Note : Les descriptions sont présentées dans la langue officielle dans laquelle elles ont été soumises.
W094/23298 21 S ~ 9 2 ~ PCT~S94/03490
DETERMINATION OF CONCENTRATION BY AFFINITY TITRATION
AND
COMPETITIVE DISPLACEMENT IN DRUG DELIVERY
Field of the Invention
5The invention in one aspect is related to
determination of concentration of an analyte in a simplified,
field-usable form suitable for use with small sample volumes.
Another application of the principles involved in such
determination relates to delivery systems wherein competition
between a portion of a complex and a releasing agent determines
the fate of a complex that includes the substance of interest.
The assays of the invention involve a systematically variable
competition immunoassay to determine concentration without the
necessity for serial dilution, either to place the concentration
of analyte within the ~1.5 log-unit range of a typical
immunoassay or to obtain multiple readings in the dynamic range.
The drug delivery aspect invention concerns use of affinity
competition to regulate the availability and/or activity of
certain binding moieties in in vivo or in in vitro environments.
Backqround Art
AssaYs
Assays to determine the concentration of analyte in
clinical, environmental, or other settings generally involve the
use of serial dilution. The purpose of such dilution is
twofold: it may be necessary to bring the concentration of the
sample within the range of the assay; if the sample is too
concentrated, a meaningful reading may not result.
Additionally, serial dilution may accommodate and span a dynamic
range wherein variable readings over a series of concentrations
is obtained, thus enhancing the precision of the result. In
other cases, the level of dilution itself may be used as an
assessment of concentration. In this instance, immunoassays or
other specific binding assays can be used to assess the quantity
of an analyte in a sample by using a multiplicity of test
regions or portions in combination with serial dilutions of the
W094/23298 ~ PCT~S94/03490
sample. A variety of test formats is used wherein the same test
format is used in these multiple test regions, but the sample
containing analyte is used in lower and lower concentrations
until a discernible response disappears. By taking account of
5 the level of dilution at which a response is no longer visible,
and comparing the results to those obtained with standards, the
concentration of analyte in the original sample can be back-
calculated.
Methods that employ serial dilution are useful, but
10 quite labor- or machine-intensive, and are not suited for
semiquantitative determinations as might be needed in testing
for analytes outside of a laboratory context. For example,
ascertaining the levels of cont~m;n~nts in soil at the location
where field testing is appropriate should be accomplished by
15 methods that require only the application of a single sample
volume, rather than the more complex and error-prone process of
obtaining multiple dilutions. Similarly, in clinical settings,
shortages of trained and reliable personnel manually to conduct
serial dilutions.for assessment is a recognized problem in
20 supplying health care; instrumentation to make such dilutions
mechanically is expensive and of limited reliability.
Furthermore, it would be desirable to conduct clinical assays on
extremely small samples so as to minimize the invasive nature of
sample taking. Conduct of serial dilution on samples in the
25 microliter range, for example, is inherently inaccurate.
One approach to this problem has been set forth in
U.S. patents 4,654,310 and 4,923,800 to Ly. In the methods
described, systematically varying amounts of test reagents in
multiple test portions are used to obtain semiquantitative
30 results for the same solution of analyte without necessity for
serial dilutions of the sample. One easily understood disclosed
approach takes advantage of two competing catalytically
controlled reactions using varying relative amounts of the two
catalysts. In its simplest form, two enzymes which utilize the J
35 analyte as the substrate compete for conversion of the substrate
to product. One of the products gives an all-or-none detectable
result; the other does not give a detectable response. If there
is a high concentration of analyte, even large amounts of
W094/23298 ~ 1~ 7 ~ PCT~S94/03490
- 3 -
competing enzyme which take away from conversion to thedetectable product don't matter; however, at low concentrations
of analyte, not enough will be left to see the result.
Therefore, high concentrations of analyte will be capable of
giving a detectable result in the presence even of high
concentrations of the competing enzyme; low concentrations of
analyte will only give a detectable result at low concentrations
of competing enzyme. In a somewhat different, but related,
approach, described in U.S. patent 4,042,329 to Hochstrasser,
variable stoichiometric reagent concentrations are used to
achieve semiquantitative results in a series of test regions.
The assay aspect of the present invention similarly
provides a method that permits quantitative analyte
concentration determination using a series of test regions
without the need for serial dilution. In contrast to the above-
mentioned techniques, the invention method takes advantage of
the variable binding affinity of a multiplicity of ligands
either with the analyte itself or with a specific binding
partner of the analyte. In either case, the invention takes
advantage of a multiplicity of ligands which react with varying
degrees of efficacy for a single substance.
Drug Delivery
Discovery of biologically active compounds or of
compounds which are capable of labeling tissues for diagnosis is
only one part of a complex problem of using these materials to
their desired ends. In order for drugs, for example, to be
effective they must remain in the environment of their intended
targets for a sufficient time to exert their effects. They also
benefit from selective delivery to these targets in many cases.
The same is true of the use of labels to localize specific
tissues. A mechanism must be found to permit such localization
to occur and then to clear the remaining background label from
the system so that the localized label can be detected. A
number of approaches are known for providing these effects. For
example, it has long been suggested that toxins which are
designed to destroy, for example, tumor cells be complexed with
specific affinity reagents such as immunoglobulins which are
W094/23298 ~ PCT~S94/03490
-- 4
capable of binding the tumors. Presumably by collecting the
combination of specific binding agent and toxin at the surface
of the tumor, the tumor is destroyed at the expense of tissues
to which the specific affinity reagent is not attracted.
Similarly, techniques are available for specific imaging of
particular tissues by combining scintigraphic labels, for
example, with agents having specific affinity for their targets.
More generally, a number of drugs have been modified to enhance
their overall half-life by coupling them to polymers such as
polyethylene glycol. However, it has not been possible
carefully to calibrate the pharmacokinetics so that the
biologically active material to be delivered is brought
exclusively to the desired target and cleared from the remaining
tissues when homing has been accomplished. Nor has it been
possible, in many cases, even to accumulate sufficient drug at
the desired location.
The drug delivery aspect of the present invention
provides strategies which improve the precision with which these
general approaches can achieve their effects, and is applicable
to delivery of substances of interest to targets in both in vivo
and in vitro environments.
Disclosure of the Invention
First, the invention provides a method which is
straightforward and quantitative for simple determination of the
concentration of an analyte of interest under field or clinical
conditions, and takes advantage of varying affinities of ligands
either for analyte or for a specific binding moiety, wherein the
specific binding moiety is reactive with the analyte. In the
preferred competitive mode, since the reagents offer varying
degrees of competition for the specific binding partner, the
competitive success of the test sample with regard to an orderly
array of competitors can be used as an index of its amount
following calibration runs with known concentrations of analyte.
Thus, in one preferred aspect, the invention is
directed to a method to determine the quantity of an analyte in
the sample, wherein the method comprises applying the sample to
a multiplicity of test regions. The test regions contain
W094/23298 21~ 7 9 2 9 PCT~S94/03490
- 5 -
ligands which have varying affinities for a specific bindingpartner that is capable of binding the desired analyte. The
test regions are arranged sequentially--i.e., in such a manner
that a monotonic increase of affinity of the contained ligands
for the specific binding partner can be discerned. Of course,
this is preferably done in the simplest possible way--e.g., in a
linear arrangement wherein ligands of increasing affinity are
arranged, for example, from left to right.
At the time the sample is contacted with the
multiplicity of test regions, a quantity of specific binding
partner is also supplied to each, and, indeed, can be contained
in each test region ab initio. After contact of the test
regions with the sample in the presence of specific binding
partner, the specific binding partner that is not bound to the
ligand coupled to the test region is washed away, if necessary.
As is known in the art, for some methods of detection, washing
is not necessary. The re~;n1ng bound specific binding partner
is then detected. The portion of test regions containing bound
specific binding partner is then used as a measure of the
analyte. Intermediate values between all-or-none in each zone
provide further precision in quantitation.
Unlike the semiquantitative data available from
conducting the methods of Ly and Hochstrasser described above,
the results obtained using the method of the present invention
are susceptible to fully quantitative analysis if desired or can
be interpreted more qualitatively. The precision of the answer
obtained can be increased by augmenting the number of and
appropriate selection of reagents for the test portions.
In the alternative, the multiplicity of test portions
is used in a direct, noncompetitive format wherein the series of
specific binding ligands are of varying affinities for the
analyte itself. At low concentrations of analyte, only those
test portions which contain ligands with high affinities for the
analyte will bind sufficient amounts of analyte to be
detectable. This form of the method of the invention can be
conducted either in a kinetically controlled or equilibrium-
controlled format. The optimization of the range of binding
affinities required to provide quantitative results will be
æs
W094l23298 PCT~S94/03490
- 6 -
different from that of the competitive format and can be
conducted using routine experimentation.
In other aspects, the invention is directed to test
devices for use in the method of the invention. These test
devices contain a multiplicity of test regions containing
ligands of varying affinity for the specific binding partner
(or, with respect to the second format, for the analyte); the
test regions are arranged in such a manner that regions
containing ligands of increasing affinity for the specific
binding partner (or analyte) can readily be discerned.
The drug delivery aspect of the invention provides a
means to control the absolute and relative amounts of materials
of interest in desired locations by effecting competition
between a releasing agent and a modulating material contained in
complexes with the substance of interest. The releasing agent
may be supplied at the appropriate time, or may be endogenous to
the targeted location. The complex is maintained intact until
the releasing agent is supplied or encountered; the component of
interest is then able to exert its effects.
Thus, the invention is directed to a method to
modulate the activity of a desired substance with respect to a
target in an environment, which method comprises supplying said
desired substance as a complex that includes a control agent;
and effecting the release of the desired substance from the
complex by interaction of the complex with a releasing agent.
The releasing agent competes with the control agent for the
desired substance or with the desired substance for the control
agent, or dissociates an aggregate effected by a control agent.
In one embodiment, the complex contains as the control
agent a size-enhancing agent which provides complexes that have
low clearance rates in in ~ivo systems. Therefore the component
of interest can be maintained in the subject as a complex of
high aggregate weight. Clearance can be effected by providing a
releasing reagent that destroys the complex and releases the
desired component. This method can be refined so that the
desired component also includes a specific affinity component
which selectively binds a target. The size-enhancing agent is
W094/23298 21~ ~ 9 ~ ~ PCT~S94/03490
not released or dissociated until sufficient desired component
has accumulated at the target.
In another embodiment, an active component is
complexed with a partner which inactivates it while it remains
in the complex. If this inactivating material is also capable
of specific binding to the target for which the biologically
active material is intended, and if binding to the target
destroys the complex, the active substance will be active only
in the presence of the target. The target or materials in its
proximity thereby function as releasing agents. Alternatively,
an external releasing agent can be used.
In a third illustrative embodiment, an active
substance is administered complexed to a competitor for its
intended substrate or pseudosubstrate. High concentrations of
its intended substrate or pseudosubstrate then release the
active material.
Brief Description of Drawings
Figures la-lc show diagrammatically the effect of
increasing concentrations of releasing agent (RA) on a complex
between active component (AC) and a control agent (CA).
Figures 2a-2b show diagrammatically the effect of
increasing concentrations of releasing agent on an aggregated
complex that includes a self-aggregating size-enhancing agent
(SEA) as the control agent.
Figures 3a-3b show diagrammatically the effect of
increasing concentrations of releasing agent on an aggregated
complex that includes size-enhancing agent and an
immunoconjugate.
Figures 4a and 4b show a more elaborate version of the
complexes of Figures 3a-3b.
Figures 5a and 5b show the effect of competition of an
epitope target at high concentration on the binding of an
immunoconjugate to a polymer containing an epitope analog.
Figure 6 shows illustrative affinity titration curves
for various reagents reactive with an analyte.
W094/~298 ~ PCT~S94/03490
- 8 -
Modes of Carryinq Out the Invention
In the invention method, titration of affinity
replaces titration of concentration for generating a readable
signal. The general principle can be described as follows:
Figure 6 illustrates three specific binding agents or ligands
(A, B and C) for which the response to di~ferent doses of
analyte are offset due to differèntial affinity of the binding
agent for the analyte or for a competitor species. As shown,
the displaced dose-response curves generate a distinguishable
response pattern for different concentration zones (numbered 1-
5) of analyte. The concentration zone of an unknown specimen
can thus be determined by comparing its response pattern with
respect to these three ligands and comparing this response to
that of standards. Quantitation within a zone can be
accomplished by standard methods, if desired, or by increasing
the number of different binding agents within this range.
U.S. Patent 5,113,866, which is incorporated herein by
reference, describes the production of diverse panels of ligands
with maximally varying characteristics. These panels are
particularly useful in supplying ligands of varying affinity for
a single moiety, including a specific binding partner for
analyte or for the analyte itself. While maximal diversity is
not required for the ligands used in the present invention, such
diversity is advantageous since it permits systematic control of
the binding of the ligands for the analyte or specific binding
partner.
While the diverse panels of ligands described in the
above-referenced patent are a preferred source of the ligands
with monotonically varying binding capability, other sources of
such ligands could also be used. A series of monoclonal
antibodies of varying affinities for their specific binding
partner, for example, could be used. Similarly, peptides having
random or systematically varied sequences can be generated using
techniques by now well known in the art.
While it is relatively inconvenient to do so, it is
not impossible to obtain a range of ligands of varying
affinities essentially by trial and error among suitable
candidates based on the nature of the target to which the ligand
W094/23298 21~ 7 ~ ~ ~; PCT~S94/03490
_ g _
is to be bound. For example, if the analyte or the specific
binding partner for analyte is an enzyme, variations on the
substrate molecule or inhibitor molecules for the enzyme may be
used. If the analyte or the specific binding partner for
analyte is a receptor, variations on the ligand known to bind to
the receptor might be used; conversely, if the target compound
is a moiety which is known to bind to a receptor, variations in
the binding site of the receptor protein can be employed.
Thus, a multiplicity of approaches for obtaining the
desired collection of ligands with systematically varying
affinities for analyte or specific binding partner for analyte
is available in the art.
The preferred competitive form of the method uses a
"specific binding partner" for analyte. As used herein,
"specific binding partner" refers to a substance which is known
to bind with considerable affinity for the desired analyte;
typical such specific binding partners would be antibodies or
immunologically reactive fragments thereof, such as Fab, Fab' or
Fab'2 fragments or, for example, when the analyte is a ligand
which matches a receptor, the receptor for the ligand, etc. In
addition, of course, if the analyte is itself an antibody, the
specific binding partner can be the antigen to which the
antibody is responsive.
The assay of the invention is preferably conducted on
a solid support matrix to which the ligands of varying affinity
for the specific binding partner or analyte are coupled.
However, it would also be possible to format the assay for
homogeneous solutions, e.g. a fluorescent energy transfer as a
detection method is available in a homogeneous solution phase;
no wash or attachment to solid support is required. The solid
matrix may be of any design, so long as discrete test regions
can be defined at its surface. Conventional substrates of this
type, such as, for example, microtiter plates, can conveniently
be used. Alternatively, flat, hydrophilic surfaces that have
been divided into test regions by application of hydrophobic
boundaries can be used. For example, a cellulose backing with
wax cross-hatchings so as to define a multiplicity of
rectangular or square regions arranged linearly could be used.
~s~
W094/23298 ~ PCT~S94/03490
- 10 -
The design of the array of test regions is a matter of
convenience and simplicity of interpreting the results.
Preferably, the regions are arranged in such a manner that a
linear array of ligands of increasing affinity for the specific
binding partner or analyte can be coupled to the backing.
Alternatively, the ligands can be arranged as a circle or a
spiral or any other convenient, orderly design pattern. A
multiplicity of a series of such ligands of monotonically
increasing affinity for specific binding partners of the same
analyte or different analytes can be arranged on the surface of
the same substrate or solid support.
The nature of the coupling of the individual ligands
to the test regions, if such coupling is desired, also varies
widely, depending on the nature of the solid substrate and the
nature of the ligand used. The binding may be covalent or by
adsorption. If peptides are used as a source of ligands having
multiple affinities, linker moieties capable of forming ester,
amide, or disulfide bonds with the peptide and suitable covalent
bonds with the substrate may be employed. However, additional
types of ligands, including nucleic acids, carbohydrates, and
other polymers could also be used. Some of these are described
in the above-referenced patent. Coupling is through
conventional procedures; for example, binding to cellulose
substrates may be effected by cyanogen bromide, alkyl
chloroformates or carbonyl diimidazole to form cyclic carbonate
or carbonate active esters.
In setting up the test device, each test region is
separately coupled to the ligand of specified affinity in a
known pattern. The coupling thus results in a series of test
regions of varying affinity for the specific binding partner or
analyte.
For conduct in the competition format, the support
containing the test regions may optionally be supplied with a
known, preferably constant, amount of specific binding partner
contained in, but not coupled to, each test region in advance of
the test itself. Alternatively, the specific binding partner
may be supplied as a separate solution at the time sample is
applied. In the conduct of the test, the sample is applied to
W094t23298 ~ g ~ PCT~S94/03490
the entire series of test regions along with a constant amount
of specific binding partner. The sample and specific binding
partner are allowed to remain in contact with the series of test
regions for sufficient time to permit competition for the
specific binding partner between the ligand and the analyte
contained in the sample. Depending on the method of detection
of binding of specific binding partners to the ligand, it may or
may not be necessary or desirable, after the incubation period,
to wash the test portions so that only specific binding partner
bound to ligand in the test region remains.
In any event, the presence or absence of ligand-bound
specific binding partner in a specific test region is detected
after incubation in competition with the sample containing
analyte. This detection can be conducted in a variety of ways.
In some methods, it is not necessary to wash away solution
containing unbound specific binding partner. For example, in a
format involving coupling of ligand to solid support, the solid
support may be provided with a fluorescence-emitting compound
wherein the fluorescence can be quenched by a moiety attached to
the specific binding partner. Only bound specific binding
partner is capable of quenching the fluorescence, and the
presence of unbound partner in the solution does not interfere
with the reading. This method can also be used in homogeneous
medium where the ligand is in solution in the test region.
Alternatively, optical devices which detect the presence or
absence of a signal, such as a fluorescence signal, only at the
surface and not elsewhere can also be employed.
More traditional methods, such as, for example, adding
a substrate solution to the series of test regions wherein the
specific binding partner is coupled to an enzyme for the
detection of bound enzyme may require prewashing of the test
regions. If the test regions are, indeed, washed free of
unbound specific binding partner, the presence or absence of the
specific binding partner is then detected for each test region
using such conventional methods. For example, if the specific
binding partner is an antibody, this antibody may itself contain
a label or may be labeled using a second antibody specifically
immunoreactive with it. Various conventional methods of
W094/23298 ~5 PCT~S94/03490
- 12 -
labeling are well known in the art, including radiolabeling,
enzymatic labeling, fluorescent labeling, and combinations of
these.
When detection has been effected, the pattern of
binding is then observed. In a single series of test regions,
samples with large concentrations of analyte will result in
failure to bind specific binding partner in the majority of test
regions. Samples containing only low amounts of analyte which
are poorly capable of competing with coupled ligand will result
in a series wherein most of the test regions show the presence
of label. Thus, the proportion of test regions showing binding
is roughly inversely proportional to the level of analyte in the
sample. The test is made semiquantitative by suitable
standardization with known amounts of analyte.
The level of precision may be varied according to the
desired need for same by adjusting the relative affinities of
the test portions for the specific binding partner. A large
number of such test portions having ligands with only moderate
increments of affinity can be used to enhance the precision of
the assay.
Alternatively, in the direct format, the multiplicity
of test portions is contacted with the sample putatively
containing analyte. The ability of the analyte to bind to
ligand in a particular test portion will depend on the affinity
of the ligand residing in that test portion for the analyte
itself. Binding can be judged on a kinetic or equilibrium
basis; if judged on a kinetic basis, short-term incubations
which terminate prior to establishment of equilibrium are used.
In this format, analyte will bind to ligand in those
test portions containing ligand for which it has the highest
affinity preferentially; at low concentrations, only test
portions having ligands with very high affinity for the analyte
will succeed in binding detectable amounts of analyte. At
higher concentrations of analyte even test portions with ligand
having relatively small affinities will be able to bind analyte.
The detection of ligand-bound analyte in this format
may also be conducted by measuring changes in characteristics of
the surface of the solid support due to binding; however, more
W094/23298 2 1 ~ 7 ~ 2 ~ PCT~S94/03490
- 13 -
conventional approaches involving removing sample containing
unbound analyte and then using a secondary binding agent
containing label are preferred. Washing, however, is generally
not indicated since this may alter the binding characteristic of
the sample. One example of this approach would employ, for
example, an antibody or fragment thereof capable of specifically
binding analyte wherein the antibody or fragment itself contains
a radioactive, fluorescent, enzyme, or other label.
The results are then read by comparing the number of
test regions binding analyte in standard concentration controls
as compared to the number of regions binding analyte in the
sample to be tested. In this format, higher concentrations of
analyte will show detectable binding in a greater number of test
regions.
Whether conducted in a direct or competitive format,
the assay is conducted using a multiplicity of test regions.
Preferably, the test regions are arranged in such a way that the
result in each can be measured and associated with a particular
ligand. For direct reading devices, some orderly arrangement
will be necessary, such as a linear arrangement of ligands with
increasing affinity for analyte or increasing affinity for the
specific binding reagent. Alternatively, other orderly
arrangements such as spirals or even two-dimensional arrays
could be used, as long as the results are intelligible. If the
test regions are simply wells of a microtiter plate or test
tubes in a rack, the arrangement is flexible and at the option
of the practitioner. Random physical arrangements may also be
used using computer processing to sort out the position of the
ligands of various affinities. In principle, however, in a
single formatted test, simply the number of test portions which
provide positive results will be determinative.
One particularly convenient method to construct a
device with the re~uired number of test regions comprises a
basic hydrophilic matrix wherein regions of the matrix are
separated from each other by hydrophobic barriers. Thus, a
cellulose mat might be subdivided into squares or other suitably
shaped regions by lines of wax or other hydrophobic barrier.
W094/~298 ~ PCT~S94/03490
- 14 -
The invention also provides methods to control the
behavior of materials in specified environments, such as
metabolic systems, by controlling the interaction of substances
of interest with other materials through a process of "affinity
competition". The material or substance of interest is included
in a complex with a "control" agent, which controls the behavior
of the substance of interest, either by inactivating it,
changing its size, or otherwise affecting its properties. The
complex is dissociated by a "releasing agent" which competes
either with the substance of interest for the control agent or
with the control agent for the substance of interest, or which
causes dissociation of a self-aggregate formed by the control
agent. The presence of the releasing agent in the proximity of
the complex between the "control agent" and the substance of
interest is achieved either arbitrarily by administering the
releasing agent to the appropriate environment or by the
location of high concentrations of the releasing agent at a
target position.
The release of the desired substance from the complex
is essentially the result of an affinity titration with respect
to the concentration of the releasing agent. As the
concentration of releasing agent becomes higher and higher, its
capacity to release the desired substance becomes substantially
greater. By adjusting the affinity of the control agent for the
desired substance or the strength of its aggregation with
respect to the affinity of the releasing agent for the substance
of interest or the control agent, the appropriate concentration
of releasing agent can be determined. If the releasing agent is
endogenous, the affinities of the components of the system must
be adjusted to accommodate the concentration of releasing agent
found in the environment.
Successful application of the present method of the
invention depends upon correct selection of the affinities of
the releasing agent for a component of the complex and of this
component for the remaining components of the complex. For
example, if the releasing agent is designed to dissociate the
complex by virtue of outcompeting the active component for
binding to a control agent, the affinity of the releasing agent
W094/23298 21~ 7 ~ ~ 9 PCT~S94/03490
- 15 -
for the control agent as compared to the affinity of the active
component to the control agent must be such that the achievable
concentration of the releasing agent is sufficient to tip the
balance of affinities in favor of the releasing agent/control
agent interaction. In order to provide releasing agents of
suitable affinities, recourse can be had to panels of substances
having systematically varied properties. This will provide a
range of affinities for the control agent so that appropriate
selection of a releasing agent can be made. Conversely, such
panels can provide candidates for a control agent suitable to
respond to a predetermined releasing agent concentration.
Such panels with systematically varying properties are
described, for example, in U.S. patent 5,133,866 as set forth
above. This patent describes panels of ligands with maximally
varying characteristics so as to minimize the number of
candidates in the panel required to span the range of desired
affinities. The use of such panels as sources for materials for
construction of the complexes as well as releasing agents is
preferred but not required. Any method for obtaining a range of
affinities so that a material of the required affinity for the
complex component (or the releasing agent) can be used.
The general principle is diagrammed in Figures la-lc.
In its simplest form, the active component (AC in the figure) is
administered in the form of a complex with a control agent (CA)
into an environment that generally has low concentrations of a
releasing agent as shown in Figure la. At these concentration
levels, the complex remains intact. However, when the releasing
agent concentration is increased, the complex is disassociated
due to the affinity of the releasing agent (RA) either for the
CA as shown in Figure lb or for the active component as shown in
Figure lc. If the releasing agent is also the target for the
active component or-is associated with it in some way, its
effective concentration will be highest at the target location.
An alternative form of this concept is diagrammed
schematically in Figures 2a and 2b. In Figure 2a, a high
aggregate weight complex is formed by virtue of the self-
association of the control agent, in this case, a size-enhancing
agent (SEA). The SEA may be covalently bound to a substance of
W094/~298 2~ PCT~S94/03490
- 16 -
interest, the active component (AC). In the absence or low
concentration of releasing agent, as shown in Figure 2a, the SEA
is aggregated to form the complex. However, in the presence of
suitable concentrations of releasing agent, as shown in Figure
2b, the size-enhancing agent is disassociated so that the active
component is present as an effectively lower aggregate weight
moiety.
The following examples are intended to illustrate but
not to limit the invention.
Example 1
AssaYs
The assays can be illustrated using materials
described by Scott, J. K. et al. Proc Natl Acad Sci USA (1992)
89:5398-5402. Briefly, the lectin Concanavalin A (ConA)
specifically binds the sugar methyl-alpha-Man (MeMan).
Multivalent ConA can also bind various bacterial-derived
dextrans such as that from strain B-1355-5, to form an insoluble
complex. Precipitation of ConA-dextran can be inhibited in a
dose-dependent manner by MeMan.
By screening a phage epitope display library, several
peptides (e.g. MYWYPY and VGRAFS) were identified which also
inhibit Con A-dextran precipitation, with distinctive 50
inhibition values.
Measurement of MeMan as an analyte can thus in
principle be accomplished using peptides as competing ligands or
for blocking the precipitation of ConA-dextran, effectively
expanding the range of MeMan concentrations that can be
measured. Dextran and peptides could also in principle be
immobilized and used to trap ConA competitively with MeMan.
Conversely, Con A could be viewed as the analyte and dextran B-
1355-5 as the specific binding agent, with MeMan and the
peptides as competing agents that allow a range of ConA
concentrations to be determined.
The drug delivery aspect of the invention as it
relates to dru~ delivery is illustrated in the following
embodiments. It will be apparent from review of these
W094/23298 2 ~ ~ ~ 9 2 9 PCT~Sg4/03490
- 17 -
embodiments that they share the general characteristics of
providing a complex between a control moiety and a substance of
interest which is dismantled in the presence of a releasing
agent.
5Example 2
Delivery of Monoclonal Antibodies Cou~led with Toxins or
Radioactive Metals for Use in Treatment or Diaqnosis
Many attempts have been made to utilize monoclonal
antibodies as specific affinity reagents to target malignant
cells and mark them either for detection or destruction. Thus,
a vast literature describing immunoconjugates for delivery of
label, toxins or other drugs, or other desired substances to
tumor targets is now available. In this embodiment of the
present invention, the immunoconjugate is, therefore, the
substance of interest. The tumor cells constitute the target.
One problem that has been encountered with the
delivery of immunoconjugates is that although a low molecular
weight, preferably less than 70 kd, and more preferably less
than 20 kd, is desired to permit penetration of the tumor mass,
immunoconjugates of such low molecular weights clear rapidly
through the kidneys. It is estimated that localization to the
tumor requires 24-36 hours; by this time, such low molecular
weight substances have mostly cleared. On the other hand, once
the appropriate amount of material has accumulated at the tumor
by virtue of the ability of the monoclonal antibody to bind,
rapid clearance becomes desirable. If the purpose of the
immunoconjugate is labeling, such clearance reduces the
background and enhances the accuracy of detection. If the
immunoconjugate is intended as a therapeutic, such clearance
prevents adverse effects to healthy tissue.
Complexes of >70 kd, however, are not cleared rapidly.
If the immunoconjugate could be made part of a complex that
exceeds this molecular weight for a sufficient time to permit
homing to the tumor tissue, and could then be released to allow
penetration and clearance, an ideal pharmacodynamic situation
would be achieved.
2~
W094/~298 PCT~S94/03490
- 18 -
This ideal situation can be achieved by supplying an
appropriate size-enhancing agent which can be maintained in a
complex with the immunoconjugate until a releasing agent is
supplied. A variety of designs of the size-enhancing agent and
the complex might be envisioned.
For example, the monoclonal antibody or preferably an
immunoreactive portion thereof -- i.e. the Fv region -- might be
constructed as a fusion protein with a peptide which can be
modified to contain both a chelating moiety for a radioisotopic
label and an extension which serves as a size-enhancing agent as
shown in Figure 3. Although a fusion protein of the Fv region
is illustrated, the control agent (in this case a size-enhancing
agent represented by the peptide extension) could be, instead, a
covalently bound moiety, including a peptide, but also including
other appropriate chemistries such as oligonucleotides.
As shown in Figure 3a, the antibody or Fv region
thereof coupled to the radiolabel or drug shown as an asterisk
in the figure) is extended by a size-enhancing agent (SEA). The
label can be placed either on the SEA or on Fv and can be added
before or after coupling. The SEA is designed so as to
encourage dimer or multimer formation. This can be effected by
using two mutually attractive extensions, SEA and SEA', which
are oppositely charged; alternatively, extensions can be used
which have a natural affinity such as hydrophobic extensions.
In general, size-enhancing agents which simply self-aggregate
are preferred since only one set of reagents needs be prepared
in this case. The affinity of the extensions for each other can
be modified by design of epitope/paratope pairs using panels of
candidates of maximal diversity as described above or by using
the screening techniques described in U.S. patent 5,217,869,
incorporated herein by reference. The releasing agent, RA, has
sufficient affinity for at least one of the SEA extensions so
that at appropriate concentrations the dimer is dissociated as
shown in Figure 3b.
A somewhat more sophisticated version of this design
is shown in Figure 4. In this illustration, the size-enhancing
agent extension has at least two binding arms formed by
branching of the fused peptide. The binding arms might then be
W094/23298 PCT~S94/03490
-- 19 --
comprised of polyionic regions, such as polylysine (poly+) or
polyglutamic or polyaspartic (poly~) which can be configured to
permit aggregation using the electrostatic attraction between
the positive and negative arms to hold the complex together.
Alternatively, the binding arms may comprise an epitope/paratope
pair to effect binding. The peptide with the branched arms
fused to the Fv or Mab is thus a size-enhancing agent which is
capable of effecting aggregation of the individual fusion
peptides. The aggregate is then administered to the subject
animal body and the complex is allowed to accumulate at target
cells or tissues. When sufficient time has elapsed to allow the
accumulation to have occurred, a releasing agent is added. If
the binding arms are polyionic, the releasing agent is in the
form of, for example, a positively or negatively charged peptide
(polylysine or polyglutamic or polyaspartic). If the binding
arms are an epitope/paratope pair, the releasing agent will be
the paratope or epitope or mimic thereof. Interaction of the
releasing agent with aggregated complexes dissociates them and
reduces the aggregate weight so that the clearance and
penetration powers of the fusion protein are realized.
The foregoing embodiment is illustrated schematically
in Figure 4. Figure 4a shows the intact aggregate assembled by
virtue of the oppositely charged polyionic arms of the branched
extension. At high concentration of a releasing agent having a
positive charge, for example, as shown in Figure 4b, the complex
is dissociated.
The active moiety may also be contained in a
controlled release setting, such as polymeric beads with
interstices bearing moieties which interact with a controlling
region of the active component. The active component, along
with its controlling region, is released from the interstices by
diffusion. The rate of release can be controlled by the
specific affinity between a portion of the active component and
a "control agent" contained in the polymer, where the effective
local concentration of the polymer-contained moiety can be
adjusted to effect the appropriate rate of diffusion. In this
instance, the releasing agent comprises the ambient conditions
W094/23298 ~ ~S~ ~ PCT~S94/03490
- 20 -
that control diffusion. Suitable polymers include porous
polymers such as polyacrylamide, collagen, and the like.
Alternatively, a radioisotope label or toxin or drug
may be coupled to the Fv unit either covalently, as in the case
of drugs or toxins, or by chelation as in the case of
radioisotopes, and the resulting immunoconjugate adsorbed to a
polymer having at least one affinity ligand (such as the
relevant antigen, or an epitope thereof) coupled to the polymer.
Suitable polymers include, for example, inert biocompatible
materials such as polyethylene glycol, polyacrylamide,
polymethacrylate, and the like. The affinity polymer size-
enhancing agent then maintains the immunoconjugate in the
environment in which the target is included until sufficient
immunoconjugate accumulates at the target. Some of the affinity
polymer size-enhancing agent may be released when the complex
reaches the target due to competition for the immunoconjugate
from the endogenous antigen (as the releasing agent); for
circulating complex, however, this can be released at an
appropriate time by administration of, for example, a peptide
representing the target epitope. The smaller epitope will then
replace the affinity polymer effectively lowering the aggregate
weight of the substitute complex so that it can be cleared
effectively.
This is illustrated in Figure 5. Figure 5 shows the
immunoconjugate complexed by affinity to an epitope or epitope
analog region contained on a polymer. At low concentrations of
the epitope as shown in Figure 5a, the complex remains intact.
When the local concentration of the epitope is increased, as
shown in Figure 5b, the complex is dissociated.
Another strategy provides, in the immunoconjugate, a
separate region of affinity for the size-enhancing agent. In
this strategy, the releasing agent is not provided by the
target. The size-enhancing agent may also be polyvalent. For
example, the immunoreactive regions of the desired antibody may
be coupled covalently to a second immunoreactive region
unrelated to that directed to the target; the immunoconjugate
will also contain the desired moiety such as toxin, drug or
radioisotope. This immunoconjugate, therefore, contains
W094/23298 2 ~ 5 7 ~ 2 9 PCT~S94/03490
- 21 -
1) immunospecific regions designed to bind the targetspecifically; 2) an immunospecific region designed to bind the
size-enhancing agent; and 3) a label or drug. The size-
enhancing agent may comprise one or more ligands specifically
immunoreactive with region of the immunoconjugate designed as
its specific binding partner covalently bound to polymer. The
complex then comprises the size-enhancing agent (such as the
derivatized polymer) complexed with the immunoconjugate through
this second immunoreactive region. The accumulation of the
complex at the target in this case does not result in any
appreciable dissociation of the complex. This is effected by
administering, as a releasing agent, the affinity ligand used to
couple the complex to the size-enhancing agent or a mimic
thereof.
Example 3
Maskinq and Unmaskina of Active Sites
The invention method may also be used to unmask the
active site of a biologically active molecule at its site of
action. This embodiment depends on either a differential
affinity of the masking agent for a releasing agent present
specifically at the site at which biological activity is to be
produced or on differential affinity of the biologically active
agent with respect to its target as compared to the masking
(control) agent.
An application of this embodiment relates to the
administration of insulin. Insulin is now administered by
injection; however, normal hepatic insulin levels are 10-50
times blood levels. In order to obtain the desired hepatic
insulin levels, it would be necessary to increase dramatically
the dose of insulin administered by injection. This has been
done, in one protocol, by administering 25 times the usual
levels of insulin along with sufficient glucose to offset the
hypoglycemic effect of these high insulin levels. This approach
is, on its face, cumbersome and less than satisfactory.
In the invention method, the active site of insulin is
masked by a ligand that has a higher affinity for hepatic-
specific cell surface antigens than for insulin. In another
W094/~298 ~5~ 22 - PCT~S94/03490
alternative, insulin could be supplied as an immunoconjugate
specific for hepatic cell antigens and released from a masking
agent-containing complex in the manner described in Example 1
above. The insulin is thus protected from activity in the blood
until it reaches the hepatic sites wherein the cell surface
antigens displace the masking agent and result in active
insulin.
Example 4
Substrate-Affinity or Pseudosubstrate Affinity Com~etition
In a third embodiment, a biologically active moiety
which operates on an endogenous substrate directly or indirectly
can be titrated out depending on the concentration of the
substrate by supplying the moiety coupled to a substitute
substrate with which the biologically active substance binds,
but does not recognize as a substrate or pseudosubstrate. For
example, insulin can be supplied as a complex with antibody or
lectin which binds to dextran; the dextran is then displaced by
high endogenous glucose levels. In this illustration, glucose
might be considered a "pseudosubstrate" for insulin since
although insulin does not interact with glucose directly, it is
responsible for its metabolism.
It is, of course, desirable to enhance the levels of
available insulin when high concentrations of glucose are
present. By coupling the insulin to an antibody or lectin which
has an affinity for dextran, a complex can be formed which
adversely affects the capacity of insulin to effect glucose
metabolism. High concentrations of glucose will release the
antibody- or lectin-bound insulin from the immobilizing dextran
and permit the insulin, though still coupled to lectin or
antibody, to perform its metabolic function. In this
illustration, the dextran plays the role of a control agent,
glucose is a releasing agent, and the desired substance is
insulin; the active component of the complex is the coupled
lectin-insulin or antibody-insulin.
W094/23298 21 S ~ 9 2 ~ PCT~S94/03490
- 23 -
Example 5
Use of Bivalent Mimotopes
In still another embodiment, advantage is taken of the
use of the combination of a bivalent mimotope, the valences of
which are complementary, with suitable affinity constants, to
the valences of a bivalent antibody or immunoreactive portion
thereof. At least one of the valences of the bivalent mimotope
has a controlled affinity for its counterpart region on the
antibody as compared to a target antigen -- e.g. a tumor
antigen. In a sense, then, this valency mimics the tumor
antigen. At low concentrations of the tumor antigen, such as
those resulting in blood from antigen-shedding, the bivalent
mimotope is complexed through both valences to its complementary
bivalent antibody and the complex is effectively inert.
However, at high concentration of tumor antigen as would be
found at the tumor site, at least the portion of the bivalent
mimotope which mimics the tumor antigen is displaced from its
binding site on the bivalent antibody effectively resulting in
binding of the bivalent antibody to the tumor. Due to the now
decreased affinity of the mimotope for antibody, wherein only
one of the antibody valences interacts with one of the mimotope
valences, the bivalent mimotope is eventually shed. Thus, the
administered complex is designed to permit the integrity of the
complex to be qualitatively affected by the difference of
endogenous concentrations of the tumor antigen -- i.e. the
releasing agent.
An alternative arrangement utilizes a chimeric protein
that contains an extension to the antibody where the extension
contains an epitope analog. The epitope analog binds to the
immunoreactive portion of another molecule of the chimera, thus
forming a dimer. The dimer is dissociated when the epitope is
~ displaced by the releasing agent, e.g. a tumor antigen.
Summary of Examples 2-5
In examples 2-5, a complex is formed that includes a
desired substance for e.g., therapy or diagnosis and a control
agent which imparts a desired characteristic to the complex --
e.g. size, inactivity of the desired substance, etc. The
W094/23298 ~ PCT~S94/03490
- 24 -
complex remains intact until dissociated when the releasing
agent is either supplied endogenously or administered separately
at an appropriate time and at an appropriate concentration.
