h and
stored in a
desiccator. During filtration and drying processes, the NCO peak %T at 2265 cm-
1 reduced to
50% compared to that measured from the reaction mixture. Moreover, upon
standing in the
atmosphere for overnight, the NCO peak disappeared completely.
Binding studies of COBALT MS71 compared to conventional MIP MS50
Polymer Initial Conc (1V1) OD (618 nm) Free (1V1) Bound (1V1) Bound (mg)
(20 mg)
MS71 0.005 0.1099 0.001376 0.003624 2.8
MS71 0.005 0.1245 0.001559 0.003441 2.66
MS50 0.005 0.187 0.002342 0.002658 2.05
MS50 0.005 0.1779 0.002228 0.002772 2.14
In these duplicate experiments it is seen that the COBALT gave more than a 30%
increase in the amount of cholesterol bound under these conditions. When the
two polymers
were refluxed with methanol (discarded) and then hydrolyzed and the
hydrolyzate tested for
cholesterol, MS50 afforded no detectable cholesterol, while MS71 afforded
nearly the amount
taken up in the binding experiment, in accord with the formation of a stable,
covalent bond in
the latter, a COBALT.
Figure 8 shows the IR Spectrum of MS71 (when the maximum conversion
into NCO is reached).
Polymer MS50 was converted in the same way to the isothiocyanate MIP, MS80.
This polymer was found to be far more stable, as expected, than the isocyanate
polymer
MS71. On standing for 48 hr open to the atmosphere, the decrease of the
characteristic NCS
peak in the infrared, at 2087 cm', indicated that only about 50% had reacted.
When treated
with cholesterol as in the case of MS71, it was possible to show significant
binding at
29
CA 02446921 2003-10-16
WO 02/083708 PCT/IL02/00307
concentrations where the 'parent' MIP was binding only very small amounts of
cholesterol.
At an initial cholesterol concentration of 0.005 M, MS80 (20 mg polymer
samples were used
in all experiments in duplicate) bound 39% more cholesterol than did MS50; at
an initial
cholesterol concentration of 0.003 M, MS80 bound 48% more cholesterol than did
MS50; at
an initial cholesterol concentration of 0.001 M, MS80 bound 76% more
cholesterol than did
MS50.
EXAMPLE 4
MIP-based COBALTS for the binding of toxic organophosphates
The overall approach followed is outlined in Figure 9. DFP was chosen as a
representative "OP (organophosphate) agent", illustrating the approach that
can be used for
sarin, soman, etc. as well as other chemical warfare toxins. Additional "OP
agents" are
described below to further generalize the method. Briefly, the design was to
produce MIPs
having complementary cavities for DFP as well as a suitably positioned
nucleophile (an active
OH group that can react with DFP). The steps included: (1) the synthesis of
functional
monomers that contain a diisopropyl phosphate group linked to oxygen (the OH
subsequently
acts as the nucleophile for reaction with DFP), providing a DFP-binding
cavity; (2)
polymerization of the functional monomer (5-15 %) with an excess (80-90%) of
crosslinker
and additional monomers in a solvent (porogen) that afforded pores as the
macroreticular
polymer formed and separated out; (3) hydrolysis of the polymer which removed
diisopropyl
phosphate groups, providing the selective cavity and exposing the oxygen
nucleophile; (4)
estimation of DFP binding by the MIPs. Control polymers (the unhydrolyzed
polymer and/or
a non-imprinted polymer prepared from the functional monomer without a
diisopropyl
phosphate group) were synthesized and characterized and used for comparison.
CA 02446921 2003-10-16
WO 02/083708 PCT/IL02/00307
Functional monomers.
The chemistry of phosphorus compounds suggested that oxygen a-nucleophiles
would
be particularly effective in allowing hydrolysis of the phosphate groups in
the activation step
of the MIPs (hydrolysis of MIP-B to MIP-C; see Fig. 9) and allow good
reactivity with DFP
and other "OP agents" in the covalent reaction step (MIP-C to MIP-D, Fig. 9
e). Figure 10
depicts representative structures of synthesized functional monomers used for
the DFP
binding MIPs that were prepared. Additional functional monomers were prepared
for other
fluorophosphates and fluorophosponates and representative syntheses are
presented below.
Each of the functional monomers was characterized (including 1H and 31P NMR,
MS, IR,
microanalysis) and tested for thermal and hydrolytic stability for subsequent
steps.
Polymerization.
Suspension polymerization (droplets of reactants dissolved in toluene
suspended in
water containing surface active agents were polymerized) and dispersion
polymerization (also
termed precipitation polymerization; solutions of reactants in a solvent -
toluene/methanol -
were polymerized under rapid stirnng) methods were used in order to obtain
relatively
uniform particles and maximize the yields (conventional bulk polymerization
methods, which
requires subsequent grinding and sieving and results in large losses, were
also used).
Thermally initiated polymerization and photochemically initiated
polymerization methods
were also used.
Divinylbenzene/styrene (DVB/S) and ethylene glycol dimethacrylate/monomethyl
acrylate (EGDMA/MMA) crosslinker/monomer mixtures were used. Mixtures of
divinylbenzene/styrene/4-vinylpyridine (DVB/SNP) and other combinations were
also used.
The organic solvent wash after polymerization was analyzed for phosphorus
containing substances (3'P-NMR) to indicate the proportion of functional
monomer (typically
>80%) that had been incorporated into the polymer. The MIP polymers were also
analyzed
31
CA 02446921 2003-10-16
WO 02/083708 PCT/IL02/00307
for phosphorus by the sensitive ICP method after a weighed sample for fully
combusted in
oxygen using a Schoeniger flask. The phosphorus content was determined by
inductively
coupled plasma atomic emission spectrometry (ICP-AES) at 178.200 nm.
All of the MIPs were tested for swelling in the solvent used for the binding
reactions.
In the divinylbenzene polymers, the size of the obtained spherical particles
varied; in some
suspension polymerizations ca. 50-85 ~ (micron) particles were obtained. In
some dispersion
polymerizations,l.5 to 5~ (micron) particles were obtained. Water, methanol
and THF
generally did not swell these polymers appreciably but toluene and 2-propanol
did.
H~ sis o~~Activation).
Polymer hydrolysis conditions were varied and optimized for MIPs. Conditions
included: (1) aq. KOH/toluene/2-propanol; (2) NHZOH-HCl/triethylamine/
toluene/2-
propanol; (3) NHZOH-HCl/triethylamine/ toluene/2-propanol/DBU; (4) NHZOH-
HCl/40% aq.
KOH/ tetrabutylammonium bromide; (5) NHZOH/ toluene/2-propanol/water.
Conditions (4)
were generally used. In addition, extensive incubation of 3 under conditions
(4) showed no
chemical change to the oxime group (e.g., hydrolysis to the aldehyde).
3'P-NMR analysis of the washes after hydrolysis of the MIP and phosphorus
analysis
of the polymers (ICP-AES) were used to estimate the cavity formation
(activation).
OP-agent binding to MIPs.
Assays for DFP and other OP agent binding were developed using acetylcholine
esterase and butyrylcholine esterase. The calibration was carried out under
conditions that
allowed detection of mM equilibrium binding constants (even without any
covalent binding
reaction). Reactions were carned out in solution, either aqueous or organic
solvent, and
controls were used to assay the effects of solvent upon the enzyme.
32
CA 02446921 2003-10-16
WO 02/083708 PCT/IL02/00307
Preuaration of functional monomer 3:
Preparation of 4-vinylbenzaldehyde oxime: Hexamethylenetetramine (70.80 g, 50
mmol), 0.6 g hydroquinone and 4-vinylbenzylchloride (36.7 ml, 250 mmol) were
mixed with
stirring with glacial acetic acid (100 ml) until a solution was obtained.
Water (100 ml) was
added and the mixture was boiled under reflux for 2h. Then concentrated HCl
(85 ml) was
added and the mixture was boiled for an additional 15 min. After cooling to
room temperature
the mixture was extracted with Et20 (3x100 ml). The organic phase was washed
with 10%
sodium bicarbonate (3x100 ml) and with water (100 ml), dried and evaporated to
yield 4-
vinylbenzaldehyde (28.30 g). The latter was added to a solution of
hydroxylamine (16.4 g) in
150 ml of water. The mixture was cooled to 5°C and lOM NaOH (24 ml) was
added
dropwise. Stirring was continued for 48h. Then the mixture was extracted with
2x100 ml of
dichloromethane. The organic layer was washed with water, dried and evaporated
to obtain
the crude product (98.4% yield). The pure oxime was obtained after
recrystallization from
hexane and chromatography on silica gel (petroleum ether/ether eluent). 1H NMR
(CDC13):
~ delta) = 5.34 (d, 1 H), 5.83 (d, 1 H), 6.73 (dd, 1 H), 7.33-7.55 (m, 4H),
8.18 (s, 1 H), 9.14 (br
s, 1 H).
Preparation of 3: Sodium (160mg, 7 mmol) was added to the solution of oxime
(1g,
6.75 mmol) in dry diethyl ether (100 ml) and reaction mixture was stirred
until all sodium
reacted (~6h). A solution of diisopropylchlorophosphate (DCP) [H.McCombie,
J.Chem.Soc.,
1945, 380] (1.3 ml, 6.75 mmol ) in dry ether (50 ml) was added dropwise and
the reaction
mixture was stirred overnight at room temperature. After filtration and
evaporation, the
residue was chromatographed on silica gel to give monomer 3 as an oily
residue. 'H NMR
(CDCI3): delta) = 1.29 (d, 6H), 4.57 (q, 1H), 5.39 (d, 1H), 5.83 (d, 1H), 6.63
(dd, 1H),
7.44 (d, 2H), 7.55 (d,2H), 9.80 (s,lH). 3~P NMR (CDCI3): ~delta) =-2.39.
33
CA 02446921 2003-10-16
WO 02/083708 PCT/IL02/00307
Preuaration of functional monomer 9: The reaction was carried out as described
for 3 using
4-vinylbenzaldehyde oxime and commercial (Aldrich) diphenylchlorophosphate;
the product
had m.p. 68-72°C (see Figure 11 for the synthetic scheme). 'H NMR
(CDCl3): delta) _
5.3 7 (d, 1 H), 5.84 (d, l H), 6.72 (dd, 1 H), 7.21-7.36 (m, l OH), 7.44 (d,
2H), 7.62 (d, 2H),
8.33(s, 1H). 3'P NMR (CDC13): delta) _ -12.17
0 0
P(OiPr)3 ~ PCIs
Br ~ ~ ~ ~C/P \ ~ 0~ ~CI~P
BICP
Preparation of monomer 10
The benzylisopropylchlorophosphate (BICP) was prepared by heating a mixture of
triisopropylphosphite (10 ml, 60 mmol) and benzylbromide (4.76 ml, 60 mmol )
for 2h at
90°C. The unreacted starting materials were distilled out and the
residue (8.7 g) was
dissolved in CCl4 (5 ml) and heated to reflux. Phosphorus pentachloride (7.0
gr ) was then
added in small portions during 1h and the reaction mixture was refluxed for an
additional
O.Sh. Distillation (O.SmmHg) provided compound BICP (65%yield) as a colorless
liquid. 'H
NMR (CDC13): Odelta = 1.25 (d, 3H), 1.31 (d, 3H), 3.44 (d, 2H), 4.83 (m, 1H),
7.28 (m, SH).
3'P NMR (CDCl3): delta) = 35.68.
The synthesis of 10 was carried out as described for 3 and 9, using BICP and 4-
vinylbenzaldehyde oxime. 'H NMR (CDC13): delta) = 1.27 (d, 3H), 1.31 (d, 3H),
3.40 (d,
2H), 5.33 (d, 1H), 5.83 (d, 1H), 6.71 (m,lH), 7.25 (m, SH), 7.41 (d, 2H), 7.62
(d, 2H), 8.23 (s,
1H). 3~P NMR (CDC13): delta) = 28.10.
Preparation of diphenylfluorophosphate (DPFP) and benzyl
isopropylfluorophosphonate (BIFP).
Sodium fluoride ( 4 equivalents ) was added to one equivalent of
34
CA 02446921 2003-10-16
WO 02/083708 PCT/IL02/00307
chlorodiphenylphosphate or benzylchloroisopropylphosphonate (BICP) dissolved
in CC14 and
the reaction mixture was heated to reflux for 24h. The mixture was then
filtered, an
additional portion of NaF was added to the filtrate which was then heated to
reflux for
additional 24h and monitored by 3'P and by 19F NMR. The reaction mixture was
filtered and
evaporated. In the case of FP(O)(OiPr)CHZPh the product was distilled.
NMR data of FP(O)(OPh)Z: 3'P NMR (CDCl3): delta) _ -23.04. '9F NMR
(CDC13): delta) _ -82.4 (d, Jp_p = 1166Hz)
NMR data of FP(O)(OiPr)CHZPh : 1H NMR (CDC13): Odelta) = 1.81 (d, 3H), 1.52
(d, 3H),
3.21 (d, 2H), 4.78 (m, 1H), 7.28 (s, SH). ~3C NMR (CDCl3): delta) = 23.63,
23.99, 31.42
33.64 (d), 73.79, 127.61, 127.66, 128.94, 129.73, 129.98. 31P NMR (CDC13):
delta) _
27.14 (d). '9F NMR (CDC13): delta) _ -70.75 (d, Jp_p = 1333Hz).
Preparation of Polymer fMIP92-421 using Monomer 9 by Solution Polymerization
Divinylbenzene 7.1 g
Styrene 0.7g
Toluene 6.3 g
Methanol 23.5 ml
Monomer 9 0.40g
Initiator 2,2'- azobisisobutyronitrile 0.08g
All ingredients were placed in a three-necked round-bottom 100m1 flask,
equipped with a
mechanical stirrer, condenser (CaCl2 tube) and dropping funnel, except for the
initiatordissolved in 2.7m1 methanol. The stirred solution was heated to
reflux and after O.Shr
the initiator solution was added. After an additional 6 hours reflux the
reaction mixture was
cooled. The colorless solid polymer beads were filtered and washed once with
toluene-
methanol (1:1) mixture (300m1) and twice with acetone (300m1 each). The
polymer was dried
under vacuum at 60°C for 18hrs. Yield = 4.5g, 55%.
CA 02446921 2003-10-16
WO 02/083708 PCT/IL02/00307
Preparation of Polymer [MIP92-42III]
MIP92-42, 1 g, was stirred with 50 ml toluene for 0.5 hr; 5 ml of 2-propanol
were then
followed by 10 ml of an aqueous NaOH solution (40 g in 100 ml water). While
stirring
vigorously, 2.5g of solid tetrabutylammonium bromide (same results were
obtained using the
chloride) were added and the stirred reaction was warmed to 70°C for 12
hr. The cooled
reaction was filtered on a sintered-glass filter (the filtrate was used to
determine phosphorus
content), and the polymer washed with water (50 ml), 0.1 N HCl (50 ml), 1 %
NaZC03
solution (50 ml), six portions of 2-propanol (each 50 ml), chloroform (50 ml),
three portions
of toluene (50 ml), and finally dried at 60°C for 12 hr.
Representative binding experiments, using MIP 92-42III as an example.
MIP92-42III is designed to bind diphenylfluorophosphate (DPFP); the
unhydrolyzed
MIP 92-42 is the control polymer.
The concentrations of DPFP present during various stages of the experiment
were
determined by measurement of percent inhibition of butyryl choline esterase
(BChe) activity.
A calibration curve was constructed, using fresh solutions of DPFP in
isopropyl alcohol, to
convert percent BChe inhibition to DPFP concentration. BChe activity was
measured in units
of change in A412 per minute based on the production of thiocholine from the
enzymatic
hydrolysis of butyryl thiocholine, which created a yellow color in the
presence of the
colorimetric reagent DTNB.
The calibration curve is shown in Figure 12.
Samples containing 10 mg polymer, either 92-42 or 92-42III, were mixed with 1
ml
isopropyl alcohol and preincubated on an oscillating shaker at room
temperature for 44 hours.
DPFP was then added to the samples to a final concentration of 5 p.M
(micromolar). The
mixture of DPFP and polymer was returned to the shaker to incubate for 24
hours. The
36
CA 02446921 2003-10-16
WO 02/083708 PCT/IL02/00307
sample containing polymer 92-42 was a control for background binding that
occurred in
locations other than the specific binding pocket formed by the template.
Additional controls
containing isopropyl alcohol only (no polymer, no DPFP) and 5 pM (micromolar)
DPFP in
isopropyl alcohol only (no polymer) were also run. These two controls were
also assayed for
percent inhibition at the beginning of the incubation (time zero).
After 24 hours the samples and controls were centrifuged at 14000 x g for 1
minute to
sediment the polymers. The concentrations of DPFP present in the supernatants
were then
determined by measurement of percent inhibition of butyryl choline esterase
(BChe) activity.
The percent inhibition for each sample and its equivalent DPFP concentration
from the
calibration curve are given in Table 1 below.
Table 1
Sample Percent Inhibition DPFP
of BChe Activity Concentration
I~ ~Mlo
t = 0 hours
Control 1 isopropyl alcohol0 0
only
Control 2 5 p.M DPFP in 62 4.5
isopropyl
alcohol
t = 24 hours
Control 1 isopropyl alcohol0 0
only
Control 2 5 ~M DPFP in isopropyl50 4
alcohol
Control3 5 ~M DPFP + 10 57 4.5
mg
polymer 92-42
Sample 5 ~M DPFP + 10 mg 23 1.5
polymer
92-42III
There was a small amount of DPFP lost due to background hydrolysis, on the
order of
0.5 pM DPFP. However, the sample containing the polymer dropped significantly,
to 1.5 p.M,
indicating that even considering a background hydrolysis of 10% of the DPFP,
60% (3 pM)
of the DPFP was taken up by the polymer, clearly demonstrating the affinity of
the imprinted
polymer for the diphenyl compound.
37
CA 02446921 2003-10-16
WO 02/083708 PCT/IL02/00307
Selectivity of MIP 92-42III. The above experiment was repeated using the
polymers 92-42
and 92-42III with a diisopropyl phosphate cholinesterase inhibitor
(diisopropyl
chlorophosphate, DCP) instead of the diphenylphosphate DPFP.
Concentrations of DCP present during various stages of the experiment were
determined by measurement of percent inhibition of butyryl choline esterase
(BChe) activity
as for DPFP above.
The calibration curve for DCP is shown in Figure 13.
Samples containing 10 mg polymer, either 92-42 or 92-42III, were mixed with 1
ml
isopropyl alcohol and preincubated on an oscillating shaker at room
temperature for 24 hours.
DCP was then added to the samples to a final concentration of S p.M. The
mixture of DCP
and polymer was returned to the shaker to incubate for 24 hours. The sample
containing
polymer 92-42 was a control for background binding that occurred in locations
other than the
specific binding pocket formed by the template. Additional controls containing
isopropyl
alcohol only and 5 pM DCP in isopropyl alcohol only (no polymer) were run as
above. In
addition, controls containing 5 p,M DPFP alone and DPFP with polymer 92-42III
were run for
comparison.
After 24 hours the samples and controls were centrifuged at 14000 x g for 1
minute to
sediment the polymers. The concentrations of DPFP present in the supernatants
were then
determined by measurement of percent inhibition of butyryl choline esterase
(BChe) activity.
The percent inhibition for each sample and its equivalent DPFP concentration
from the
calibration curve are given in Table 2 below.
38
CA 02446921 2003-10-16
WO 02/083708 PCT/IL02/00307
Table 2
Sample % InhibitionDCP
of BChe Concentrat
Activity ion
t = 0
hours
Control isopropyl alcohol only 0 0
1
Control 5 pM DPFP in isopropyl 69 -----
2 alcohol
Control 5 ~M DCP in isopropyl alcohol83 S.0
4
t = 24
hours
Control isopropyl alcohol only 0 0
1
Control 5 pM DPFP in isopropyl 52 -----
2 alcohol
Control S ~M DPFP + 10 mg polymer 6 -----
3 92-42III
Control 5 pM DCP in isopropyl alcohol72 4.0
4
Control S ~M DCP + 10 mg polymer 70 4.0
92-42
Sample 62 3.0
5 ~M
DCP +
mg
polymer
92-42III
There was some DCP loss due to background hydrolysis, on the order of 1.0 ~M.
About an equivalent amount bound to the polymer. This amounted to 20% of the
initial
5 concentration of DCP. In comparison, almost all the DPFP (control 3) was
taken up by the
same amount of polymer. This lack of DCP uptake clearly demonstrates the
specificity of this
MIP for the diphenyl analog.
EXAMPLE 5
10 APPLICATIONS OF COBALTS
In addition to the previously described illustrative applications of the
COBALT
compounds according to the present invention, other applications are also
possible. These
applications may optionally include any application in which highly specific
binding to a
particular target molecule, followed by the formation of an irreversible
covalent bond
between the COBALT compound and the target molecule, is both desirable and
possible. The
previous description includes methods for designing and creating these COBALT
compounds,
which may be used according to the illustrative, exemplary applications given
below.
Bile acid sequestrants. A number of polymers, such as cholestyramine, are used
as
bile acid sequestrants. Their action is based on the presence of strongly
basic groups in the
39
CA 02446921 2003-10-16
WO 02/083708 PCT/IL02/00307
polymer (typically, ion exchange resin type of polymers) and they are used for
cholesterol
lowering and bile-related diseases. These materials are limited because they
have limited
potency and they also remove (bind) other required substances such as
nutrients, drugs, etc.
Selective COBALTs according to the present invention which bind bile acids and
salts
do not remove needed nutrients, drugs or other substances and will be more
potent.
Importantly, the COBALTs can be made so that they are selective to the more
hydrophobic
bile acids such as deoxycholic acid. While bile acids and salts serve
important functions in the
body, such as promoting digestion of fat, researchers have found that the more
hydrophobic
(water-resistant) bile acids, such as deoxycholic acid (DCA), chenodeoxycholic
acid (CDCA)
and lithocholic acid (LCA) facilitate higher absorption of lipids such as
cholesterol and fats
into the blood stream and are toxic, causing damage to cells and promoting
cancer. Current
research indicates that these more hydrophobic bile acids are highly
significant disease-
causing agents.
The COBALTs of the present invention with their irreversible binding provide
more
efficient removal of bile sequestrants, with more specific binding, than
compounds which are
known in the art.
Environmental Detection, Removal and Protection
There is a need for detection of toxic chemicals used as weapons by the
military or
terrorists such as sarine or soman nerve gases. Existing biological based
detectors lack
stability or require special conditions for storage. This limits their
application in the field.
Alternatively, systems based on materials such as the cholinesterase enzymes
lack selectivity
to specific organo-phosphate chemical weapons (see for example USA DOD CDB02-
106
Request for Proposal "Improved Field Biosensor For Organophosphates").
Similarly there is a
need for compounds that remove the chemical weapons or can be used for
protection e.g. a
CA 02446921 2003-10-16
WO 02/083708 PCT/IL02/00307
topical skin protectant (see The U.S. Army Medical Research Institute of
Chemical Defense,
Bioscience 2002 Medical Defense Review Conference).
MIPs have been developed to target organophosphate insecticides in water but
they
lack selectivity and are not usable for clean-up and protection,(see Jenkins
AL et al. Analyst
126, 798-802 (2001)). The COBALTs of this invention overcome these
limitations.
EXAMPLE 6
Cvclodextrin-based COBALTS selected from combinatorial libraries
Illustrative, non-exclusive examples of approaches for obtaining COBALTs based
on
cyclodeextrins where the covalent bond forming group on the COBALT is an
isothiocyanate
are shown in Figure 14. The beta-cyclodextrin is illustrated but alpha- or
gamma
cyclodextrin based combinatorial libraries can also be used. The degree of
substitution on the
cyclodextrin can be varied widely; in Figure 14 one combinatorial library
(Comb.Lib.A) is
illustrated with one varying substituent, R~, while the second combinatorial
library
(Comb.Lib.B) is illustrated with two varying substituents, Rland Rz. Each
could contain
three, four or more substitutents, where each substituent can be a large
number of different
groups. In each case the final library is shown as trityl protected
structures; when testing for
binding the trityl groups are removed, as shown in the upper right hand corner
of Figure 14.
Such COBALTs can be used for irreversible binding to small and molecules, to
peptides and
proteins, if these targets contain a hydroxyl or amino group.
In one example of an application of this approach, a Comb.Lib.B was prepared
containing at least 1000 members. The preparation was carned out by reacting
one equiv. of
ten different R1I substances with tritylated-mono-4-isocyanato-benzyl-beta-
cyclodextrin, in
ten different tubes, each tube containing a different R1I. Each tube was
further divided into
ten different tubes and each reacted separately with one equiv. of R2I, again
using the same
set of ten different substituted benzyl iodides. After detritylation and
treatment of each of the
41
CA 02446921 2003-10-16
WO 02/083708 PCT/IL02/00307
1,000 tubes with one equivalent of activated fluorescent coumarincarboxylic
acid, the
individual substances (not totally pure materials on the basis of TLC analysis
and HPLC of
selected wells) were applied at different concentrations (10 microM, 1 microM,
and 0.1
microM) to solutions of various proteins, including BSA, lysozyme and purified
mouse IgG,
and incubated at room temperature for periods of 1, 8 and 15 hrs. at various
pH buffers.
There were significant differences in some of the tubes, the maximum being ca.
100-
to 1000-fold differences in binding on the basis of the fluorescence. Covalent
reaction, i.e.,
COBALT activity, was indicated by minor loss of fluorescence (< 1 S%) after
dialysis of the
fluorescent protein solutions.
15
42
