WO2003035911A1 - Clone identification by direct detection of nucleic acid binding proteins from vertebrate cell expression systems and applications thereof - Google Patents

Abstract

A general vertebrate cloning and screening process for identification of genes encoding nucleic acid binding proteins is disclosed, including steps of library division and protein expression, followed by a nucleic acid binding assay as a clone screening step that is analyzed by either an Electrophoretic Mobility Shift Assay (EMSA) or a chromatography assay, preferably in a high throughput assay format. The disclosed technology provides a complete clone selection system for genes of DNA and RNA binding proteins expressed in vertebrate cells. Included are ways to optimize vertebrate cell transformation or transfection conditions by measuring the nucleic acid binding function of an expressed control protein. Also included is a high throughput electrophoretic mobility shift assay (EMSA) for detection of nucleic acid binding proteins, using a novel application of thin layer agarose gels for EMSA, combining a high vacuum and high temperature blotting technique for agarose gel desiccation with the use of a high efficiency transfer matrix (for instance, quaternary aminated nylon membranes) for blotting nucleic acid-protein complexes from agarose gels. Compensation for sample-to-sample variations in signal-to-noise ratio in clone screening is provided by introducing a reference probe to detect binding of a known protein expressed in the vertebrate host cell, along with the tester probe with the target sequence for which new clones encoding binding proteins are sought. High throughput chromatography screening methodology for nucleic acid binding proteins is also disclosed, allowing automation of screening. The disclosed technology includes combinatorial applications providing assays for functional gene linkage groups and can also be applied in connection with cDNA microarray and protein chip technologies.

Description

TITLE OF THE INVENTION

CLONE IDENTIFICATION BY DIRECT DETECTION OF NUCLEIC ACID BINDING PROTEINS FROM VERTEBRATE CELL EXPRESSION SYSTEMS AND APPLICATIONS THEREOF

This application claims priority from U.S. Provisional Application Serial

No. 60/330,633 filed October 26, 2001. The entirety of that provisional

application is incorporated herein by reference.

BACKGROUND OF THE INVENTION

Field of the Invention

The present invention relates generally to methods for screening,

identifying and selecting DNA clones encoding polypeptides which, as expressed

in vertebrate cells, specifically bind nucleic acid (DNA and RNA) sequences, or

DNA clones expressing epigenetic factors that affect protein-nucleic acid

interactions. The present invention further relates to methods of combinatorial

expression of known genes in an array format for the purpose of assessing

functional linage by nucleic acid binding assays.

Background of the Technology

For both DNA binding protein clone screening and for effective

polypeptide drug selection, it is desirable to express vertebrate polypeptides in

vertebrate systems because the binding constant and binding specificity of an expressed polypeptide is largely influenced by post-translational modifications such as phosphorylation, glycosylation, cleavage and folding. These properties are

cell-type specific, rarely inter-changeable and tightly attributable to the functional phenotype of a living cell. Therefore, significant efforts have been made to develop a vertebrae clone expression and selection system for expression cloning

of vertebrate genes.

A pioneer technique that describes the subdivision of positive cDNA pools opened the possibility of clone selection in vertebrate cells. Wong, et al.. Science,

228, 810-815 (1985); U. S. Patent No. 4,675,285 to Clark, et al. However, this method lacks a sensitive high throughput clone-selection application for identifying

the cDNA of DNA or RNA binding proteins. Application of this technology for

cloning of a DNA binding protein has been successful only when high level

transient gene expression was examined (Tsai. et al.. Nature, 339, 446-451 (1989).

Previous applications of polyacrylamide gel electrophoresis for detection of

nucleic acid binding proteins as a clone selection method have suffered numerous

disadvantages. See, e.g., Varshavskv, "Electrophoretic assay for DNA-Binding

Proteins", Methods in Enzymol., 151, 551-565 (1987). Alternative technologies which are suitable for selection of cDNA clones coding for nucleic acid binding proteins either completely avoid the expression of the protein of interest in living

cells (U.S. Patent No. 5,654,150) or apply indirect, reporter detection technologies

(U.S. Patent Nos. 5,989,814, 5,610,015, and 6,261,772, SenGupta. et al. Proc.

Natl. Acad. Sci. USA, 93, 8496-8501 (1996), and Inouve. et al.. DNA Cell Biol., 7, 731-742 (1994)). Indirect detection technologies have been developed mainly because of the lack of a direct, sensitive high throughput detection method that would decrease false-positive clone identifications. The present invention provides a vertebrate clone expression, screening and selection system for expression

cloning vertebrate genes encoding nucleic acid binding proteins and for selection of clones for epigenetic factors that influence interactions between proteins and nucleic acids. The present invention further provides a method for identification of

a majority of genes in a combinatorial gene expression format that is compatible with cDNA arrays and protein arrays. In this format of the invention functional

linkages of nucleic acid binding proteins can be efficiently assessed.

Some of the work disclosed herein is reported in Dobi. et al.. "Mammalian

Expression Cloning of Nucleic Acid Binding Proteins by Agarose Thin-Layer

Gelshift Clone Selection", BioTechniques, 33, 868-872 (2002).

SUMMARY OF THE INVENTION

According to a first aspect of the invention, there is provided a method for detecting a cDNA clone encoding a vertebrate nucleic acid binding protein that

specifically binds to a tester nucleotide sequence. This method comprises the steps

of: (a) obtaining a library of vertebrate cDNA clones to be tested for encoding the

vertebrate nucleic acid binding protein, wherein each of the cDNA clones is

operably inserted into a vertebrate expression vector; (b) introducing a portion of the library of vertebrate cDNA clones into a culture of vertebrate host cells under

conditions such that a polypeptide encoded by a cDNA clone operably inserted into the vertebrate expression vector is expressed; and (c) contacting a polypeptide encoded by a cDNA clone that is expressed in step (b) with the tester nucleotide sequence under conditions such that the nucleic acid binding protein, if present, specifically binds to the tester nucleotide sequence, thereby forming a specific protein-nucleic acid complex.

The invention method of detecting a cDNA clone encoding a vertebrate nucleic acid binding protein further comprises electrophoretically separating the specific protein-nucleic acid complex from protein and nucleic acid not in that

complex, using high-resolution gel electrophoresis according to invention methods

disclosed herein. In particular, the method of detecting a cDNA clone encoding a vertebrate nucleic acid binding protein further comprises the steps of (d)

electrophoretically separating the specific protein-nucleic acid complex from

protein and tester nucleotide sequence not present in the specific protein-nucleic

acid complex in a thin-layer polysaccharide hydrogel under conditions such that

migration of the specific protein-nucleic acid complex is detectably retarded compared to migration of the tester probe and to migration of the protein; and (e)

detecting the specific protein-nucleic acid complex, if any, that was separated from

the protein and the tester nucleotide sequence in step (d), by detecting retarded migration of protein or tester nucleotide sequence. In this step, detection of

specific protein-nucleic acid complex indicates that a cDNA clone encoding the

vertebrate nucleic acid binding protein has been detected in the tested portion of

the library of vertebrate cDNA clones.

In the invention method of detecting a cDNA clone encoding a vertebrate

nucleic acid binding protein, preferably the tested portion of the library of vertebrate cDNA clones consists of about 500 cDNA clones that are expressed in the culture of vertebrate host cells. In another aspect, the invention method of detecting a cDNA clone encoding a vertebrate nucleic acid binding protein further comprises steps of subdividing the portion of the library of vertebrate cDNA clones in which a cDNA clone encoding said vertebrate nucleic acid binding protein has been detected into subportions and separately testing each of said subportions for the presence of a cDNA clone encoding the vertebrate nucleic acid binding protein by performing steps (b) through (e).

In a preferred embodiment of the invention method of detecting a cDNA clone encoding a vertebrate nucleic acid binding protein, the method further comprises the following steps to control for varying concentrations of vector DNA in separately amplified portions of a library: step (f), prior to step (b), mixing a control DNA operably inserted into the vertebrate expression vector with the library of vertebrate cDNA clones at a known molar ratio, where this control DNA encodes a control DNA binding protein that specifically binds to a control nucleotide sequence; step (g), prior to step (b) and after step (f), amplifying the portion of the cDNA library by replication in a host cell; and (h) providing the control nucleotide sequence during the contacting of the polypeptide encoded by a cDNA clone that is expressed in step (b) with the tester nucleotide sequence. This contacting of the protein and nucleotide sequences is performed under conditions such that the control nucleic acid binding protein, if present, specifically binds to the control nucleotide sequence, thereby forming a specific control protein-nucleic acid complex. After this, the method includes step (i), detecting the specific control protein-nucleic acid complex, if any, that was separated from the control protein and the control nucleotide sequence in step (d), by detecting retarded migration of the control protein or said control nucleotide sequence. Thereafter, this form of the method includes repeating steps (b)-(i) with a second portion of the library, after mixing the control DNA with the library, and determining the ratio of the amount of specific control protein-nucleic acid complex detected using proteins expressed from the portion of the library to the amount of the specific control protein-nucleic acid complex (detected under the same conditions) using proteins expressed from the second portion of the library. In this method, the ratio of the amount of control protein-nucleic acid complex in the two portions of the library indicates the ratio of cDNA clones in those two portions, after separate amplification of each portion as in step (g).

In the invention method, due to the sensitivity of detection methods of the invention (see Example 1 , FIG. 2), for detecting a cDNA clone encoding a vertebrate nucleic acid binding protein one may use a vertebrate host cell that endogenously expresses the same vertebrate nucleic acid binding protein or another vertebrate nucleic acid binding protein that specifically binds to the tester nucleotide sequence.

In yet another aspect, the invention relates to a method for identifying a polynucleotide encoding a nucleic acid binding protein that specifically binds to a tester nucleotide sequence. This method comprises the steps of: (a) obtaining a protein encoded by a polynucleotide to be tested for encoding the nucleic acid binding protein; (b) contacting the protein with the tester nucleotide sequence under conditions such that the nucleic acid binding protein, if present, specifically binds to the tester nucleotide sequence, thereby forming a specific protein-nucleic acid complex; and (c) electrophoretically separating the specific protein-nucleic acid complex from protein and tester nucleotide sequence not present in the specific protein-nucleic acid complex. This separation uses the invention method

employing a thin-layer polysaccharide hydrogel under conditions such that migration of the specific protein-nucleic acid complex is detectably retarded compared to migration of the tester probe and to migration of the protein. This method for identifying a polynucleotide encoding a nucleic acid binding protein

further comprises a step (d), detecting the specific protein-nucleic acid complex, if

any, that was separated from the protein and the tester nucleotide sequence in step

(c), by detecting retarded migration of the protein or the tester nucleotide sequence. In this step, detection of the specific protein-nucleic acid complex identifies the

polynucleotide as a polynucleotide encoding the nucleic acid binding protein.

Still another aspect of the invention re relates to a method for identifying a

nucleic acid binding protein that specifically binds to a tester nucleotide sequence.

This method method comprises the steps of: (a) obtaining a polypeptide preparation to be tested for the presence of the nucleic acid binding protein; (b)

contacting the polypeptide preparation with the tester nucleotide sequence under

conditions such that the nucleic acid binding protein, if present, specifically binds

to the tester nucleotide sequence, thereby forming a specific protein-nucleic acid complex; and (c) electrophoretically separating the specific protein-nucleic acid complex from protein and tester nucleotide sequence not present in the specific

protein-nucleic acid complex, according to the invention method, using a thin-layer

polysaccharide hydrogel; and (d) detecting the specific protein-nucleic acid complex, if any, that was separated from the protein and the tester nucleotide sequence in step (c), as above. In this step, detection of the specific protein-nucleic acid complex indicates that the nucleic acid binding protein was present in the

polypeptide preparation.

In this method, the specific protein-nucleic acid complex may be detected

by detecting retarded migration of the tester nucleotide sequence in the specific protein-nucleic acid complex, for instance, where the tester nucleotide sequence

includes a labeling moiety and the specific protein-nucleic acid complex is detected

by detecting retarded migration of that labeling moiety. Alternatively, or in addition, the specific protein-nucleic acid complex may be detected by detecting

retarded migration of the protein in the specific protein-nucleic acid complex, for instance, where protein in the specific protein-nucleic acid complex is detected by

an antibody that specifically binds to the protein. When antibody is used to detect

protein in the complex, the antibody that specifically binds to the protein in the

specific protein-nucleic acid complex may be bound to the specific protein-nucleic

acid complex prior to the step (c) of electrophoretically separating the specific

protein-nucleic acid complex from protein and tester nucleotide sequence.

A variety of polypeptide sources may be tested for nucleic acid binding

proteins according to the invention methods. For instance, the polypeptide

preparation may be obtained by extraction of polypeptides from a vertebrate cell expressing either an endogenous or exogenous gene for the protein. In preferred

embodiments, the polypeptide preparation is obtained by expression of a cDNA

clone library in a vertebrate expression system, particularly a mammalian expression system. Preferably, the vertebrate expression system comprises an expression vector and a vertebrate host cell that expresses polypeptides from cDNA clones carried in the expression vector. Alternatively, the vertebrate expression system comprises a vector and a cell-free coupled transcription-

translation system that expresses polypeptides from cDNA clones carried in the

vector.

Another aspect of the invention relates to a method for separating a specific

protein-nucleic acid complex from protein and nucleic acid not present in that specific protein-nucleic acid complex. This method comprises a step of

electrophoretically separating the specific protein-nucleic acid complex from the

nucleic acid and the protein in a thin-layer polysaccharide hydrogel under conditions such that migration of the specific protein-nucleic acid complex is detectably retarded compared to migration of the protein and to migration of the

nucleic acid. In this method, preferably the thin-layer polysaccharide hydrogel

comprises an agarose gel and has a thickness of less than or equal to about 3 mm.

More preferably the thin-layer polysaccharide hydrogel has a thickness in the range

of about 2.0 mm to about 2.5 mm. The thin-layer polysaccharide hydrogel

comprises about 0.5 % to about 2.5 % (w/v) agarose, preferably about 1.0 % (w/v)

agarose. This separation method of the invention is useful for detecting complexes

of proteins with DNA or with RNA.

Yet another aspect of the invention relates to a method for identifying an agent that inhibits or enhances formation of a specific protein-nucleic acid complex, using the separation method of the invention. More in particular, the

invention provides a method for identifying an agent that inhibits or enhances

formation of a specific protein-nucleic acid complex comprising a nucleic acid binding protein and a tester nucleotide sequence, which method comprises the steps of: (a) contacting the nucleic acid binding protein with the tester nucleotide sequence in a first sample and a second sample, in the presence and absence,

respectively, of a putative agent to be tested for inhibiting or enhancing formation

of the specific protein-nucleic acid complex. This step is performed under conditions such that, in the absence of the putative agent, the nucleic acid binding protein specifically binds to the tester nucleotide sequence, thereby forming the specific protein-nucleic acid complex. Next is step (c), electrophoretically

separating the specific protein-nucleic acid complex formed in each of the first sample and the second sample, from protein and tester nucleotide sequence not

present in the specific protein-nucleic acid complex, according to the invention

method, above; and (d) determining the amount of the specific protein-nucleic acid

complex, if any, in each of the first sample and the second sample, that was

separated from the protein and the tester nucleotide sequence in step (c), as

described above. In this method, a decrease in the amount of the specific protein- nucleic acid complex in the presence of the putative agent in the first sample

compared to the amount of the specific protein-nucleic acid complex in the absence

of the putative agent in the second sample indicates that the putative agent inhibits

formation of said specific protein-nucleic acid complex; and an increase in the

amount of the specific protein-nucleic acid complex in the presence of the putative agent in the first sample compared to the amount of the specific protein-nucleic

acid complex in the absence of the putative agent in the second sample indicates

that the putative agent enhances formation of the specific protein-nucleic acid

complex. BRIEF DESCRIPTION OF THE DRAWINGS

The present invention may be better understood with reference to the

accompanying drawings in which:

FIG. 1 shows a flowchart of a preferred embodiment of the invention method for identification of a DNA clone that expresses a nucleic acid binding protein for a selected nucleotide sequence ("tester probe").

FIG. 2 illustrates the sensitivity of the invention method for detection of

clones expressing two different DNA binding proteins, including detection of expression of a protein from a clone in a host cell that endogenously produces the

same protein. Expression vectors for either the Mus musculus SATB1 gene or the

Rattus norvegicus HMGI(Y) embryonic brain-specific cDNA were mixed with

cDNA library pools in the indicated ratios, and complexes of DNA with DNA

binding proteins were detected by electrophoresis as described in Example 1. The

autoradiograph shows that both proteins were clearly detected by the invention

method even at the lowest ratio (1 :500) of target clone to library clones, producing the smallest number cells expressing the target clones. (B) A clone expressing

mouse SATB1 protein was mixed into an expression library at a 1 : 500,000 ratio,

and the expressed protein pools were tested for the presence of SATB1 DNA-binding activity according to the invention method. A band of complexes of

the SATB1 protein with labeled target DNA is visible in the analysis of plate 7,

row E, lane 13.

FIG. 3 illustrates identification of the SATB1 protein-expressing clone in a library pool at a ratio of 1:500,000 using an antibody-protein-DNA complex in a

"supershift" assay, as described in Example 2, in which anti-SATBl antibody is added to the protein-DNA complexes before electrophoresis, resulting in

"supershifting" of the tester DNA to antibody-protein-DNA complexes that migrate more slowly than DNA or protein-DNA complexes.

FIG. 4 illustrates the capability of the invention methods for identifying an unknown binding protein, as described in Example 3. (A) Complexes of DNA and an unknown binding protein are detected in plate 6, row G, pool 8 (upper gel). The

clone was isolated and sequenced and identified as an isoform of the hnRNPDO

gene. Inclusion of a reference DNA for a known binding protein in each sample,

which produces a relatively constant intensity at the appropriate gelshift position

(lower gel), shows that the higher intensity at the hnRNPDO gelshift position in the upper gel is not a false-positive clone identification resulting from a higher protein

concentration than in surrounding samples. (B) The effect of antibody specific to

the hnRNPDO gene product in retarding migration of the hnRNPDO protein-DNA complexes is demonstrated in a supershift assay.

DETAILED DESCRIPTION OF THE INVENTION

Recent advancements in gene discovery emphasize the importance of gene identification. However, for effective drug discovery and development, what

matters is only the ultimate functional phenotype resulting from gene expression.

The present invention emphasizes gene function as the measure of biological

relevance. The present inventor has recognized the importance of both genetic and epigenetic elements in the ultimate gene function and has provided technology to identify aspects of both of these elements, by developing a concept which the inventor calls the "Functionomics" paradigm. This term refers to the identification of a gene, a group of genes, an epigenetic factor, a group of epigenetic factors or

any combination of the above, by detection of functional changes. In contrast, the

previously developed Genomics and Proteonomics paradigms identify changes in mRNA and protein levels respectively. Another previously described paradigm, "Functional Genomics", is a gene discovery paradigm. The purpose of the

Functional Genomics paradigm is to identify a gene by a functional assay.

The Functionomics paradigm of the present invention further develops the

Functional Genomics concept. The central hypothesis of the Functionomics

paradigm is that genetic or epigenetic function is the ultimate measure of

phenotype. Therefore, the aim of the Functionomics paradigm is identification of

both the epigenic and genetic denominators of biological function. The

Functionomics aspect of the present invention provides the ability to assess

changes in RNA and DNA binding functions of a given protein and the ability to

identify both genetic and epigenetic factors underlying those functional changes.

The present Functionomics paradigm is capable of addressing the genotype- phenotype paradox. For example, 31 ,000 human genes have been identified by the

Human Genome Consortium, close to the figure of 19,000 of a nematode worm.

However, the phenotypic and functional differences of the compared species are

significant. The reason for these differences is that the product of a single gene is

diversified by epigenetic factors. First, mRNA editing and splicing can lead to

diversification. RNA-binding proteins are essential in this process. Second,

post-translational modification is a powerful mechanism to achieve functional diversification. In living cells, proteins and exogenous polypeptide drugs are subjected to phosphorylation, glycosylation, cleavage, folding, degradation, protein-protein interaction and ligand interactions that determine biological activity, half-life and intracellular transport. Epigenetic processing is sharply different among various cell types and among various species. Therefore, the

functional properties of a gene product can be assessed objectively only if the gene

product is subjected to proper epigenetic modifications.

The present invention operates by measuring RNA and DNA binding

functions of gene products. These functions are central to RNA editing and transport, transcriptional regulation and viral life cycles. Moreover, RNA and

DNA binding functions are determined by both genetic and epigenetic elements specific to a given cell type and are fundamental functions. The invention provides

technology for directly assessing nucleic acid binding activity of a gene product,

thereby advantageously avoiding any indirect detection approaches.

Applications of the present invention technology may be considered either

"divisional" or "combinatorial", depending on whether a library of clones to be

examined are divided into separately tested samples or are combined from individual clones or less complex clone mixtures. Divisional applications are

suitable for selection of an unknown gene. Combinatorial applications are designed for selection of both clones and epigenetic factors.

In divisional applications the only initial requirements to perform a clone selection are a clone library and an RNA or DNA sequence that serves as a specific

binding site for a desired protein. To identify a DNA sequence coding for a binding protein, a DNA library is cloned in an expression format (for instance, in a plasmid or viral vector). The library is divided into pools of small numbers of

clones and desired vertebrate cells are transformed or transfected with each pool separately to express each product of a recombinant gene in the post-translationally

modified form(s) characteristic of the desired host cells, according to principles well known in the art. After expression of the divided clones, the cells are lysed, protein extracts are tested by RNA or DNA binding assay, and clone selection is further performed as described below.

In combinatorial applications of the present invention, a gene of a known

RNA or DNA binding protein, in an expression-ready format, combined with

either divided pools of an expression library, pre-selected expression clones of

post-translational modification enzymes, polypeptide drug variants, and/or with expression clones of suspected dimerization partners. Alternatively, the

transformed expression host cells are simply treated by ligand variants. Desired

host cells are co-transformed or co-transfected by any of the above combinations of

clones, and the protein products are co-expressed. Then an RNA or DNA-binding

assay is performed and analyzed by high throughput electrophoretic mobility shift assay ("EMSA"), or by column chromatography as described below.

Combinatorial applications of the invention include a combinatorial matrix

arrangement and co-expression of functionally linked genes which are purposely

organized in functional groups to allow high throughput, simultaneous assay of proteins of already isolated genes, or of native or artificial ligands, in an organized

fashion. Moreover, the elements of a matrix may correspond to elements of a

cDNA microarray or to a protein chip manifesting the functional assay format of the array technology. This "Functionom" matrix arrangement provides significant capacity for representation of spatial and temporal factors when it is applied to different cell types or organotypic cell cultures. This feature of the combinatorial application allows identification of different sets of modification enzymes at

different stages of development or in different cell types, thereby permitting

identification of the particular factor(s) that lead to activation or inhibition of a particular DNA or RNA binding activity. Combinatorial applications are

compatible with both cDNA microarray and Proteonomics technologies and can provide fundamental evidence for the functional activity of genes or proteins

detected by microarray technologies.

More particularly, the present inventor has developed technology for direct

clone screening and selection in a vertebrate host for genes encoding nucleic acid

binding polypeptides or polypeptide drugs that act on either DNA or RNA binding

proteins. This invention permits rapid and efficient identification of such a clone

via the polypeptide product which is expressed in vertebrate cells and therefore

exhibits vertebrate post-translational modifications.

Selected embodiments of the present invention originally were developed

for cloning the gene for a particular gene repressor protein that binds to a specific

target sequence and thereby represses an associated gene during brain development. Proper phosphorylation of the repressor is necessary for specific

binding to the target DNA sequence. Previously, this cloning was attempted using

a phage display library, a phage expression library and the so-called "Yeast One-

Hybrid System" without success. Using the present invention, the cloning was

successfully and efficiently completed. The invention has therefore been shown to provide efficient cloning and screening of vertebrate genes for nucleic acid binding proteins, through expressing the gene product in its natural environment that

provides the proper post-translational modifications whereby the protein is fully active. This feature is essential for precise functional characterization of a

vertebrate nucleic acid binding protein and for identification of its coding

sequence.

One main feature that the invention provides, therefore, is a complete clone screening and selection system for genes of DNA and RNA binding proteins expressed in vertebrate cells. In one aspect, the invention provides means for

optimizing vertebrate cell transformation or transfection conditions, by detecting or

measuring the nucleic acid binding function of an expressed control protein. In another aspect, the invention provides a high throughput electrophoretic mobility

shift assay (EMSA) for detection of complexes between a nucleic acid and a

nucleic acid binding protein, using a novel application of layer agarose gels for

EMSA. This gel system combines a high vacuum and high temperature blotting

technique for agarose gel desiccation with the use of high efficiency transfer matrix (preferably, quaternary aminated nylon membranes) for blotting nucleic

acid-protein complexes from agarose gels. Another aspect of the invention

provides improvements in overcoming sample-to-sample variations in

signal-to-noise ratio during clone screening, by introducing a reference probe to

detect a known nucleic acid binding protein that is expressed in the vertebrate host cells, thereby providing a functional indication of the total protein level in each cell

lysate sample as a basis for normalization of those levels in comparing different

samples. The invention also provides high throughput chromatography screening methodology for nucleic acid binding proteins, permitting automated operation of the screening system. The general vertebrate cloning and screening process for genes encoding

nucleic acid binding proteins includes steps of library division and protein expression, followed by a nucleic acid binding assay as a clone screening and selection step, where binding is analyzed by either EMSA or a chromatography

assay, preferably in a high throughput assay format.

1. Library division.

The genetic library for screening may be any DNA library, for example, a random permutation library, subtractive library or any desired homogenous or mixed DNA sequences. Preferably, the library is a primary library, to provide the

most accurate representation of the source genetic material.

The library is constructed in a vector, preferably a shuttle vector comprising

a strong promoter/enhancer construct that controls expression of the cloned DNA

sequence in the vertebrate cell or cell line desired to be the host for expression and

post-translational modification. Advantageously, the promoter is a strong

promoter. Preferably, the promoter and other elements of the expression system

will produce recombinant proteins at a minimum level of at least about 1%, more preferably, about 10% of the total protein content. The vector may contain a

replication origin that allows replication of the vector in vertebrate cells, as well as preferably another replication origin for amplification of the DNA in a bacterial

host, and one or more selective marker genes, such as an antibiotic resistance gene.

Alternatively, the vector can be a phage/viral vector system and the divided library

amplification may be performed in non-bacterial cells or in vitro. For amplification in bacteria, cells are transformed with the library-vector construct. The transformation efficiency, extent of representation and average insert length

may be determined to assess library quality.

Preferably, the vector DNA representing the primary library that is used to

transform vertebrate cells is produced under conditions that prevent methylation of the vector DNA. Numerous vertebrate genes are subjected to gene silencing by the

methylation of GC-nucleotide rich sequences. Therefore, it is preferable to

transform or transfect vertebrate cells with a non-methylated form of the vector

DNA. For example, it is advantageous to grow the library vector in dcm- and dam- bacterial strains.

The number of clones per polypeptide test sample is a critical parameter of

the direct vertebrate screening method of the invention. For instance, a library

aliquot of about 500,000 cfu (colony formation unit), may be divided into 1000

wells with about 500 cfu/well clone representation. Greater library division of such an aliquot (into more cultures) is preferred, such as 2,500, 5,000 or 10,000

wells, that would result in 200, 100 or 50 clones/well representations respectively.

For libraries of better quality (in terms of representativeness and insert length, for

instance), the initial library size may be decreased below 500,000. In addition, if

robotic processing is available, then screening of larger libraries, over 500,000 clones, may be readily accomplished using methods of the present invention.

Accurate estimation of the nucleic acid concentration is critical for efficient

transformation and screening of vertebrate expression host cells. For DNA vectors,

it is difficult to determine vector DNA concentrations with the required accuracy since common techniques for measuring bacterial DNA content also detect contaminating RNA which can cause 50-80% inaccuracy. During the transfection step, a 50% error in DNA concentration can result in an order of magnitude drop in protein expression level. Therefore, the samples may be further treated (e.g., with

RNase) and/or further purified to remove contaminating RNA.

Alternatively, in another aspect the invention provides a method for determination of optimal nucleic acid concentration for cell transformation or

transfection, by empirically testing a control combination of nucleic acid binding

protein and "tester" nucleic acid sequence that binds to the control protein. The

gene encoding the control binding protein, for instance a protein that specifically binds to a particular tester DNA sequence, is cloned into the same vector carrying

the library clones to be screened. The vector DNA carrying the control gene is mixed with an excess of the library-vector DNA, for instance, at a 1 :500 molar

ratio, respectively.

The mixture of control and library vector DNA is amplified, for instance, in

a bacterial host. Preferably, multiple (for example, 1-10) control cultures are

inoculated with about 500 cfu/culture of the control transformant mixture. The

transformant cells are incubated under conditions that select for transformants, for example, antibiotic selection conditions. Vector DNA is isolated from the bacterial

cell cultures and is purified by conventional methods, for instance, gel filtration

chromatography, or any purification technique that eliminates small molecule and

toxic protein contaminants.

After transfection of multiple cell cultures (as described below) with a

range of concentrations of the test DNA mixture, transformed cells are incubated to

allow protein expression. Care should be taken to avoid incubating a large number of cell cultures concentrated in a small area as such crowding produces locally high concentrations of toxic gases that slow cell growth, alter cell metabolism and lower

the expression rate of a recombinant protein. Thus, the expression level of proteins is improved by either more distant relative positioning of cell culture plates or by moderate ventilation of the cell cultures during the incubation time. After incubation to allow library expression, the transformed vertebrate

cells are resuspended in lysis buffer to test for DNA binding. For example, sample

mixtures of control DNA:library vector DNA with ratios bracketing 1:500 are

carried through the amplification and purification procedure for use as test DNA mixtures in optimization of transfection of vertebrate host cells which library

division produces about 500 clones/culture.

Conventional lysis buffers may be used, but the KCl concentration of the

buffer is preferably adjusted for every cell type tested, since coding regions frequently contain signal sequences that guide a recombinant protein into the cell

nucleus and, therefore, conditions that lyse the nucleus must be used. In general, 450-600 mM of KCl is sufficient to release the nuclear content of the cell.

However, it is preferablef to verify that the nuclei of the selected host cells are

properly lysed to make the binding proteins accessible for the binding assay, for

instance, by visual inspection or biochemical assay (e.g., for DNA).

The cognate tester DNA for the control binding protein is mixed with

protein lysates of the control DNA-transformed cells and tested for DNA binding,

for instance, in an Electrophoretic Mobility Shift Assay (EMSA), such as the EMSA described below. The control vector DNA concentration that results in the greatest amount of protein-DNA complex (for instance, the highest band intensity in EMSA) is used as a concentration standard during a fluorometric DNA concentration determination for all separately amplified portions of the library.

For instance, each of the 1000 aliquots of purified DNA of the library described above is diluted with the highest possible accuracy to the concentration of DNA

recorded for the control DNA dilution that gave optimum detection of the control DNA binding protein.

2. Protein expression.

Any vertebrate cell which can be genetically transformed or transfected to

express a desired library is suitable for use in the present invention. For optimal

post-translational modification of the expressed protein, any kind of vertebrate cells or cell lines may be used such as stem cells, glioma cell lines, epithelial cell

lines, primary cell cultures, organotypic cultures, provided that the expression host

provides a high yield of the recombinant protein relative to total cell protein

content. For delivery of the library clones, any technology may be used that provides efficient transformation and expression, including transformation,

transfection or, if a viral delivery system is used, infection. Preferably, the

combination delivery system and expression host provides an expression level of a

control protein in the total cellular protein of at least about 1%, more preferably at

least about 3% and most preferably, at least about 10%.

For example, to test the direct vertebrate screening method of the invention, human kidney epithelial cells were plated in 24 well plates at a density of about

25% confluence and were incubated under conditions that produced about 50%

confluence within 24 hr. Then, 1000 cell culture wells were transfected with the 1000 purified library vector DNA dilutions described above, adjusted for optimal DNA concentration by reference to the optimal control DNA concentration for screening. Following transformation or transfection, the cells go through a short

recovery period (typically 3-5 hrs), and then recombinant proteins are expressed, for instance, during a 48 hr incubation period. Culture conditions were selected such that by the end of the incubation for protein expression the cells reach a density of about 200,000 cells per culture well. Then the cells were resuspended in

lysis buffer designed to keep proteins intact, support the subsequent DNA binding

reaction and not interfere with the EMSA, according to principles well known in

the art. A typical lysis buffer composition to meet these requirements is 20 mM HEPES (pH 7.9), 450 mM KCl, 1.5 mM MgCl₂, 0.2 mM EDTA, 0.1% IGEPAL,

25% glycerol, 0.5 mM AEBSF, 10 mg/L Leupeptin, and 10 mg/L Aprotinin.

However, the required concentrations of lysis buffer components, especially the KCl and non-ionic detergent concentrations, may be optimized empirically for a

given cell type prior to the actual experiment by microscopically monitoring the lysis process.

To select conditions that provide sufficient DNA binding protein for

detection, the following exemplary calculations apply, which can be adapted to

cells of different sizes and varying expression levels, as needed. For example, the protein weight in 200,000 cells (in one well, under the exemplary conditions above), is 100 microgram. The lysis buffer volume is 50 microliter. Therefore, the

protein concentration is 2 microgram/microliter. For the binding assay described

below, the maximum applied volume ratio is 1/10, due to the necessary salt and

detergent dilution, and the binding reaction volume is 10 microliter. Therefore, by the addition of 1 microliter of cell lysate into the binding mixture, the added total protein amount is 2 microgram. If the transfection efficiency is 50% and the representation of a positive clone is 1 :500 (according to the clone input ratio), then 200 cells are expressing the positive clone (50% of 200,000/500=200). However,

the 10% recombinant protein expression level may be only a theoretical optimum. In practice this level may be lower, for instance, only about 1 % recombinant

protein, because living cells may activate pathways to degrade and eliminate excessively represented proteins.

In summary, in the above example, the positive clone is assumed to be

expressed in transformed 200 cells at a level of 1% of the total protein content of

those 200 cells. Therefore, the representation of the desired DNA binding protein

is 1 :100,000 relative to the total protein mass. The applied total protein mass in each binding reaction is 2 micrograms. One-hundred thousandth fraction of 2

microgram of protein is 20 picogram, which represents the amount of DNA- binding protein in the binding reaction mixture. Assuming that this protein has a

molecular weight of 100 kDa, 20 picogram of the protein corresponds to a

molecular number of 200 attomol (200-quintillionth of Avogadro's number). The molecular number of the tester DNA is 10 femtomol; therefore, using conventional

radiolabeling and detection technology for DNA binding protein assays, a signal

would be detectable only with DNA that is labeled with a specific activity of at

least about 5000 cpm/femtomol, where the quantitative binding of 200 attomol typically results in a detectable about 1000 cpm signal.

However, within the 1 : 100,000 binding protein/background protein ratio assumed here, the abundance of non-specific binding proteins is likely to be high.

Thereby, the signal-to-noise ratio is likely to be decreased, and the specific signal is likely to drop below 200 attomol of bound DNA. In addition, abundant non- specific binding proteins may distort the gel migration characteristics of the

specific complex and thereby cause a gel migration anomaly.

In summary, if a library that represents 500,000 individual clones is divided

into 1000 culture wells under the above typical conditions, then a nucleic acid binding polypeptide can be detected only by gel electrophoretic conditions that are highly sensitive by conventional standards. If a library of 500,000 clones is divided into 2,500, 5,000 or 10,000 wells, then binding protein detection may not

require sophisticated EMSA conditions or, for a primary screening, may not require gel electrophoresis at all. However, for any screening using a division of a

500,000 clone library into over 1,000 samples, an automated process using robotic

manipulators would be preferred.

3. Clone Screening by High Throughput Electrophoretic Mobility Shift Assay (EMSA).

For high throughput testing according to the invention screening method,

the generally used polyacrylamide gel electrophoresis system is inadequate.

Migration in this gel system is described by the "reptation with stretch" model which predicts increased migrational stress, presumably due to the pore size of the

gel matrix being similar to the size of migrating DNA-protein complexes. This

stress can lead to decreased stability of the complex during the gel migration.

Considering the modified Ogston sieving model, the application of a large pore gel

matrix is necessary. Agarose gels have a large pore matrix, but the resolution of conventional agarose gels is poor and transfer of the DNA-protein complex from

the gel to a membrane for analysis of results may take several hours. For instance,

U. S. Patent No. 4,983,268 issued to Kirkpatrick. et al. discloses high gel strength low electroendosmosis agarose preparations suitable for use in the present

invention.

In addition to agarose, certain other polysaccharide are suitable for use in gel electrophoresis in the practice of the present invention. As used herein, the

term "polysaccharide" refers to a native or derivatized thermoreversible and/or

pH-reversible hydrogel-forming polysaccharide, preferably selected from among: agar, agarose, curdlan, gellan, konjac, pectin, pullulan, and to a lesser degree

alginate and carrageenan, and the like; as well as thermoreversible or pH-reversible

hydrogel combinations of any of the foregoing with either another polysaccharide

or a non-polysaccharide polymer which associates firmly with the gel-forming polymers so that it does not significantly dissociate under electrophoretic

conditions. (Linear polyacrylamide does not associate firmly with the gel-forming

polymers of this invention.) Methods for making and using such polysaccharide

hydrogels for electrophoresis are described, for instance, in U. S. Patent No. 5,143,646.

In another aspect, therefore, the present invention provides a method of separating nucleic acids, and particularly a method for detection of protein-nucleic

acid complexes, using thin-layer polysaccharide hydrogels, preferably agarose gels,

by combining the application of high temperature and high vacuum for gel drying

with the use of highly efficient binding substrate (for example, quaternary amine derivatized membrane) for nucleic acid transfer, for use in a high throughput

EMSA method that facilitates application of robotic technology.

To overcome the problems with conventional agarose and other polysaccharide hydogels, the gel is prepared in a thin-layer format (for example, less than or equal to about 3 mm, preferably about 2 to 2.5 mm), which improves

the resolution of the gels. Preferably, and more importantly for efficient detection, the transfer after electrophoresis of the protein-nucleic acid complex from the gel matrix onto a membrane surface is carried out by the simultaneous application of high vacuum and high temperature. At high temperature, the agarose gel material normally would collapse. However, if a thin layer agarose is prepared with, for

instance, a three square centimeter/milliliter surface/volume ratio, then by applying

a vacuum (for instance, less than about 20 torr), the resulting high evaporation rate keeps the agarose at low temperature despite use of a relatively high environmental

temperature for blotting (e.g., preferably about 80°C). A high blotting temperature

is necessary because, if high vacuum is applied without high temperature, then the

gel structure suffers lateral distortion. Also, high temperature facilitates a higher evaporation rate that keeps the gel temperature low and thereby prevents the gel

matrix from deformation and ensures a rapid and accurate transfer. This blotting process reduces the blotting time substantially, for instance to about 20 minutes,

allowing the completion of a gel running and blotting procedure within about

90 minutes

In addition, the quality of the blotting is ensured by the use of an efficient

DNA binding substrate for the transfer membrane, preferably a quaternary amine

derivatized nylon membrane, that retains quantitatively the blotted nucleic acids

and complexes. By reducing the loaded sample volume, for instance to about 5 microliter, and by using wedge-shaped loading wells, the resolution of the gel is also markedly improved. As a general practice, maintaining the gel temperature at below ambient temperature (preferably about 8°C) throughout the electrophoresis also improves complex detection, by increasing their stability during migration.

The invention also provides means to reduce the effects of variability in the

signal to noise ratio that results from unequal protein concentrations among individual cell lysates, by introducing two radioactively labeled DNA probes that are tested in parallel for each cell lysate binding assay. One of these DNA probes is called a "reference" and the other one is called a "tester". The reference DNA

sequence is different from the sequence of the tester and binds a known DNA

binding protein that is expressed in the vertebrate cloning host. The binding level

of the reference DNA is proportional to the total protein concentration in a sample

and therefore provides a baseline binding intensity to which the binding of the

tester sequence can be compared and normalized. Thus, the ratio of free/bound

reference probe serves as a quantitative normalization value to assess the

significance of free/bound ratio differences of the tester in different lysate samples.

The DNA binding reaction mixture typically contains various inhibitors of non-specific binding, such as about 1 mg/ml of non-specific competitor DNA and

about 0.1% by weight of carrier protein (e.g., Serum Albumin), as well as

radioactively labeled tester or reference DNA (preferably about 10 femtomol) and

preferably about 2 microgram of protein lysate. It is preferable to keep the bound/free DNA ratio at or below about 50% during the reaction. Therefore, prior

to large scale analyses (for instance, 1000 assays with the reference and 1000

assays with the tester DNA, in the example above), optimal protein concentration

of the reaction mixture is assessed by titration against the reference DNA. The above EMSA screening procedure can be performed under the

described conditions by manual processing by one person at a rate of 300 assays/day. Robotic technology can also be applied for this step. 4. Screening by high throughput chromatography. The invention also provides methods for screening samples for DNA

binding proteins by high throughput chromatography, for instance, either instead of, or as a pre-screening step for, the above high throughput EMSA assay. Thus, if

a culture well from the library division contains a relatively small number of clones

(e.g., only about 100 cfu), then chromatography screening procedures can be

introduced. For chromatographic screening, the binding reaction should be

adjusted to provide approximately 50% saturation of the reference DNA, to ensure the optimum bound/free probe ratio of about 50%. Then the binding mixture is

passed through a chromatography column that separates DNA-protein complexes

from free probe, under conditions well known in the art. The amount of the free

probe and/or the bound fraction is assessed by either radioactive or non-radioactive

technologies. A decreased concentration of free probe is an indication of a positive clone (expressing a DNA binding protein) because the DNA binding protein of the

positive clone will increase DNA retention in the protein-bound fraction.

A titration reaction was performed using the above model system

conditions, to assess what representation of a clone of a DNA-binding protein in a

primary library distribution well would provide sufficient DNA retention to detect a positive signal by measuring only the bound fraction and/or the free probe. The results suggest that about a 1 :100 representation of a DNA-binding clone per culture is sufficient to detect either an increase in the protein-bound DNA or a

decrease in free probe, allowing screening by chromatography.

Application of reference and tester binding reactions is also recommended for chromatography screening. For instance, the chromatography procedure can be performed in a 96 well format with a non-radioactive detection method and can be operated automatically. Clone candidate wells selected by chromatography are

preferably further analyzed by EMSA, to confirm the presence and amount of DNA binding protein.

5. Data analysis and single clone isolation. Both conventional autoradiography and known phospho-imaging recording

methods are suitable for recording of EMSA results. The results are preferably

evaluated independently by two investigators and then compared. If by both

detection techniques a positive clone-containing well is identified, then the clone candidate may be further tested, for instance, in three independent EMSA tests

comparing tester and reference bound/free ratios.

The present invention methods have been successfully applied to detect a

gene expressed in vertebrate cells which encodes a protein which binds to a tester

DNA that is similar to sequences known to be recognized by a particular binding protein. This test of the technology, conducted under the above conditions,

resulted in five culture wells (out of 1000 library portions) that contained positive

clone candidates. One out of the five clone candidates was further processed. The representation of a single positive clone in the total clones in the well was 1 :500; therefore, the resulting clone pool required further division. Bacterial cells were transformed with plasmid DNA mixture from the pool

of 500 different clones and were divided among 100 wells with a 5 clone/well

representation. The procedures described under sections 1, 2, 3 and 5, above, were repeated, and one positive well containing at least one clone expressing the DNA binding protein was identified. The EMSA signal was clearly detectable and the

DNA-protein complex was clearly distinct from the background.

Bacterial cultures were again transformed with the plasmid mixture of the

positive culture well, and single colonies representing single clones were isolated.

Plasmid DNA was purified from the bacterial cells, and the plasmids were analyzed for insert size by PCR and by digestion with restriction endonucleases.

Kidney epithelial cells were transfected with the individual plasmids.

Finally, a single positive clone was identified by EMSA. The DNA

sequence of the positive clone was determined and a homology search was performed against deposited sequences in the GenBank. A single matching gene

was identified, with a 99% homology. The identified gene encodes a protein with

known affinity to nucleic acid sequences strikingly similar to the tester DNA

sequence.

The test operation of the invention was-performed by EMSA. However, a 96 well format chromatography screening for nucleic acid binding polypeptides

also has been assessed and found suitable for high throughput screening.

The invention technology also can be applied for cloning of RNA-binding

proteins, by using RNA sequences in the binding tests, under conditions well known in the art. The invention also is suitable for screening clones to isolate polypeptide

drugs that are effective against a DNA or RNA binding protein, for instance, a regulator protein of a virus or a cancer cell. Moreover, the effective polypeptide can be selected in the cell that is the natural target for the viral infection or can be

selected in the cancer cell of interest. The cloning technology can be applied for a viral delivery system, when the library is infected into the cells by a carrier virus such as HIV or adenovirus. Then a regulator of the viral replication or viral protein

expression (for example the TAR-binding protein in HIV) can be monitored in the

divided wells by RNA or DNA binding assay. The positive well where the viral regulator protein binding activity is decreased is further divided, and the clone

responsible for the inhibition is identified from the library.

The invention also can be applied to screening of receptor blocker drugs

that interfere with the DNA binding properties of a particular receptor. For

instance, the invention can be used for cloning of protein kinases, protein

glycosylases, proteolytic enzymes and enzymes assisting proper folding. If an

enzyme activates or inhibits a target protein with respect to its DNA or RNA

binding activity, then the gene of the target protein can be co-transfected with a

library. After co-expression of the divided library with the target protein, a binding assay will identify any positive well where the DNA or RNA binding activity of the

target protein is impaired. With further dilution the effective clone can be isolated and then characterized.

In summary, the invention can be used for cloning DNA and RNA binding proteins, to clone genes of anti-viral drugs, anti-cancer drugs, receptor blockers, kinases, glycosylases, and peptidases, so long as the function of the desired gene

can be detected by monitoring a protein with DNA or RNA binding activity.

Advantages of the invention methods over presently known systems for cloning and screening of genes for nucleic acid binding proteins include the fact that the polypeptides are expressed in vertebrate cells or cell lines which provide

proper post-translational processing, such as phosphorylation, glycosylation,

cleavage and folding, appropriate for vertebrate cells. The nucleotide binding

characteristics of these polypeptides are therefore optimal. In contrast, conventional phage expression library screening technology operates in bacterial

cells, whereas another known technology, the so-called "Yeast One-Hybrid

System", operates in yeast cells, thereby providing post-translational modifications

representative only of a bacterial or yeast cell, respectively. A yeast system is less than ideal for direct screening of clones for DNA-binding receptors, activators, suppressors and transcription factors, for example, from the human brain.

Accordingly, efforts to introduce genes for mammalian post-translational

modifications (for instance, protein kinases) into yeast cells have been made, to

compensate for this serious disadvantage, but reproduction in yeast of the total vertebrate environment provided by the present invention would presumably

require introduction of hundreds or thousands of kinases, glycosylases, proteolytic

enzymes and enzymes that assist in protein folding. Although mammalian

expression systems, such as the "Mammalian One-Hybrid System", have recently

been developed, these technologies rely on uncertain "reporter" technologies for clone detection and therefore identify a significant number of false positive clones. The traditional way to identity a DNA binding protein is DNA-affinity chromatography, requiring grams of cell or tissue samples that are impossible to

access in many cases, for example, from human specimens. Once the protein is purified, there is still the challenge of clone isolation ahead. In contrast, the

present invention can operate in human cell lines of choice to clone the desired gene. Therefore, another advantage is the flexibility of the invention system. The procedure can use any kind of library such as subtractive library, cDNA library or

combinatorial random peptide library. The only initial information needed to start

a clone selection is a DNA or an RNA sequence for which identification of specific

binding proteins is desired. Alternatively, the system can be used to identify clones expressing any polypeptide that activates or inhibits a DNA or RNA binding

function of another protein.

Another advantage of the invention is that a broad range of vectors can be

chosen or constructed, and the approach is not limited to any particular components

with inherent limitations. Further, the invention methods can be completely

performed by automated systems, using conventional robotic stations for DNA manipulations, such as those used in current human genome analysis projects.

Thus, the invention technology is remarkably cost and time effective, and the pay

off is multiplied because, once (and only once) a cell lysate collection of divided library pools (for instance, 1000 portions) is prepared, then those portions can be

tested for many different DNA binding proteins. Also, "lysate libraries" may be generated from multiple tissue samples at different stages of development, for

instance, from embryonic brain, newborn brain and adult brain, and used

comparatively to select the most effective polypeptide drugs for each patient age range. The same approach can be extended to different tissue types or brain areas,

thereby permitting comparative drug selection to be performed in a spacio-

temporal fashion.

The invention provides efficient clone selection in vertebrate cells expressing both DNA and RNA binding proteins, as well as proteins that inhibit or

activate nucleic acid binding polypeptides. Thus, the invention can be used for

screening for polypeptide drugs that inhibit or facilitate the nucleic acid binding

properties of a given binding factor. For example, to identify a peptide drug that blocks the binding of a DNA-binding receptor (Vitamin D Receptor) in a brain

tumor cell, one can transform a random library into that particular cell line and

perform clone selection to identify clones of polypeptide that blocks the VDR

binding to its cognate element.

Qualitative and quantitative assessments of relative binding inhibition of

multiple positive clones can also be assessed using the invention methods. The

effectiveness of a polypeptide drug, for instance, an anti-tumor drug, is assessed

"on target". Moreover, the clone is on hand right away. If the drug is now

introduced for therapeutic purposes into the same tumor cells as used for screening, the post-translational modification of the cells will support the effectiveness of the

drug. Also, only those drugs will be selected which escape the proteolytic

inactivation sufficiently to provide an adequate half-life in the host cell. Therefore,

the cloning system of the invention gives an indication of complete drug effectiveness in a given cell type. This is a unique feature of the invention.

Another problem addressed by the invention is the cloning of RNA-binding proteins by direct detection. Hybrid systems recently applied for cloning RNA binding proteins in mammalian cells rely on reporter detection, by definition indicating that detection is "indirect". The acquired information includes both signal and meaningless noise from effects related only to the reporter system.

Moreover, indirect detection technologies rely on artificial protein fusion constructs. Disadvantageous fusion of two proteins alone may result in a large

number of false positive clones and may hinder detection of real positives. In contrast, the present technology does not use fusion proteins or reporter systems,

but directly detects and follows the visible DNA or RNA-protein complex

throughout the entire procedure, focusing solely on the basic function, the protein

nucleic acid interaction. One ultimate benefit of the approach presented here is

therefore the high reliability of positive signals.

The discovery and availability of the majority of genes in the human

genome presents a different application of the invention technology for clone

selection. This application uses combinatorial co-expression of previously

identified genes. In this combinatorial application, genes in expression-ready format are arranged in a matrix according to functional linkage groups. These

groups are then combined with genes coding for RNA or DNA binding proteins to

form combinatorial expression pools. Vertebrate cells are transfected separately with each combinatorial expression pool, and the proteins are co-expressed. The

tested binding proteins are assayed by the invention technology. A partner element

in an expression matrix may be represented on a microarray. Moreover, an

expression matrix element may represent the combination of multiple functionally linked partners in connection with an assayed DNA or RNA binding protein. All of these elements may be represented on a cDNA array. For example, assume an unknown member of a protein family forms a heterodimer with a known protein in

order to bind an RNA element. Then, vectors coding for individual members of

this protein family are co-expressed with the vector expressing the known nucleic acid binding protein. This combination may be achieved with one complete combination of all involved expression vectors or by individual combinations of

the known binding protein and one or a few members of the protein family, or may

be achieved with a partially complete combination to assess functional

complementation.

A benefit of the sensitive detection technology is that it allows high clone representation for detection of protein-nucleic acid binding (that is, high library

pool complexity). Therefore, numerous clones may be combined in one expression

matrix element. This number is lower however if co-expression of multiple proteins in the same cell is required for the detection of a change in nucleic acid

binding. The combinatorial matrix arrangement will reveal the coding sequence of

an unknown interaction partner.

Vectors representing cDNAs of complete signaling pathways may also be

combined and co-expressed in expression matrix elements to identify interacting members. In this arrangement the invention technology identifies a functionally important epigenetic signal that has an effect on DNA or RNA binding properties

of a protein. Furthermore, in replicated forms of a combinatorial matrix the

transfected cells may be stimulated by different epigenetic ligands, and the signaling pathway that leads to DNA or RNA binding of a protein can be identified by simply monitoring the RNA or DNA binding activity of that protein. In addition, this same matrix may be produced in multiple replicated formats, each representing the functional characteristics of different cell or tissue types. All the combinatorial arrangements may be utilized to assay for functional verification of

down or up regulated gene expression detected by a cDNA microarray or by proteomics technology.

In another aspect the invention provides a method for detection of protein-

DNA complexes using antibodies for a specific DNA binding protein. Thus, the thin-layer agarose gel electrophoresis method of the invention is highly suitable for

small or large-scale analysis of antibody-protein-DNA complexes in A so-called

"supershift assay." The pore-size of traditionally applied polyacrylamide gels does

not allow the larger sized particles of antibody-protein-DNA complexes to migrate into the matrix and consequently fails to resolve these complexes. Instead, these complexes appear as a large unresolved aggregate at the loading well of the

polyacrylamide gel. Therefore, the interpretation of specific antibody recognition

of a DNA-protein complex is ambiguous using a polyacrylamide gel. By contrast,

the larger pore size of agarose thin layers allow a large antibody-protein-DNA

complex to migrate into the gel during electrophoresis and resolves the migrating

complex without ambiguity. Use of the gel shift in an array application for testing

large numbers of antibodies for specific protein recognition provides a high

throughput method for the proteomics paradigm. For purposes of illustrating preferred embodiments of the present invention,

the following, non-limiting examples are included. These results demonstrate the

feasibility of DNA-binding protein gene detection using the invention methods. EXAMPLE 1 : Detection of a desired clone in a library of 500.000 clones

Overview. FIG. 1 shows a flowchart of a preferred embodiment of the

invention method for identification of a DNA clone that expresses a nucleic acid binding protein for a selected nucleotide sequence ("tester probe"), using electrophoresis in thin layer agarose to perform a "gel shift" assay of the invention. Typically, the first stage, "selection of a positive pool" containing a

desired clone, begins with preparing or otherwise obtaining a cDNA expression library containing about 500,000 clones (clone-forming-units = "cfu"), and

separating the library into about 1000 subfractions or "pools" (in practice, the use

often 96-well microculture plates allows convenient processing of 960 separate pools). Each pool is separately amplified in a microculture of a conventional

microbial host, typically E. coli. Plasmids are extracted and purified from each microculture, and each plasmid preparation is separately introduced (e.g.,

transfected) into a microculture of an appropriate vertebrate or mammalian

expression host. Lysates of transfected expression hosts are then tested for DNA

binding proteins by agarose thin-layer electrophoresis, according to the invention

method, using a reference (control) probe and a tester probe. The former contains a nucleotide sequence that binds a reference protein known to be expressed in the

expression host, either endogenously or by inclusion in the expression vector used

to prepare the clone library, while the latter contains a sequence for which a clone

encoding a cognate binding protein is being sought. Positive pools containing a clone expressing the desired binding protein are identified by detection of a band of tester probe whose migration is retarded due to complex formation with a binding protein, in a sample where the band of reference probe complexed with reference protein appears at about the same level as other samples not showing tester probe complexes, thereby providing a control for consistency of protein and probe levels

across multiple samples.

After selection of positive pools, each such pool (containing about 500

clones) is further sub-divided, for instance, into about 100 subfractions (e.g., into

96 wells of a microculture plate), amplified in the microbial host and tested for

expression in separate microcultures of an appropriate vertebrate expression host. Positive subfractions or pools containing a clone expressing a desired binding

protein are again identified by complex formation with the tester probe, with or

without comparison to reference probe which is less likely to be needed for these subfractions which contain an average of only five clones each and, therefore, less

prone to some artifacts that the reference probe is designed to detect. Individual

positive clones are then subjected to clone verification, including various tests to

identify and characterize the protein forming complexes with the tester probe.

Gel Preparation. The gel matrix is preferably agarose, but also may

comprise dextrins or any other non-polyacrylamide-based gel matrix with suitably

large pores, in which a nucleic acid molecule and a protein or the complex of the

two (or a complex of the two bound to an antibody) migrate freely under

electrophoretic conditions, free of any ionic or covalent immobilization or retardation activities. The gel preferably has a thickness of about 3 mm or less,

more preferably about 2 to 2.5 mm.

After melting agarose in 1 x TBE buffer (45 mM Tris-borate and 10 mM

EDTA, pH 8.0), the gel matrix is allowed to solidify for 12-24 hrs at 22°C. Then the gel is melted again by bringing the gel to a boil, and the liquid is placed at room temperature and allow ed to cool to 65 °C. The agarose thin-layers are then

prepared by pouring the still liquid gel solution into gel casettes. After solidification at room temperature, the gel is placed at 4°C for at least 2 hr, or sealed to prevent evaporation and stored at 4°C for a maximum of four weeks.

The above repeated cycle of melting and solidification increases the firmness of the gel matrix and improves the sharpness and resolution of the migrating

DNA-protein complexes.

FIG. 2 illustrates the sensitivity of the invention method for detection of

clones expressing two different DNA binding proteins, including detection of

expression of a protein from a clone in a host cell that endogenously produces the

same protein. Expression vectors for either the Mus musculus S ATB 1 gene (Dickinson, et al.. "A tissue-specific MAR/SAR DNA-binding protein with unusual binding site recognition", Cell, 70, 631-645 (1992)) or the Rattus

norvegicus HMGI(Y) embryonic brain-specific cDNA (Dobi unpublished) were

mixed with library pools in 1 :50, 1 :150 and 1 :500 ratios and complexes of DNA

and DNA binding proteins were detected by electrophoresis as described above. As shown in the autoradiograph (panel A), both proteins were clearly detected by

the invention method even at the lowest 1 :500 ratio of desired clone to library

clones. Expression of the HMGI(Y) protein above the endogenous level in the host

kidney epithelial cell line (BCGR) is clearly detectable, even at the 1 :500 ratio of

desired to library clones. For panel (B), a clone expressing the mouse SATB1

protein was mixed into an expression library in a 1 :500,000 ratio. After dividing the library into 960 pools, proteins of each pool were expressed in kidney epithelial

cells (PEAKrapid, Edge BioSystems), and the protein pools were tested for the presence of SATB 1 DNA-binding activity according to the invention method. A band of complexes of the SATBl protein with labeled tester DNA is visible in the

analysis of plate 7, row E, lane 13. The analysis also shows that the intensity of reference probe binding in this sample does not deviate from other reference reaction intensities, indicating that the pool identified as containing the SATB 1 clone is a "true-positive" pool (that is, not an artifact due to different amounts of protein extracts or DNAs in different samples).

EXAMPLE 2: Detection and characterization of a desired clone using antibodies

This example illustrates the use of an antibody to detect and identify a particular nucleic acid binding protein, for instance, where a library is known to

contain at least one protein that binds to a tester probe but may also contain one or

more additional binding protein(s) specific for that same tester probe.

In particular, FIG. 3 demonstrates identification of a SATB 1 protein-

expressing clone added to a cDNA clone library at a ratio of 1 :500,000, using an antibody-protein-DNA complex in a "supershift" gel electrophoresis assay of the

invention. The protein pool that contained the expressed SATBl protein was

mixed with anti-SATBl antibody and the tester probe. Upon specific recognition,

an antibody-protein-DNA complex is formed that is observed as a "supershifted" band that migrates in the thin-layer agarose gel (lane 5) more slowly than the

SATBl protein-DNA complext alone ("complex 1"). Antibody that does not

recognize the SATBl protein does not produce this supershift (lane 6). As a control, a specific DNA competitor (W, "wildtype") is shown to compete for SATBl protein (complex 1). This competitor is specific for SATBl and, therefore, does not compete for endogenous HMGI(Y) (complex 2). The letter M indicates

mutant competitor DNA that does not bind to SATB 1 or HMGI(Y) proteins.

EXAMPLE 3 : Detection and characterization of a novel rat forebrain binding protein. FIG. 4 illustrates the capability of the invention methods for identifying a

clone for an unknown binding protein from an expression library of cDNA clones

from a selected tissue. The cDNA library originated from embryonic rat forebrain and, therefore, embryonic rat forebrain nuclear proteins were tested in the gel shift

and supershift assays to determine whether an antibody can specifically bind to the

protein-DNA complex of the tester DNA and rat forebrain nuclear proteins.

Complexes of tester DNA and an unknown binding protein are detected in plate 6, row G, pool 8 (panel A, upper gel). The clone was isolated and sequenced and

identified as an isoform of the hnRNPDO gene. Dimethyl-arginilation and specific

phosphorylation and splicing significantly alter the nucleic acid binding

characteristics and subcellular localization (availability) of hnRNPDO gene products. Therefore, successful expression cloning of this isoform in a non-

mammalian expression system is unlikely. Inclusion of the reference DNA probe

for a known binding protein in each sample, which produces a relatively constant

intensity at the appropriate gelshift position (lower gel), shows that the higher intensity at the hnRNPDO gelshift position in the upper gel is not a "false-positive"

identification resulting from a higher protein concentration than in surrounding samples. After identification of the coding sequence of the hnRNPDO gene, antibody specific to this protein was obtained and formation of antibody-protein- DNA complex was tested in a supershift assay (panel B). The specific antibody caused "supershifting" of protein-probe complex bands into several slower moving

bands, perhaps indicating complexes containing multiple binding protein and/or antibody molecules, while control antibody that is not specific for the binding protein does not produce any such supershift.

* * *

All publications and patents cited herein are hereby incoφorated by

reference herein in their entirety.

While the foregoing specification teaches the principles of the present invention, with examples provided for the purpose of illustration, it will be

appreciated by one skilled in the art from reading this disclosure that various

changes in form and detail can be made without departing from the concept, spirit

and scope of the invention.

Claims

WHAT IS CLAIMED IS:

1. A method for detecting a cDNA clone encoding a vertebrate nucleic acid

binding protein that specifically binds to a tester nucleotide sequence, said method comprising the steps of:

(a) obtaining a library of vertebrate cDNA clones to be tested for encoding

said vertebrate nucleic acid binding protein, wherein each of said cDNA clones is operably inserted into a vertebrate expression vector;

(b) introducing a portion of said library of vertebrate cDNA clones into a culture of vertebrate host cells under conditions such that a polypeptide encoded by

a cDNA clone operably inserted into said vertebrate expression vector is expressed;

(c) contacting said polypeptide encoded by a cDNA clone that is expressed

in step (b) with said tester nucleotide sequence under conditions such that said

nucleic acid binding protein, if present, specifically binds to said tester nucleotide sequence, thereby forming a specific protein-nucleic acid complex;

(d) electrophoretically separating said specific protein-nucleic acid complex

from protein and tester nucleotide sequence not present in said specific protein- nucleic acid complex in a thin-layer polysaccharide hydrogel under conditions such

that migration of said specific protein-nucleic acid complex is detectably retarded

compared to migration of said tester probe and to migration of said protein; and

(e) detecting said specific protein-nucleic acid complex, if any, that was

separated from said protein and said tester nucleotide sequence in step (d), by detecting retarded migration of said protein or said tester nucleotide sequence,

wherein detection of said specific protein-nucleic acid complex indicates that a cDNA clone encoding said vertebrate nucleic acid binding protein has been detected in said portion of said library of vertebrate cDNA clones.

2. The method of Claim 1, wherein said portion of said library of

vertebrate cDNA clones consists of about 500 cDNA clones that are expressed in said culture of vertebrate host cells.

3. The method of Claim 2, further comprising the steps of subdividing said portion of said library of vertebrate cDNA clones in which said cDNA clone

encoding said vertebrate nucleic acid binding protein has been detected into

subportions and separately testing each of said subportions for the presence of

cDNA clone encoding said vertebrate nucleic acid binding protein by performing steps (b) through (e).

4. The method of Claim 1, wherein further:

(f) prior to step (b), mixing a control DNA operably inserted into said

vertebrate expression vector with said library of vertebrate cDNA clones at a

known molar ratio, said control DNA encoding a control DNA binding protein that

specifically binds to a control nucleotide sequence;

(g) prior to step (b) and after step (f), amplifying said portion of said cDNA

library by replication in a host cell;

(h) providing said control nucleotide sequence during said contacting said

polypeptide encoded by a cDNA clone that is expressed in step (b) with said tester

nucleotide sequence, wherein said contacting is under conditions such that said

control nucleic acid binding protein, if present, specifically binds to said control nucleotide sequence, thereby forming a specific control protein-nucleic acid complex; (i) detecting said specific control protein-nucleic acid complex, if any, that

was separated from said control protein and said control nucleotide sequence in step (d), by detecting retarded migration of said control protein or said control nucleotide sequence;

(j) repeating steps (b)-(i) with a second portion of said library after mixing said control DNA with said library; and

(k) determining the ratio of the amount of said specific control protein- nucleic acid complex detected using proteins expressed from said portion of said

library to the amount of said specific control protein-nucleic acid complex detected

under said conditions using proteins expressed from said second portion of said

library, wherein said ratio indicates the ratio of cDNA clones in said portion to

cDNA clones in said second portion after said amplifying in step (g).

5. The method of Claim 4, wherein said vertebrate host cell used for

detecting a cDNA clone encoding said vertebrate nucleic acid binding protein

endogenously expresses said vertebrate nucleic acid binding protein or another

vertebrate nucleic acid binding protein that specifically binds to said tester

nucleotide sequence.

6. A method for identifying a polynucleotide encoding a nucleic acid

binding protein that specifically binds to a tester nucleotide sequence, said method

comprising the steps of:

(a) obtaining a protein encoded by a polynucleotide to be tested for

encoding said nucleic acid binding protein;

(b) contacting said protein with said tester nucleotide sequence under

conditions such that said nucleic acid binding protein, if present, specifically binds to said tester nucleotide sequence, thereby forming a specific protein-nucleic acid

complex; and

(c) electrophoretically separating said specific protein-nucleic acid complex from protein and tester nucleotide sequence not present in said specific protein- nucleic acid complex in a thin-layer polysaccharide hydrogel under conditions such that migration of said specific protein-nucleic acid complex is detectably retarded compared to migration of said tester probe and to migration of said protein; and

(d) detecting said specific protein-nucleic acid complex, if any, that was separated from said protein and said tester nucleotide sequence in step (c) by detecting retarded migration of said protein or said tester nucleotide sequence, wherein detection of said specific protein-nucleic acid complex identifies said polynucleotide as a polynucleotide encoding said nucleic acid binding protein.

7. A method for identifying a nucleic acid binding protein that specifically binds to a tester nucleotide sequence, said method comprising the steps of: (a) obtaining a polypeptide preparation to be tested for the presence of said nucleic acid binding protein;

(b) contacting said polypeptide preparation with said tester nucleotide sequence under conditions such that said nucleic acid binding protein, if present, specifically binds to said tester nucleotide sequence, thereby forming a specific protein-nucleic acid complex; and

(c) electrophoretically separating said specific protein-nucleic acid complex from protein and tester nucleotide sequence not present in said specific protein- nucleic acid complex in a thin-layer polysaccharide hydrogel under conditions such that migration of said specific protein-nucleic acid complex is detectably retarded

compared to migration of said tester probe and to migration of said protein; and (d) detecting said specific protein-nucleic acid complex, if any, that was separated from said protein and said tester nucleotide sequence in step (c) by detecting retarded migration of said protein or said tester nucleotide sequence,

wherein detection of said specific protein-nucleic acid complex indicates that said

nucleic acid binding protein was present in said polypeptide preparation.

8. The method of Claim 7, wherein said specific protein-nucleic acid

complex is detected by detecting retarded migration of said tester nucleotide

sequence in said specific protein-nucleic acid complex.

9. The method of Claim 8, wherein said tester nucleotide sequence

includes a labeling moiety and said specific protein-nucleic acid complex is detected by detecting retarded migration of labeling moiety.

10. The method of Claim 7, wherein said specific protein-nucleic acid

complex is detected by detecting retarded migration of said protein in said specific

protein-nucleic acid complex.

1 1. The method of Claim 10, wherein said protein in said specific protein-

nucleic acid complex is detected by an antibody that specifically binds to said

protein.

12. The method of Claim 1 1, wherein said antibody that specifically binds

to said protein in said specific protein-nucleic acid complex is bound to said

specific protein-nucleic acid complex prior to said step (c) of electrophoretically separating said specific protein-nucleic acid complex from protein and tester

nucleotide sequence.

13. The method of Claim 7, wherein said polypeptide preparation is obtained by extraction of polypeptides from a vertebrate cell.

14. The method of Claim 7, wherein said polypeptide preparation is obtained by expression of a cDNA clone library in a vertebrate expression system.

15. The method of Claim 14, wherein said vertebrate expression system is a mammalian expression system.

16. The method of Claim 14, wherein said vertebrate expression system

comprises an expression vector and a vertebrate host cell that expresses

polypeptides from cDNA clones carried in said expression vector.

17. The method of Claim 14, wherein said vertebrate expression system

comprises a vector and a cell-free coupled transcription-translation system that

expresses polypeptides from cDNA clones carried in said vector.

18. A method for separating a specific protein-nucleic acid complex from

protein and nucleic acid not present in said specific protein-nucleic acid complex, said method comprising a step of electrophoretically separating said specific

protein-nucleic acid complex from said nucleic acid and said protein in a thin-layer

polysaccharide hydrogel under conditions such that migration of said specific

protein-nucleic acid complex is detectably retarded compared to migration of said protein and to migration of said nucleic acid.

19. The method of Claim 18, wherein said thin-layer polysaccharide

hydrogel comprises an agarose gel and has a thickness of less than or equal to

about 3 mm.

20. The method of Claim 19, wherein said thin-layer polysaccharide

hydrogel has a thickness in the range of about 2.0 mm to about 2.5 mm.

21. The method of Claim 18, wherein said thin-layer polysaccharide

hydrogel comprises about 0.5 % to about 2.5 % (w/v) agarose.

22. The method of Claim 21 , wherein said thin-layer polysaccharide

hydrogel comprises about 1.0 % (w/v) agarose.

23. The method of Claim 18, wherein said nucleic acid is DNA.

24. The method of Claim 18, wherein said nucleic acid is RNA.

25. A method for identifying an agent that inhibits or enhances formation

of a specific protein-nucleic acid complex comprising a nucleic acid binding

protein that specifically binds to a tester nucleotide sequence and said tester

nucleotide sequence, said method comprising

(a) contacting said nucleic acid binding protein with said tester nucleotide

sequence in a first sample and a second sample, in the presence and absence,

respectively, of a putative agent to be tested for inhibiting or enhancing formation

of said specific protein-nucleic acid complex, under conditions such that, in the

absence of said putative agent, said nucleic acid binding protein specifically binds to said tester nucleotide sequence, thereby forming said specific protein-nucleic

acid complex; (c) electrophoretically separating said specific protein-nucleic acid complex

formed in each of said first sample and said second sample, from protein and tester

nucleotide sequence not present in said specific protein-nucleic acid complex in a thin-layer polysaccharide hydrogel under conditions such that migration of said specific protein-nucleic acid complex is detectably retarded compared to migration of said tester probe and to migration of said protein;

(d) determining the amount of said specific protein-nucleic acid complex, if

any, in each of said first sample and said second sample, that was separated from said protein and said tester nucleotide sequence in step (c) by detecting retarded migration of said protein or said tester nucleotide sequence,

wherein a decrease in the amount of said specific protein-nucleic acid complex in the presence of said putative agent in said first sample compared to the

amount of said specific protein-nucleic acid complex in the absence of said putative agent in said second sample indicates that said putative agent inhibits

formation of said specific protein-nucleic acid complex, and

an increase in the amount of said specific protein-nucleic acid complex in

the presence of said putative agent in said first sample compared to the amount of

said specific protein-nucleic acid complex in the absence of said putative agent in said second sample indicates that said putative agent enhances formation of said

specific protein-nucleic acid complex.