WO2003062831A1

WO2003062831A1 - Method for determining differences in molecular interactions and for screening a combinatorial library

Info

Publication number: WO2003062831A1
Application number: PCT/US2002/002308
Authority: WO
Inventors: Kit S. Lam; Alan L. Lehman
Original assignee: The Regents Of The University Of California
Priority date: 2002-01-24
Filing date: 2002-01-24
Publication date: 2003-07-31

Abstract

The invention includes a method for determining the differences between the molecular interactions of two different mixtures of molecules and identifying ligands specific for molecules in one mixture. The method utilizes a combinatorial library to compare the molecular interactions of the two mixtures and eliminates those interactions that are common to both mixtures and those that are unique to the first mixture, such that interactions essentially unique to the target mixture are identified. Ligands specific for molecules in the target mixture can then be identified. The invention also includes a method of screening a combinatorial library to distinguish between true positive beads and false positive beads and to provide for the identification of ligands specific for target molecules.

Description

METHOD FOR DETERMINING DIFFERENCES IN

MOLECULAR INTERACTIONS AND FOR

SCREENING A COMBINATORIAL LD3RARY

Kit S. Lam and Alan L. Lehman

The United States Government has a paid-up license in this invention and the right in limited circumstances to require the patent owner to license others on reasonable terms as provided for by the terms of Grant Nos. R21 CA78909 and R33 CA86364, awarded by the National Institutes of Health/National Cancer Institute.

BACKGROUND OF THE INVENTION

1. Field of the Invention

The invention relates generally to methods for determining the differences

between the molecular interactions of two different mixtures of molecules and

identifying ligands specific for molecules of one mixture. The invention also

relates to screening methods for one-bead-one-compound combinatorial libraries.

2. Description of Related Art

Combinatorial libraries can be used to study interactions between the

molecules of a cell. In particular, the one-bead-one-compound library method

(Lam, Kit S. et al. "A new type of synthetic peptide library for identifying ligand-

binding activity. " Nature 354 (1991): 82-84) has been used to create libraries of

compounds such as peptides, chemical oligomers, and small molecules directed

against targets such as antibodies, proteases, streptavidin and other enzymes, as

well as bacteria and whole cells. The screening of a bead library with a target

molecule or target mixture of molecules results in the identification of beads, referred to as true positive beads, that have bound the target molecules. The

chemical structures of the compounds on the true positive beads can then be

identified and the compounds confirmed as ligands for the target molecules. The

ligands can then be used to determine the interactions with, and structures of, the

target molecules.

There is a need for a method for comparing, and determiriing the differences

between, the molecular interactions of two different mixtures of molecules, in

particular, functionally-related molecules derived from normal cells and cancer

cells. There is also a need for a method for determining ligands specific for

molecules in one of the two mixtures, referred to as the target mixture.

There is also a need for a method for accurately screening a combinatorial

library to determine which solid phase supports are true positives. A problem with

most existing screening methods, which include the use of enzymes, radionuclides,

fluorescent probes, and color dyes, is that they result in some beads, referred to as

false positives, that directly bind chemicals or reagents used in the screening

process. This problem is significant where millions of beads are being screened,

as there may be only a small number of true positive beads among a larger number

of false positive beads. Thus, a screening method is needed that will accurately

identify a small number of true positive beads out of a large number of false

positive beads. SUMMARY OF THE INVENTION

The present invention is directed to a quick and efficient method for

comparing and determining the differences between the molecular interactions of

two different mixtures of molecules and of identifying ligands specific for

molecules in one of the mixtures, the target mixture. The method compares the

molecular interactions of the two mixtures and eliminates those interactions that are

common to both mixtures and those that are unique to the first mixture, such that

the interactions essentially unique to the target mixture are identified. Then,

ligands specific for molecules in the target mixture can be identified.

The method comprises: preparing first and second mixtures of molecules

with a tag or label, where the second mixture is the target mixture; introducing the

first mixture to a combinatorial library of solid phase supports; incubating the

library with the first mixture of molecules; performing a first marking step to mark

the solid phase supports that have molecules of the first mixture bound to them;

introducing the target mixture to the library; incubating the library with the target

mixture of molecules; immobilizing the library; obtaining a first image, referred

to as image "A," before the marking of any solid phase supports that have

molecules of the target mixture bound to them, showing as marked only those solid

phase supports that were marked in the first marking step; performing a

second marking step to mark the solid phase supports that have molecules of the

target mixture bound to them; obtaining a second image, referred to as image "B, " showing as marked those solid phase supports that were marked in the first marking

step and those that were marked in the second marking step; performing image

analysis on the first and second images to create a third image referred to as "C,"

showing, for each solid phase support, (B-A)/A, such that image "C" identifies the

solid phase supports that were marked only in the second marking step; isolating

a solid phase support identified in image "C"; and determining the chemical

structure of the compound on that solid phase support.

The invention is also directed to a method for screening a combinatorial bead

library that quickly and accurately distinguishes between a small number of true

positive beads and a large number of false positive beads and provides for the

identification of ligands specific for a target molecule.

BRIEF DESCRIPTION OF THE DRAWINGS

Fig. 1 is a diagram showing an example of images "A, " "B, " and "C" of the

method of the first embodiment of the invention.

Fig. 1A depicts an example of image "A," showing as stained the beads

marked in the first marking step after screening with the first mixture of molecules .

Fig. IB depicts an example of image "B," showing the beads of Fig. 1A,

showing as stained the beads marked in the first marking step after screening with

the first mixture of molecules and the beads marked in the second marking step

after screening with the target mixture of molecules. Fig 1 C depicts image " C , " created by the application of the formula (B-A)/A

to eachpair of corresponding pixels present in images "A" and "B" of Figs. lA and

IB, respectively.

Fig. 2 is a table comparing the appearance of beads shown in images "A,"

"B," and "C. "

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

Method for Determining Differences between Molecular Interactions of Two Different Mixtures of Molecules

The first embodiment of the invention is a method for determining the

differences between the molecular interactions of two different mixtures of

molecules and of identifying ligands specific for molecules in one of the

two mixtures, the target mixture. The method is not limited to mixtures of

molecules and can be used to determine the differences between the molecular

interactions of two different molecules (e.g. two different forms of a molecule) and

to identify ligands specific for one form of the molecule. For purposes of this

description, the phrase "mixture of molecules" shall be understood to mean a single

type of molecule or a mixture of molecule types.

The mixtures to be compared are preferably functionally related mixtures of

molecules, for example, protein extracts from normal cells and cancer cells of the

same tissue or organ, from different stages in the cell cycle, from wild type and

mutant organisms, or from wild type and mutant proteins, or plasma and serum. The method includes the following steps. A one-bead-one-compound

combinatorial library is synthesized, preferably by the "split synthesis" method.

Lam et al., "A new type of synthetic peptide library," 82-84. The library may

consist of peptides, chemical oligomers, oligonucleotides, or other small molecules.

For example, a peptide library containing cXXXXXc peptides, where "X" is any

L-amino acid, excluding cysteine, arginine, and lysine, and "c" is D-cysteine, can

be used. A solid phase support, such as beads or discs made of polystyrene,

agarose, acrylamide, glass, plastic, or paramagnetic substances, is used. For

purposes of this description, the term "beads" shall be understood to mean any type

of solid phase support. For a peptide library, a standard synthetic solid phase

peptide synthesis method, such as fluorenylmethyoxycarbonyl (Fmoc) chemistry

or t-butyloxycarbonyl (Boc) chemistry, is used.

The first mixture of molecules and the second mixture of molecules (the

target mixture) are tagged or labeled in a fashion that will allow for subsequent

identification of positive beads. Various types of tagging systems can be used, such

as biotinylation where biotin is the label and streptavidin-alkaline phosphatase

conjugate is the label binder, flag antigen and antiflag antibody, antigen and

corresponding antibody, glutathione-S- transferase and glutathione alkaline

phosphatase conjugate, or other systems known to those skilled in the art. Enzyme

reporting systems other than alkaline phosphatase may be used, such as horseradish peroxidase and glucose oxidase. The tagging is performed according to standard

methods pertaining to the system that is used.

Unless otherwise indicated, the screening of the beads and the bead reactions

and incubations are performed in a reaction vessel, preferably a column, although

a test tube or other container can be used. The number of beads is preferably about

100,000 to 1,000,000, although the method will work for a smaller or greater

number of beads . The size of the column depends on the size and number of beads

used. Columns from about 1 ml to 10 ml in size are appropriate for about 100,000

to 1,000,000 beads of about 80 microns in diameter, respectively. The level of

solution in the column or other vessel is maintained above the level of the beads,

so that the beads do not become dry. All of the bead reactions are incubated at

about room temperature, preferably keeping the beads in motion. All of the bead

reactions are incubated, and the beads are washed, in a solution that is suitable for

the target mixture. Such solutions may include HEPES, Tris, and phosphate-based

buffers, such as PBS.

The beads are pretreated with a solution of a blocking agent in buffer to

block non-specific binding. Suitable blocking agents include gelatin and bovine

serum albumin, for example.

The beads are screened with the first mixture of molecules by incubating the

beads with a solution of the first mixture of molecules. The amount of the first

mixture should be sufficient to saturate potential binding sites on the beads. The incubation period should be sufficient for the molecules of the first mixture to bind

to the beads, and is typically about one hour. An incubation period of more or less

than an hour may be used; however, an incubation period that is too long may

result in degradation of the proteins or other molecules in the mixture, while an

incubation period that is too short may result in insufficient time for the molecules

to bind to the beads.

After the incubation period, the beads are washed to remove unbound or

loosely associated proteins or other molecules.

If the tagging system utilizes a label and a label binder, then, before the

first marking step is performed, the beads are incubated with a solution of the label

binder for about one hour. If, for example, the label is biotin, then the label

binder is a streptavidin-alkaline phosphatase conjugate in buffer. After the

incubation period, the beads are first washed to remove any unbound label binder,

for example, streptavidin-alkaline phosphatase conjugate, and then washed in a

solution suitable for the tagging system used. If the tagging system does not utilize

a label and a label binder, then the first marking step follows directly.

The next step is the first marking step. The purpose of this step is to mark

the beads that have molecules of the first mixture bound to them. Marking agents

include an enzyme substrate, such as 5-bromo-4-chloro-3-indolyl-phosphate (BCIP)

for alkaline phosphatase, and radioactive, color, or fluorescent compounds. The

beads are marked by incubating them in a solution of the marking agent, such as BCIP in alkaline phosphatase buffer, for about one hour. A longer incubation

period, such as overnight, may be used if there are small amounts of molecules

bound to the beads. This incubation results in the marking of the beads that have

molecules of the first mixture, or chemicals or reagents used in the preceding steps,

such as streptavidin-alkaline phosphatase conjugate, bound to them, leaving the

beads with no bound molecules unmarked. If BCIP was used as the marking agent,

the marked beads will now be stained blue because, in the presence of alkaline

phosphatase, BCIP is converted to indigo which appears blue and precipitates as an

insoluble marker on the beads that have proteins or other molecules bound to them.

The unmarked beads will remain colorless.

After the incubation period, the beads are washed to remove any unreacted

marking agent.

The beads that are now marked, or stained blue if BCIP was used as the

marking agent, consist of beads that bound molecules present in the first mixture

and false positive beads. The former category may include beads that bound

molecules essentially unique to the first mixture and beads that bound molecules

common to both the first mixture and the target mixture. In addition to beads,

there may be artifacts or contaminants that are marked, such as dust or impurities.

The beads are then screened with the target mixture of molecules. The

target mixture has been tagged or labeled in the same fashion as the first mixture

of molecules, for example, by biotinylation. The beads are incubated with a solution of the target mixture. The amount of the target mixture is preferably an

amount sufficient to saturate potential binding sites on the beads. The incubation

period is, for example, about one hour, or longer, preferably about the same time

period as was used for the incubation of the beads with the first mixture.

After the incubation period, the beads are washed to remove unbound or

loosely associated proteins or other molecules.

If the tagging system utilizes a label and a label binder, then the beads are

incubated with a solution of the label binder, for example, streptavidin-alkaline

phosphatase conjugate (1.5 mg/ml) in buffer if the label is biotin, diluted between

about 1 : 1000 and 1 : 100,000, for about one hour. After the incubation period, the

beads are first washed to remove any unbound label binder, for example

streptavidin-alkaline phosphatase conjugate, and then washed in a solution suitable

for the tagging system used. If alkaline phosphatase and BCIP are being used, the

buffer is preferably alkaline phosphatase buffer.

If the marking agent that is being used is a substrate that will turn color,

such as BCIP, then, after washing, a number of permanently colored beads may be

added to the bead mixture to serve as reference points in subsequent steps. The

reference beads should be of a color different from that of the color product. If

BCIP, which marks the beads by causing them to turn blue, is being used, then red

reference beads work well. About one reference bead should be added for every

500 beads of the library. The beads are immobilized by adding a solution of a suitable support matrix

to the beads. The support matrix should be porous enough to allow diffusion of the

marking agent. If BCIP is being used, agarose or acrylamide can be used as the

support matrix. The solution should be one that is appropriate for the tagging

system used; if alkaline phosphatase and BCIP are being used, the buffer is

preferably alkaline phosphatase buffer.

The bead-matrix solution is distributed on a surface that permits imaging,

such as a thin tray or dish on a flatbed scanner, and allowed to gel. The beads

should now be immobilized. A solution appropriate for the tagging system, such

as alkaline phosphatase buffer if alkaline phosphatase and BCIP are being used,

without the marking agent, is then added to the tray and incubated for enough time

to allow the support matrix to reach an equilibrium of size, as the matrix may pull

away from the edges of the tray upon addition of the solution. This may take about

five minutes. The solution is then removed.

The tray or dish is prepared for imaging to record the position and color

intensity of the beads. If the marking agent being used is a substrate that will turn

color, then the imaging is preferably accomplished using a flatbed scanner. The

tray is placed on the flatbed scanner and prepared for transparency scanning at

about 600 to 2400 dpi, preferably at least 1200 dpi, with no sha ening or color

adjustments. Next, the second marking step is begun. The marking agent is added to the

immobilized beads in order to mark the beads that have molecules of the target

mixture bound to them. The marking agent is preferably the same one that was

used in the first marking step. A solution of the marking agent in an appropriate

buffer, such as BCIP in alkaline phosphatase buffer, is gently added to the tray on

top of the support matrix. The tray is then immediately scanned, before the

marking agent has time to react with the tagging system. The resulting graphical

image, designated "A," is saved. The purpose of this step is to obtain an image

that shows as marked or stained only the beads that were marked in the

first marking step after screening with the first mixture: the false positive beads and

the beads that bound molecules present in the first mixture . (Unmarked beads will

also appear in image "A," but they will be colorless.) If the scanning is not done

quickly enough and the marking agent reacts with the tagging system, then some

of the beads that have molecules of the target mixture bound to them may also be

shown on image "A," and the resulting image analysis will not be as accurate.

Fig. 1 shows examples of images "A," "B," and "C" in Figs. 1A, IB, and

1C, respectively. In each of Figs. 1A, IB, and 1C, the beads are indicated by

circles and are shown as if they were immobilized in a support matrix. Each bead

has multiple copies of the same compound, indicated by the straight line projecting

from the circle. In each of Figs. 1A, IB, and 1C, one of the beads has bound a

reagent used in the screening process, streptavidin-alkaline phosphatase conjugate, shown as "Strep AP. " Other beads are shown with bound proteins, indicated by

zigzag lines with a different number for each different protein.

Fig. 1A shows an example of image "A" for eight beads. In this example,

the first mixture of molecules contains proteins 1, 2, and 3, and the target mixture

contains proteins 3, 4, and 5. Four beads are stained (shown as shaded) after the

first marking step: one false positive bead with streptavidin-alkaline phosphatase

conjugate bound to it and three beads with proteins 1, 2, or 3 bound to them.

Four beads are unstained (shown as unshaded): two beads that have not bound any

molecules and two beads that have bound proteins unique to the target mixture

(proteins 4 or 5).

The scanning step is preferably performed after the addition of the marking

agent, rather than before, in order to minimize any changes that may occur due to

the addition of the marking agent, including size changes in the support matrix,

slight changes in the position of the tray on the scanner, and changes in the

refractive index of the solution in and on the matrix. This will enhance the ability

to correctly align the images in later steps. Alternatively, the scanning step may

be performed before the addition of the marking agent, but corrections for size and

refractive changes would need to be made.

After image "A" is obtained, the second marking step is completed by

incubating the immobilized beads for sufficient time to allow the marking agent to

react with the tagging system to mark the beads that have molecules of the target mixture bound to them. The tray is periodically rescanned every hour or so, as

needed, to ensure that there is sufficient time for the beads with bound molecules

to become marked. If BCIP was used as the marking agent, the marked beads will

now be stained blue. The unmarked beads will remain colorless.

One of the images is saved and designated "B," preferably an image that

shows sufficient differences in intensity of the beads shown as marked in

image "A. " Image "B" shows as marked or stained the beads that were marked in

the first marking step after screening with the first mixture, which, if BCIP was

used, may now appear a darker blue, and the beads that were marked in the

second marking step which were colorless after screening with the first mixture but

became stained after screening with the target mixture. It is the latter category of

beads that have bound molecules essentially unique to the target mixture, and it is

the molecular interactions between these beads and the molecules bound to them

that are essentially unique to the target mixture. (Unmarked beads will also appear

in image "B, " but they will be colorless.)

Fig. IB shows ήnage "B" for the eight beads shown in Fig. 1A. Six of the

beads are stained (shown as shaded) after the second marking step: the four beads

that were stained in Fig. 1A and the two beads that have bound proteins unique to

the target mixture (proteins 4 or 5). Two of the beads are unstained (shown as

unshaded): the two beads that have not bound any molecules. The reaction of the marking agent with the tagging system is then stopped.

If the marking agent is BCIP, a solution of acid is added to the tray, with gentle

shaking, to lower the pH and stop the alkaline phosphatase reaction. The scanned

trays can be stored covered, in a humid environment, until needed. Distilled water

can be added to prevent dehydration of the support matrix.

Graphical image analysis is performed to compare images "A" and "B. "

The images are preferably of the same size and are aligned at all common points.

If colored reference beads were added earlier, they may now be used to assist in

aligning images "A" and "B" so that the two images can be precisely overlaid.

Image manipulation, including alignment and conversion to gray scale if desired,

is accomplished by the use of software such as Adobe Photoshop.

The comparison of images "A" and "B" is accomplished by applying the

formula (B-A)/A to each pair of corresponding points or pixels present in

images "A" and "B. " Errors due to division by zero are eliminated by adding one

where necessary. The numerical results of (B-A)/A are used to create a third

image, designated "C, " shown in Fig. 1C. This data analysis is performed through

the use of custom programming.

Image "C" shows the beads that were marked in image "B" that were not

marked in image "A. " These are the true positive beads - those that bound

molecules essentially unique to the target mixture. Thus, the effect of the

application of the formula (B-A)/A is to compare the molecular interactions of the first mixture to those of the target mixture, eliminate those interactions common to

both mixtures and those interactions unique to the first mixture, and identify the

interactions essentially unique to the target mixture.

In Fig. 1C, the two true positive beads are shown as solid dark circles and

have bound proteins 4 or 5 that are unique to the target mixture. Image "C" may

also show the beads that were marked in image "A" and also marked in

image "B" - those beads that bound molecules present in the first mixture and the

false positive beads - but they are distinguishable from the true positive beads. If

BCIP was used, these beads appear as doughnuts with a clear inside area and a dark

outside ring. In Fig. 1C, these beads are shown as shaded rings. (Image "C" will

not show the beads that were not marked in either image "A" or image "B. ") The

application of (B-A)/A emphasizes the true positive beads by giving substantially

more weight to beads that were not marked in image "A" but marked in image "B . "

Less weight is given to beads that were marked in image "A" and also marked in

image "B. "

Fig. 2 is a table comparing the appearance of the beads in Figs. 1A, IB, and

lC. A bead with no bound molecules appears colorless in both images "A" and

"B" and does not appear at all in image "C. " A false positive bead, such as a bead

that has bound streptavidin-alkaline phosphatase conjugate, appears stained in

images "A" and "B" and as a doughnut-shaped ring in image "C. " A bead that has

bound a molecule present in the first mixture (which may be either unique to the first mixture or also present in the target mixture) appears stained in images "A"

and "B" and as a doughnut-shaped ring in image "C. " A bead that has bound a

molecule unique to the target mixture appears colorless in image "A," stained in

image "B," and as a dark circle in image "C. "

Image "C" is annotated (for example, with arrows) to indicate the true

positive beads of interest. This annotated image "C" can then be used, by

overlaying it on image "B, " to create a picking template to remove the true positive

beads from the support matrix. The tray with the bead support matrix is placed on

top of the picking template, and the beads of interest are removed using a pipette

tip.

After being removed from the matrix, the true positive beads are stringently

washed to separate the beads from the target molecules bound to the beads.

The chemical structure of the compound or ligand on each true positive bead

is then determined. In the case of a peptide library, the amino acid sequence can

be determined with an automated peptide sequencer.

Method for Screening Combinatorial Bead Library

The second embodiment of the invention is a method of screening a one-

bead-one-compound combinatorial library to distinguish between true positive beads

and false positive beads and to provide for the identification of ligands specific for

a target molecule. The method includes the following steps. A one-bead-one-compound

Lam et al., "A new type of synthetic peptide library," 82-84. The library may

For example, a peptide library containing cXXXXXc peptides, where "X" is any

be used. A solid phase support, such as beads or discs made of polystyrene,

agarose, acrylamide, glass, plastic, or paramagnetic substances, is used. For

of solid phase support. For a peptide library, a standard synthetic solid phase

peptide synthesis method, such as fluorenylmethyoxycarbonyl (Fmoc) chemistry

or t-butyloxycarbonyl (Boc) chemistry, is used.

The target molecule or mixture of molecules is tagged or labeled in a fashion

that will allow for subsequent identification of positive beads. Various types of

tagging systems can be used, such as biotinylation where biotin is the label and

streptavidin- alkaline phosphatase conjugate is the label binder, flag antigen and

antiflag antibody, antigen and corresponding antibody, glutathione-S-transferase,

and glutathione alkaline phosphatase conjugate, or other systems known to those

skilled in the art. Enzyme reporting systems other than alkaline phosphatase may

be used, such as horseradish peroxidase and glucose oxidase. The tagging is

performed according to standard methods pertaining to the system that is used. Unless otherwise indicated, the screening of the beads and the bead reactions

100,000 to 1,000,000, although the method will work for a smaller or greater

to 1,000,000 beads of about 80 microns in diameter, respectively. The level of

so that the beads do not become dry. All of the bead reactions are incubated at

about room temperature, preferably keeping the beads in motion. All of the bead

the target mixture. Such solutions may include HEPES, Tris, and phosphate-based

buffers, such as PBS.

The beads are pretreated with a solution of a blocking agent in buffer to

block non-specific binding. Suitable blocking agents include gelatin and bovine

serum albumin, for example.

If the tagging system utilizes a label and a label binder, then, before the

binder for about one hour. If, for example, the label is biotin, then the label binder

is a streptavidin-alkaline phosphatase conjugate in buffer. After the incubation

period, the beads are first washed to remove any unbound label binder, for example, streptavidin-alkaline phosphatase conjugate, and then washed in a solution

suitable for the tagging system used. If the tagging system does not utilize a label

and a label binder, then the first marking step follows directly.

The next step is the first marking step. The purpose of this step is to mark

the beads that have reagents, such as the label binder, or chemicals from the

reaction, bound to them. Marking agents include an enzyme substrate, such as

BCIP for alkaline phosphatase, and radioactive, color, or fluorescent compounds.

The beads are marked by incubating them in a solution of the marking agent, such

as BCIP in alkaline phosphatase buffer, for about one hour. A longer incubation

period, such as overnight, may be used if necessary. This incubation results in the

marking of the beads that have chemicals or reagents used in the preceding steps,

such as streptavidin-alkaline phosphatase conjugate, bound to them, leaving the

beads with no bound molecules unmarked. If BCIP was used as the marking agent,

the marked beads will now be stained blue because, in the presence of alkaline

The unmarked beads will remain colorless.

After the incubation period, the beads are washed to remove any unreacted

marking agent. The beads that are now marked, or stained blue if BCIP was used as the

marking agent, consist of false positive beads. In addition to beads, there may be

artifacts or contaminants that are marked, such as dust or impurities.

The beads are then screened with the target molecule or mixture of

molecules. The target mixture has been tagged or labeled in a manner compatible

with the chemicals and reagents used previously, for example, by biotinylation.

The beads are incubated with a solution of the target mixture. The amount of the

target mixture is preferably an amount sufficient to saturate potential binding sites

on the beads. The incubation period is preferably about one hour.

After the incubation period, the beads are washed to remove unbound or

loosely associated proteins or other molecules.

If the tagging system utilizes a label and a label binder, then the beads are

beads are first washed to remove any unbound label binder, for example

buffer is preferably alkaline phosphatase buffer.

If the marking agent that is being used is a substrate that will turn color,

such as BCIP, then, after washing, a number of permanently colored beads may be added to the bead mixture to serve as reference points in subsequent steps. The

reference beads work well. About one reference bead should be added for every

500 beads of the library.

The beads are immobilized by adding a solution of a suitable support matrix

to the beads . The support matrix should be porous enough to allow diffusion of the

marking agent. If BCIP is being used, agarose or acrylamide can be used as the

support matrix. The solution should be one that is appropriate for the tagging

system used; if alkaline phosphatase and BCIP are being used, the buffer is

preferably alkaline phosphatase buffer.

The bead-matrix solution is distributed on a surface that permits imaging,

such as a thin tray or dish on a flatbed scanner, and allowed to gel. The beads

should now be unmobilized. A solution appropriate for the tagging system, such

as alkaline phosphatase buffer if alkaline phosphatase and BCIP are being used,

without marking agent, is then added to the tray and incubated for enough time to

allow the support matrix to reach an equilibrium of size, as the matrix may pull

five minutes. The solution is then removed.

The tray or dish is prepared for imaging to record the position and color

intensity of the beads. If the marking agent being used is a substrate that will turn color, then the imaging is preferably accomplished using a flatbed scanner. The

tray is placed on the flatbed scanner and prepared for transparency scanning at

about 600 to 2400 dpi, preferably at least 1200 dpi, with no sharpening or color

adjustments.

Next, the second marking step is begun. The marking agent is added to the

immobilized beads in order to mark the beads that have the target molecule or

molecules of the target mixture bound to them. The marking agent is preferably

the same one that was used in the first marking step. A solution of the marking

agent in an appropriate buffer, such as BCIP in alkaline phosphatase buffer, is

gently added to the tray on top of the support matrix. The tray is then immediately

scanned, before the marking agent has time to react with the tagging system. The

resulting graphical image, designated "A, " is saved. The purpose of this step is to

obtain an image that shows as marked or stained only the beads that were marked

in the first marking step after screening with the chemicals and reagents: the false

positive beads. (Unmarked beads will also appear in image "A," but they will be

colorless . ) If the scanning is not done quickly enough and the marking agent reacts

with the tagging system, then some of the beads tliat have molecules of the target

mixture bound to them (i.e. the true positive beads) may also be shown on image

"A," and the resulting image analysis will not be as accurate.

The scanning step is preferably performed after the addition of the marking

agent, rather than before, in order to minimize any changes that may occur due to the addition of the marking agent, including size changes in the support matrix,

slight changes in the position of the tray on the scanner, and changes in the

refractive changes would need to be made.

After image "A" is obtained, the second marking step is completed by

react with the tagging system to mark the beads that have the target molecule or

molecules of the target mixture bound to them. The tray is periodically rescanned

every hour or so, as needed, to ensure that there is sufficient time for the beads

with bound molecules to become marked. If BCIP was used as the marking agent,

the marked beads will now be stained blue. The unmarked beads will remain

colorless.

One of the images is saved and designated "B," preferably an image that

shows sufficient differences in intensity of the beads shown as marked in

image "A. " Image "B" shows as marked or stained the beads that were marked in

the first marking step after screening with the chemicals and reagents, which, if

BCIP was used, may now appear a darker blue, and the beads that were marked in

the second marking step which were colorless after screening with the chemicals

and reagents but became stained after screening with the target mixture. It is the latter category of beads that have bound molecules present in the target mixture.

(Unmarked beads will also appear in image "B," but they will be colorless.)

The reaction of the marking agent with the tagging system is then stopped.

can be added to prevent dehydration of the support matrix.

Graphical image analysis is performed to compare images "A" and "B. "

aligning images "A" and "B" so that the two images can be precisely overlaid.

is accomplished by the use of software such as Adobe Photoshop.

The comparison of images "A" and "B" is accomplished by applying the

formula (B-A)/A to each pair of corresponding points or pixels present in

where necessary. The numerical results of (B-A)/A are used to create a third

image, designated "C. " This data analysis is performed through the use of custom

programming.

Image "C" shows the beads that were marked in image "B" that were not

marked in image "A. " These are the true positive beads - those that bound molecules present in the target mixture. Image "C" may also show the beads that

were marked in image "A" and also marked in image "B" - the false positive

beads - but they are distinguishable from the true positive beads. If BCIP was

used, these beads appear as doughnuts with a clear inside area and a dark outside

ring. (Image "C" will not show the beads that were not marked in either

image "A" or image "B . ") The application of (B-A)/A emphasizes the true positive

beads by giving substantially more weight to beads that were not marked in

image "A" but marked in image "B. " Less weight is given to beads that were

marked in image "A" and also marked in image "B. "

Image "C" is annotated (for example, with arrows) to indicate the true

positive beads of interest. This annotated image "C" can then be used, by

tip.

After being removed from the matrix, the true positive beads are stringently

washed to separate the beads from the target molecules bound to the beads.

The chemical structure of the compound or ligand on each true positive bead

be determined with an automated peptide sequencer. Example

The following is an example of the first embodiment of the invention.

General Method. The two mixtures of molecules to be compared were

human plasma, referred to as "the target mixture, " and human serum, referred to

as "the first mixture." Both the plasma and serum were biotinylated with

EZ-Link™ Sulfo-NHS-LC-Biotin from Pierce. Bead libraries were generated on

TenteGel from Rapp Polymere. Amino acids used in bead library synthesis were

from Synpep. Bead library screenings were done in disposable polypropylene

columns from Perkin Elmer Life Sciences. All buffer reagents were from Sigma

unless otherwise noted. PBS, TBS and BCIP buffers were as described in Lam,

Kit S. and Michal Lebl, "Synthesis of a one-bead one-compound combinatorial

peptide library," Methods in Molecular Biology , vol. 87, Combinatorial Peptide

Library Protocols . Edited by Shmuel Cabilly (Humana Press: 1997), 1-6. Bead

staining utilized BCIP from Bio Synth AG which was hydrolyzed with streptavidin-

alkaline phosphatase conjugate from Zymed. Bead immobilization was done in

SeaPlaque agarose from Bio Wittaker Molecular Applications in Omni tray lids

from Nunc/Nalgene. Scanned images were generated on a flatbed transparency

scanner from Umax. Computer imaging was accomplished with Adobe Photoshop

and compiler software was from Metrowerks. Amino acid sequencing was done

on an Applied Biosystems Procise 494 Protein Sequencer. Library Synthesis. "One-bead-one-compound" cXXXXXc peptide libraries

were synthesized essentially as described (Lam and Lebl, Methods in Molecular

Biology, vol. 87) using the "split synthesis approach, " where "X" denotes any L-

amino acid except for cysteine, arginine, and lysine, and "c" is D-cysteine.

TenteGel was used as a solid support and standard Fmoc chemistry was used for

the solid phase peptide synthesis reactions. (Stewart and Young 1984. Solid-

Phase Peptide Synthesis. Pierce Chemical Co. Atherton, E. and R. C. Sheppard.

1989. Solid Phase Peptide Synthesis: A Practical Approach. Practical Approach

Series.)

Extract Preparation and Biotinylation. Human plasma and serum were

prepared by the collection of human blood into stoppered vacuum tubes in the

presence of Na-heparin as an anti-coagulant for plasma and in the absence of Na-

heparin for serum. Blood was allowed to clot in the tubes for 25 minutes after

which the top layer of plasma was collected. Na-heparin was added to the serum

to equalize the amount of Na-heparin in plasma and serum after which both samples

were centrifuged at 1000 rpm for 20 minutes in a Sorvall SS34 centrifuge rotor.

After centrifugation, samples were aliquoted, dialyzed against PBS for four hours

at 4° C, quantitated and stored frozen at -80° C. The concentrations of the plasma

and serum protein extracts after dialysis were 66.5 and 57.8 mg/ml, respectively.

Plasma and serum were tagged using biotinylation where biotin was the label

and streptavidin alkaline phosphatase conjugate was the label binder. Biotinylation was performed according to the manufacturer's instructions. About 0.2 mg of

protein was incubated in PBS with a 20-fold molar excess of sulfo-NHS-LC-biotin

for 30 minutes at 23° C. The reaction was quenched by the addition of Tris-HCl

pH 7.9 to 100 mM. For calculations of molarity, the average molecular weight of

the proteins in the plasma and serum solutions was assumed to be 50 kD.

Library Screening. Four experiments were conducted. On average,

250,000 "one-bead-one-compound" beads were screened in each experiment. All

bead reactions were performed in 1.5 ml polypropylene disposable micro-columns

with inside dimensions of 37 m x 6 mm. Unless otherwise noted, all bead

reactions were incubated on a Lab quake bench top rotator (Barnstead/Thermolyne)

for 1 hour at 23° C. Bead libraries were stored in PBS with 0.05% sodium-azide.

All incubation buffers and components were thoroughly mixed before addition to

the column containing the bead library.

The beads were pretreated by placing 250,000 beads in a 1.5 ml column and

blocking non-specific protein binding in PBS + 0.1 % Tween-20, 0.1 % gelatin

(Sigma) and 0.05% sodium-azide (PBSTGNaN₃) for one hour. Following the

preblock, the beads were washed three times by the addition of 1.5 ml of

PBSTGNaN₃, followed by vacuum assisted removal of the wash buffer. The level

of the buffer solution in the column was maintained above the level of the beads at

all times. At no time were the beads allowed to become dry. After preblocking, the beads were screened with the first mixture of

molecules, by incubating the beads with 4 μg of biotinylated serum in 1 ml of

PBSTGNaN₃ for one hour. The beads were then washed three times in

PBSTGNaN₃ to remove unbound or loosely associated proteins. The beads were

then incubated for one hour with 1 μ\ streptavidin-alkaline phosphatase conjugate

in 1 ml PBSTGNaN₃. Following the streptavidin-alkaline phosphatase conjugate

(1.5 mg/ml) incubation, the beads were washed two times in TBS and once in

alkaline phosphatase buffer (100 mM Tris-HCl pH 8.8, 100 mM NaCl, and

2.34 mM MgCl₂) to remove unbound streptavidin-alkaline phosphatase conjugate

and to replace the phosphate-based buffer with a Tris-based buffer.

The first marking step was performed by incubating the beads with a

solution of BCIP at a final concentration of 0.165 mg/ml in one ml of alkaline

phosphatase buffer. After incubation for one hour, beads with bound protein were

stained varying shades of blue through the action of the attached streptavidin-

alkaline phosphatase conjugate. The BCIP reaction was terminated by washing

three times with PBSTGNaN₃ to remove unreacted BCIP.

After screening with the serum and the first marking step, the beads were

screened with the target mixture by incubating the beads with 4 μg of biotinylated

plasma in 1 ml of PBSTGNaN₃ for one hour. The beads were then washed

three times in PBSTGNaN₃ to remove unbound or loosely associated proteins . The

beads were incubated a second time, as above, for one hour with streptavidin- alkaline phosphatase conjugate and then washed two times in TBS and once in

alkaline phosphatase buffer.

After washing, approximately 2,500 red reference beads were added to serve

as reference points in subsequent steps.

Next, the beads were immobilized in a support matrix. A solution of 1 %

low-melt SeaPlaque agarose in 5 ml of alkaline phosphatase buffer was prepared.

The agarose solution was cooled to about 45° C, and successive 1 ml aliquots were

added to the column, withdrawn with beads, and placed in the lid of a sterile

polystyrene OmniTray. The dimensions of the lid were 86 mm wide x 128 mm long

x 5 mm high. The bead-agarose solution was spread evenly across the surface of

the lid through a gentle shaking action before being allowed to cool and harden.

After the bead-agarose solution had hardened, 5 ml of alkaline phosphatase buffer

was added to the lid and allowed to equilibrate for five minutes. During this time,

the agarose pulled slightly away from the edge of the lid. After five minutes, the

alkaline phosphatase buffer was removed with gentle suction, and the lid was

placed on a flatbed scanner set up for transparency scanning.

The second marking step was begun. Five ml of alkaline phosphatase buffer

containing 0.165 mg/ml BCIP was gently added drop-wise on top of the agarose

in the lid. Then the lid and beads were immediately scanned at 1200 dpi with no

sharpening or color adjustments using the transparency scanning mode. The

resultant RGB image was saved and designated image "A. " The second marking step was completed by incubating the beads for one hour. Then, the beads were

scanned again and the second scanned image designated image "B. " Images "A"

and "B" were compared, and if needed, the beads were scanned again after

one more hour. After scanning, the BCIP reaction was stopped by the addition of

10-20 drops of 1 N HC1 and gentle rocking of the lid to mix the acid with the

alkaline phosphatase buffer. The addition of acid lowers the pH below the

optimum for alkaline phosphatase and essentially stops the reaction. Scanned plates

were stored covered, in a humid environment, until needed. When necessary,

distilled water was added drop-wise to prevent agarose dehydration.

Image and Data Analysis. Proper analysis of images "A" and "B" required

that the images be of the same size and that all common points in the two images

be aligned so that they could be precisely overlaid. This image manipulation was

accomplished with Adobe Photoshop. After alignment, the images were converted

to grayscale and saved in raw image format. The RGB image can either be

converted directly to a grayscale image or the red channel in a RGB image can be

used if increased sensitivity is needed.

Data analysis of the image pairs was accomplished using custom

programming, done in C code. The C program created three arrays. The size of

each of the arrays, in bytes, was determined by multiplying the height of the image

by the width. Images "A" and "B" were then loaded into the first two arrays, and

the formula (B-A)/A was applied to each pair of corresponding points. Division- by-zero errors were avoided by adding one where necessary . The numerical results

of (B-A)/A were placed in the third array at the same relative point in the array at

which the "A" and "B" values were resident. Following these calculations, the

values in the third array were written out in raw data image format.

The contents of the third data file were opened as a raw image file,

designated image "C. " Image "C" was used as a template to identify the true

positive beads, those that bound molecules essentially unique to the plasma. This

was accomplished by a visual analysis of all points corresponding to a particular

bead in image "C, " as well as images "A" and "B. " Beads that appeared whole in

image "C" were confirmed as unique to image "B" by side-by-side comparison of

images "A" and "B. " Arrows pointing towards beads of interest were drawn on

image "C. " These arrows were then overlaid on image "B" ; the combined image

was printed and used as a picking template for retrieval of the beads of interest.

To isolate beads of interest, the lid containing the beads immobilized in agarose

was placed on top of the picking template and then both were placed under a

dissecting microscope. Gel-loading pipette tips were used to retrieve those beads

indicated by arrows.

Determination of primary amino acid sequence. After retrieval, beads of

interest were washed sequentially in distilled water, 8M guanidine-HCl,

8M guanidine-HCl, and distilled water prior to submission for automated peptide

sequencing. The following ligands were identified: cHTLHQc, cFHNNHc, cAHVWHc, cHVHPWc, cHYHVSc, cHGHTIc, cMHGHFc, cYGHFSc,

cNLTHIc, cHYQTGc, cGIHYLc, and cPFHHSc, where each amino acid is an

L-amino acid.

The invention has been described above with reference to the preferred

embodiments. Those skilled in the art may envision other embodiments and

variations of the invention that fall within the scope of the claims.

Claims

CLAIMSWe claim:

1. A method for determining the differences between the molecular interactions

of two different mixtures of molecules, comprising:

labeling a .first mixture of molecules and a target mixture of

molecules;

introducing said first mixture of molecules to a combinatorial library

of solid phase supports;

incubating said combinatorial library with said first mixture of

molecules;

performing a first marking step to mark those of said solid phase

supports that have a molecule of said first mixture bound to

them;

introducing said target mixture of molecules to said combinatorial

library;

incubating said combinatorial library with said target mixture of

molecules;

obtaining a first image showing as marked those of said solid phase

supports that have a molecule of said first mixture bound to

them; performing a second marking step to mark those of said solid phase

supports that have a molecule of said target mixture bound to

them;

obtaining a second image showing as marked those of said solid

phase supports that have a molecule of said target mixture

bound to them; and

creating a third image identifying those of said solid phase supports

that have a molecule of said target mixture bound to them,

wherein said third image is created by comparing said

first image and said second image.

2. The method of Claim 1, further comprising:

isolating one of said solid phase supports identified in said

third image; and

determining the chemical structure of a ligand on one of said isolated

solid phase supports.

3. The metliod of Claim 1 , wherein said first mixture of molecules is a protein

extract from normal cells and said target mixture of molecules is a protein

extract from cancer cells.

4. The method of Claim 1, wherein said combinatorial library is a one-bead-

one-compound peptide library.

5. The method of Claim 1 , wherein said labeling is performed by biotinylation.

6. The method of Claim 5, further comprising:

before said first marking step is performed, incubating said

combinatorial library with a solution of streptavidin-alkaline

phosphatase conjugate; and,

after said incubating said combinatorial library with said target

mixture of molecules is performed, but before said obtaining

said first image is performed, incubating said combinatorial

library with a solution of streptavidin-alkaline phosphatase

conjugate.

7. The method of Claim 1 , wherein said labeling is performed by the use of an

antigen and corresponding antibody.

8. The method of Claim 1 , wherein said first and said second marking steps are

performed by incubating said combinatorial library in a solution of 5-bromo-

4-chloro-3-indolyl-phospate .

9. The method of Claim 1, wherein said combinatorial library is immobilized

in a support matrix before said first image is obtained.

10. The method of Claim 1, wherein said first and said second images are

graphical images, and said third image is created by comparing said first and

said second images on a pixel-by-pixel basis.

11. The method of Claim 10, wherein said first image is image "A, " said second

image is image "B, " and said third image is created by applying the formula

(B-A)/A on a pixel-by-pixel basis.

12. A method for determining the differences between the molecular interactions

of two different mixtures of molecules and identifying a ligand specific for

a molecule in one of the mixtures, comprising:

labeling a first mixture of molecules and a target mixture of

molecules;

introducing said first mixture of molecules to a combinatorial library

of solid phase supports;

incubating said combinatorial library with said first mixture of

molecules;

performing a first marking step to mark those of said solid phase

supports that have a molecule of said first mixture bound to

them;

introducing said target mixture of molecules to said combinatorial

library;

incubating said combinatorial library with said target mixture of

molecules; obtaining a first image showing as marked those of said solid phase

supports that have a molecule of said first mixture bound to

them;

performing a second marking step to mark those of said solid phase

supports that have a molecule of said target mixture bound to

them;

obtaining a second image showing as marked those of said solid

phase supports that have a molecule of said target mixture

bound to them;

creating a third image identifying those of said solid phase supports

that have a molecule of said target mixture bound to them,

wherein said third image is created by comparing said

first image and said second image;

isolating one of said solid phase supports identified in said

third image; and

determining the chemical structure of a ligand on one of said isolated

solid phase supports.

13. The method of Claim 12, wherein said first and said second images are

graphical images, and said third image is created by comparing said first and

said second images on a pixel-by-pixel basis.

14. The method of Claim 13, wherein said first image is image "A, " said second

image is image "B, " and said third image is created by applying the formula

(B-A)/A on a pixel-by-pixel basis.

15. A method for identifying a ligand specific for a molecule in one of

two different mixtures of molecules, comprising:

labeling a first mixture of molecules and a target mixture of

molecules;

introducing said first mixture of molecules to a combinatorial library

of solid phase supports;

incubating said combinatorial library with said first mixture of

molecules;

performing a first marking step to mark those of said solid phase

supports that have a molecule of said first mixture bound to

them;

introducing said target mixture of molecules to said combinatorial

library;

incubating said combinatorial library with said target mixture of

molecules;

obtaining a first image showing as marked those of said solid phase

supports that have a molecule of said first mixture bound to

them; performing a second marking step to mark those of said solid phase

supports that have a molecule of said target mixture bound to

them;

obtaining a second image showing as marked those of said solid

phase supports that have a molecule of said target mixture

bound to them;

creating a third image identifying those of said solid phase supports

that have a molecule of said target mixture bound to them,

wherein said third image is created by comparing said

first image and said second image;

isolating one of said solid phase supports identified in said

third image; and

determining the chemical structure of a ligand on one of said isolated

solid phase supports.

16. A method for identifying a ligand specific for a target molecule , comprising :

labeling a target molecule;

incubating a combinatorial library of solid phase supports with a

label binder;

performing a first marking step to mark those of said solid phase

supports that have a molecule of said label binder bound to

them; introducing said target molecule to said combinatorial library;

incubating said combinatorial library with said target molecule;

obtaining a first image showing as marked those of said solid phase

supports that were marked in said first marking step;

performing a second marking step to mark those of said solid phase

supports that have a target molecule bound to them;

obtaining a second image showing as marked those of said solid

phase supports that have a target molecule bound to them;

creating a third image identifying those of said solid phase supports

that have a target molecule bound to them, wherein said

third image is created by comparing said first image and said

second image;

isolating one of said solid phase supports identified in said

third image; and

determining the chemical structure of a ligand on one of said isolated

solid phase supports.

17. The method of Claim 16, wherein said labeling is performed by

biotinylation, and further, wherein said label binder is streptavidin-alkaline

phosphatase conjugate.

18. The method of Claim 16, wherein said labeling is performed by the use of

an antigen and corresponding antibody.

9. A method of screening a combinatorial library, comprising:

labeling a target molecule;

incubating a combinatorial library of solid phase supports with a

label binder;

performing a first marking step to mark those of said solid phase

supports that have a molecule of said label binder bound to

them;

introducing said target molecule to said combinatorial library;

incubating said combinatorial library with said target molecule;

obtaining a first image showing as marked those of said solid phase

supports that were marked in said first marking step;

performing a second marking step to mark those of said solid phase

supports that have a target molecule bound to them;

obtaining a second image showing as marked those of said solid

phase supports that have a target molecule bound to them; and

creating a third image identifying those of said solid phase supports

that have a target molecule bound to them, wherein said

third image is created by comparing said first image and said

second image.

0. The method of Claim 19, further comprising:

isolating one of said solid phase supports identified in said

third image; and

determining the chemical structure of a ligand on one of said isolated

solid phase supports.