Multivalent single chain antibodies

CA 02117477 1997-11-03

WO 94/13806 PCT/US93/12039

CA. i1 7477

MULTIVALENT SINGLE CHAIN ANTIBODIES

The present invention relates to single chain multivalent antibodies.

Antibodies are proteins belonging to a group of immunoglobulins elicited by the immune system in response to a specific antigen or substance which the body deems foreign.

There are five classes of human antibodies, each class having the same basic structure. The basic structure of an antibody is a tetramer, or a multiple thereof, composed of two identical heterodimers each consisting of a light and a heavy chain. The light chain is composed of one variable (V) and one constant (Q domain, while a heavy chain is composed of one variable and three or more constant domains. The variabledomainsfromboththelightand heavychain, designated V_L and Vu respectively, determine the specificity of an immunoglobulin, while the constant (C) domains carry out various effector functions.

Amino acid sequence data indicate that each variable domain comprises three complementarity determining regions (CDR) flanked by four relatively conserved framework regions (FR). The FR are thought to maintain the structural integrity of the variable region domain. The CDR have been assumed to be responsible for the binding specificity of individual antibodies and to account for the diversity of binding of antibodies.

As the basic structure of an antibody contains two heterodimers, antibodies are multivalent molecules. For example, the IgG classes have two identical antigen binding sites, while the pentameric JgM class has 10 identical binding sites.

Monoclonal antibodies having identical genetic parentage and binding specificity have been useful both as diagnostic and therapeutic agents. Monoclonal antibodies are routinely produced by hybridomas generated by fusion of mouse lymphoid cells with an appropriate mouse myeloma cell line according to established procedures. The administration of murine antibodies for in vivo therapy and diagnostics in humans is limited however, due to the human anti-mouse antibody response illicited by the human immune system.

Chimeric antibodies, in which the binding or variable regions of antibodies derived from one species are combined with the constant regions of antibodies derived from a different species, have been produced by recombinant DNA methodology. See, for example,

Sahagenetal.,/ Immunol., 137:1066-1074 (1986): Sun et al., Proc. Natl. Acad. Sci USA,

82:214-218(1987); Nishimura etal.. Cancer Res., 47:999-1005(1987); and Lie etal. ProcNatl. Acad. Sci. USA, 84:3439-3443 (1987) which disclose chimeric antibodies to tumor-associated antigens. Typically, the variable region of a murine antibody is joined with the constant region of a human antibody. It is expected that as such chimeric antibodies are largely human in composition, they will be substantially less immunogenic than murine antibodies.

Chimeric antibodies still carry the Fc regions which are not necessary for antigen binding, but constitute a major portion of the overall antibody structure which affects its pharmacokinetics. For the use of antibodies in immunotherapy or immunodiagnostics, is it

-1CA 02117477 1997-11-03 desirable to have antibody-like molecules which localize and bind to the target tissue rapidly and for the unbound material to quickly clear from the body. Generally, smaller antibody fragments have greater capillary permeability and are more rapidly cleared from the body than whole antibodies.

Since it is the variable regions of light and heavy chains that interact with an antigen, single chain antibody fragments (scFvs) have been created with one and one V_H, containing all six CDR's, joined by a peptide linker (U.S.

Patent 4,946,778) to create a V_L-L-V_H polypeptide, wherein the

L stands for the peptide linker. A scFv wherein the V_L and Vfj domains are oriented V_h-L-Vl is disclosed in U.S. Patent 5,132,405.

As the scFvs have one binding site as compared to the minimum of two for complete antibodies, the scFvs have reduced avidity as compared to the antibody containing two or more binding sites.

It would therefore be advantageous to obtain constructions of scFvs having more than one binding site to enhance the avidity of the polypeptide, and retain or increase their antigen recognition properties. In addition, it would be beneficial to obtain multivalent scFvs which are bispecific to allow for recognition of different epitopes on the target tissue, to allow for antibody-based recruitment of other immune effector functions, or allow antibody capture of a therapeutic or diagnostic moiety.

It has been found that single chain antibody fragments, each having one V_H and one V_L domain covalently

-2A

64693-5016

CA 02117477 1997-11-03 linked by a first peptide linker, can be covalently linked by a second peptide linker to form a multivalent single chain antibody which maintains the binding affinity of a whole antibody. In one embodiment, the present invention is a multivalent single chain antibody having affinity for an antigen wherein the multivalent single chain antibody comprises two or more light chain variable domains and two or more heavy chain variable domains? wherein, each variable domain is linked to at least one other variable domain.

In another embodiment, the present invention is a multivalent single chain antibody which comprises two or more single chain antibody fragments, each fragment having affinity for an antigen wherein the fragments are covalently linked by a first peptide linker which contains an amino acid sequence of Leu Ser Ala Asp Asp Ala Lys Lys Asp Ala Ala Lys Lys Asp Asp Ala Lys Lys Asp Asp Ala Lys Lys Asp Leu and each fragment comprising:

(a) a first polypeptide comprising a light chain variable domain?

(b) a second polypeptide comprising a heavy chain variable domain ? and (c) a second peptide linker linking the first and second polypeptides into a functional binding moiety.

In another embodiment, the invention provides a DNA sequence which codes for a multivalent single chain antibody, the multivalent single chain antibody comprising two or more

-364693-5016

CA 02117477 1997-11-03 single chain antibody fragments, each fragment having affinity for an antigen wherein the fragments are covalently linked by a first peptide linker and each fragment comprising:

(a) a first polypeptide comprising a light chain variable domain;

(b) a second polypeptide comprising a heavy chain variable domain ; and (c) a second peptide linker linking the first and second polypeptides into a functional binding moiety.

The multivalent single chain antibodies allow for the construction of an antibody fragment which has the specificity and avidity of a whole antibody but are smaller in size allowing for more rapid capillary permeability. Multivalent single chain antibodies also allow for the construction of a multivalent single chain antibody wherein the binding sites can be two different antigenic determinants.

BRIEF DESCRIPTION OF THE DRAWINGS

Figure 1 illustrates covalently linked single chain antibodies having the configuration V_L-L-V_H-L-V_L-L-V_H(LHLH) and Vk-L-VjpL-Vjj-L-VLfLHHL) and a noncovalently linked Fv single chain antibody (Fv2).

<td></td><td></td><td> Figure</td><td> 2</td><td> illustrates</td><td> the</td><td> nucleotide</td><td> sequence</td><td> of</td><td> CC49</td>

<td> V_L</td><td> (SEQ</td><td> ID NOî 1)</td><td> •</td><td></td><td></td><td></td><td></td><td></td><td></td>

<td></td><td></td><td> Figure</td><td> 3</td><td> illustrates</td><td> the</td><td> amino acid</td><td> sequence</td><td> of</td><td> CC49</td>

<td>^VL</td><td> (SEQ</td><td> ID NO: 2)</td><td> •</td><td></td><td></td><td></td><td></td><td></td><td></td>

<td></td><td></td><td> Figure</td><td> 4</td><td> illustrates</td><td> the</td><td> nucleotide</td><td> sequence</td><td> of</td><td> CC49</td>

<td>^VH</td><td> (SEQ</td><td> ID NO: 3)</td><td> .</td><td></td><td></td><td></td><td></td><td></td><td></td>

-3a64693-5016

A

CA 02117477 1997-11-03

Figure 5 illustrates the amino acid sequence of CC49 V_H (SEQ ID NOs 4).

Figure 6 illustrates the nucleotide sequence and amino acid sequence of the CC49 single chain antibody LHLH in p49LHLH (SEQ ID NO: 6).

Figure 7 illustrates the nucleotide sequence and amino acid sequence of the CC49 single antibody LHHL in

P49LHHL (SEQ ID NO: 8).

Figure 8 illustrates construction of plasmids

PSL301T and pSL301HT.

Figure 9 illustrates construction of plasmid

P49LHHL.

Figure 10 illustrates construction of plasmid

P49LHLH.

Figure 11 illustrates the results of a competition assay using CC49 IgG, CC49 scFv2, and CC49 scFv using biotinylated CC49 IgG as competitor.

Nucleic acids, amino acids, peptides, protective groups, active groups and such, when abbreviated, are abbreviated according to the IUPAC IUB (Commission on Biological Nomenclature) or the practice in the fields concerned.

The term single chain antibody fragment (scFv) or antibody fragment as used herein means a polypeptide containing a V_L domain linked to a domain by a peptide linker (L), represented by V_L-L-V_H. The order of the V_L and V_H domains can be reversed to obtain polypeptides represented as V_H-L-V_L. Domain is a segment of protein that assumes a

-3b64693-5016

<img file="CA2117477C_D0001.tif" />

CA 02117477 1997-11-03 discrete function, such as antigen binding or antigen recognit ion.

A multivalent single chain antibody means two or more single chain antibody fragments covalently linked by a peptide linker. The antibody fragments can be joined to form bivalent single chain antibodies having the order of the V_Land domains as follows:

-3c64693-5016

A

CA 02117477 1997-11-03

WO 94/13806

PCT/US93/12039

<img file="CA2117477C_D0002.tif" />

<img file="CA2117477C_D0003.tif" />

Single chain multivalent antibodies which are Vivaient and greater have one or more antibody fragments joined to a bivalent single chain antibody by an additional interpeptide linker. In a preferred embodiment, the number of Vl and Vh domains is equivalent.

The present invention also provides for multivalent single chain antibodies which can be designated Vh-L-Vh-L-V_l-L-Vl or V_l-L-V_l-L-Vh-L-Vh·

Covalently linked single chain antibodies having the configuration V_l-L-Vh-L-V_l-L-Vh (LHLH) and V_l-L-Vh-L-V_h-L-Vl (LHHL) are illustrated in Figure 1. A noncovalently linked Fv single chain antibody &lt;Fv2) is also illustrated in Figure 1.

The single chain antibody fragments for use in the present invention can be derived from the light and/or heavy chain variable domains of any antibody. Preferably, the light and heavy chain variable domains are specific for the same antigen. The individual antibody fragments which are joined to form a multivalent single chain antibody may be directed against the same antigen or can be directed against different antigens.

To prepare a vector containing the DNA sequence for a single chain multivalent antibody, a source of the genes encoding for these regions is required. The appropriate DNA sequence can be obtained from published sources or can be obtained by standard procedures known in the art. For example, Kabat et al.. Sequences of Proteins of Immunological Interest 4thed._t(W9y), published by The U.S. Department of Health and Human Services, discloses sequences of most of the antibody variable regions which have been described to date.

When the genetic sequence is unknown, it is generally possible to utilize cDNA sequences obtained from mRNA by reverse transcriptase mediated synthesis as a source of DNA to clone into a vector. For antibodies, the source of mRNA can be obtained from a wide range of hybridomas. See, for example, the catalogue ATCC Cell Lines and Hybridomas, American

Type Culture Collection, 20309 Parklawn Drive, Rockville Md., USA (1990). Hybridomas secreting monoclonal antibodies reactive with a wide variety of antigens are listed therein, are available from the collection, and usable in the present invention. These cell lines and others of similar nature can be utilized as a source of mRNA coding for the variable domains or to obtain antibody protein to determine amino acid sequence of the monoclonal antibody itself.

Variable regions of antibodies can also be derived by immunizing an appropriate vertebrate, normally a domestic animal, and most conveniently a mouse. The immunogen will be the antigen of interest, or where a hapten, an antigenic conjugate of the hapten to an antigen such as keyhole limpet hemocyanin (KLH). The immunization may be carried out conventionally with one or more repeated injections of the immunogen into the host mammal, normally at two to three week intervals. Usually, three days after the last challenge, the spleen is removed and dissociated into single cells to be used for cell fusion to provide hybridomas from which mRNA can readily be obtained by standard procedures known in the art.

-4CA 02H7477 1997-11-03

When an antibody of interest is obtained, and only its amino acid sequence is known, it is possible to reverse translate the sequence.

The V_L and domains for use in the present invention are preferably obtained from one of a series of CC antibodies against tumor-associated glycoprotein 72 antigen (TAG-72) disclosed in published PCT Application WO 90/04410 on May 3, 1990, and published PCT Application WO 89/00692 on January 26, 1989. More preferred are the V_L and domains from the monoclonal antibody designated CC49 in PCT

Publications WO 90/04410 and WO 89/00692. The nucleotide sequence (SEQ ID NO: 1) which codes for the V_L of CC49 is substantially the same as that given in Figure 2. The amino acid sequence (SEQ ID NO: 2) of the V_L of CC49 is substantially the same as that given in Figure 3. The nucleotide sequence (SEQ ID NO.- 3) which codes for the of

CC49 is substantially the same as that given in Figure 4. The amino acid sequence (SEQ ID NO: 4) for the V_H of CC49 is substantially the same as that given in Figure 5.

To form the antibody fragments and multivalent single chain antibodies of the present invention, it is necessary to have a suitable peptide linker. Suitable linkers for Joining the Vjj and domains are those which allow the V_Hand V_L domains to fold into a single polypeptide chain which will have a three dimensional structure very similar to the original structure of a whole antibody and thus maintain the binding specificity of the whole antibody from which antibody fragment is derived. Suitable linkers for linking the scFvs

-5A

64693-5016

CA 02117477 1997-11-03 are those which allow the linking of two or more scPvs such that the and V_L domains of each immunoglobulin fragment have a three dimensional structure such that each fragment maintains the binding specificity of the whole antibody from which the immunoglobulin fragment is derived. Linkers having the desired properties can be obtained by the method disclosed in U.S. Patent 4,946,778. Prom the polypeptide sequences generated by the methods described in the 4,946,778, genetic sequences coding for the polypeptide can be obtained.

Preferably, the peptide linker joining the and V_L domains to form a scPv and the peptide linker joining two or more scPvs to form a multivalent single chain antibody have substantially the same amino acid sequence.

It is also necessary that the linker peptides be attached to the antibody fragments such that the binding of the linker to the individual antibody fragments does not interfere with the binding capacity of the antigen recognition site.

A preferred linker is based on the helical linker 20 designated 205C as disclosed in Pantoliano et al. Biochem.,

10117-10125 (1991) but with the first and last amino acids changed because of the codon dictated by the Xho I site at one end and the Hind III site at the other. The amino acid sequence (SEQ ID NO: 5) of the preferred linker is as follows: Leu-Ser-Ala-Asp-Asp-Ala-Lys-Lys-Asp-Ala-Ala-Lys-Lys-AspAsp- Ala-Lys -Lys -Asp- Asp- Ala-Lys -Lys- Asp-Leu .

-5a64693-5016

PCT/US93/12039

WO 94/13806

Α2ί17477

The linker is generally 10 to 50 aminoacid residues. Preferably, the linker is 10 to 30 amino acid residues. More preferably the linker is 12 to 30 amino acid residues. Most preferred is a linker of 15 to 25 amino acid residues.

Expression vehicles for production of the molecules of the invention include 5 plasmids or other vectors. In general, such vectors contain replicon and control sequences which are derived from species compatible with a host cell. The vector ordinarily carries a replicon site, as well as specific genes which are capable of providing phenotypic selection in transformed cells. For example, E. co//is readily transformed using pBR322 [Bolivar etal., Gene, 2,95- (1977), or Sambrook et al., Molecular Cloning, Cold Spring Harbor Press, New York, 2nd

Ed. (1989)].

Plasmids suitable for eukaryotic cells may also be used. S. cerevisiae, or common baker's yeast, is the most commonly used among eukaryotic microorganisms, although a number of other strains, such as Pi chia pastoris, are available. Cultures of cells derived from multicellular organisms such asSP2/0 or Chinese Hamster Ovary (CHO), which are available from the ATCC, may also be used as hosts. Typical of vector plasmids suitable for mammalian cells are pSV2neoand p$V2gpt(ATCC); pSVLand pKSV-10 (Pharmacia), pBPV-1/pML2d (International Biotechnology, Inc.).

The use of prokaryotic and eukaryotic viral expression vectors to express the genes for polypeptides of the present invention is also contemplated.

It is preferred that the expression vectors and the inserts-which code for the single chain multivalent antibodies have compatible restriction sites at the insertion junctions and that those restriction sites are unique to the areas of insertion. Both vector and insertare treated with restriction endonucleases and then ligated by any of a variety of methods such as those described in Sambrook et al., supra.

Preferred genetic constructions of vectors for production of single chain multivalent antibodies of the present invention are those which contain a constitutively active transcriptional promoter, a region encoding signal peptide which will direct synthesis/secretion of the nascent single chai n polypeptide out of the cel I. Preferably, the expression rate is commensurate with the transport, folding and assembly steps to avoid accumulation of the polypeptide as insoluble material. In addition to the replicon and control sequences, additional elements may also be needed for optimal synthesis of single chain polypeptide.

These elements may include splice signals, as well as transcription promoter, enhancers, and termination signals. Furthermore, additional genes and their products may be required to facilitate assembly and folding (chaperones).

Vectors which are commercially available can easily be altered to meet the above criteria for a vector. Such alterations are easily performed by those of ordinary skill in the art in light of the available literature and the teachings herein.

-6PCT/US93/12039

WO 94/13806 ^2117477

Additionally, it is preferred that the cloning vector contain a selectable marker, such as a drug resistance marker or other marker which causes expression of a selectable trait bythe host cell. Host cell* refers to cells which can be recombinantly transformed with vectors constructed using recombinant DNA techniques. A drug resistance or other selectable marker is intended in part to facilitate in the selection of transformants. Additionally, the presence of a selectable marker, such as a drug resistance marker, may be of use in keeping contaminating microorganisms from multiplying in the culture medium. In this embodiment, such a pure culture of the transformed host cell would be obtained by culturing the cel Is under conditions which require the induced phenotype for survival.

IQ Recovery and purification of the present invention can be accomplished using standard techniques known in the art. For example, if they are secreted into the culture medium, the single chain multivalent antibodies can be concentrated by ultrafiltration. When the polypeptides are transported to the periplasmic space of a host cell, purification can be accomplished by osmotically shocking the cells, and proceeding with ultrafiltration, antigen ] 5 affinity column chromatography or column chromatography using ion exchange chromatography and gel filtration. Polypeptides which are insoluble and present as refractile bodies, also called inclusion bodies, can be purified by lysisof the cells, repeated centrifugation and washing to isolate the inclusion bodies, solubilization, such as with guanidtne-HCI, and refolding followed by purification of the biologically active molecules.

The activity of single chain multivalent antibodies can be measured by standard assays known in the art, for example competition assays, enzyme-linked immunosorbent assay (ELISA), and radioimmunoassay (RIA).

The multivalent single chain antibodies of the present invention provide unique benefits for use in diagnostics and therapeutics. The use of multivalent single chain antibodies afford a number of advantages over the use of largerfragmentsorentire antibody molecules. They reach their target tissue more rapidly, and are cleared more quickly from the body.

For diagnostic and/or therapeutic uses, the multivalent single chain antibodies can be constructed such that one or more antibody fragments are directed against a target tissue and one or more antibody fragments are directed againsta diagnostic ortherapeutic agent.

The invention also concerns pharmaceutical compositions which are particularly advantageous for use in the diagnosis and/or therapy of diseases, such as cancer, where target antigens are often expressed on the surface of cells. For diagnostic and/or therapeutic uses, the multivalent single chain antibodies can be conjugated with an appropriate imaging or therapeutic agent by methods known in the art. The pharmaceutical compositions of the invention are prepared by methods known in the art, e.g., by conventional mixing, dissolving or lyophilizing processes.

-7WO 94/13806

PCT/US93/12039

CA21 17477

The invention will be further clarified by a consideration of the following examples, which are intended to be purely exemplary of the present invention.

-8WO 94/13806

PCT/US93/12039

C'?i17477

ABBREVIATIONS

BCIP bp

Bis-Tris propane

BSA

COR

ELISA

Fv2

IEF

Kbp

LB

Mab

MES

ΜΗ

NBT

Oligo

PAG

PAGE

PBS

PCR pSCFV

RIGS

RIT

SCFv scFv2

SOS

TBS

Tris

TTBS

Vh

V_L

5-bromo-4-chloro-3-indoyl phosphate base pair (1,3-bis[tris(hydroxymethyl)-methylamino]propane) bovine serum albumin

Complementarity determining region enzyme linked immunosorbent assay nan-covalent single chain Fv dimer isoelectric focusing kilo base pair

Luria-Bertani medium monoclonal antibody

2-(N-Morpholino)ethane sulfonic acid molecular weight nitro blue tétrazolium chloride

Oligonucleotides polyacrylamide gel polyacrylamide gel electrophoresis phosphate buffered saline polymerase chain reaction plasmid containing DNA sequence coding for SCFV radioimmunoguided surgery radio immuno therapy single chain Fv immunoglobulin fragment monomer single chain Fv immunoglobulin fragment dimer covalently linked sodium dodecyl sulfate

Tris-buffered saline (Tris[hydroxymethyl]aminomethane)

Tween-20 wash solution immunoglobulin heavy chain variable domain immunoglobulin light chain variable domain

-9«

CA 02117477 1997-11-03

Ant ibodies

CC49i A murine monoclonal antibody specific to the human tumor-associated glycoprotein 72 (TAG-72) deposited as

ATCC NO. HB9459.

CC49 FABi An antigen binding portion of CC49 consisting of an intact light chain linked to the N-terminal portion of the heavy chain.

CC49 scFVi Single chain antibody fragment consisting of two variable domains of CC49 antibody joined by a peptide linker.

CC49 Fv2; Two CC49 scFv non-covalently linked to form a dimer. The number after Fv refers to the number of monomer subunits of a given molecule, e.g., CC49 Fv6 refers to the hexamer multimers.

CC49 scFv2: Covalently-linked single chain antibody fragment consisting of two CC49 V_L domains and two domains joined by three linkers. Six possible combinations for the order of linking the V_L(L) and the V_H(H) domains together are: LHLH, LHHL, LLHH, HLLH, HLHL, and HHLL.

Plasmids pSCFV UHM; Plasmid containing coding sequence for scFv consisting of a CC49 variable light chain and a CC49 variable heavy chain joined by a 25 amino acid linker.

P49LHLH or p49LHHL: Plasmids containing the coding sequence for producing CC49 scFv2 LHLH or LHHL products, respectively.

64693-5016

-10A

CA 02117477 1997-11-03

EXAMPLES

General Experimental

Procedures for molecular cloning are as those described in Sambrook et al., Molecular Cloning, Cold Spring Harbor Press, New York, 2nd Ed. (1989) and Ausubel et al., Current Protocols In Molecular Biology, John Wiley and Sons,

New York (1992).

All water used throughout was deionized distilled water.

Oligonucleotide Synthesis and Purification

All oligonucleotides (oligos) were synthesized on either a Model 380A or a Model 391 DNA Synthesizer from

Applied Biosystems (Poster City, CA) using standard βcyanoethyl phosphoramidites and synthesis columns. Protecting groups on the product were removed by heating in concentrated ammonium hydroxide at 55°C for 6 to 15 hours. The ammonium hydroxide was removed through evaporation and the crude mixtures were resuspended in 30 to 40 pL of sterile water. After electrophoresis on polyacrylamide-urea gels, the oligos were visualized using short wavelength ultraviolet (UV) light. DNA bands were excised from the gel and eluted into 1 mL of

100 mM Tris-HCl, pH 7.4, 500 mM NaCl, 5 mM EDTA over 2 hours at 65°C. Pinal purification was achieved by applying the DNA to Sep-Pac™ C-18 columns (Millipore, Bedford, MA) and eluting the bound oligos with 60 percent methanol. The

-10a64693-5016

WO 94/13806

CA2H 7477

PCT/US93/12039 solution volume was reduced to approximately 50 pL and the DNA concentration was determined by measuring the optical density at 260 nm (OD₂₆₀).

Restriction Enzyme Digests

All restriction enzyme digests were performed using Bethesda Research

Laboratories (Gaithersburg, MD), New England Biolabs, Inc, (Beverly, MA) or Boehringer Mannheim (BM, Indianapolis, IN) enzymes and buffers foil owing the manufacturer's recommended procedures. Digested products were separated by polyacrylamide gel electrophoresis (PAGE). The gels were stained with ethidium bromide, the DNA bands were visualized using long wavelength UV light and the DNA bands were then excised. The gel slices were placed In dialysis tubing (Union Carbide Corp., Chicago) containing 5 mM Tris, 2.5 mM acetic acid, 1 mM EDTA, pH 8.0 and eluted using a Max Submarine electrophoresis apparatus (Hoefer Scientific Instruments, CA). Sample volumes were reduced on a Speed Vac Concentrator (Savant Instruments, Inc., NY). The DNA was ethanol precipitated and redissolved in sterile water.

Enzyme Linked Immunosorbent Assay (ELISA)

TAG-72 antigen, prepared substantially as described by Johnson et al. Can. Res.,

46,850-857 (1986), was adsorbed onto the wells of a polyvinyl chloride 96 well microtiter plate (Dynatech Laboratories, Inc., Chantilly, VA) by drying overnight. The plate was blocked with 1 percent BSA in PBS for 1 hour at 31°Cand then washed 3 times with 200pLof PBS,

2Q 0.05 percent Tween-20. 25 pL of test antibodies and 25 pL of biotinylated CC49 (1/20,000 dilution of a 1 mg/mLsolution) were added to the wells and the plate incubated for 30 minutes at 31 °C. The relative amounts of TAG-72 bound to the plate, biotinylated CC49, streptavidinalkaline phosphatase, and color development times were determined empirically in order not to have excess of either antigen or biotinylated CC49, yet have enough signal to detect competition by scFv. Positive controls were CC49at 5 pg/mL and CC49 Fab at 10 pg/mL.

Negative controls were 1 percent BSA in PBS and/or concentrated LB. Unbound proteins were washed away. 50 pL of a 1:1000 dilution of streptavidin conjugated with alkaline phosphatase (Southern Biotechnology Associates, Inc., Birmingham, AL) were added and the plate was incubated for 30 minutes at 31 °C. The plate was washed 3 more times. SOpLof a para-nitrophenyl-phosphate solution (Kirkegaard &amp; Perry Laboratories, Inc., Gaithersburg, MD) were added and the color reaction was allowed to develop fora minimum of 20 minutes. The relative amount of scFv2 binding was measured by optical density scanning at 404-450 nm using a microplate reader (Molecular Devices Corporation, Manlo Park, CA). Binding of the scFv2 species resulted in decreased binding of the biotinylated CC49 with a concomitant decrease in color development.

SDS-PAGE and Western Blotting

Samples for SDS-PAGE analysis (20 pL) were prepared by boiling in a non-reducing sample preparation buffer-Seprasol I (Integrated Separation Systems (ISS), Natick, MA) for

-11CA 02117477 2000-07-13

64693-5016 minutes and loaded on 10-20 percent gradient polyacrylamide Daiichi Minigels as per the manufacturer's directions (ISS).

★

Electrophoresis was conducted using a Mini 2-gel apparatus (155) at 55 mA per gel at constant current for approximately 75 minutes. Gels were stained in Coomassie Brilliant Blue R-250 (Bio-Rad, Richmond, CA) for at least 1 hour and destained. Molecular weight standards were prestained (Mid Range Kit, Diversified Biotech, Newton Center, MA) and included the following proteins: Phosphorylase b, glutamate dehydrogenase, ovalbumin, lactate dehydrogenase, carbonic amhydrase, B-lactoglobuiin and cytochrome C The corresponding MWsare: 95,500,55,000,43,000, 36,000, 29,000, 18,400, and 12,400, respectively.

When Western analyses were conducted, a duplicate gel was also run. After electrophoresis, one of the gels was equilibrated for 15-20 minutes in anode buffer #1 (0.3 M *

Tris-HQ pH 10.4). An Immobilon-P PVDF (polyvinylidenedichlorine) membrane (Millipore, Bedford, MA) was treated with methanol for2 seconds, and immersed in water for 2 minutes. The membrane was then equilibrated in anode buffer #1 for 3 minutes. AMilliblot-SDE apparatus (Millipore) was utilized to transfer proteins in the gel to the membrane. A drop of anode buffer #1 was placed in the middle of the anode electrode surface. A sheet of Whatman 3MM filter paper was soaked in anode buffer #1 and smoothly placed on the electrode surface. Another filter paper soaked in anode buffer #2 (25 mM tris pH 10.4) was placed on top of the first one. A sandwich was made by next adding the wetted PVDF membrane, placing the equilibrated gel on top of this and finally adding a sheet of filter paper soaked in cathode buffer (25mM Tris-HCl, pH 9.4 in 40 mM glycine). Transfer was accomplished in 30 minutes using 250 mA constant current (initial voltage ranged from 8-20 volts).

After blotting, the membrane was rinsed briefly in water and placed in a dish with 20 mL blocking solution ( 1 percent bovine serum albumin (BSA) (Sigma, St. Louis, MO) in Tris-buffered saline (TBS)). TB5 was purchased from Pierce Chemical (Rockford, IL) as a preweighed powder such that when 500 mL water is added, the mixture gives a 25 mM Tris,

0.15 M sodium chloride solution at pH 7.6. The membranes were blocked for a minimum of hour at ambient temperature and then washed 3 times for 5 minutes each using 20 mL ★

0.5 percent Tween-20 wash solution (TTBS). To prepare the TTBS, 0.5mLof Tween 20 (Sigma) was mixed per liter of TBS. The probe antibody used was 20 mL biotinylated FAJD14 solution (10 pgper20mL antibody buffer). Antibody buffer was made by adding 1 gBSAper IQQmLof TTBS. After probing for 30-60 minutes at ambient temperature, the membrane was washed 3 times with TTBS, as above.

Next, the membrane was incubated for 30-60 minutes at ambient temperature with 20 mL of a 1:500dilution in antibody buffer of streptavidin conjugated with alkaline phosphatase (Southern Biotechnology Associates, Birmingham, AL). The wash step was again repeated afterthis, as above. Prior to the color reaction, membranes were washed for minutes in an alkaline carbonate buffer (20 mL). This buffer is 0.1 M sodium bicarbonate, ★

Trade-mark

-12CA 02117477 2000-07-13

64693-5016 mM MgCl₂-H_;0, pH 9.8. To make up the substrate for alkaline phosphatase, nitroblue tétrazolium (NBT) chloride (50 mg. Sigma) was dissolved in 70 percent dimethylformamide.

5-Bromo-4-chloro-3-indoyl phosphate (BC1P) (25 mg, 5igma) was separately dissolved in 100 percent dimethylformamide. 5-Bromo-4-chioro-3-indoyl phosphate (BC1P) 25 mg, Sigma) was separately dissolved in 100 percent dimethylformamide. These solutions are also commercially available as a Western developing agent sold by Promega. For color development, 120 pLof each were added to the alkaline solution above and allowed to react for 15 minutes before they were washed from the developed membranes with water. Biotinylated FAID14

FAID14 is a murine anti-idiotypic antibody (lgG2a, K isotype) deposited as ATCC

No. CRL 10256 directed against CC49. FAID14 was purified using a Nygene Protein A affinity column (Yonkers, NY). The manufacturer's protocol was followed, exceptthat 0.1 Msodium citrate, pH 3.0 was used as the elution buffer. Fractions were neutralized to pH —7 using 1.0 M

Tris-HCI pH 9.0. The biotinylation reaction was set up as follows. FAID14(1mg, lOOpLin water) was mixed with 100 pL of 0.1 M Na_;CO₃ pH 9.6. Biotinyl-e-amino-caoroic acid N-hydroxy succinimide ester (Biotin-X-NH5) (Calbiochem, LaJolla, CA) (2.5 mg) was dissolved in 0.5 mL dimethylsuifoxide. Biotin-X-NHS solution (20 pL) was added to the FAID14 solution and allowed to react at 22*C for 4 hours. Excess biotin and impurities were removed by gel * filtration, using a Pharmacia Superose 12 HR10/30 column (Piscataway, NJ), At a flow rate of 0.8ml/min, the biotinylated FAID14 emerged with a peak at 16.8 min. The fractions making up this peak were pooled and stored at 4®C and used to detect the CC49 idiotype as determined by theCC49 V, and V CDRs.

Isoelectric Focusing (IEF)

Isoelectric points (pi's) were predicted using a computer program called PROTEIN-TiTRATE, available through DNA5TAR (Madison, WI). Based on amino acid composition with an input sequence, a MW value is given, in addition to the pi. Since Cys residues contribute to the charge, the count was adjusted to 0 for Cys, since they are all involved in disulfide bonds.

Experimentally, pi's were determined using Isogel agarose IEF plates, pH range

3-10(FMC Bioproducts, Rockland, ME). A Biorad Sio-phoresis horizontal electrophoresis cell was used to run the IEF, following the directions of both manufacturers. The electrophoresis conditions were: 500 volts (limiting), at 20 mA current and 10 W of constant power. Focusing was complete in 90 min. IEF standards were purchased from Biorad; the kit included phycocyanin, β-lactoglobulin B, bovine carbonic anhydrase, human carbonic anhydrase, equine myoglobin, human hemoglobins A and C, 3 lentil lectins and cytochrome C with pi values of 4.65, 5.10, 6.00, 6.50, 7.00, 7.10and 7.50, 7.80, 8.00, and 8.20 and 9.60, resoectively. Gels were stained and destained according to the directions provided by FMC.

★

Trade-mark

-13CA 02117477 2000-07-13

64693-5016

Quantitation of CC49 Antibody Species

All purified CC49 antibodies including the IgG, scFv2 species and the monomeric scFv were quantitated by measuring the absorbence of protein dilutions at 28Q mm using matching 1.0 cm pathiength quartz cuvettes (Heilma) and a Perkin-Elmer UV/VIS Spectrophotometer, Model 552A. Molar absorptivities (E_m) were determined for each antibody by using the following formula:

E_m = (number Trp)X 5,500 + (number Tyr) X 1,340 + (number(Cys)2)X 150 + (number Phe) X 10

The values are based on information given by D. B. Wetlaufer, Advances in Protein Chemistry, 12,375-378).

High Performance Liquid Chromatography

All high performance liquid chromatography (HPLC) was performed for CC49 scFv2 purification using an LKB HPLC system with titanium or tefion tubing throughout. The system consists of the Model 2150 HPLC pump, model 2152 controller, UV CORD5II model 2238 detection system set at an absorbence of 276 nm and the model 2211 SuperRac fraction collector.

PCR Generation of Subunits

All polymerase chain reactions (PCR) were performed with a reaction mixture consisting of: 150 picograms (pg) plasmid target (p5CFVUHM); 100 pmoles primers; 1 pL Perkin-Elmer-Cetus (PEC, Norwalk, CT) Ampli-Taq polymerase; 16 pL of 10 mM dNTPs and 10 pL of 10X buffer both supplied in the PEC kit; and sufficient water to bring the volume to total volume to 100 pL The PCR reactions were carried out essentially as described by the manufacturer. Reactions were done in a PEC 9600 thermocycler with 30 cycles of: dénaturation of The DNA at 94^eC for 20 to 45 sec, annealing from between 52 to 60’C for 0.5 to 1.5 min., and elongation at 72°C for 0.5 to 2.0 min. Oligonucleotide primers were synthesized on an Applied Biosystems (Foster City, CA) 380A or 391 DNA synthesizer and purified as above.

Ligations

Ligation reactions using 100 ng of vector DNA and a corresponding 1:1 stoichiometric equivalent of insert DNA were performed using a Stratagene (La Jolla, CA) T4 DNA ligase kit following the manufacturer's directions. Ligation reactions (20 pL total volume) were initially incubated at 18’Cand allowed to cool gradually overnight to4°C.

Transformations

Transformationswereperformed utilizing 100 pL of Stratagene E. coli AGI competent cells (5tratagene, La Jolla, CA) according to the directions provided by the manufacturer. DNA from the ligation reactions ( 1 -5 pL) were used. After the transformation step, cells were allowed to recover for 1 hr in Luria broth (LB) at 37^eCwith continuous mixing and subsequently plated onto either 20 pg/mL chloramphenicol containing (CAM 20) Luria agar for pSCFVUHM, p49LHLH or p49LHHL or 100 pg/mL ampicillin (AMP 100) Luria agar plates *

Trade-mark

-14CA 02117477 2000-07-13

64693-5016 (LB-AMP 100) for clones containing the plasmid pSL301 or subsequent constructions derived from pSL301.

Screening of E. coli Clones

Bacterial plasmids were isolated from LB broth 5 culture containing the appropriate drug to maintain selection pressure using Promega (Madison, WI) Magic mini-prep plasmid preparation kits. The kit was used per the manufacturer's specifications .

Plasmid Constructions

Two plasmids, designated p49LHLH and p49LHHL, were constructed to produce multivalent single chain antibodies.

The host cell containing p49LHLH produced a polypeptide which can be designated by Vl-L-Vh-L-VI-L-Vh where V_L and V_H are the light and heavy chain variable regions of CC49 antibody and linker (L) is a 25 amino acid linker having the sequence (SEQ ID NO: 5).

Leu-Ser-Ala-Asp-Asp-Ala-Lys-Lys-Asp-Ala-Ala-Lys-LysAsp-Asp-Ala-Lys-Lys-Asp-Asp-Ala-Lys-Lys-Asp-Leu.

The host cell containing p49LHHL produced a polypeptide which can be designated by Vl-L-Vh-L-Vh-L-Vl where V_Land V_H are the light and heavy chain variable domains of the CC49 antibody and L is a peptide linker having the amino acid sequence indicated above.

The nucleotide sequence (SEQ ID NO : 6) and amino acid sequence (SEQ ID NO: 7) of the CC49 Vl-L-Vh-L-V^-L-Vh (p49LHLH) are given in Figure 6. The nucleotide sequence (SEQ ID NO: 8) and amino acid sequence (SEQ ID NO: 9) of the CC49 V_l-L-V_h-L-V_hL-V_l (p49LHHL) are given in Figure 7.

CA 02117477 2000-07-13

64693-5016

Construction of pSL301 HT

The construction of pSL301 HT is illustrated in Figure 8. The Bacillus lichiformis pencillinase P (pen?) terminator sequence was removed from the plasmid designated pSCFV UHM by a 45 minute digest with Nhe I and BamH I, excised from a 4.5 percent polyacrylamide gel after electrophoresis, electroeluted, ethanol precipitated and ligated into the same sites in the similarily prepared vector: pSL301 (Invitrogen,

San Diego, CA). A procedure for preparing pSCFV UHM is given in U.S. Patent No. 6, 071, 515. In general, pSCFV UHM contains a nucleotide sequence for a penP promoter; a unique Neo I restriction site; CC49 V_L region; Hind III restriction site; a 25 amino acid linker; a unique a Xho I restriction site; CC49 V_H region; Nhe I restriction site; penP terminator; and BamH I restriction site (see, Figure 8). The penP promoter and terminator are described in Mezes, et al. (1983), J. Biol.

Chem., 258, 11211-11218(1983).

An aliquot of the ligation reaction (3 μΐυ) was used to transform competent E. coli AGI cells which were plated on

LB-AMP100 agar plates and grown overnight. Potential clones containing the penP terminator insert were screened using a Pharmacia (Gaithersburg, MD) T7

15a

CA 02117477 2000-07-13

64693-5016

Quickprime ^UP DNA labeling kit in conjunction with the microwave colony iysis procedure outlined in Buluwela et al.. Nucleic Add Research, 17,452 (1989). The probe, which was the pen P-Nhe l-8amH I terminator fragment itself was prepared and used according to the it directions supplied with the Quickprime kit. A done which was probe positive and which contained the 207 base pair inserts from a BamH 1 and Nhe I digest (base pairs (bp) 1958 to 2165, Figure 6) was designated pSL301 T and chosen to construct p5L301 HT which would contain the nucleotide sequence for CC49 Vh- The reason the Nhe Ι-BamH I penP terminator was placed intopSL301 was to eliminate the Eco47 III restriction endonuclease site present in the polylinker region between its Nhe I and BamH I sites. This was designed to accommodate the subsequent build-up of the V_L and V_H domains where the Eco47 III site needed to be unique for the placement of each successive V domain into the construction. As each V domain was added at the Eco47 lll-Nhe I sites, the Eco47 III was destroyed in each case to make the next Eco47 III site coming in on the unique insert.

The V_H sequence was made by PCR with oligos 5* SCP1 and 3'oligo SCP5 using pSCFV UHM asthe target forPCR amplification. The DNA sequence forSCP1 (SEQ ID NO: 10) and SCP5 (SEQ ID NO: 11 ) are as follows:

SCP1 : 5’-TAAA CTCGAG GTT CAG TTG CAG CAG -3'

SCP5: 5’-TAAA GCT AGC ACCA AGC GCT TAG TGA GGA GAC GGT GAC TGA GGT-3'

The underlined portion indicates the endonuclease restriction sites.

The amplified Vh DNA was purified from a 4 percent PAÇ, electroeiuted ethanol precipitated and dissolved in 20 pL water. The V_H sequence was digested with Xho I and Nhe I restriction enzymes and used as the insert with the pSL301 T vector which had been digested with the same restriction enzymes and subsequently purified. A standard ligation reaction was done and an aliquot (4 pL) used to transform competent E. coli AG 1 cells. The transformed cells were plated onto LB AMP100 agar plates. Candidate dones were picked from a Nhe I and Xho I digest screen that revealed that the CC49V_H insen had been obtained.

DNA sequencing was performed to verify the sequence of the CC49V_H with United States Biochemical (USB) (Cleveland, Ohio) Sequence kit and sequencing primers pSL30lSEQ8 (a 21 bp sequencing primer which annealed in the p5L301 vector 57 bp upstream from the Xho I site) and CC49VHP, revealed clones with the correct CC49Vh sequence in P5L301HT. This plasmid was used as the starting point in the construction of both pSL301-HHLT and pSL301-HLHT. The sequencing oligos used are shown here.

The nucleotide sequence of pSL301 SEQ B (SEQ ID NO: 12) and CC49V_H (SEQ ID No: 13) are as follows:

pSL301SEQB: S'-TCG TCC GAT TAG GCA AGC TTA-3'

CC49VHP: 5'-GAT GAT TTT AAA TAC AAT GAG-3' *

Trade-mark

-16CA 02117477 1997-11-03

Example 1 p49LHHL Construction

Using pSL301 HT (5 pg) as the starting material, it was digested with Eco47 III and Nhe I and the larger vector fragment was purified. A CC49V_H insert fragment was generated by PCR using SCP6B as the 5' oligo and SCP5 as the 3' oligo. The nucleotide sequence (SEQ ID NO: 14) of SCP6B is as

<td colspan="14"> fOllOWS:</td>

<td rowspan="2"></td><td rowspan="2"> SCP6B:</td><td rowspan="2"> 5'-TAAA AAA GAC</td><td colspan="3"> TGC GCA GAT</td><td colspan="4"> GAC GCA AAG AAA</td><td colspan="4"> GAC GCA GCT AAA</td>

<td> GAT</td><td> GCC</td><td> AAA</td><td> AAG</td><td> GAT</td><td> GAC</td><td> GCC</td><td> AAG</td><td> AAA</td><td> GAT</td><td> CTT</td>

<td> 10</td><td></td><td> GAG GTT</td><td> CAG</td><td> TTG</td><td> CAG</td><td> CAG</td><td> TCT-</td><td> -G'</td><td></td><td></td><td></td><td></td><td></td>

<td></td><td> The oligo</td><td colspan="2"> SCP6B also ι</td><td colspan="2"> contains</td><td> part</td><td> of</td><td> the</td><td colspan="2"> coding</td><td colspan="3"> region for</td>

the linker (bp 8-76 of SEQ ID NO: 14). The portion of the oligo designed to anneal with the CC40VH target in pSCFV UHM is from bp77-90 in SEQ ID NO: 14.

The underlined sequence corresponds to the Fsp I site. The resulting PCR insert was purified, digested with Fsp I and Nhe I and used in a ligation reaction with the pSL301 HT Eco47 III-Nhe I vector (Figure 7). Competent E.

coll AGI cells were used for the transformation of this ligation reaction (3 pL) and were plated on LB-AMP100 agar plates. Two clones having the correct size Xho Ι-Nhe I insert representative of the pSL301 HHT product were sequenced with the oligo SQP1 and a single clone with the correct sequence (nucleotides 1124-1543 of Figure 7) was chosen for further construction. The nucleotide sequence of SQP1 (SEQ ID NO: 15) is as follows:

SQP1: 5'-TG ACT TTA TGT AAG ATG ATG T-3'

-17A

64693-5016

CA 02117477 1997-11-03

The final linker-V_L subunit (bp 1544-1963, Figure 7) was generated using the 5'oligo, SCP7b and the 3' oligo,

SCP8a, using pSCFV UHM as the target for the PCR. The nucleotiode sequence of SCP7b (SEQ ID NO: 16) is as follows:

SCP7b: 5'-TAAA TGC GCA GAT GAC GCA AAG AAA GAC GCA GCT AAA

AAA GAC GAT GCC AAA AAG GAT GAC GCC AAG AAA GAT CTT

GAC ATT GTG ATG TCA CAG TCT CC

The underlined nucleotides correspond to an Fsp I site. The nucleotide seguence of SCP8a (SEQ ID NO: 17) is as follows:

SCP8a: 5'-TAAA GCT AGC TTT TTA CTT AAG CAC CAG CTT GGT

CCC-3'

The first set of underlined nucleotides correspond to an Nhe I site, while the other corresponds to an Afl II site. Nucleotides 8-76 of SCP70 code for the linker (nucleotides 1544-1612 of Figure 7) while nucleotides 77-99 which anneal to the V_L correspond to 1613-1635 of Figure 7.

The primer SCP8a has a short tall at its 5' end, a Nhe I restriction site, a stop codon, an Afl II restriction site and the last 21 bases of the V^. After Fsp I and Nhe I digestion, this resulting 420 bp insert was purified and ligated into the

Nhe I and Eco47 III sites of the purified pSL301HHT vector, candidate clones were screened with Nhe I and Xho I, the correct size insert verified and sequenced with 49LFR2(-) and

SQP1 to confirm the newly inserted sequence in pSL301HHLT.

The nucleotide sequence (SEQ ID NO: 18) is as follows:

49LFR2I-): 5'-CTG CTG GTA CCA GGC CAA G-3'

-18A

64693-5016

CA 02117477 1997-11-03

The plasmid pSL301HHLT was digested with Xho I and Nhe I, purified, and the resulting 1179 bp V_H-linker-V_Hlinker-V_L segment ligated into pSCFV UHM, which had been cut with the same restriction enzymes and the larger vector fragment purified, to form p49LHHL. The ligation reaction (4 jjI aliquot) was used to transform competent E. coli AGI cells (Stratagene) and plated onto LBCAM20 agar plates. A single clone which had a plasmid with the correct restriction enzyme map was selected to contain p49LHHL. The p49LHHL contains a penP promoter and a nucleotide sequence for the CC49 multivalent single chain antibody scFv2:

^VL^L“^VH^L-VH“^L-^VL ^{or CC49} SCFv2 (LHHL).

Example 2 : p49LHLH Const ruct ion

The construction of p49LHLH is schematically represented in Figure 10. A linker-V_L subunit was generated with the 5' oligo SCP7b and the 3' oligo SCP9 (SEQ ID NOs 19).

SCP9i 5'-TAA AGC TAG CAC CAA GCG CTT AGT TTC AGC ACC AGC

TTG GTC CCA G-3'

The SCP7b oligo (nucleotides 8-76) codes for the 20 linker in Figure 6 (corresponding to nucleotides 1124-1192) and annealed to the pSCFV UHM target for the PCR (nucleotides 77-99) corresponding to nucleotides 1193-1215 of the V_L in Figure 6.

SCP9 has a Nhe I site (first underlined nucleotides) and an Eco47 III site (second underlined nucleotides) which are restriction sites needed for making the pSL301HLT ready to accept the next V domain. Nucleotides 18-23 of SCP9 correspond to nucleotides 1532-1537 of nucleotides 1508-1531

-1964693-5016

A

CA 02117477 1997-11-03 of Figure 6 which was also the annealing region for SCP9 in the PCR. The plasmid pSL301 HT was digested with Eco47 III and Nhe I and the larger vector fragment was purified for ligation with the linker-CC49Vj_j DNA insert fragment from the PCR which had been treated with Fsp I and Nhe I and purified.

The ligation mixture (3 pL) was used to transform E. coli AGI competent cells and one colony having the correct Xho Ι-Nhe I size fragment was sequenced using the oligo PENPTSEQ2. The nucleotide sequence (SEQ ID NO· 20) is as follows:

5'-TTG ATC ACC AAG TGA CTT TAT G-3'

The sequencing results indicated that there had been a PCR error and deletion in the resulting pSL301HT clone. A five base deletion, corresponding to nucleotides 1533-1537 as seen in Figure 6 had been obtained and nucleotide 1531 which should have been a T was actually a G, as determined from the

DNA sequence data. The resulting sequence was ' . . . G AAGC GCT T... etc.

wherein the underlined sequence fortuitously formed an

Eco47 III site. The AGCGCT sequence in Figure 6, would 20 correspond to nucleotides 1530, 1531, 1532, 1538, 1539 and

1540. This error was corrected in the next step, generating pSL301 HLHT, by incorporating the 5 base deletion at the end of oligo SCP6C (SEQ ID NO: 21).

SCP6C: 5 ' -TAAGCGCTGATGATGCTAAGAAGGACGCCGCAAAAAA

GGACGACGCAAAAAAAGATGATGCAAAAAAGGATCTGG

AGGTTCAGTTGCAGCAGTCTGAC-3'

The underlined sequence in SCP6C corresponds to an

Eco47 III site. SCP6C was used as the 5' oligo, with SCP10 as

-19a64693-5016

A

CA 02117477 1997-11-03 the 3' oligo in a PCR to generate a linker CC49 Vl segment.

The nucleotide sequence (SEQ ID NO: 22) is as follows:

SCP1O: 5'TTG TGC TAG CTT TTT ATG AGG AGA CGG TGA CTG AGG

TT-3'

The underlined sequence in SCP1O corresponds to the

Nhe I site found at nucleotides 1958-1963 in Figure 6. The

PCR insert was digested this time only with Nhe I and purified. The vector (pSL301 HLT) was digested at the Eco47

III site (that had been formed) and Nhe I and purified. The insert and vector were ligated and an aliquot (3 pL) used to transform competent E. coli AGI cells. This was plated on LBAMP100 plates and candidate clones screened with Xho I and Nhe

I. Three clones having the correct size DNA were obtained.

Two of these clones were sequenced using the oligo 49VLCDR3(+) and SQP1. The nucleotide sequence (SEQ ID NO: 23) of

49VLCDR3(+) is as follows :

49VLCDR3(+):

5'-CAG CAG TAT TAT AGC TAT-3'

One clone, with the correct sequence was obtained and the sequence from nucleotides 1533 to 1963 in Figure 6 were verified, giving a correct pSL301 HLHL clone.

To generate the final plasmid, p49LHLH for expression in E. coli, pSL301 HLHT (5 pg) was digested with Nhe I and Xho I, and the smaller insert fragment containing the V_H-L-V_L-L-V_H sequence purified. It was ligated with the larger purified vector fragment from a digest of pSCFV UHM

-19bA

64693-5016

CA 02117477 1997-11-03 (5 ug) with Xho I and Nhe I. An aliquot of the ligation mix (4 μΐι) was used to transform competent E. coli AGI cells. The transformation mix was plated on LB-CAM20 plates, and a representative clone for p49LHLH was selected on the basis of a correct restriction enzyme map (see Figure 10) and biological activity toward TAG-72.

Example 3; Purification of CC49 scFv2 LHLH and LHHL

Covalently Linked Dimers

For the purification of the CC49 covalently linked 10 single chain dimers, (scFv2), E. coli periplasmic fractions were prepared from 1.0 L overnight cultures of both p49LHLH and p49LHHL. Briefly, the culture was divided into 4 x 250 mL portions and centrifuged at 5,000 rpm for 10 minutes in a Sorvall GS-3 rotor. The pelleted cells were washed and resuspended in 100 mL each of 10 mM tris-HCl pH 7.3 containing 30 mM NaCI. The cells were again pelleted and washed with a total of 100 mL 30 mM tris-HCl pH 7.3 and pooled into one tube. To this, 100 mL of 30 mM tris-HCl pH 7.3 containing 40 percent w/v sucrose and 2.0 mL of 10 mM EDTA pH 7.5 was added. 20 The mixture was kept at room temperature, with occasional shaking, for 10 minutes. The hypertonic cells were then pelleted as before. In the next step, the shock, the pellet was quickly suspended in 20 mL ice cold 0.5 mM MgCl₂ and kept on ice for 10 minutes, with occasional shaking. The cells were pelleted as before and the supernatant

-19cA

64693-5016

CA 02117477 2000-07-13

64693-5016 containing the E coli periplasmic fraction was clarified further by filtration through a 0.2 pm Nalge (Rochester, NY) filter apparatus and concentrated in Amicon (Danvers, MA) Centriprep 30 and Centricon 30 devices to a vol ume of less than 1.0 mL

The concentrated periplasmic shockates from either the p49LHLH or p49LHHL clones were injected onto a Pharmacia (Piscataway, NJ) Superdex 75 HR 10/30 HPLC column that had been equilibrated with PBS. At a flow rate of 0.5 mL/minute, the product of interest, as determined by competition ELISA, had emerged between 21 through 24 minutes. The active fractions were pooled, concentrated as before and dialyzed overnight using a system 500 Microdialyzer Unit (Pierce Chemical) against 20 mM Tris-HCJ pH 7.6 with 3-4 changes of buffer and using an 8,000 MW cut-off membrane. The sample was injected on a Pharmacia Mono Q* HR 5/5 anion exchange HPLC column. A gradient program using 20 mM Tris-HCl pH 7.6 as buffer A and the same solution plus 0.5 M NaCl as buffer B was employed at a flow rate of 1.5 mL/min. The products of interest in each case, as determined by competition ELISA, emerged from the column between 3 and 4 minutes. Analysis of the fractions at this point on duplicate SDS-PAGE gels, one stained with Coomassie Brilliant Blue R-250 and the other transferred for Western analysis (using biotinylated PAID 14 as the probe antibody) revealed a single band at the calculated molecular weight for the scFv2 (LHLH or LHHL) species at 58,239 daltons. The active fractions were in each case concentrated, dialysed against 50 mM MES pH it

5.8 overnight and injected on a Pharmacia Mono S HR 5/5 cation exchange column. The two fractions of interest from this purification step, as determined by SDS-PAGE and ELISA, fractions 5 and 6, eluted just before the start of the gradient, so they had not actually bound to the column. Fractions 5 and 6 were consequently pooled for future purification.

A Mono Q column was again run on the active Mono S fractions but the buffer used was 20 mM Tris-HG, pH 8.0 and the flow rate was decreased to 0.8 mL/minute. The products emerged without binding, but the impurity left over from the Mono 5 was slightly more held up, so that separation did occur between 5 and 6 minutes. After this run, the products were homogeneous and were saved for further characterization.

Isoelectric Focusing

The isoelectric points (pi) of the constructs was predicted using the DNASTAR (Madison, Wl) computer program Protein-titrate. Based on amino acid composition, a MW and pi value was calculated.

Experimentally, pis were determined using FMC Bioproducts (Rockland, ME) lsogel*ÎEF plates, pH range 3-10. A Biorad (Richmond, CA) electrophoresis unit was used to run the IEF, following the directions of both manufacturers. The electrophoresis conditions were as follows: 500 V (limiting) at 20 mA and at 10 W of constant power. Focusing was complete in minutes. Biorad IEF standards included phycocyanin, beta lactoglobulin B, bovine carbonic anhydrase, human carbonic anhydrase, equine myoglobulin, human hemoglobins A and C, 3 lentil lectin, and cytochrome C with pi value of 4.65, 5.10,6.00,6,50,7.00, 7.50,7.8, 8.00,8.20 ★

Trade-mark

-20WO 94/13806

PCT/US93/12039

C * 21 17477 and 9.6, respectively. Gels were stained and destained according to directions provided by FMC. The DNASTAR program predicted values of 8.1 forthe pi for both scFv2 species. A single, homogeneous band forthe pure products was observed on the gel at pi values for both at 6.9.

Purified CC49 antibodies such as the IgG, scFv2 (LHLH and LHHL) were quantitated by measuring the absorbence spectrophotometricaliy at 280 nm. Molar absorbtivity values, cm, were determined for each using the formula cited above by Wetlaufer.

Based on the amino acid composition, the E^01% (280 nanometers) values for CC49 IgG, CC49 scFv2 LHLH, CC49 scFv2 LHHL and CC49 scFv were 1.49,1.65, 1.65 and 1.71, respectively.

jq Example 4

Relative activities of the CC49 scFv2 species LHLH and LHHL, were compared with the tgG and a monomer scFv form with a FLAG peptide at the COOH terminus.

Percent competition was determined from the ELISA data by the following equation:

Zero competition - sample reading (QD405-450 nm) _x1Qa¹ zero competition -100 percent competition

The zero competition value was determined by mixing (1:1) one percent BSA with the biotinylated CC49(3X 10-14 moles) while the 100 percent competition value was based on a 5 pg/mL sample of CC49 IgG mixed with the biotinylated CC49 IgG. The data are presented in Figure 11. Absorbence values for the samples were measured at 405 nm -450 nm.

²⁰ The average of triplicate readings was used. Initially samples (25 pL) were applied to the TAG-72 coated microliter plates at 1.0 X 10-10 moles of binding sites/mL. Biotinylated CC49 (4 pg/pL diluted 1:20,000-used 25 pL) diluted the samples by a factor of 2. Serial dilutions (1:2) were performed. Both forms of the scFv2 are approximately equivalent to the IgG (see Figure 11). In a separate experiment, a CC49scFv monomer was compared to a Fab fragment, ²5 both of which are monovalent and these were also shown to be equivalent in their binding affinity for TAG-72. These results indicate that both forms of the covalently linked dimers have 2 fully functional antigen binding sites. This is the same increase in avidity as observed with the whole IgG, relative to a monomeric species.

These data also indicate that the scFv2 molecules, like their CC49 IgG parent are candidates for immunotherapeutic applications, but with the benefit of increased capillary permeability and more rapid biodistribution pharmacokinetics. The advantage should allow multiple injections of compounds of the present invention and give higher tumorztissue ratios in immunotherapeutic treatment regimens for cancer treatment, relative to the existing IgG molecules.

Other embodiments of the invention will be apparent to those skilled in the art from a consideration of this specification or practice of the invention disclosed herein. It is

-214^13806 . _ „

PCTZÜS93/12039 intended that the specification and examples be considered as exemplary only, with the true scope and spirit of the invention being indicated by the following claims.