US 6,071,515 AGrant

Dimer and Multimer Forms of Single Chain Polypeptides

Issue Date:2000-06-06

•5 Claims

•73 Drawing Sheets

Abstract

The present invention discloses novel proteins which are dimers and multimers of single chain polypeptides. The single chain polypeptides having two domains derived from the immunoglobulin superfamily which are joined by a peptide linker. The dimers and multimers being formed by non-covalent linking of the single chain polypeptides.

Metadata

Assignee

The Dow Chemical Company

Inventors

Peter S. Mezes
Ruth A. Richard
Joseph A. Affholter
Nicolas J. Kotite

Application Information

Application Number:US 4639030

Filing Date:1995-06-05

Priority Date:1992-08-21

Art Unit:162

Classifications

IPC:

A61K 39395G01N 3353G01N 33574

Field of Search:

530424435388.1130.1;178.1;1.11;277.1;133.1;138.1;141.1;152.1;155.1;158.1;172.1;174.1;181. 17.1;7.2;7.23

Patent Drawings (73 sheets)

Description

BACKGROUND OF THE INVENTION

This invention relates to the field of molecular biology and production of single chain polypeptides.

The immunoglobulin (Ig) superfamily is a group of proteins having as a common denominator an adhesion function or binding function that triggers a subsequent event at a cell surface. The molecules of the Ig superfamily have a diversity of functions, including involvement in immune response and in controlling the behavior of cells in various tissues. See, for example, Williams et al., Ann. Rev. Immunol., 6, 381-405 (1988); Williams, Immunology Today, 8, 298-303 (1987); and Matsunaga, Immunology Today, 6, 260-263 (1985).

The Ig superfamily includes such functionally-important members as the antigen receptors of lymphocytes [e.g., immunoglobulins, T-cell receptor (TCR) molecules, and Class I and Class II molecules of the major histocompatability complex (MHC)]. Lymphocytes are divided broadly into cells that mature in the thymus (T-lymphocytes or T-cells) and cells that mature without passing through the thymus (B lymphocytes or B cells). T-cells can be divided into a number of subtypes such as T-helper cells or cytotoxic T-cells. Helper T-cells respond only to antigen that is on the surface of an antigen presenting cell that also expresses a class II MHC. Cytotoxic T-cells destroy other cells that have been infected by an organism, such as a virus, by recognizing foreign antigen presented in the context of the class I MHC molecule on the surface of the infected cell.

A structural feature that characterizes members of the Ig superfamily is the presence of one or more regions homologous to the basic structural unit of immunoglobulins, the immunoglobulin homology unit. These regions are usually between 10 and 110 amino acid residues in length and are characterized by a series of anti-parallel .beta. sheets that generate a compact sandwich of two .beta. sheets or a single large sheet in the case of MHC proteins. This structure is stabilized by a small series of conserved residues, particularly, two virtually invariant cysteine residues that generate a signature disulfide bond holding the faces of the .beta. sheet sandwich together. The loop sequences connecting the .beta. strands are less significant for the basic homology unit structure and therefore, can accommodate extensive variability.

The Ig superfamily contains distinct classes of immune receptors, molecules that interact directly with an antigen including: (1) antibodies, (2) T-cell receptors, (3) MHC class I, (4) MHC class II; (5) Immunoglobulin Fc receptors (including CD16, CD23, CDw32, etc.) and (6) some cell-surface adhesion molecules (including but not limited to CD4 and CD8 and .beta.1 integrin family-of proteins). Antibodies can be expressed either on the cell surface or humorally; that is secreted into the blood to carry out their functions at a distance from the cell of origin. T-cells synthesize only a surface-bound antigen receptor. There are two classes of receptors encoded within the MHC. Class I MHC molecules are present in varying amounts on virtually all somatic cells. Class II MHC molecules are expressed only on particular antigen-presenting cells such as macrophages and B cells. Class I and class II MHC play a role in presenting antigens in a form which can be recognized by the T-cell receptors. Immunoglobulin Fc receptors are expressed on a variety of hematopoietic cells and function in antibody-dependent T-cell killing and in mast cell degranulation. Adhesion molecules are cell surface proteins involved in regulating the adhesion, migration and overall trafficking of lymphocytes necessary for their immunological function.

The basic structure of an antibody is a tetramer composed of two identical heterodimers each consisting of a light and a heavy chain. The antibody molecule is divided into variable (V) domains that are responsible for binding specificity and constant (C) domains that carry out various effector functions. The V domain is constructed of one V homology unit from both the light and heavy chains, designated V.sub.L and V.sub.H respectively. The variable region can be further subdivided into framework regions (FR) and complementarity determining regions (CDR), also designated as hypervariable regions. The FR maintain the structural integrity of the variable region domain. The CDR are the polypeptide segments within the variable region that mediate binding to the antigen. As the antibody contains two heterodimers, it has two antigen binding sites (divalent).

The T-cell antigen receptor is a heterodimer with one V domain involved in antigen recognition and one C domain that interacts with a membrane-bound protein complex, CD3, presumably involved in signal transduction. See for example Davis, Annu. Rev. Immunol., 3, 537- (1985); Kronenberg et al., Annu. Rev. Immunol., 4, 529- (1986). The majority of T-cells responsible for general antigen recognition use .alpha. and .beta. chains. A small subset of T-cells employ T-cell receptors composed of .gamma. and .delta. chains. As with antibodies, the V domain contains hypervariable regions which are the primary sequences responsible for antigenic interaction. Unlike antibodies, however, these hypervariable regions are not subject to somatic mutation during T cell selection and maturation.

Class I MHC molecules are transmembrane, glycosylated polymorphic polypeptide chains in close, non-covalent association with a .beta.2 microglobulin. The glycosylated polypeptide chain is divided into five distinct regions: three extracellular domains, a transmembrane region, and a cytoplasmic domain; the three extracellular domains designated .alpha.1, .alpha.2 and .alpha.3. Most of the sequence diversity of class I MHC molecules lies in the .alpha.1 and .alpha.2 domains with little sequence diversity in the .alpha.3 domain. The .alpha.1 and .alpha.2 domains interact to form a tertiary structure that is capable of binding a wide array of peptides. In this way, the .alpha.1/.alpha.2 interface is analogous to the V.sub.H /V.sub.L interface of antibodies of the V.alpha./V.beta. interface of the T-cell receptor.

Class II MHC molecules including the I-E, I-A molecules in the mouse, the HLA-DP, DQ and DR molecules in man and equivalent molecules in other mammals) are heterodimeric cell surface proteins composed of .alpha. and .beta. chain polypeptides. With the exception of I-E, HLA-DR and their equivalent genes in other organisms, both the .alpha. and .beta. chains of class II MHC proteins are highly polymorphic. Both chains have two extracellular domains, .alpha.1 and .alpha.2 for the .alpha. chain, and .beta.1 and .beta.2 for the .beta. chain. In each chain, there is also a transmembrane segment and a short intracytoplasmic domain. The .alpha.2 and .beta.2 resemble immunoglobulin constant region domains. The highest degree of sequence variation occurs in the .beta.1 domain with the .alpha.1 and .beta.1 domains forming the recognition site for binding antigen.

By using current techniques in molecular biology, it is possible to clone polypeptide chains. The more complex the molecules, e.g., the multichain immune receptors of the Ig superfamily, the more difficult they are to produce in a single foreign host. Genes coding for each chain of a complex molecule can be cloned and expressed in separate hosts but the aggregation and refolding of the resultant polypeptides into a biologically active entity is difficult to achieve. Multiple chains expressed by multiple genes within the same host have an advantage in aggregating and refolding into native structure but their expression in stoichiometric amounts is difficult to regulate. Thus, neither approach has proven to be efficient.

Since it is the variable regions of light and heavy chain antibodies that interact with an antigen, single chain polypeptides have been created with one V.sub.L and one V.sub.H joined by a peptide linker (U.S. Pat. No. 4,946,778). This single chain fragment (Fv) is univalent and is one of the smallest structures necessary for antigen binding activity having all 6 CDR regions of a V.sub.L -V.sub.H combination. Similarly, the formation of single chain polypeptides containing the variable domains of .alpha..beta. T-cell antigen receptors have also been reported. Novotny et al., Proc. Natl. Acad. Sci. USA, 88, 8646-8650 (1991); Soo Hoo et al., Proc. Natl. Acad. Sci. USA, 89, 4759-4763 (1992). The ability to produce as single chain polypeptides the antigenic or ligand binding portions of antibodies and .alpha..beta. T-cell antigen receptors reduces the difficulties discussed above which are encountered when producing multichain polypeptides.

The use of the functional smaller polypeptides which maintain an antigen recognition site are also advantageous when used for therapeutic purposes in mammals in that the smaller molecules are capable of rapidly localizing in the target tissue, such as single chain antibodies localizing in cancerous tissue or single chain T-cell receptors localizing to the inflammatory sites of patients suffering from such T-cell dependent pathological conditions as acute or chronic graft rejection, graft-vs-host disease, rheumatoid arthritis and other autoimmune diseases. Similarly, soluble single chain MHC molecules (perhaps bound to the disease-propagating antigens) could localize to and block pathologic antigen presentation at sites of T-cell dependent inflammation. However, while being able to penetrate the desired tissue, polypeptides with a molecular weight of less than about 50,000 daltons have the disadvantage of being retained within the glomerulus of the kidney. This is particularly disadvantageous when the polypeptide is bound with a carrier, such as a radioisotope or toxin, as the isotope or toxin will also accumulate in the kidneys.

Single chain polypeptides also suffer from the disadvantage of having only one binding site, thereby reducing their avidity.

It would therefore be advantageous to obtain constructions of the Ig superfamily which retain their antigen or ligand recognition properties, are of sufficient molecular weight to enhance retention at the target site with reduced retention in the kidney, and have a multiplicity of binding sites to enhance the avidity of the polypeptide. In addition, it would be beneficial if the single chain immunoglobulin superfamily proteins could be rendered bispecific to allow for recognition of different epitopes on the target tissue (or lymphocyte) or to allow for antibody-based recruitment of other immune effector functions (i.e., complement proteins, cytotoxic lymphocytes, etc.).

SUMMARY OF THE INVENTION

Surprisingly it has been found that single chains of the Ig superfamily will non-covalently associate to form dimers and multimers. In one embodiment, the invention is a protein comprising a dimer or multimer of single chain polypeptides wherein the polypeptides are non-covalently linked and each polypeptide comprises (a) a first polypeptide comprising an antigen binding portion of a variable domain from a member of the immunoglobulin superfamily; (b) a second polypeptide comprising an antigen binding portion of a variable domain from a member of the immunoglobulin superfamily; and (c) a peptide linker linking the first and second polypeptides (a) and (b) into a single chain polypeptide having affinity for the antigen. In particular, the invention is a protein comprising a dimer or multimer of single chain polypeptides wherein the polypeptides are non-covalently linked and each polypeptide comprises

(a) a first polypeptide comprising an antigen binding portion of a light chain variable region of an antibody, an .alpha.-chain variable region of a T-cell receptor or the .alpha.1 domain of either a class I or class II major histocompatibility complex molecule;

(b) a second polypeptide comprising the antigen binding portion of a heavy chain variable region of an antibody; a .beta.-chain variable region of a T-cell receptor or the .beta.1 domain of a class I or class II major histocompatibility complex molecule; and

In another aspect the present invention includes the aforementioned dimer or multimers conjugated to an imaging marker or therapeutic agent. The invention also includes a composition comprising the dimer or multimer conjugated to an imaging marker or therapeutic agent in a pharmaceutically acceptable carrier.

The invention is also directed to a method for in vivo diagnosis of cancer which comprises administering to an animal containing a tumor expressing a recognizable marker, a pharmaceutically effective amount of a composition containing a dimer or multimer of single chain polypeptides wherein the first and second polypeptide are specific for the recognizable marker.

In another embodiment, the invention is a method for intraoperative therapy which comprises (a) administering a pharmaceutically effective amount of the aforementioned composition containing di- or multimers of a single chain antibody against a recognizable marker to a patient containing a tumor expressing this recognizable marker, whereby the tumor is localized and (b) excising the localized tumor.

In another aspect, the invention is related to a method for the prevention and/or treatment of pathological T-cell dependent inflammatory disorders comprising administering to a patient a pharmaceutically effective amount of the aforementioned composition containing dimers or multimers of single chain antibodies, T-cell receptor or MHC molecules.

In still another embodiment, the invention concerns a process for preparing and expressing the aforementioned dimer and multimers of single chain polypeptides. Such an exemplary process comprises:

(i) providing a genetic sequence coding for a single chain polypeptide;

(ii) transforming a host T-cell with the genetic sequence;

(iii) expressing the genetic sequence in the host T-cell; and

(iv) recovery of single chain polypeptides which have non-covalently linked to form dimers and multimers.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 shows the digestion of pRW83 to produce vectors used in the construction of plasmid pSCFV31.

FIG. 2 shows the construction of plasmid pSCFV31.

FIG. 3, i.e. FIGS. 3A 3B, 3C, illustrates the nucleotide sequence (SEQ ID NO:1) and amino acid sequence (SEQ ID NO:2) of SCFVT.

FIG. 4 shows the construction of plasmid pSCFV UNIH.

FIG. 5, i.e. FIGS. 5A and 5B, 5C, 5D, illustrates the nucleotide sequence (SEQ ID NO:3) and amino acid sequence (SEQ ID NO:4) of Hum4 V.sub.L gene Cla I-Hind III segment in pRL1001.

FIG. 6, i.e. FIGS. 6A, 6B, 6C, illustrates the plasmid pSCFV VHH and plasmid pSCFV UHMH. (B) and (C) illustrate a schematic representation of a single chain Fv monomer and single chain Fv dimer, respectively. The darkened areas represent the complimentarily determining regions.

FIG. 7, i.e. FIGS. 7A and 7B, 7C, 7D, illustrates the nucleotide sequence (SEQ ID NO:5) and amino acid sequence (SEQ ID NO:6) of a Human Subgroup 4 variable light chain (H4V.sub.L) and a murine variable having chain single chain Fv immunoglobulin fragment.

FIG. 8, i.e. FIGS. 8A and 8B, 8C, 8D, illustrates the DNA sequence and amino acid sequence of CC49 SCFV species with a pI of 8.1(S1) (SEQ ID NO:7 and NO:8), 5.8(S2) (SEQ ID NO:9 and NO:10)and 5.2(S3) (SEQ ID NO:11 and NO:12).

FIG. 9, i.e. FIGS. 9A, 9B, 9C, shows the results of a cation exchange chromatograph activity, SDS-PAGE and Western Blot of E. coli pSCFV UHM8.1 periplasmic fraction.

FIG. 10, i.e. FIGS. 10A, 10B, 10C, shows the results of an anion exchange chromatograph, SDS-PAGE and Western Blot of Pool 2 (FIG. 9) of E coli pSCFV UHM8.1 periplasmic fraction.

FIG. 11, i.e. FIGS. 11A, 11B, and 11C, shows the results of a gel filtration chromatograph SDS-PAGE and isoelectric focusing (IEF) from fraction 1 of the anionic exchange (FIG. 10).

FIG. 12, i.e. 12A, 12B, 12C, shows the results of a gel filtration chromatograph, activity SDS-PAGE and Western Blot of E. coli pSCFV UHM5.2 periplasmic fraction.

FIG. 13 illustrates the results of a CC49 antibody competition based on moles of antibody binding sites.

FIG. 14 illustrates the results of a competition ELISA read one hour after addition of substrate.

FIG. 15, i.e.FIGS. 15A, 15B, 15C, 15D, 15E, and 15F, shows the construction of plasmids pP41, pP42, pP421, pP422, pP431 and pP432.

FIG. 16, i.e. FIGS. 16A and 16B, 16C, 16D, illustrates the DNA sequence and amino acid sequence of CC49 SCFV in pPY21(S1) (SEQ ID NO:14 and NO:15)and pPY22(S2) (SEQ ID NO:16 and NO:17).

FIG. 17 shows the relative activity of CC49 SCFV species obtained from Pichia pP431 and pP432 clones.

FIG. 18, i.e. FIGS. 18A, and 18B, shows the results of gel filtration and SDS-PAGE analysis of Pichia CC49 scFv species having a pI of 5.2.

FIG. 19, i.e. FIGS. 19A, and 19B, shows the results of an anion exchange chromatography and SDS-PAGE of Pichia CC49 scFv species having a pI of 5.2.

FIGS. 20A and 20B, shows the results of an isoelectric focusing and Western Blot of Pichia scFv species having a pI of 5.2.

FIG. 21, i.e. FIGS. 21A and 21B, illustrates the results of a competition ELISA based on (A) moles of antibody sites and (B) moles of antibody.

FIG. 22, i.e. FIGS. 20A and 20B, shows the construction of plasmids pATDFLAG and pSC49FLAG.

FIG. 23, i.e. FIGS. 23A and 23B, 23C, illustrates the nucleotide sequences (SEQ ID NO:19) and amino acid sequences (SEQ ID NO:20) in plasmid pATDOFLAG.

FIG. 24, i.e. FIGS. 24A and 24B, 24C, 24D, illustrates the nucleotide sequences (SEQ ID NO:21) and amino acid sequences (SEQ ID NO:22) in plasmid pSC49FLAG.

FIG. 25, i.e. FIGS. 25A, 25B, and 25C, shows the results of an M2 affinity chromatograph (A), analytical gel filtration (B), SDS-PAGE (C) and Western Blot of H4L 49HF scFv's(D).

FIG. 26, i.e. FIGS. 26A, 26B, 26C, and 26D, illustrates the results of binding of H4L49HF single chain species by TAG ELISA (A, B and C) or BSM ELISA (D).

FIG. 27 illustrates the results of a competition assay based on moles of antibody binding sites of H4L49F single chain species.

FIG. 28 illustrates the results of a competition assay based on moles of H4L49F single chain species.

FIG. 29 illustrates the plasmid pscDRB/.alpha.-FLAG.

FIG. 30, i.e. FIGS. 30A and 30B, 30C, 30D, illustrates the nucleotide sequence (SEQ ID NO:23) and amino acid sequence (SEQ ID NO:24) of the pscDR.beta.11/.alpha.-FLAG transcriptional unit.

FIG. 31, i.e. FIGS. 31A and 31B, 31C, 31D, illustrates the nucleotide sequence (SEQ ID NO:25) and amino acid sequence (SEQ ID NO:26) of the pscDR.beta.42/.alpha.-FLAG transcriptional unit.

FIG. 32, i.e. FIGS. 32A and 32B, 32C, 32D, illustrates the nucleotide sequence (SEQ ID NO:27) and amino acid sequence (SEQ ID NO:28) of the pscDR.beta.48/.beta.-FLAG transcriptional unit.

FIG. 33, i.e. FIGS. 33A and 33B, 33C, 33D, illustrates the nucleotide sequence (SEQ ID NO:29) and amino acid sequence (SEQ ID NO:30) of the pscDR.beta.41/.alpha.-FLAG transcriptional unit.

FIG. 34 illustrates the results of an ELISA showing the relative binding of 11 hybridoma supernatants to purified sdDRB11-FLAG protein.

DETAILED DESCRIPTION OF THE INVENTION

The entire teaching of all references cited herein are hereby incorporated by reference.

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.

As used herein, the term "non-covalently linked" means a polypeptide that will be separated on the basis of its molecular weight by size exclusion chromatography when done under non-reducing or reducing conditions, but will dissociate into two or more polypeptides when subjected to denaturing conditions such as heating and sodium dodecyl sulfate.

The term "single chain polypeptide" as used herein means a polypeptide from two original peptides linked by a peptide linker where the two original peptides are domains from the immunoglobulin superfamily and is capable of binding an antigenic determinant or receptor. "Domain" is a segment of protein that assumes a discrete function, such as antigen binding or antigen recognition.

The "immunoglobulin superfamily" means a family of proteins containing one or more domains that resemble an immunoglobulin domain. Proteins of the immunoglobulin superfamily include antibodies, T-cell receptors, CD1, CD2, CD3, CD7, CD8, Thy-1, Tp44, class I MHC molecules such as human leukocyte antigen (HLA)-A, -B, -C, -E; class II MHC molecules such as HLA-DP, -DZ, -DX, -DQ and -DR; immunoglobulin Fc receptors such as CD16, CD23, CDw32; and cell-surface adhesion molecules such as CD4, CD8 and the .beta.1 integrin family of proteins.

A "dimer" is two single chain polypeptides which are non-covalently linked and the term "multimer" means an even number of diners which are non-covalently linked. "Vector" means a DNA element used as a vehicle for cloning a fragment of foreign DNA.

The single chain polypeptides which form dimers and multimers in the present invention can be obtained from any member of the Ig superfamily. Preferably, the two original peptides in the single chain are selected from a light and heavy chain variable region of an antibody, and .alpha.1 and .alpha.2 domains of a class I MHC molecule or .alpha.1 and .beta.1 domains of a class II MHC molecule. More preferred are single chain polypeptides where the two original peptides in the single chain are selected from a light and heavy chain variable region of an antibody. Alternatively, it may be desirable to produce dimers and multimers of single polypeptides where the two original peptides are derived from the same domain (e.g., V.sub.L -linker-V.sub.L or V.sub.H -linker-V.sub.H). The present invention also includes the formation of dimers and multimers wherein one single chain polypeptide is directed against one antigen and the second or additional polypeptides are directed against a different antigen.

The dimers and multimers of the present invention formed by domains of the Ig superfamily are advantageous for therapeutic purposes in mammals as they retain their antigenic recognition properties, contain multiple binding sites and are of larger molecular weight than the single chain monomers thereby reducing their retention in the kidneys.

Isolation of an Appropriate DNA Sequence

To prepare a vector containing the DNA sequence for an antigen binding portion of an Ig superfamily protein, a source of the genes encoding for these regions will be 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 4th ed., (1991), published by The U.S. Department of Health and Human Services, discloses sequences of most of the Ab and TCR variable regions which have been described to date. Kabat et al. also discloses nucleotide and amino acid sequences of class I and class-II MHC proteins, CD4, CD8 and some Fc receptor sequences. A list of known human class II HLA sequences is contained in the report by Marsh, S. G. E. and Bodmer, J. G., entitled "HLA Class II Nucleotide Sequences" in Immunobiology, 182; 369-403 (1991). Up-to-date lists of other MHC sequences are also found in volume 182 of Immunobiology. WO 90/04410 published May 3, 1990 discloses a DNA sequence for variable regions of antibodies against tumor associated antigen (TAG).

When the genetic sequence is unknown, it is generally possible to utilize cDNA sequences obtained from mRNA by reverse transcriptase medicated 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 region or to obtain protein to determine amino acid sequence from the monoclonal antibody itself. The specificity of the antibody to be engineered will be determined by the original selection process. The class of antibody can be determined by criteria known to those skilled in the art.

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.

Unlike the immunoglobulin locus which contains in excess of 100 V.sub.L and V.sub.H germline genes and which undergoes extensive somatic mutation, the TCR V.beta. and V.alpha. loci are much less diverse (especially true in humans) and do not undergo somatic mutation. For this reason, the T-cell receptor V region genes which have been studied to date fall into discrete subfamilies which differ significantly (<80 percent sequence identity between different subfamilies) in sequence. For human V.beta., there appears to be approximately 25-30 such subfamilies. For human V.alpha. there appears to be slightly fewer subfamilies. One skilled in the art can use this sequence information to devise a number of assays for determining which T-cell receptor subfamilies are being expressed in the lymphocytes of a given sample. For example, oligonucleotide probes can be made which are specific for each of the subfamilies. By isolating lymphocytes from either peripheral blood or from a lymphocyte-rich infiltrate at an inflammatory site (i.e., synovial fluid in rheumatoid arthritis patients), one can isolate mRNA, reverse transcribe it to cDNA and use the subfamily-specific probes to detect the presence or absence of various T-cell receptor V regions in the cDNA from the site of interest. By such an approach it becomes possible to identify T-cell receptor subfamilies which may be important in the disease process.

After identifying a target T-cell receptor subfamily (or subfamilies), the V regions can be isolated using strategies essentially identical to those used for antibodies. For example, oligonucleotide primers can be synthesized which will allow the V regions of interest to be amplified, digested and ligated into an appropriate expression vector.

Class I and class II MHC molecules of interest are generally defined in the literature by epidemiological studies linking the presence of a given MHC molecule with a high incidence of a given disease. For example, rheumatoid arthritis is associated with a high incidence of HLA-DR.beta.1*0401, 0408 and 0101 alleles. In addition, one skilled in the art can use mixed lymphocyte reactions and similar assays to define MHC incompatibilities (i.e., target MHC molecules) in patients experiencing graft rejection or graft vs. host disease. Again, the MHC class II .alpha.1 and .beta.1 (or class 1 .alpha.1 and .alpha.2) domains can be selectively amplified either from cDNA of cell lines expressing the appropriate alleles or from peripheral blood of patients. Given the appropriate restriction sites on the PCR primers, the amplified cDNA can be inserted into the expression vector as with antibody V.sub.H s and V.sub.L s.

T-cell receptor cDNA and cDNA of Class I and Class II MHC can be obtained in a manner similar to that used for obtaining cDNA of the light and heavy variable regions of an antibody. For example, peripheral blood, various lymphoid tissues, or inflammatory filtrate samples from an appropriate vertebrate can be obtained, lymphocytes isolated and either used directly or allowed to proliferate transiently (24 hours to 6 weeks) in culture before use. Isolation of the desired lymphocytes can be accomplished by standard techniques known in the art, for example, affinity purification, density gradient centrifugation, or by use of a flourescence-activated cell sorter. From the cultured lymphocytes, mRNA can be obtained which in turn is utilized to produce cDNA. From the pool of cDNA, the desired cDNA can be amplified using such techniques as polymerase chain reaction. Polymerase chain reaction (PCR) or PCR with splicing by overlap extension (PCR-SOE). PCR in essence involves exponentially amplifying DNA in vitro using sequence specified 5' and 3' oligonucleotides. PCR is described in Mullins et al., Meth. Enz., 155, 335-350 (1987); and PCR Technology, Erlich (ed.) (1989).

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

To form the single chain polypeptides which form dimers and multimers it is necessary to have a suitable peptide linker which links the first and second polypeptides. Suitable linkers for joining the polypeptides are those which allow the separate polypeptide chains to fold into a single polypeptide chain which will have a three dimensional structure very similar to the original structure made of two polypeptide chains, and allows non-covalent linkage of single chains to form dimers and multimers of the single polypeptide chain. The linker must thus be (1) sufficiently large to span the distance between the two polypeptides; (2) sufficiently flexible to allow the association of the two polypeptides; and (3) relatively hydrophilic, being on the water-accessible surface of the molecule. Obtaining a similar three dimensional structure allows the resulting single chain polypeptide, dimer or multimer to recognize and react with the initial antigen or ligand molecule. Generally, a linker which allows separate polypeptide chains to fold into a single polypeptide chain with a three dimensional structure similar to the original structure made of two polypeptide chains should also allow formation of dimer and multimers. Linkers having the desired properties can be obtained by the method disclosed in U.S. Pat. No. 4,946,778, the disclosure of which is hereby incorporated by reference. From the polypeptide sequences generated by the methods described in the U.S. Pat. No. 4,946,778, genetic sequences coding for the polypeptide can be obtained.

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

A preferred linker is the helical linker designated 205 as disclosed in Pantoliano et al. Biochem., 30, 10117-10125 (1991).

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

Expression vehicles for production of the molecules of the invention include 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. coli is readily transformed using pBR322 [Bolivar et al., Gene, 2, 95- (1977), or Sambrook et al., Molecular Cloning, Cold Spring Harbor Press, New York, 2nd Ed. (1989)], a plasmid derived from an E. coli species which contains genes for ampicillin and tetracycline resistance, and thus provides an easy means for identifying transformed cells. Expression vectors compatible with procaryotic cells are well known in the art and are available from commercial sources, e.g., pBR325 from Gibco Biological Research Laboratories. Typical of vector plasmids suitable for procaryotic cells are pUC8, pUC9, pBR322 and PBR329 available from Biorad Laboratories (Richmond, Calif.) and pPL and pKK222 available from Pharmacia-LKB (Piscataway, N.J.), and the Bluesript SK/KS vectors from Stratagene (LaJolla, Calif.).

Plasmids suitable for eukaryotic microbes 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 Pichia pastoris, are available. Cultures of cells derived from multicellular organisms such as Sp2/0 or Chinese Hamster Ovary (CHO), which are available from the ATCC, may also be used as hosts. Typical of vector plasmids suitable for eukaryotic cells are pSV2neo and pSV2gpt (ATCC); pSVL and pKSV-10 (Pharmacia), pBPV-1/pML2d (International Biotechnology, Inc.).

The use of viral expression vectors to express the genes for polypeptides of the present invention is also contemplated. As used herein, the term "viral expression vector" refers to a DNA molecule that includes a promoter sequence derived from the long terminal repeat (LTR) region of a viral genome. Exemplary phage include .lambda.phage and fd phage (See, Sambrook et al., supra; McCafferty et al., Nature, 6301, 552-554 (1990).

It is preferred that the expression vectors and the inserts which code for the single chain polypeptides have compatible restriction sites at the insertion junctions and that those restriction sites are unique to the areas of insertion. Both vector and insert are 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 dimers and multimers 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 chain polypeptide out of the cytoplasm into either an intracellular compartment (in eukaryotes) or the cell exterior (or periplasm in prokaryotes). 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.

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.

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 by the host microorganism. Such 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 microorganisms would be obtained by culturing the microorganisms under conditions which require the induced phenotype for survival.

Recovery and purification of the dimers and multimers of the present invention can be accomplished using standard techniques known in the art. For example, if the polypeptides are excreted into the culture medium, the polypeptides can be concentrated by ultrafiltration and then separated from monomers by gel filtration. 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 affinity column chromatography or column chromatography using an ion exchange chromatography and gel filtration. Polypeptides which are insoluble and present as refractile bodies, also called inclusion bodies, can be purified by lysis of the cells, repeated centrifugation and washing to isolate the inclusion bodies, solubilization, such as with guanidine-HCl, and then passage of the solubilized material through a gel filtration column.

While not wishing to be bound by theory, it is believed the dimers and multimers of the present invention form as a result of two discrete physicochemical states (isoforms) of the single chain polypeptide. In one transition state (A) or folding intermediate, only monomers will be formed. In the second transition state (B) only dimers, or multiples thereof will be formed. The transition states are believed to be related to a conformational or charge equilibrium at a single residue or region in either polypeptide of the single chain polypeptide or even in the linker joining the two. During synthesis, since the single chain polypeptide has not yet assumed a final stable three dimensional structure, the single chain polypeptide is thought to be fairly random in configuration. Single chain polypeptides in this later transition state are capable of forming interchain dimers, a dimer being capable of non-covalently linking to another dimer or multimer.

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

Uses

The dimers and multimers of the present invention provide unique benefits for use in therapeutics, such as for use in a variety of cancer treatments and treatment of autoimmune diseases such as rheumatoid arthritis. In particular, the dimers and multimers are useful for, but not limited to, in vivo and in vitro uses in diagnostics, therapy, imaging and biosensors.

The dimers and multimers may be incorporated into a pharmaceutically acceptable carrier. Injectable compositions of the present invention may be either in suspension or solution form. In solution form the complex (or when desired the separate components) is dissolved in a pharmaceutically acceptable carrier. Such carriers comprise a suitable solvent, preservatives such as benzyl alcohol, if needed, and buffers. Useful solvents include, for example, water, aqueous alcohols, glycols, and phosphonate or carbonate esters. Such aqueous solutions generally contain no more than 50 percent of the organic solvent by volume.

Dimers and multimers composed of antigen specific antibody domains 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 dimers or multimers can be conjugated with an appropriate imaging or therapeutic agent. Examples of antibodies from which the V.sub.L and V.sub.H domains can be obtained to produce dimers or multimers are CC antibodies, such as CC49, disclosed in published PCT Application WO 89/00692 on Jan. 26, 1989 and published PCT Application WO 90/04410 on May 3, 1990.

Methods for preparing and administering conjugates of the dimers and multimers are accomplished by methods well known or readily determined, as described, for example, in Goldenberg et al., New England J. Med., 298, 1384-1388 (1978); Goldenberg et al.,Gastroenterol., 84, 524-532 (1983); Siccardi et al., Cancer Res., 46, 4817-4822 (1986); Keenan et al., J. Nucl. med., 25, 1197-1203 (1984); Meares et al., Anal. Biochem., 142, 68-78 (1984). Moreover, suitable dosages will depend on the age and weight of the patient and the therapeutic agent employed and are well known or readily determined.

Conjugates of the dimers and multimers and an imaging marker are administered in a pharmaceutically effective amount for the in vivo diagnostic assays, and then detecting the presence of the imaging marker by appropriate detection means. Generally, the dosage should be effective to visualize or detect tumor sites, distinct from normal tissues. Preferably, a one-time dosage will be between 0.1 mg to 200 mg of the conjugate of the dimer or multimer and imaging marker per patient.

Examples of imaging markers which can be conjugated to the dimers and multimers are well known and include substances which can be detected by diagnostic imaging using a gamma scanner or hand held gamma probe, and substances which can be detected by nuclear magnetic resonance imaging using a nuclear magnetic resonance spectrometer.

Suitable but not limiting examples of substances which can be detected using a gamma scanner include .sup.125 I, .sup.131 I, .sup.123 I, .sup.111 In, .sup.105 Rh, .sup.153 Sm, .sup.67 Cu, .sup.67 Ga, .sup.166 Ho, .sup.177 Lu, .sup.186 Re, .sup.188 Re and .sup.99m Tc. An example of a substance which can be detected using a nuclear magnetic resonance spectrometer is gadolinium.

A pharmaceutically effective amount of a composition containing dimers and multimers of antibodies is that which should be sufficient to achieve effective binding with the antigens against which the antibodies have specific affinity.

Examples of antibody-therapeutic agent conjugates which can be used in therapy include dimers or multimers of antibodies coupled to radionuclides, such as .sup.131 I, .sup.90 Y, .sup.105 Rh, .sup.47 Sc, .sup.67 Cu, .sup.212 Bi, .sup.211 At, .sup.67 Ga, .sup.125 I, .sup.186 Re, .sup.188 Re, .sup.177 Lu, .sup.99m Tc, .sup.153 Sm, .sup.123 I and .sup.111 In; to drugs such as methotrexate, adriamycin; to biological response modifiers, such as interferon and to toxins, such as ricin.

Methods of preparing and administering conjugates of the dimers and multimers of antibodies and a therapeutic agent are well known or readily determined. The pharmaceutical composition may be administered in a single dose or multiple dosage form. Moreover, suitable dosages will depend on the age and weight of the patient and the therapeutic agent employed and are well known or readily determined.

Dimers and multimers containing single chain antibodies are particularly suitable for radioimmunoguided surgery (RIGS). In RIGS, an antibody labeled with an imaging marker is injected into a patient having a tumor that expresses a target antigen. The antibody localizes to the tumor and is detected by a hand-held gamma detecting probe (GDP). The tumor is then excised (see, Martin et al. Amer. J. Surg., 156, 386-392 (1988); Martin et al., Hybridoma, 5, S97-S108). An exemplary GDP is the Neoprobe.TM. scanner, commercially available from Neoprobe Corporation, Columbus, Ohio.

Dimers and multimers of T cell receptors, MHC molecules and other members of the immunoglobulin gene superfamily are particularly suitable for immunotherapy of chronic inflammatory diseases such as chronic graft rejection, graft-vs-host disease, rheumatoid arthritis and other autoimmune diseases. In these diseases, dimers and multimers of the single chain proteins would have extended serum half-lives which would enhance the overall bioavailability of the therapeutic agent while simultaneously reducing the frequency with which the drug must be administered. In addition, the increased avidity achieved by multiple antigen binding sites should allow for prolonged duration of action (i.e., increased stability of the therapeutic at its desired site of action at the surface of the lymphocyte or antigen presenting cell).

The administration of compositions containing dimers and multimers of T cell receptors, MHC molecules and other members of the immunoglobulin gene superfamily are the same as those described above for single chain antibodies.

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

Antibodies

CC49: A murine monoclonal antibody specific to the human tumor-associated glycoprotein 72 (TAG-72) deposited as ATCC No. HB9459.

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

ChCC49: Chimeric CC49 has the murine constant regions of the CC49 antibody replaced with the corresponding human constant regions.

CC49 scFv(X): Single chain antibody fragment consisting of two variable domains of CC49 antibody joined by a peptide linker. X refers to the theoretical pI of the scFv (e.g., CC49 scFv8.1).

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.

Hum4 V.sub.L or H4VL: Variable domain of the human Subgroup IV kappa light chain.

SCFV UHM(X): SCFV consisting of a CC49 variable light chain and a CC49 variable heavy chain joined by a 25 amino acid linker named UNIHOPE (SEQ ID NO:31). (X) refers to the theoretical pI of the SCFV which changes as amino acids are added to the C terminus, e.g., scFv UHM8.1; scFv UHM5.2 and scFv UHM5.2.

SCFV UHH: SCFV consisting of a Hum4 V.sub.L and a CC49 variable heavy chain joined by a 25 amino acid linker named UNIHOPE (SEQ ID NO:31).

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), the disclosures of which are hereby incorporated by reference.

All water used throughout was deionized distilled water.

Oligonucleotide Synthesis and Purification

All oligonuclotides (oligos) were synthesized on either a Model 380A or a Model 391 DNA Synthesizer from Applied Biosystems (Foster City, Calif.) using standard .beta.-cyanoethyl phosphoramidites and synthesis columns. Protecting groups on the product were removed by heating in concentrated ammonium hydroxide at 55.degree. C. for 6 to 15 hours. The ammonium hydroxide was removed through evaporation and the crude mixtures were resuspended in 30 to 40 .mu.L 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-HC1, pH 7.4, 500 mM NaCl, 5 mM EDTA over 2 hours at 65.degree. C. Final purification was achieved by applying the DNA to Sep-Pac.TM. C-18 columns (Millipore, Bedford, Mass.) and eluting the bound oligos with 60 percent methanol. The solution volume was reduced to approximately 50 .mu.L and the DNA concentration was determined by measuring the optical density at 260 nm (OD.sub.260)

Bacterial Strains

Both Eschericia coli (E. coli) RR1 and E. coli GM 161 (dam.sup.31 ), containing the plasmid pRW83, were obtained from Dr. J. O. Lampen (Rutgers University, New Brunswick, N.J.) E. coli DHl was purchased from Invitrogen (San Diego, Calif.) and competent E. coli AGl was purchased from Stratagene (La Jolla, Calif.). The plasmid pCGS515 was obtained from Collaborative Research, Inc. (Bedford, Mass.).

Restriction Enzyme Digests

All restriction enzyme digests were performed using Bethesda Research Laboratories (Gaithersburg, Md.), or New England Biolabs, Inc. (Beverly, Mass.) enzymes and buffers following 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.degree. C. and then washed 3 times with 200 .mu.L of PBS, 0.05 percent Tween-20. 25 .mu.L of test antibodies and 25 .mu.L of biotinylated CC49 (1/20,000 dilution of a 1 mg/mL solution) were added to the wells and the plate incubated for 30 minutes at 31.degree. C. The relative amounts of TAG-72 bound to the plate, biotinylated CC49, streptavidin-alkaline 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 CC49 at 5 .mu.g/mL and CC49 Fab at 10 .mu.g/mL. Negative controls were 1 percent BSA in PBS and/or concentrated LB. Unbound proteins were washed away. 50 .mu.L of a 1:1000 dilution of streptavidin conjugated with alkaline phosphatase (Southern Biotechnology Associates, Inc., Birmingham, Ala.) were added and the plate was incubated for 30 minutes at 31.degree. C. The plate was washed 3 more times. 50 .mu.L of a para-nitrophenyl-phosphate solution (Kirkegaard & Perry Laboratories, Inc., Gaithersburg, Md.) were added and the color reaction was allowed to develop for a minimum of 20 minutes. The relative amount of scFv binding was measured by optical density scanning at 404-450 nm using a microplate reader (Molecular Devices Corporation, Manlo Park, Calif.). Binding of the scFv resulted in decreased binding of the biotinylated CC49 with a concomitant decrease in color development.

Amino-Terminal Sequence Analysis

Amino terminal amino acid sequencing was performed by the Edman degradation procedure using an Applied Biosystems model 470A Protein Sequencer (Foster City, Calif.). Fractions were analyzed with an interface to an Applied Biosystems Model 120A Analyzer (Rewrick, M. W. et al.). Protein samples (about 10 .mu.g each) were prepared by removing PBS present in the sample by 2 cycles of water dilution and concentration using a Centrican 30 device (Amicon).

SDS-PAGE and Western Blotting

Samples for SDS-PAGE analysis (20 .mu.L) were prepared by boiling in a non-reducing sample preparation buffer-Seprasol I (Integrated Separation Systems (ISS), Natick, Mass.) for 5 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 (ISS) at 55 mA per gel at constant current for approximately 75 minutes. Gels were stained in Coomassie Brilliant Blue R-250 (Bio-Rad, Richmond, Calif.) for at least 1 hour and destained. Molecular weight standards were prestained (Mid Range Kit, Diversified Biotech, Newton Center, Mass.) and included the following proteins: Phosphorylase b, glutamate dehydrogenase, ovalbumin, lactate dehydrogenase, carbonic amhydrase, B-lactoglobulin and cytochrome C. The corresponding MWs are: 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-HCl pH 10.4). An Immobilon-P PVDF (polyvinylidene dichlorine) membrane (Millipone, Bedford, Mass.) was treated with methanol for 2 seconds, and immersed in water for 2 minutes. The membrane was then equilibrated in anode buffer #1 for 3 minutes. A Milliblot-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 (25 mM Tris-HCl, pH 9.4 in 40 mM glycine). The screws for the cathode were firmly secured. 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)). TBS was purchased from Pierce Chemical (Rockford, Ill.) 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 1 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, 0.5 mL of Tween 20 (Sigma) was mixed per liter of TBS. The probe antibody used was 20 mL biotinylated FAID14 solution (10 .mu.g per 20 mL antibody buffer). Antibody buffer was made by adding 1 g BSA per 100 mL of 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 1:500 dilution in antibody buffer of streptavidin conjugated with alkaline phosphatase (Southern Biotechnology Associates, Birmingham, Ala.). The wash step was again repeated after this, as above. Prior to the color reaction, membranes were washed for 2 minutes in an alkaline carbonate buffer (20 mL). This buffer is 0.1 M sodium bicarbonate, 1 mM MgCl.sub.2. H.sub.2 O, pH 9.8. To make up the substrate for alkaline phosphatase, nitroblue tetrazolium (NBT) chloride (50 mg, Sigma) was dissolved in 70 percent dimethylformamide. 5-Bromo-4-chloro-3-indoyl phosphate (BCIP) (25 mg, Sigma) was separately dissolved in 100 percent dimethylformamide. 5-Bromo-4-chloro-3-indoyl phosphate (BCIP) 25 mg, Sigma) was separately dissolved in 100 percent dimethylformamide. These solutions are also commercially available as a Western developing agent sold by Promega. These solutions were stored at 4.degree. C. in the dark for up to 3 months. For color development, 120 .mu.L of 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 (IgG2a, K isotype) deposited as ATCC No. CRL 10256 directed against CC49. FAID14 was purified using a Nygene Protein A affinity column (Yonkers, N.Y.). The manufacturer's protocol was followed, except that 0.1 M sodium citrate, pH 3.0 was used as the elution buffer. Fractions were neutralized to pH .about.7 using 1.0 M Tris-HCl pH 9.0. The biotinylation reaction was set up as follows. FAID14 (1 mg, 100 .mu.L in water) was mixed with 100 .mu.L of 0.1 M Na.sub.2 CO.sub.3 pH 9.6. Biotinyl-.epsilon.-amino-caproic acid N-hydroxy succinimide ester (Biotin-X-NHS) (Calbiochem, LaJolla, Calif.) (2.5 mg) was dissolved in 0.5 mL dimethylsulfoxide. Biotin-X-NHS solution (20 .mu.L) was added to the FAID14 solution and allowed to react at 22.degree. C. for 4 h. Excess biotin and impurities were removed. by gel filtration, using a Pharmacia Superose 12 HR10/30 column (Piscataway, N.J.). At a flow rate of 0.8 mL/min, the biotinylated FAID14 emerged with a peak at 16.8 min. The fractions making up this peak were pooled and stored at 4.degree. C. and used to detect the CC49 idiotype as determined by the CC49 V.sub.L and V.sub.H CDRs.

Isoelectric Focusing (IEF) and Western Analysis

Isoelectric points (pI's) were predicted using a computer program called PROTEIN-TITRATE, available through DNASTAR (Madison, Wis.). 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 Bio-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, .beta.-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, 7.10 and 7.50, 7.80, 8.00, and 8.20 and 9.60 respectively. Gels were stained and destained according to the directions provided by FMC.

For IEF Westerns, duplicate gels were electrophoresed. Focused proteins in the second gel were transferred to PVDF membranes, as above, using a Biorad Model 483 slab dryer hooked up to a dry ice trap and vacuum pump. A piece of filter paper wetted in water was placed on the drier surface, followed by a prepared PVDF membrane, and then by the gel itself, with its solid support facing up. The plastic cover of the drier was then placed on top of this "sandwich". The drier was operated under vacuum without heat for 20 min to transfer the proteins in the gel to the membrane. Following the transfer, the blotted membrane was blocked and subsequently probed with biotinylated FAID 14 as described above for SDS-PAGE.

Quantitation of CC49 Antibody Species

Purified CC49 antibodies were quantitated by measuring the absorbance of protein dilutions at 280 mm using matching 1.0 mL 1.0 cm pathlength quartz cuvettes (Hellma) and a Perkin-Elmer UV/VIS Spectrophotometer, Model 552A. Molar absorptivities (E.sub.m) were determined for each antibody by using the following formula:

The values are based on information given by D. B. Wetlaufer, Advances in Protein Chemistry, 17, 375-378).

High Performance Liquid Chromatography

All high performance liquid chromatography (HPLC) was performed using an LKB HPLC system with titanium or teflon tubing throughout. The system consists of the Model 2150 HPLC pump, model 2152 controller, UV CORD SII model 2238 detection system set at an absorbance of 276 nm and the model 2211 SuperRac fraction collector.

Example 1

Formation of Single Chain Antibody Diners in E. coli

A. Preparation of pSCFV 31

A vector was prepared from plasmid pRW 83 containing a chloramphenicol resistance (Cam.sup.r) gene for clone selection, and a penP gene with a penP promoter and terminator (see Mezes, et al. (1983), J. Biol. Chem., 258, 11211-11218) and the pelB signal sequence [see Lei et al., J. Bact., 169, 4379-4983 (1987)]. The vector designated Fragment A (see FIG. 1) was prepared by removal of the penP gene with a Hind III/Sal I digest.

The penP promoter and pelB signal sequence were obtained by a PCR using pRW 83 as a template and oligonucleotides penP1 and penP2 as primers. The fragment was designated Fragment B (see FIG. 1). A Nco I enzyme restriction site was introduced at the 3' end of the signal sequence region by the penP2 oligonucleotide.

penP1 (SEQ ID NO:32):

5'-CGATAAGCTTGAATTCCATCACTTCC-3'

penP2 (SEQ ID NO:33):

5'-GGCCATGGCTGGTTGGGCAGCGAGTAATAACAATCCAGCG GCT GCCGTAGGCAATAGGTATTTCATCAAAATCGTCTCCCTCCGTTTGAA-3'

A scFv comprised of a Hum4 V.sub.L, a CC49 V.sub.H, and an 18 amino acid linker (Lys Glu Ser Gly Ser Val Ser Ser Glu Gln Leu Ala Gln Phe Arg Ser Leu Asp) was obtained from pCGS515/SCFV1 by PCR using oligonucleotides penP3 and penP6. This fragment was designated Fragment D (see FIG. 2). The complete nucleotide sequence (SEQ ID NO:1) and amino acid sequence (SEQ ID NO:2) of pCGS515/SCFV1 is given in FIG. 3. A Bcl I site was introduced at the 3' end of the V.sub.H region by the penP6 oligonucleotide.

penP3 (SEQ ID NO:34):

5'-GCTGCCCAACCAGCCATGGCCGACATCGTGATGACCCAGTCTCC-3'

penP6(-)(SEQ ID NO:35):

5'-CTCTTGATCACCAAGTGACTTTATGTAAGATGATGTTTTG ACG GATTCATCGCAATGTTTTTATTTGCCGGAGACGGTGACTGAGGTTCC-3'

Fragments B and D were joined by SOE-PCR (splicing by overlap extention-polymerase chain reaction) using oligonucleotides penP1 and penP6, following the procedures of Horton et al., Gene 77, 61-68 (1989). The new fragment was designated E (See FIG. 2).

Fragment C, containing the penP termination codon, was isolated by digesting pRW 83 with Bcl I and Sal I. pRW 83 was isolated from E. coli strain GM161, which is DNA methylase minus or dam.sup.-.

Plasmid pSCFV 31 (see FIG. 2) was created with a three part ligation of Fragments A, C, and E.

B. Preparation of pSCFV 31b

The Nco I restriction enzyme site within the Cam.sup.r gene and the Hind III site located at the 5' end of the penP promoter in pSCFV 31 were destroyed through a PCR DNA amplification using oligonucleotides Nco1.1 and Nco1.3(-) to generate an Eco RI to Nco I fragment and oligonucleotides Nco1.2 and Nco1.4c(-) to generate a Nco I to Eco RI fragment. These two fragments were joined by PCR-SOE using oligonucleotides Nco1.1 and Nco1.4c(-). The oligonucleotides are set forth below:

Nco1.1 (SEQ ID NO:36):

5'-TCCGGAATTCCGTATGGCAATGA-3'

Nco1.3(-)(SEQ ID NO:37):

5'-CTTGCGTATAATATTTGCCCATCGTGAAAACGGGGGC-3'

Nco1.2 (SEQ ID NO:38):

5'-ATGGGCAAATATTATACGCAAG-3'

Nco1.4c(-)(SEQ ID NO:39):

5'-CACTGAATTCATCGATGATAAGCTGTCAAACATGAG-3'

pSCFV 31 was digested with Eco RI and the larger fragment was isolated by polyacrylamide gel electrophoresis. To prevent self ligation, the DNA was dephosphorylated using calf intestinal alkaline phosphatase according to the teachings of Sambrook et al., supra.

A two part ligation of the larger pSCFV 31 digested fragment and the PCR-SOE fragment, described above, resulted in the creation of pSCFV 31b (see FIG. 4).

C. Preparation of pSCFV 33H

pSCFV 31b was digested with Nco I and Sal I and a fragment containing the Cam.sup.r gene was isolated.

The Hum4 V.sub.L was obtained by PCR DNA amplification using pCGS515/SCFV1 (See FIG. 3) as a template and oligonucleotides 104BH1 and 104BH2(-) as primers.

104BH1 (SEQ ID NO:40 ):

5'-CAGCCATGGCCGACATCGTGATGACCCAGTCTCCA-3'

104BH2 (-)(SEQ ID NO:41):

5'-AAGCTTGCCCCATGCTGCTTTAACGTTAGTTTTATCTGCTGG AGACAGAGTGCCTTCTGCCTCCACCTTGGTCCCTCCGCCGAAAG-3'

The CC49 V.sub.H was obtained by PCR using p49y1-2.3 (PCT WO 90/04410) as a template and oligonucleotides 104B3 and 104B4(-) as primers. A Nhe I enzyme restriction site was introduced just past the termination codon in the 3' end (before the Bcl I site) by oligonucleotide 104B4(-).

104B3 (SEQ ID NO:42):

5'-GTTAAAGCAGCATGGGGCAAGCTTATGACTCAGTTGCAGCAGTCTGACGC-3'

104B4 (-)(SEQ ID NO:43):

5'-CTCTTGATCACCAAGTGACTTTATGTAAGATGATGTTTTGACGGATT CATCGCTAGCTTTTTATTTGCCATAATAAGGGGAGACGGTGACTGAGGTTCC-3'

In the PCR which joined these two fragments using oligonucleotides 104BH1 and 104B4(-) as primers, a coding region for a 22 amino acid linker was formed.

A fragment C (same as above) containing the penP termination codon was isolated from pRW 83 digested with Bcl I and Sal I.

Plasmid pSCFV 33H (see FIG. 4) was created with a three part ligation of the vector, fragment C, and the SCFV fragment described above.

D. Preparation of pSCFV UNIH

To create a vector for any SCFV with the desired restriction sites in place, a plasmid designated UNIH was created by digesting pSCFV 33H with NcoI and NheI, and isolating the DNA fragment containing the Cam.sup.r gene.

Hum4 V.sub.L was obtained by PCR DNA amplification using pRL1001 as a template and oligonucleotides UNIH1 and UNIH2(-) as primers. The DNA sequence (SEQ ID NO:3) in pRL1001 coding for the Hum4 V.sub.L is given in FIG. 5. Oligonucleotides for the PCR were:

UNIH1 (SEQ ID NO:44):

5'-CAGCCATGGCCGACATTGTGATGTCACAGTCTCC-3'

The Nco I site is in bold and the hybridizing sequence is underlined.

UNIH2(-)(SEQ ID NO:45):

5'-GAGGTCCGTAAGATCTGCCTCGCTACCTAGCAAA AGGTCCTCAAGCTTGATCACCACCTTGGTCCCTCCGC-3'

The Hind III site is in bold.

The CC49 V.sub.H was obtained by a PCR using p49y1-2.3 as a template and oligonucleotides UNI3 and UNI4(-) as primers.

UNI3 (SEQ ID NO:46):

5'-AGCGAGGCAGATCTTACGGACCTCGAGGTTCAGTTGCAGCAGTCTGAC-3'.

The Xho I site is in bold and the hybridizing sequence is underlined.

UNI4(-)(SEQ ID NO:47):

5'-CATCGCTAGCTTTTTATGAGGAGACGGTGACTGAGGTTCC-3'.

The Nhe I site is in bold and the hybridizing sequence is underlined.

Oligonucleotides UNIH1 and UNI4(-) were used in the PCR-SOE amplification which joined the Hum4 V.sub.L and CC49 V.sub.H fragments and formed a coding region for a negatively charged fifteen amino acid linker. The DNA was digested with Nhe I and Nco I and ligated with the vector fragment from the Nco I-Nhe I digest of pSCFV 33H. The resultant plasmid was designated pSCFV UNIH (see FIG. 4).

With the construction of pSCFV UNIH, a universal vector for any SCFV was created with all the desired restriction enzyme sites in place.

E. Preparation of pSCFV UHH and pSCFV UHM

pSCFV UNIH was digested with Hind III/Xho I, and the large DNA fragment containing the Cam.sup.r gene, Hum4 V.sub.L and CC49 V.sub.H was isolated.

A fragment coding for a 25 amino acid linker, UNIHOPE, was made by annealing the two oligonucleotides shown below. Annealing conditions were as follows: 10 .mu.g of each oligo were added to Bethesda Research Laboratories (Gaithersburg, Md.) enzyme Buffer #2 for a total volume of 53 .mu.L, the mixture was heated to 98.degree. C. for one minute, cooled slowly to 40.degree. C. and placed at 4.degree. C. overnight. Following digestion with Hind III/Xho I, the linker was electrophoresed through a 6 percent PAG. The DNA bands were excised from the gel and the DNA was electroeluted for five minutes at 200 volts. The DNA was pelleted by ethanol precipitation and dissolved in water. The linker UNIHOPE is based on 205C SCA.TM. linker (see Whitlow et al., Methods: A Companion to Methods in Enzymology, 2, 97-105 (1991), but the first amino acid was changed from serine to leucine and the twenty-fifth amino acid were changed from glycine to leucine, to accommodate the Hind III and Xho I restriction sites, respectively. The nucleotide sequence encoding the linker UNIHOPE is set forth below:

UNIHOPE (SEQ ID NO:48)(Top Strand):

5'-TATAAAGCTTAGTGCGGACGATGCGAAAAAGGATGCTGCGAAG AAGGATGACGCTAAGAAAGACGATGCTAAAAAGGACCTCGAGTCTA-3'

UNIHOPE(-)(SEQ ID NO:49) Bottom Strand:

5' TAGACTCGAGGTCCTTTTTAGCATCGTCTTTCTTAGCGT CAT CCTTCTTCGCAGCATCCTTTTTCGCATCGTCCGCACTAAGCTTTATA-3'

The resulting annealed strand was digested with Hind III/Xho I and ligated into the vector, thus generating the plasmid pSCFV UHH (shown in FIG. 6). Plasmid pSCFV UHH expresses a biologically active, TAG-72 binding SCFV consisting of the Hum4 V.sub.L and CC49 V.sub.H. The expression plasmid utilizes the .beta.-lactamase penP promoter, pectate lyase pelB signal sequence and the penP terminator region. Different immunoglobulin light chain variable regions can be inserted in the Nco I-Hind III restriction sites, different SCFV linkers can be inserted in the Hind III-Xho I sites and different immunoglobulin heavy chain variable regions can be inserted in the Xho I-Nhe I sites.

F. Preparation of pSCFV UNIM and pSCFV UHM

pSCFV UNIM was created in conjunction with pSCFV UNIH. It differs in that the DNA sequence coding for the CC49V.sub.L is incorporated into UNIM as opposed to the DNA sequence coding for the Hum4 V.sub.L. The sequence for murine CC49V.sub.L is disclosed in PCT WO 90/04410, published May 3, 1990, and was produced by PCR using the oligos UNIM1 and UNIM2(-) as primers. The CC49 V.sub.H and the vector fragment were obtained by the manner described above for pSCFV UNIH. The SCFV insert was generated by PCR-SOE using oligos UNIM1 and UNI4(-) to join the light and heavy chains and form a coding region for a negatively charged fifteen amino acid linker. Following NheI and NcoI digestion and purification of the vector and insert, a ligation of the two pieces produced pSCFV UNIM.

UNIM1 (SEQ ID NO:50):

5' CAGCCATGGCCGACATTGTGATGTCACAGTCTCC-3'

UNIM2(-)(SEQ ID NO:51):

5' GAGGTCCGTAAGATCTGCCTCGCTACCTAGCAAAA GGTCCTCAAGCTTCAGCACCAGCTTGGTCCCAGCAC-3

pSCFV UHM was generated by digesting pSCFV UNIM with HindIII and XhoI, isolating the vector fragment with the Cam.sup.r gene by 4 percent PAG, and ligating the twenty five amino acid linker, UNIHOPE, into the vector DNA fragment piece. The procedure was the same as described for pSCFV UHH. pSCFV UHM contains DNA sequence coding for an scFv having the CC49 V.sub.L and CC49 V.sub.H with a computer predicted pI of 8.1. Hence, this plasmid was named pSCFV UHM8.1. Hence, this plasmid was named pSCFV UHM8.1

The plasmid map of pSCFV UHH and pSCFV UHM are illustrated in FIG. 6A. For pSCFV UHH, V.sub.L is Hum4 V.sub.L, the linker (L) is UNIHOPE and V.sub.H is CC49 V.sub.H. For the pSCFV UHM series, V.sub.L is CC49 V.sub.H, the linker is UNIHOPE and V.sub.H is CC4.sup.9 V.sub.H. For the pSCFV UHM(X) series, X indicates the pI of the monomer product (pI=8.1, 5.8 or 5.2) as described herein. FIGS. 6B and 6C illustrate the monomeric (scFv) and dimeric (Fv2) forms, respectively, of the products. The CDR regions are indicated schematically by the black bands.

The DNA (SEQ ID NO:5) and amino acid (SEQ ID NO:6) sequences of the scFv obtained from pSCFV UHH constructions are given in FIG. 7. The UNIHOPE linker sequence is indicated in bold and underlined, while the CDR regions are in bold and italicized. The amino terminus of the mature scFv product is indicated by +1.

E. coli AG1 (Statagene, La Jolla, Calif.) cells were transformed with pSCFV UHH and pSCFV UHM8.1 by following the manufacturer's instructions. Competent cells were thawed on ice and gently mixed by hand before 50 .mu.L aliquots were dispensed into 15 mL polypropylene tubes. A fresh dilution of .beta.-mercaptoethanol was added to the tubes for a final concentration of 25 mM. The cells were swirled gently every two minutes for ten minutes. Either 1 .mu.L or 5 .mu.L of the prepared DNA was added to the tubes. The cells and DNA sat on ice for 30 minutes, were heat pulsed in a 42.degree. C. water bath for 45 seconds and put back on ice for two minutes. SOC medium (450 .mu.L) (medium containing bacto-tryptone, bacto-yeast extract, NaCl MgCl.sub.2 and glucose) was added to each tube and the tubes were incubated at 37.degree. C. for one hour with shaking. Either 100 .mu.L or 200 .mu.L of transformed cells were plated on LB agar containing 20 .mu.g/ml chloramphenicol. Plates were incubated at 37.degree. C. overnight.

From each transformation, twelve to eighteen bacterial colonies were selected for overnight growth at 37.degree. C. in 2 mL of LB/CAM 20. A portion of the overnight culture (1.2 mL) was used for a small scale plasmid preparation following the boiling method adapted from Holmes and Quigley (Sambrook et al, 1989, supra). Plasmid DNA was digested with restriction enzymes and the resultant fragments were size analyzed by agarose gel electrophoresls using ethidium bromide and long wavelength UV for visualization. The remaining 0.8 mL of overnight culture was sonicated (Soniprep 150, MSE Ltd., Sussex), then microcentrifuged for 3 minutes to pellet the debris. The supernatant was reduced in volume by 50 to 80 percent using a Centricon 10 microconcentrator device (Amicon, Beverly, Mass.) and tested for biological activity against TAG-72.

(G) Construction of pSCFV UHM5.8 and pSCFV UHM5.2

Two Plasmids, pSCFV UHM5.8 and pSCFV UHM5.2 were constructed to express CC49 scFvs and CC49 Fv2s having computer predicted pI's of 5.8 and 5.2, respectively. Extra negatively charged amino acids were added to the carboxyl terminus of the CC49 V.sub.H, by synthesizing two oligonucleotides (oligos) and using them for a polymerase chain reaction (PCR) amplification of the CC49 V.sub.L -linker-CC49 V.sub.H DNA present in pSCFV UHM8.1. These oligos for the 3' end of the target are as follows. For the 5.8 construction, a 58-mercalled SCUHPM 5.8 was made. For the 5.2 construction, a 64-mercalled SCUHPM 5.2 was made.

SCUHPM 5.8 (SEQ ID NO:52)

5'-AACA GCTAGC TTT TTA GGA GTC ATA GTC CTC AGG GGA GAC GGT GAC TGA GGT TCC TTG-3'

SCUHPM 5.2 (SEQ ID NO:53)

5'- AACA GCTAGC TTT TTA CTC ATA CTC TTC AGG GTC TTC AGG GGA GAC GGT GAC TGA GGT TCC TTG -3'

The nucleotides representing Nhe I restriction sites are indicated in bold. These oligos were purified from a 14 percent polyacrylamide gel containing urea (Sambrook et al., supra). The 5' oligo for the PCR was a 34 mer designated UNIMl (sequence given above).

The nucleotides representing an Nco I restriction site are indicated in bold. For the PCR's, 100 pmol of the 5' (UNIMl) and 3' (SCUHPM 5.8) oligos were used to generate the CC49 scFv 5.8 DNA product insert, while 100 pmol of each of UNIM1 and SCUHPM 5.2 were used to generate the CC49 scFv 5.2 DNA product insert. The target DNA for the oligos in the PCR reaction (100 .mu.L) was 1.0 ng of pSCFV UHM8.1. The reaction was performed using a programmable thermal controller (MJ Research Inc., Watertown, Mass.) and Taq polymerase according to the directions provided by Perkin Elmer Cetus (Norwalk, Conn.). There were 30 cycles consisting of 40 seconds at 94.degree. C., 70 seconds at 45.degree. C. and 75 seconds at 72.degree. C. For the final cycle, the temperature was held at 72.degree. C. for 2 minutes. The PCR products were analyzed on a 1.0 percent agarose gel containing ethidium bromide and found to be the expected sizes of 788 bp and 794 bp for the 5.8 and 5.2 CC49 scFv constructions, respectively. The DNA was preparatively purified from a 4 percent polyacrylamide gel, digested with Nco I (from BRL, Gaithersburg, Md.) and Nhe I (from New England Biolabs) using New England BioLabs 10.times. buffer #2 (10 mM Tris-Hcl, 10 mM MgCl.sub.2, 50 mM NaCl, 1 mM OTT, pH 7.9).

The resulting trimmed Nco I and Nhe I oligonucleotides fragment of the correct size were purified away from the DNA inserts by electrophoresis in a 4 percent polyacrylamide gel. The CC49 scFv inserts were electroeluted and ethanol precipitated (Sambrook). The vector for ligation with these inserts had been prepared from pSCFV33H digested with Nco I and Nhe I. Competent E. coli AG1 cells (Strategene, La Jolla, Calif.) were transformed with the ligated DNA and plated on LB plates containing 20 .mu.g/mL chloramphenicol.

Four clones were chosen from each transformation to prepare miniprep plasmid DNA (Sambrook) for restriction enzyme analysis. Based on these results single clones were chosen; E. coli AG1/pSCFV UHM5.8 E. coli and AG1/pSCFV UHM5.2.

The DNA and amino acid sequences of the CC49 species, pSCFV UHM8.1 (SEQ ID NO:7 and NO:8) (S1), 5.8 (SEQ ID NO:9 and NO:10) (S2) and 5.2 (SEQ ID NO:11 and NO:12) (S3), are given in FIG. 8.

Preparation of Total Cellular Protein and Escherichia Coli Periplasimic Shockates

For analysis of total cell CC49 scFv activity or preparation of samples for sodium dodecyl sulfate polyacrylamide gel electrophonesis (SDS-PAGE), E. coli (pSCFV UHM8.1; pSCFV 5.8; or pSCFV 5.2) liquid cultures (2.5 mL LB supplemented with 20 .mu.g/mL chloramphenicol) were grown overnight at 37.degree. C. Cells from 1.5 mL of culture were pelleted in a microcentrifuge, the supernatant discarded, and the cell pellets resuspended in 0.5 mL phosphate buffered saline (PBS) (Sigma Chemical). Sample tubes were placed in an ethanol ice bath before sonication (MSE Soniprep 150, UK) using 3 cycles of disruption at 14 microns amplitude for 15 seconds each and cool down for 45 seconds in between. The sonicated cell debris was pelleted in a microcentrifuge, and the supernatant, representing the total, soluble protein fraction, was filtered through a Millex-GV 0.22 .mu. filter disc (Millipore, Bedford, Mass.).

For purification of the CC49 scFv.sup.(X) derivatives, where x=8.1, 5.8 and 5.2, E. coli periplamsic fractions were prepared from 1.0 L overnight cultures. A 1.0 L culture was divided into 4.times.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 NaCl. The cells were again pelleted at 5,000 rpm for 10 minutes. Cells were washed with a total of 100 mL of 30 mM Tris-HCl (pH 7.3) and pooled into 1 tube. To this was added 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. The mixture was kept at ambient temperature, with occasional shaking for 10 minutes. The hypertonic cells were centrifuged as above. The cells where then quickly resuspended in ice cold 0.5 mM MgCl.sub.2 and the suspension kept on ice for 10 minutes with occasional shaking. The cells were centrifuged as above and the shockate, containing the E. coli periplasmic fraction was further clarified by passing through a 0.2 .mu. filter apparatus (Nalge Co., Rochester, N.Y.). This material was concentrated using Centriprep 30 and Centricon 30 ultrafiltration devices (Amicon Beverly, Mass.) to a final volume of 1-2 mL.

For isolation and analysis of just cytoplasmic proteins, previously shocked E. coli cells (0.5-1.0 mg--wet weight) were resuspended in 0.5 mL PBS and sonicated as above. The cytosol was recovered by microcentrifugation and filtered through a 0.2 .mu. filter device.

Prior to further purification, the shockate was dialyzed against 50 mM sodium acetate buffer, pH 4.95, using a System 500 microdialyzer unit (Pierce Chemical, Rockford, Ill.), with 3-4 exchange of buffer. An 8,000 molecular weight (MW) cut-off membrane was used.

Purification of CC49 Single Chain Monomer and Dimer Species

The general purification of the single chain antibody species is given in Scheme 1. The procedure under column 1 was initially used for purification of antibodies from pSCFV UHM8.1 and was later modified as given under column 2. Single chain antibody species derived from pSCFV UHM5.2 were purified as given under column 3. Final purification of the pSCFV UHM5.8 products was not completed.

A. CC49 scFv 8.1

For single chain antibody species from pSCFV UHM8.1, about half of the E. coli AG1/pSCFV UHM8.1 periplasmic shockate prepared from 1 L of culture as described above was dialyzed overnight against 50 mM sodium acetate buffer, pH 4.95. Precipitated proteins were pelleted by microcentrifugation and the supernatant was applied to a cation exchange column (Mono S HR5/5 HPLC column, Pharmacia, Piscataway, N.J.). A linear gradient program utilizing 50 mM sodium acetate pH 4.95 as buffer A and 50 mM sodium acetate pH 4.95 containing 0.3 M NaCl as buffer B was run as follows:

When the periplasmic fraction was chromatographed over Mono S, two areas of activity were observed (FIG. 9) and the two pools kept separate for the next step. The lanes 1-9 for SDS-PAGE correspond to a crude periplasmic fraction (lane 1) and fractions 1, 2, 3, 12, 16, 21, 26 and 43, lanes 2-9 respectively, with a Western transfer of the samples from a duplicate gel, indicate product in fractions 26 and 43 (lanes 17 and 18).

Pool 2 from the Mono S purification was concentrated to below 250 .mu.L using a Centricon 30 device and dialyzed against buffer A (20 mM Tris-HCl, pH 7.60) for rechromatography on the Mono Q HR5/5 anion exchange column (Pharmacia, Piscataway, N.J.). Buffer B was the same as buffer A, but in 0.5 M NaCl. The gradient program used was the same as for the Mono S run above, but for this Mono Q run the Mono S pool 2 sample was changed to 100 percent B the gradient was at 16 percent B.

The anion exchange chromatography and SDS-PAGE of the Mono S Pool 2 sample is given in FIG. 10. The arrow on the chromatogram indicates the point in the gradient of Buffer B (16 percent) where the percent of B was changed to 100 and held for 5 minutes. Lanes 1-9 correspond to fractions 1-3 of a similar chromatography (not shown) done on Pool 1 (FIG. 9); a sample of Pool 1; a sample of Pool 2; and fractions 1-4 from the anion exchange chromatography of Pool 2. Lanes 10-18 represent the same fractions as in lanes 1-9 in a Western transfer from a duplicate gel.

The final step in the purification utilized a Superdex 75 HR10/30 gel filtration column (Pharmacia). Potassium phosphate buffer (0.175 M, pH 6.85) was used as the mobile phase, at a flow rate of 0.2 mL/min. Fraction 1 from the Mono Q purification was applied to the Superdex column and the resulting chromatogram is shown in FIG. 11A-1. Both peaks emerging at 42 and 50.5 minutes were active in the competition assay, with considerably more activity being observed in the first peak (42 minutes). On SDS-PAGE and Western transfer (FIG. 11B) a band for the correct molecular weight of monomeric single chain antibody was observed. Lanes B1 and 8 are ChCC49 Fab; lanes 2 and 9, fraction 14; lanes 3 and 10, fraction 15; lanes 4 and 11 fraction 16; lanes 5 and 12, fraction 17,; lanes 6 and 13, fraction 18; and lanes 7 and 14, fraction 19. Application of a ChCC49 Fab fragment to the gel filtration column (FIG. 11A-2), using the same conditions as for the scFv sample, indicated that the scFv peak emerging at 42 minutes contained a protein that was larger than the Fab, which emerged at 45.8 minutes. When the peak emerging at 42 minutes was subjected to the reducing condition of SDS-PAGE, a monomeric unit on the gel and Western was seen, indicating a con-covalently linked dimeric (Fv2) form of the single chain antibody. Further evidence of a unique dimeric entity was obtained when an IEF gel was run.

IEF of fractions 15 (lanes C1 and C3) and 17 (lanes C2 and C4), indicates there is a distinct difference between the pI of the Fv2 (lanes 1 and 3) and the monomeric scFV (lanes 2 and 4). Samples in lanes 1 and 2 were stained with Coomassie Brilliant Blue R-250, while lanes 3 and 4 represent a Western transfer of the corresponding lanes of a duplicate IEF gel. The faint FAID14 positive band in FIG. 11C, lane 3, above the main Fv2 band is not known. It could be a trace degradation product or some higher multimer.

A second purification system was then developed to optimize the purification of the monomer, dimer and multimer species of the pI 8.2 single chain antibodies. The procedure is outlined in column two of Scheme 1. The buffer system for purification with the Mono S cation exchange column was also changed to 50 mM sodium acetate pH 5.2 and 50 mM 2-(N-morpholino)ethane sulfonic acid (MES).

The protocol which afforded pure CC49 scFv and Fv2 8.2 was as follows: A periplasmic fraction was dialyzed and applied to mono S using a 50 mM sodium acetate buffer system, as before, but at pH 5.20. At this pH less protein precipitated than at pH 4.95 but the results were much the same in terms of the chromatography. Next, the Mono Q column was utilized as above. After this, a second Mono S run was introduced for the active Mono Q fractions. The buffer system used was: buffer A (50 mM MES, pH 5.6) and buffer B being the same as buffer A in 0.5 M NaCl. The monomer and dimer were finally purified using the Superdex 75 HR10/30 column and PBS as the buffer.

B. CC49 scFv5.2 and Fv2-5.2

The periplasmic shockate from a 1.0 liter culture was prepared as described above. The strategy for purification used here was to separate the monomer from the dimer in a gel filtration step first. A Superdex 75 prep grade HR 26/60 preparative column was used. The concentrated crude sample (0.5 mL) was chromatographed at a flow rate of 1.3 mL/min. using 0.175 M potassium phosphate buffer pH 6.8. See FIG. 12 for the chromatogram. The elution times for the molecular weight standards: (1) bovine IgG (Pharmacia), MW=153 Kd; (2) chicken ovalbumin, MW=44 Kd; and (3) horse myoglobin, MW 17 Kd from a separate run are as indicated by the arrows.

SDS-PAGE analysis of the fractions was as follows: lanes 1 and 14, ChCC49 Fab; lanes 2-13 correspond to fractions 3, 8, 10, 12, 14, 16, 18, 19, 20, 22, 23 and 24 respectively and lanes 15-24 correspond to fractions 25-33 and 35, respectively.

Both the resulting dimer and monomer pools were concentrated to 300 .mu.L with centricon 30 devices and dialyzed overnight against 20 mM (1,3-bis[tris(hydroxymethyl)-methylamino]propane (Bis-Tris propane)-Cl pH 6.8 buffer and applied to the Mono Q HR5/5 column for the monomer and dimer purification. The same gradient program was used as described earlier with buffer B being 20 mM Bis-Tris propane-Cl pH 6.8 in 0.5 M NaCl. The active Mono Q fractions were pooled and concentrated to 130 .mu.L. The sample was applied to the Superdex 75 HR 10/30 column to verify purity of the monomer and exchange the buffer to PBS.

Antibody Competition

Using purified monomer and dimer, it was possible to quantitate them accurately (Table 1) and compare them directly in a competition assay to a Fab, which is monovalent, and the CC49 whole antibody, which is divalent. The results of this study are given in FIG. 13. The results indicate the diner and whole IgG have similar activity. These results further support the data of the existence of the dimer, indicating the dimer is folding in a way that gives two active, accessible binding sites.

The activity of the purified CC49 scFv5.2 was compared with the ChCC49 Fab in a competition assay (FIG. 14) and the scFv was found to be equivalent to the Fab. This indicates one can add certain features to the carboxyl terminus of a scFv construction, as in this case the addition of Pro-Glu-Asp-Pro-Glu-Asp-Tyr-Asp, that do not affect the binding site. Therefore, other modifying sequences can be contemplated which alter the overall characteristics of the molecule.

Amino-Terminal Amino Acid Analysis

The amino terminal sequencing data gave the following result: NH.sub.2 -?-I-V-M-S-Q-S-P-S?-T?-L. The first residue couldn't be unequivocally assigned, but its positioning indicates that signal peptidase had cleaved at the end of the pel B signal sequence, and released a correctly processed mature product. The residues indicated at positions 2-11 correspond to the sequence predicted by the DNA sequence (See FIG. 8). The yields for Ser and Thr at positions 9 and 10 were low, but in general, Ser and Thr yields are known to be lower than for the other amino acids. Based on these data, one can conclude that the pel B signal sequence is effective in directing the secretion of the single chain products to the E. coli periplasm. It is apparent that one need not be limited to using the pel B leader, and that the same results, as far as processing and secretion of the product is concerned, can be obtained by using other signal sequences such as omp A and pen PC, and others. One can include these sequences in the PCR oligo when making the construction upstream from the mature protein.

Temperature Stability Study of CC49 Fv2 8.1

Approximately 3 .mu.g aliquots of CC49 Fv2 8.1 kept at 37.degree. C. were analyzed at intervals over 24 hours by gel filtration chromatography. A Superdex 75 HR10/30 column was utilized, using PBS as the mobile phase, at a flow rate of 0.5 mL/min.

The results of samples taken at various time points over 24 hours at 37.degree. C. indicated that the Fv2 remained intact (eluted at the same position as the starting material), and that no degradation products were observed.

Example 2

Cloning of CC49 Single Chain Antibody Species into the Yeast Pichia pastoris

For expression of single chain antibody species in yeast, the strain Pichia pastoris GTS115 HIS4 (NRRL Y-15851) carrying a defective histidine dehydrogenase gene was used as a host cell. The plasmid pHIL-S1, available from Phillips Petroleum Company, was used as the expression plasmid modified as described below. The original sequence (SEQ ID NO:54) and features in the 5' region of pHIL-S1 are as follows:

DNA Sequencing

The dideoxy chain termination method of Sanger (Sambrook et al. supra) was used for DNA sequencing of the CC49 scFv genes present in pPY21 and pPY22. A Sequenase.TM. sequencing kit (United States Biochemical, Cleveland, Ohio) was used for the sequencing reactions, according to the manufacturer's directions. The radiolabel used was a-P32 dATP (3,000 Ci/mmol) obtained from Amersham (Arlington Heights, Ill.), lot no. AC9048. Double loadings were done using a BRL Model S2 sequencing apparatus and a premixed 6% Gel-Mix sequencing gel (BRL). After electrophoresis, gels were dried for 30 min using a Model 483 Slab Gel Dryer (Biorad, Richmond, Calif.) and exposed to Kodak X-OMat film (Sigma Chemical, St. Louis, Mo.) for 1-24 h. The sequencing oligonucleotides used for priming (0.5 pmol each) were:

1) AOX1P (SEQ ID NO:13) (23-mer), 5'-TTTAACGACAACTTGAGAAGATC-3';

2) TAGVLFR2 (SEQ ID NO:55) (20-mer), 5'-TGGTACCAGCAGAAACCAGG-3';

3) TAGVLCDR3 (SEQ ID NO:56) (Mixed 22-mer), 5'-GTCAGCA(AG)TATTATAG(CT)TATCC-3' and;

4) TAGVHCDR2 (SEQ ID NO:57) (21-mer), 5'-ATGGATTGGATATTTTTCTCC-3'

Construction of pPY1

A clone of E. coli AG1 transformed with pHIL-S1 was used to isolate purified pHIL-S1 plasmid using Qiagen's midi plasmid prep procedure (Chatsworth, Calif.). Three micrograms of purified pHIL-S1 were treated with Sst I (Biological Research Laboratories, BRL) (10 units) in a volume of 40 .mu.L for 45 minutes and then 2 .mu.L of 1.5 M NaCl and 20 units (2 .mu.L) of Sal I (BRL) were added (see FIG. 15A). After the reaction proceeded for about 45 minutes, the sample was loaded onto a (16.times.18 cm) 4 percent polyacrylamide gel (PAG).

In a separate reaction, pBluescript II KS+ (5 .mu.g) purchased from Stratagene, was treated with Sst I (12.5 units) in a volume of 40 .mu.L for 45 min. (see FIG. 15B). The buffer was then modified to favor Sal I activity, by adding 2.lambda. of 1.5 M NaCl, and then 25 units of Sal I were added. After another 45 min at 37.degree. C., this sample was also loaded on the 4 percent PAG. DNA fragments were separated by electrophoresis at 90 volts over 2.5 h.

pHIL-S1 gave 2 fragments, a larger one at 5.9 Kbp containing the ampicillin resistance gene and a smaller fragment at 2.5 Kbp containing most of the 5'AOX1 region and multiple cloning site. Both fragments were cut out of the gel and isolated by electroelution and ethanol precipitation.

For the pBluescript II digest, 2 bands were visualized, one corresponding to the vector piece at about 2.8 Kbp and a small fragment from the multiple cloning site. The 2.8 Kbp fragment was recovered as above. To make pPY1 (FIG. 15C), the pBluescript II Sst I/Sal I 2.8 Kbp fragment (.about.100 mg) was used in a ligation reaction with the 2.5 Kbp pHIL-S1 fragment (.about.100 mg). The ligation kit components were from Stratagene, while the T4 DNA ligase was a New England Biolabs product. Ligation commenced at 18.degree. C., with a cooling gradient down to 4.degree. C. overnight. E. coli SURE (Stratagen, LaJolla, Calif.) was transformed with this ligation mix (1 .mu.L and 5 .mu.L samples). After overnight growth on Luria Broth (Sambrook et al., Supra) plates containing 100 .mu.g/mL ampicillin, uncolored colonies (plasmids containing inserts) out-numbered blue colonies (without inserts) by roughly 10:1. Five of the uncolored colonies were picked for plasmid screening with Sst I and Sal I. All 5 gave the restriction pattern expected based on the plasmid construction (2 bands at 2.8 Kbp and 2.5 Kbp). The plasmid from a clone designated #4 was tested further and found to have the correct sites for BstB I, Xho I (2 sites), EcoR V and BamH I, and was thus picked as the representative clone for pPY1.

Construction of pPY2

Using the plasmid mini prep DNA of pPY1 from above, 2.5 .mu.g were treated with 20 units of EcoR V (BRL) in a reaction volume of 45 .mu.L at 37.degree. C. for 30 min. A single linearized DNA fragment at 5.3 Kbp was purified from a 3.75 percent PAG as described above. A linker DNA fragment with single EcoR I and Not I sites was then ligated into the blunt ended EcoR V site as follows.

The oligos ECONOT (SEQ ID NO:58), a 20-mer: 5'-GAATTCTTAGCGGCCGCTTG-3' for the top strand and TONOCE (SEQ ID NO:59), a 20-mer: 5'-CAAGCGGCCGCTAAGAATTC-3' for the bottom strand were synthesized and purified for annealing. A 60 fold molar excess of annealed linker over the EcoR V treated pPY1 fragment (5.7.times.10.sup.-14 moles -20 mg) were ligated using T4 DNA ligase and temperature conditions as above. Competent E. coli AG1 cells were transformed with aliquots (1 and 5 .mu.L) of this ligation mix and plated (75 or 150 .mu.L) onto LB agar plates containing 100 .mu.g/mL ampicillin. Twelve clones were picked for plasmid screening with EcoR I. Out of these, 10 clones had the correct two EcoR I fragments of roughly 600 bp and 4.7 Kbp. Four of these clones were selected for DNA sequencing of their plasmids for verification and orientation of the linker sequence. The oligo used for the sequencing (0.5 pmol) was ECOVNHESEQ (SEQ ID NO:60), a 19-mer: 5'-TGCGCATAGAAATTGCATC-3'. Two of the plasmids had the correct sequence and desired orientation, one was chosen as the representative pPY2 clone (see FIG. 15D).

Construction of pPY21 and pPY22

A common vector derived from pPY2 was used to generate both pPY21 and pPY22. Two micrograms of pPY2 mini-prep DNA were treated with 24 units of BstB I (New England Biolabs) in a 20 .mu.L reaction volume at 65.degree. C. for 15 min. At this point, 1.5 M NaCl was added (1.2 .mu.L) with 12 units (1.2 .mu.L) of BamH I (New England Biolabs) and the reaction allowed to proceed for 15 min at 37.degree. C. This sample was loaded on a 4 percent PAG, and the pPY2 fragment with BstB I and BamH I ends of roughly 5.2 Kbp was recovered by electroelution and ethanol precipitation.

The PCRs to generate 2 CC49 scFv inserts for the pPY2 vector were performed as follows. Oligos (100 pmol each) targeting pSCFV UHM5.2, perpared as described in Example 1 (5 ng) were:

1) PY49VLI (SEQ ID NO:61), a 92-mer, 5'-ACATTTCGAAACGATGCTTTTGCAAGCTTTCCTTTTCCTTTTGGCTG GTTTTGCAGCTAAGATATCTGCTGACATTGTGATGTCACAGTCTC-3'; and

2) pPY523 (SEQ ID NO:62), a 36-mer, 5'-AAATGGATCCTATTAGTCATAGTCTTCAGGGTCTTC-3'.

Taq polymerase was purchased from Perkin Elmer Cetus (Norwalk, Conn.) and used in conjunction with the GeneAmp.TM. kit, according to the manufacturer's directions. The conditions for the PCR were: 30 cycles of 94.degree. C. for 30 sec, 52.degree. C. for 30 sec and 72.degree. C. for 1 min. On the final cycle, the last elongation step (72.degree. C.) was 2 min. An MJ Research (Watertown, Mass.) thermal cycler was used.

For the PCR of the second CC49 scFv insert, oligos (100 pmol each) targeting pSCFV UHM8.1 (see example 1) (5 ng) were:

1) PY49VL1 (SEQ ID NO:61), as above; and

2) PSC49VH (SEQ ID NO:63), a 40-mer, 5'-ATGTGGATCCCTATTAGTAGGAGACGGTGACTGAGGTTCC-3'.

The reaction was set up and carried out as described for pSCFV UHM5.2, above. The PCR products were purified free of reactants by electrophoresis in a 4 percent PAG. The desired bands from the pSCFV UHM5.2 and pSCFV UHM8.1 targets were visualized at the expected 875 bp and 854 bp sizes respectively. The DNA was recovered by electroelution and ethanol precipitation. The PCR inserts were prepared for ligation by trimming the ends off of each with BstB I at the 5' ends and BamH I at the 3' ends, using the same reaction conditions as described above for the pPY2 vector. After this treatment, the DNA inserts were again purified using a 4 percent PAG, subsequent electroelution and ethanol precipitation. The DNA inserts generated from pSCFV UHM5.2 and pSCFV UHM8.1 were used to make pPY21 and pPY22, respectively (FIG. 15E). The ligation reactions regenerating pPY21 and pPY22 were set up (20 .mu.L total volume) using 100 mg of pPY2 vector for each, and a 1:1 stoichiometric equivalent of the prepared PCR inserts from pSCFV UHM5.2 and pSCFV UHM8.1 respectively. The ligation kit components and T4 DNA ligase were from Stratagene and used according to their directions. The ligation reactions were incubated at 15.degree. C. for 2 hours, and 2 .mu.L from each used to transform E. coli AG1 competent cells. Samples were plated onto LB agar containing 100 .mu.g/mL ampicillin. Seven clones each for pPY21 and pPY22 were picked for screening of their plasmid DNAs, using the BstB I-BamH I double digest described above. A single clone (#6) was obtained with the correct DNA fragments for pPY21, while four clones (#11-14) were obtained for pPY22. The clone for pPY21 and pPY22 clone #12 were selected for DNA sequencing of the insert DNAs containing the CC49 scFv constructions. Both clones contained the correct sequence for CC49 scFv, 5.2 (pPY21) and 8.1 (pPY22), respectively, and were used for the final expression plasmid constructions.

The DNA and amino acid sequences of CC49 scFv in pPY21 (SEQ ID NO:14 and NO:15) and pPY22 (SEQ ID NO:16 and NO:17) are given in FIG. 16. In these scFv constructions, the V.sub.L gene came first followed by the V.sub.H and was joined by a 25 amino acid linker which is in bold and underlined. The oligonucleotides used as DNA sequencing primers are shown above the DNA sequence lines. Restriction enzyme sites are as indicated, and underlined. The SUC2 signal peptide is underlined; the putative first amino acid of these mature products is indicated by (+1). Complementarity determining CDRs regions segments for the CC49V.sub.L and CC49V.sub.H are italicized in bold. Sequences for pPY21, giving negatively charged proteins (CC49, scFv5.2, Fv(2n)-5.2) are indicated by sequence S1 and sequences for pPY22, giving close to neutrally charged scFv products (CC49, scFv8.1P, Fv-(2n)-8.1P) are represented by sequence S2.

Construction of pPY31 and pPY32

The Sst I-Sal I DNA fragments of pPY21 and pPY22 containing the CC49 scFv constructions were isolated and purified, to be used as inserts with the pHIL-S1 vector fragment (5.9 Kbp) already described above. Three micrograms of pPY21 and pPY22 plasmid mini-prep DNAs from above (same clones as were sequenced) were treated with 9 units each of Sst I(BRL) in a 30 .mu.L reaction at 37.degree. C. for 20 min. The samples were electrophoresed on a 4 percent PAC. The resulting bands at 3.2 Kbp (containing the DNA of interest) and at 2.8 Kbp were isolated together, since the resolution was not adequate for good resolution on the PAG. These pairs of DNA inserts were then ligated with the pHIL-S1 Sst I-Sal I vector fragment (100 mg) in a 1:1 stoichiometric ratio, in a 20 .mu.L reaction volume for 2 h at 15.degree. C. to form pPY31 and pPY32 (FIG. 15F). Competent E. coli AG1 cells were transformed with 2.5 .mu.L aliquots of the ligation mixes and plated onto LB agar plates containing 100 .mu.g/mL ampicillin. Six potential clones each of pPY31 and pPY32 were selected for plasmid screening. The resulting mini-prep DNAs were digested with an Sst I-Sal I double digest, as described above for the insert preparation. Three clones had the correct 2 bands (3.2 Kbp and 5.9 Kbp) for pPY31, one was chosen as the pPY31 representative. Two clones had the correct bands for pPY32 and one of them was chosen as the representative for transformation into Pichia.

Transformation of Pichia pastoris GTS115 with pPY31 and pPY32

The procedure of Becker and Guarente for [Methods in Enzymology, 194, 182 (1990)]. electroporation of yeast cells was followed with some modifications, as described below, to introduce the CC49 scFv genes into Pichia. Yeast extract peptone glucose (YEiD) medium (125 mL) was inoculated for overnight culture at 30.degree. C. from a stock of P. pastoris GTS115 stored at 4.degree. C. The cells were harvested the next day when the O.D..sub.600nm was between 0.8-0.9. After washing with sterile water and 1M sorbitol, the cells were resuspended in ice-cold 1 M sorbitol (150 .mu.L) and held on ice. Aliquots of pPY31 or pPY32 (2 .mu.g/.mu.L) linearized with Bgl II, totaling 5 .mu.g and 10 .mu.g were added to 40 .mu.L each of the Pichia electro competent cells and transferred to pre-chilled (on ice) 0.2 cm sterile electroporation cuvettes (Biorad). Each sample was pulsed at 1.5 kV, 25 .mu.F and 200 ohms, using a Biorad Gene Pulser with Pulse Controller. The time interval of the pulsers was 4.7-4.8 msec. Immediately after each pulse, 1.0 mL of cold 1M sorbitol was added to the cuvette, and three RDB plates for each of the 4 transformations (see U.S. Pat. Nos. 4,855,231, 4,879,231 and 4,808,537) were plated (200 .mu.L each). Plates were incubated at 30.degree. C. Four days later, 36 colonies (numbered 1-36) transformed with pPY31 and 35 colonies (numbered 37-71) transformed with pPY32 were randomly picked and inoculated into 2.0 mL of BMGY media and streaked on MD and MM plates. All were grown at 30.degree. C. BMGY is buffered minimal glycerol-complex medium containing 1 m potassium phosphate (pH 6.0); 13.4 g/L yeast nitrogen base with ammonium sulfate; 400 .mu.g/L biotin; 10 g/L yeast extract; and 20 g/L peptone. MD is a medium containing 13.4 g/L yeast nitrogen base; 400 .mu.g/L biotin; and 20 g/L dextrose. MM is the same as MD with 5 mL/L of methanol substituted for the dextrose.

Clones able to grow on MM pates are not AOX.sup.-, implying that the Bgl II fragment did not incorporate at the AOX1 locus, but at some other chromosomal location. The MM plates have only methanol as a carbon source. After 2 days of growth in the BMGY medium, 9 potential AOX.sup.- mutant lines (observed by slow growth on the MM plate, 3 for PPY31, designated Nos. 4, 31, 32 and 6 for pPY32, designated Nos. 39, 44, 60, 61, 65 and 71 and an AOX.sup.+ control (Clone No. 11, from the pPY31 set) were transferred to BMMY medium (2.0 mL each). BMMY medium is the same as BMGY with 5 ml/L methanol substituted for glycerol. Aliquots of methanol (20 .mu.L) were added to each tube, roughly every 10-14 h. Samples (100 .mu.L) from each culture were taken 32 h after methanol induction and the biological activity determined by a competition assay described in Example 1. SDS-PAGE 10-20 percent gradient gels were run on 10 .mu.L aliquots taken 32 and 100 h post induction. Western transfer and detection by biotinylated FAID14 was also done for the 100 hour samples.

The relative anti TAG-72 activities were evaluated for the methanol slow (AOX-) clones, along with some normal growing clones on MM medium. Biotinylated CC49 was used as the competitor in a competition assay. The results are presented in FIG. 17, where the percent competition for a given sample was calculated based on OD.sub.405nm -OD.sub.450nm readings of the average of triplicate samples assayed using the following formula: ##EQU1## The zero competition value was obtained from a 0.2 micron filtered BMMY medium sample, while the 100 percent competition value was obtained using a 25 .mu.L sample of CC49 IgG at 5 .mu.g/mL. The Pichia samples for assay were prepared by microcentrifugation of the culture sample and filtration through a 0.22 .mu. filter. For pPY31 the results indicated that the methanol slow clones 4 and 32 competed very well (FIG. 17) as did clone 31, which by observation of growth on the MM plate could not be categorized as methanol slow. Clones 11 and 17 were producing little or no scFv. For pPY32, the methanol slow clones 39, 60, 61, 65 and 71 all gave excellent competition. Clones 38, 44 and 50 gave little or no competition.

The SDS-PAGE-Western data corroborated the competition ELISA results. The main product of the crude Pichia culture supernatant observed on SDS-PAGE-Western was the scFv product. The Pichia products correlated well with the previously obtained E. coli product equivalents in terms of size on the SDS-PAGE and activity in the competition assay. In the case of the pI 8.1 Pichia products, a very slightly slower mobility was observed relative to the Western positive band representing the E. coli CC49 scFv I8.1 product. This is due to the increase in molecular weight of the Pichia product due to the carboxyl terminal tyrosine residue weighing approximatly 78 daltons more than the corresponding serine residue at the same position in the E. coli product.

Purification of Single Chain Antibody Species from P. pastoris GTS115 pPY31 Culture

Pichia pastoris carrying the plasmid pPY31 produces a CC49 single chain antibody having a pI of about 5.2. Starter cultures (20 mL BMGY medium) was inoculated (Clone #4 above) and grown for 26.5 hours at 30.degree. C. This was used then to inoculate 800 mL BMGY medium in a 2.8 L Fernbach flask. The culture was grown at 30.degree. C. for 24 hours with shaking at 150 rpm. The Pichia cells were pelleted at 10.degree. C. using 2 tubes and a G53 rotor (Sorvall) run at 5,000 rpm for 10 minutes. The cells from each tube were aseptically resuspended in BMMY medium (40 mL) to which 75 .mu.L of methanol were added and continued to shake at 30.degree. C. Methanol (200 .mu.L) was added approximately every 12 hours over a 36 hour growth period.

Culture supernatants (containing the single chain antibody products) were harvested by spinning the cultures at 5,500 rpm in a SS34 rotor (Sorvall) for 5 minutes. The supernatant (about 66 mL) was further clarified by filtering through a 0.22 .mu. pore size filter device (Nalge) and then concentrated to 10 mL using a Centriprep 10 device (Amicon). Half of this material was applied to a Superdex 75 HR26/60 gel filtration column (Pharmacia), at a flow rate of 1.3 mL/min using PBS as the buffer. The gel filtration profile (See FIG. 18) gave three pools of single chain species. indicating the formation of monomeric single chain antibody (scFv, fraction 19-23), dimers (scFv2, fractions 13-15) and multimeric forms (scFv(2n) where n>1, fractions 8-10). On the SDS-PAGE insert, lane "S" indicates the identical purified protein standard CC49 scFv5.2 derived from E. coli pSCFV UHM5.2 (Example 1). In the processing of running SDS-PAGE, the non-covalent interactions in a dimer or multimer are caused to break apart, showing that the dimer and multimer forms correspond to the same size as the basic monomeric unit.

Each of the single chain antibody pools was further purified by use of a Mono Q HR5/5 anion exchange column as described in Example 1. Each of the various species was then re-chromatographed on an analytical gel filtration column (Superdex 75 HR10/30, Pharmacia-LKB). Table II gives the elution time, elution volume, Kav values and molecular weights determined for the monomeric and various multimeric forms. The predicted molecular weights were predicted by DNASTAR.TM. protein titrate computer program using the amino acid composition of the polypeptide.

The Anion Exchange Chromatography and SDS-PAGE are given in FIG. 19. Relevant parts A, B, and C correspond to the monomer, dimer and higher multimer forms of the scFv species, respectively. The buffer system used and the gradient of Buffer B are also shown. The SDS-PAGE insert (D) has the following samples: molecular weight markers (lane 1); monomer and dimer (lanes 2 and 3); and E. coli derived CC49 scFv 5.2 standard (lane 4); tetramer, hexamer and putative octamer (lanes 5, 6, and 7, respectively) and an approximately 34 Kd impurity (lane 8) derived from fraction 45.

The IEF gel and Western, shown in FIG. 20 indicate that the monomer and various multimer forms have distinct pI values. Both the E. coli and Pichia derived monomer forms, have a pI of 4.92 (derived from a graph of distance from cathode vs. pI of the standards), with a minor band at pI 4.97, which is the same as the pI for all multimers. That 2 bands seen in the monomer (lanes 3 and 9) support the hypothesis that there are two charge forms for the monomer in equilibrium, where the pI 4.97 form has a tendency to dimerize. In FIG. 20, lanes 1-7 represent bands stained with Coomassie Brilliant Blue R-250 and lanes 8-13 represent a Western transfer of samples corresponding to lanes 2-7. The pI markers (lane 1) were from Sigma: soybean trypsin inhibitor, pI 4.55; bovine .beta.-lactoglobulin A, pI 5.13; and human erythrocyte carbonic anhydrase B, pI 6.57. Lanes 2 and 8 show purified CC49 scFv5.2 derived from E. coli pSCFV UHM5.2; lanes 3 and 9 show the same product derived from Pichia. Lanes 4 and 10 show a sample of crude culture supernatant from Pichia containing all the CC49 single chain antibody species present. Lanes 5 and 11, 6 and 12, and 7 and 13 indicate the Pichia dimer, tetramer and hexamer species, respectively.

Purification of Single Chain Antibody Species of Pichia pastoris GTS115/pPY32

P. pastoris having the plasmid pPY32 produces CC49 single chain antibodies having a pI of about 8.1. The growth, harvesting and induction of the AOX1 promoter were the same as for the pPY31 clone above.

The purity of the 8.1 monomer and dimer (multimeric forms were not analyzed) was verified by SDS-PAGE. The IEF pattern of these molecules indicated that except for carboxyl terminal Ser residue present in the E. coli derived CC49 scFv8.1 and Fv28.1, the corresponding Pichia products have the same pI values. In Pichia the carboxyl terminal residue is Tyr.

The N-terminal amino acid sequencing results indicated that the invertase signal sequence used in constructing these single chain antibodies was accurately being removed by signal peptidase. For the CC49 scFv8.1P and Fv2-8.1P proteins, the following NH.sub.2 terminal residues were determined: Asp-Ile-Val-Met-Ser-Gln-Ser-Pro-Ser-Ser, which matches the predicted amino acid sequence of the mature form, from the DNA sequence (see FIG. 16). For the CC49 Fv2-5.2 product, the same results were obtained as above, except that the first and tenth amino acids could not be unequivocally assigned. Based on the above results with the 8.1 pI system, it is reasonable to assume for the 5.2 pI products that the unassigned residues are the same.

From the competition ELISA results (FIG. 21A and 21B), it can be observed that the avidity of the single chain species, increases with higher multimer forms. It is possible that the effects seen for the tetramer and hexamer are underestimated and that different concentrations or densities of TAG-72 on the ELISA plate may influence the results. The availability of antigen to some of the binding faces of the hexamer and octamer, for example, may not be sterically available.

Example 3

Biodistribution Characteristics of .sup.125 I Labeled CC49 Single Chain Fv Species

The biodistribution of four CC49 single chain antibody species, two monomeric species: CC49 scFv with a pI of 8.1 obtained from Pichia (scFv 8.1P) and CC49 with a pI of 5.2 also obtained from Pichia (scFv 5.2); and two dimeric species: CC49 Fv2 with a pI of 8.1 obtained from Pichia (Fv2-8.1P) and CC49 Fv2 with a pI of 5.2 obtained from E. coli (Fv2-5.2), was determined. This was compared with ChCC49 Fab and IgG. The single chain species were labeled with .sup.125 I using the Iodo-Beads.TM. (Pierce, Rockford, Ill.) method using N-chloro-benzenesulfonamide derivatited polystyrene beads (Iodo-Beads.TM., Pierce, Rockford, Ill.). To label the antibodies, initially 3 Iodo-Beads.TM. were incubated at room temperature for 5 minutes in PBS and 150 .mu.Ci carrier-free I-125 (Nordion, 100.9 .mu.Ci/.mu.L). At this point, the antibody (0.3 mg in PBS) was added and allowed to react for 8 minutes. Unincorporated isotope and reaction by-products were removed by gel filtration chromatography (Pharmacia Superdex 75 HR10/30 column), using PBS as the buffer at a flow rate of 0.6 mL/minute. An aliquot of the purified protein peak was measured in each case at A.sub.280nm and quantitated according to the E.sup.1%.sub.280nm values as described in Example 1 for the respective proteins. Levels of radioactivity were determined by measurement in a Capintec Counter, set for reading I-125.

SDS-PAGE and IEF analyses were performed as previously described. Gels were dried and radioactive bands detected by exposure to Kodak XAR X-OMAT film. The observed pI values for the 8.1P species, as determined by IEF are 7.1 and 7.25 for monomer and dimer, respectively, and for the 5.2 species, 4.9 and 5.0 for the monomer and dimer, respectively. Differences between the calculated values (after which the species are officially called) and the observed values indicate that the tertiary structure of these molecules play a role in charge-charge interactions and/or hydrogen bonding which will determine the observed pI values.

Five test groups, each containing 5 female nu/nu (CD-1) mice (obtained from Charles River Breeding Laboratories) bearing LS174T human tumor xenografts, were used to measure the biodistribution of .sup.125 I labeled antibody species. The biodistribution of the compounds was measured at various time intervals over 48 hours, after intravenous (tail vein) injection of the compounds. Animals were housed and identified by group as outlined in Section 4.0 of Standard Operation Manual for the Care of Laboratory Animals (SOP-PTG). At the appropriate time, the mice were euthanatized and the radioactivity remaining in samples of blood, liver, spleen, kidney, tumor, lung, GI tract, tail, and remaining carcass were quantified as outlined in Section 4.0 and Section 5.0 of SOP-PTG. The LS174T tumors were produced in the mice according to the procedure set forth in Section 2.0 of SOP-PTG.

For metabolite studies, groups of Balb/c mice were injected intravenously with 1 .mu.Ci of one of the antibody species. Mice from each group were anesthetized after 30 minutes, 2 h and 5 h, urine removed from the bladder, urine counted for radioactivity and then kept frozen at -70.degree. C. until HPLC analysis. A Pharmacia Superdex 75 gel filtration HR10/30 column was used for chromatography, using PBS as the eluent at 1.5 mL/minute. The volumes of the injected urine varied from 20-80 .mu.L. The column eluent passed through a dual detector for absorbance at 280 nm and for radioactivity.

Gamma Counter #2 (Searle/ND-66) was used to measure radioactivity in standards and tissue samples. Biodistribution of radioactivity in the collected tissues was analyzed using template TEMPL as outlined in Section 5.0 of SOP-PTG. Experimental outliers were treated as outlined in Section 6.0 of SOP-PTG.

Table IV gives the amount of protein and activity of each species injected into the mice for biodistribution studies.

Tables IV, V, VI, VII, VIII, and IX give the biodistribution of .sup.125 I CC49 species of scFv5.2, Fv2-5.2, and scFv8.1P, Fv2-8.1P, ChFab and IgG, respectively, based on percent injected dose per organ. Percent injected dose per organ is calculated by determining the .sup.125 I in counts per minute (CPM) in each tissue, divided by the injected dose (CPM) multiplied by 100 to give percent.

The differences in pharmacokinetics for the various sizes and charges of the CC49 antibody species are demonstrated by the data in tables IV-IX. In general, the smaller the molecule, the faster it clears the blood. The 25-50 K dalton size molecules (all except the IgG) clear the blood within 24 hours. The negatively charged Fv2-5.2 dimer has the greatest amount in the blood at 2 hour (Table V). Both dimers accumulate in the tumor to the largest extent over a 24 hour time period (Tables V and VII), excluding the IgG.

A major difference between the antibody species is observed in retention by the kidneys. The negatively charged monomer species, CC49 scFv5.2, clears the most slowly (Table IV). This is surprising in that the negative charge should have prevented electrostatic interaction with the negatively charged glomerular cells of the kidney. The nominally neutral charged species, scFv8.1P cleared more quickly (Table VI). This trend was also true for the two dimers, but clearance was much better for these relative to the monomers. It is unexpected that the negatively charged dimer would have the best tumor localization values (Table V). The data also shows that a relatively small molecule (58 Kd) has staying power on the tumor. Both dimers out performed the monomers and chimeric CC49 Fab molecule in this regard.

Results from metabolic studies indicated that after injection with .sup.125 I labeled scFv 5.2 and Fv2-5.2, only one radioactive metabolite was observed in the urine, free .sup.125 I, as determined by gel filtration chromatography. No larger peptides are observed at any of the time points. These data indicate that dehalogenation is occurring within the body and that free I-125 is able to be readily excreted into the urine.

The tumor to tissue ratios 8 to 24 hours post injection indicated the dimers would be effective in radioimmunoguided surgery procedures. Surgery may be performed in a more timely fashion, rather than 2 to 3 weeks post injection, as is presently necessitated by the biodistribution kinetics of whole antibody.

Example 4

Construction, Purification and Characterization of Anti TAG-72 Single Chain Antibody Species with the Human Subgroup IV V.sub.L

To provide for a rapid and convenient procedure to determine the effectiveness of the single chain dimers and multimers of the present invention, a unique octapeptide sequence -NH.sub.2 -Asp-Tyr-Lys-Asp-Asp-Asp-Asp-Lys-COOH, called the FLAG.TM. peptide (a trademark of International Biotechnologies Inc., CT) was used. This sequence is recognized by International Biotechnology Incorporated's (IBI) mouse M2 monoclonal antibody. Placing the FLAG.TM. sequence on two annealing oligonucleotides in pSCFV UHH leads to a single chain antibody having the human subgroup 4 V.sub.L and CC49 V.sub.H with the FLAG peptide attached at the carboxyl terminus.

The final plasmid derived from pSCFV UHH containing FLAG.TM. is designated pSC49FLAG. The nomenclature for the single chain proteins produced from pSC49FLAG is as follows: for the monomer-H4L49HF scFv, for the dimer-H4L49HF Fv2 and for the tetramer-H4L49HF Fv4. The "H4L" portion refers to the human Subgroup IV light chain V.sub.L region, "49H" refers to the CC49 V.sub.H and "F" indicates that the FLAG.TM., peptide is attached.

Construction of pSC49FLAG

A general outline for obtaining the plasmid pSC49FLAG is shown in FIG. 22 and the procedure described below.

a) Isolation of pSCFVUHH Xho I/Nhe I Vector Fragment

Approximately 5 .mu.g of plasmid pSCFV URH (Example 1) in 15 .mu.L of water was used from the Magic Mini-prep system (Promega). To this was added 5.4 .mu.L of 10.times.Buffer #2 (New England Biolabs), 45 units of Xho I (New England Biolabs), 15 units of Nhe I and 24 .mu.L of water. The reaction proceeded for 1 hour at 37.degree. C. The sample was then loaded with running dyes on a 4 percent polyacrylamide gel, electrophoresed and purified by electroelution (Sambrook et al, supra). The obtained DNA pellet was dissolved in 20 .mu.L of water.

b) Construction of pATDFLAG

Plasmid pSCFV UHH treated with Xho I and Nhe I from above was used in a ligation reaction with annealed FLAG and FLAGNC oligos.

FLAGC (SEQ ID NO:64):

5'-TCGAGACAATGTCGCTAGCGACTACAAGGACGATGATGACAAATAAAAAC-3'

FLAGNC (SEQ ID NO:65):

5'-CTAGGTTTTTATTTGTCATCATCGTCCTTGTAGTCGCTAGCGACATTGTC-3'

Equimolar amounts (3 .mu.M of each of the oligos) FLAGC and FLAGNC in a total volume of 62.5 .mu.L) were mixed together using a ligation buffer diluted to 1.times.(Stratagene). The sample was heated to 95.degree. C. for 2 min and was allowed to cool to below 35.degree. C. over 10 min before use in the ligation reaction below.

The reaction was carried out using the components and amounts indicated above. Starting at 18.degree. C., and then being cooled gradually to 4.degree. C. overnight. Competent E. coli Agl cells (Stratagene) were transformed with 3 .mu.l of the above ligation reaction and colonies

selected using LB/CAM 20 plates. A clone having the appropriate Nhe I/XhoI and Nhe I/Xho I restriction patterns was selected for DNA sequencing.

The oligonucleotide used to verify the sequence of the FLAG.TM. linker in pATDFLAG was called PENPTSEQ (SEQ ID NO:66): 5'-CTTTATGTAAGATGATGTTTTG-3. This oligonucleotide is derived from the non-coding strand of the penP terminator region (See FIG. 23). FIG. 23 gives the amino acid (SEQ ID NO:20) and DNA sequence (SEQ ID NO:9) in pATDFLAG. The amino acid sequence corresponding to the FLAG.TM. peptide is given in bold italics. The FLAG.TM. peptide is out of frame relative to the H4V.sub.L -linker sequence which ends at the Xho I site. The DNA sequencing primer, PENPTSEQ, is shown in bold. Nucleotides 824-478 inclusive were verified using the PENPTSEQ primer. At position 624, a `T` nucleotide present in the original H4V.sub.L sequence has been changed to a `C` nucleotide resulting in a Val to Ala amino acid residue 89 substitution. These differences are indicated in shadow. The CDR amino acid sequences of H.sub.4 VL are shown in bold, while the UNIHOPE linker sequence is in bold and underlined. DNA sequencing was performed using the Sequenase.TM. sequencing kit (U.S. Biochemical, Cleveland, Ohio) following the manufacturer's directions.

c) Generation of pSC49FLAG

The plasmid pATDFLAG (approximately 5 .mu.g, purified from a 2.5 mL culture by the Magic Miniprep system (Promega) was treated at 37.degree. C. for 1 h as indicated in the protocol given below:

DNA--17 .mu.L

10.times.buffer (New England Biolabs #2)--6 .mu.L

Xho I (New England Biolabs)--2.5 .mu.L (25 units)

Nhe I (New England Biolabs)--3.5 .mu.L (17.5 units)

H.sub.2 O--31 .mu.L

The resulting vector fragment was purified from a 4 percent PAG, electroeluted and ethanol precipitated. The obtained DNA was dissolved in 20 .mu.L water.

The CC49 V.sub.H insert DNA was obtained from a PCR amplification protocol (Perkin Elmer Cetus) as instructed by the manufacturer. The target DNA (uncut) containing the CC49 V.sub.H was pSCFV UHM 8.1 (Example 1) (0.5 mg). The 5' oligo used for the PCR (100 pmol) was UNI-3 (SEQ ID NO:46) (Example 1). The 3' oligo used (100 pmol) was called SC49FLAG (SEQ ID NO:67), a 33-mer as follows:

Nhe I

5'-TATTGCTAGCTGA GGA GAC GGT GAC TGA GGT TC-3'.

The resulting amplified DNA was purified from a 4 percent PAG, dissolved in 20 .mu.L water, and then treated at 37.degree. C. for 1 h as indicated in the protocol given below:

DNA--10 .mu.L (.about.2 .mu.g)

10.times.buffer (New England Biolabs #2)--6 .mu.L

Xho I (New England Biolabs)--2.5 .mu.L (25 units)

Nhe I (New England Biolabs)--3.5 .mu.L (17.5 units)

water--38 .mu.L

The resulting insert fragment was purified and dissolved in 20 .mu.L water. The vector and insert fragments from above were set up in a ligation reaction at 18.degree. C. going down to 4.degree. C. overnight as follows:

DNA vector (pATDFLAG-XhoI/Nhe I)--1.5 .mu.L

DNA insert (PCR amplified CC49 V.sub.H -XhoI/Nhe I)-0.8 .mu.L

10.times.ligation buffer (Stratagene)--2 .mu.L

10 mM ATP--2 .mu.L

T4 DNA ligase--1 .mu.L

water--12.7 .mu.L

Competent E. coli AGl cells (Stratagene) were transformed using the ligation reaction from above, as directed by the manufacturer. Clone number 1, was selected for further work, and for verification of the constructed DNA sequence. The oligos used as primers for the DNA sequencing were as follows:

1) PENPR1 (SEQ ID NO:68), a 21-mer: 5'-AAC ACT GTA GGG ATA GTG GAA-3';

2) PENPR2 (SEQ ID NO:69), a 20-mer: 5'-GTC TCC CTC CGT TTG AAT AT-3';

3) TAGVLFR2 (SEQ ID NO:55), a 20-mer: 5'-TGG TAC CAG CAG AAA CCA GG-3;

4) UHVHSEQ (SEQ ID NO:70), a 22-mer; 5'-GAT GCT GCG AAG AAG GAT GAC G-3';

5) TAGVHCDR2 (SEQ ID NO:57), a 21-mer: 5'-ATGG ATT GGA TAT TTT TCT CC-3' and;

6) TAG VH FR4 (SEQ ID NO:71), a 21-mer: 5'-ACT GGG GTC AAG GAA CCT CAG-3'.

The DNA (SEQ ID NO:21) and amino acid sequence (SEQ ID NO:22) of pSC49FLAG is given in FIG. 24. Asp=1 indicates the first amino acid from H4V.sub.L of the mature single chain antibody species. The UNIHOPE linker is in bold and underlined. The complimentarity determining region sequences are shown in bold, while the FLAG peptide is in bold italics. Nucleotides corresponding to the Nhe I site in pSCFV UHH are in bold and underlined (position 1154-1159). The `C` at position 1155 and the `A` at position 1157 are italicized within this sequence and indicate differences from the published penP terminator sequence. The DNA sequencing was carried out using the Sequenase.TM. kit mentioned above.

The predicted pI and molecular weight (MW) of each of the protein species produced was obtained from the DNASTAR.TM. Protein-Titrate computer program (Madison, Wis.).

Purification of H4L49HF Single Chain Products

A 10 mL M2 antibody affinity column was prepared (gel purchased from IBI) and incorporated on the HPLC system described under general experimental procedures. The H4L49HF crude protein sample was prepared as follows. One liter (2.times.500 mL) of E. coli AGl/pSC49FLAG was grown overnight at 37.degree. C. in LB broth containing 20 .mu.g/mL chloramphenicol. The cells were recovered by centrifugation of the culture at 5,000 rpm for 10 min at 5.degree. C. in a Sorvall GS-3 rotor (duPont). The cell pellet was resuspended in 100 mL of PBS, and subjected to sonication (MSE Soniprep 150) for 5 bursts at 14 microns amplitude lasting for 15 seconds each, and cooling in an ethanol-ice bath for 1 minute in between. The sonicate was ultracentrifuged at 26,000 rpm for 2 hours in a Type Ti30 rotor (Beckman Instruments). The resulting supernatant was filtered through a 0.22 membrane device (Nalge Co., Rochester, N.Y.) and finally concentrated to 22 mL using Centriprep 30 (Amicon) devices. The sonicate was pumped onto the affinity column at 2 mL/min with stops lasting for 1 min twice during the application, to allow interaction with the affinity ligand.

PBS was used to wash the column of any unbound material, and when the absorbance at 276 nm returned substantially to the baseline, the buffer was changed to 0.1M glycine-HCI, at pH 3.1. Elution of a peak began 4 minutes later, the fractions of which were immediately neutralized using 1 M Tris pH 8.8 and PANPEHA indicator (range 0-14) pH paper strips (Schleicher Schull). The eluted protein was concentrated to 200 .mu.L using Centriprep 30 and Centricon 30 (Amicon) devices.

The resulting single chain antibody species were separated by gel filtration using a Pharmacia Superdex 75 Prep Grade HR 26/60 column and PBS as the eluent. The flow rate was initially 0.4 mL/min, and then increased to 1.2 mL/min during the chromatography. Three pools of fractions from this run were made, and then tested analytically on a Superdex 75 HR10/30 column for purity and characteristic elution time. Biorad gel filtration standards were also chromatographed on this column using the same buffer (PBS) and a 0.6 mL/min flow rate, to generate a standard curve for comparison with the elution times of the single chain antibody species. The three species obtained were concentrated, quantitated at OD 280 nm and stored frozen at -20.degree. C.

The results of the purification and characterization of the H4L49HF single chain products is given in FIG. 25. Panel A shows the M2 affinity chromatography profile for E. coli pSC49FLAG sonicate preparation. Stars (*) indicate points in the application of sample when the flow was stopped for 30 seconds. The arrow indicates the point at which the buffer was changed and the flow chart increased to 2.0 mL/min. Fractions were 2 minutes long.

Panel B represents an analytical gel filtration chromatogram obtained from application of 0.5 percent of the total pooled sample from fractions 19-21 shown in A.

Panel C shows a 10-20 percent gradient SDS-PAGE of samples stained with Coomassie Brilliant Blue R-250 (Lanes 1-6) or Western transferred (lanes 7-11) and probed with biotinylated FAID14. Lane 1: prestained molecular weight standards (Diversified Biotech); lanes 2-4 and 7-9 are purified H4L49HF scFv, Fv2 and Fv4, respectively; lanes 5 and 10; E. coli pSC49FLAG sonicate sample, the same as that as applied to M2 affinity column; and lanes 6 and 11, Pichia CC49scFv 5.2.

Amino Terminal Sequencing

Samples of the three protein species of H4L49HF (10 .mu.g each) were desalted by 3 successive water washings using Centrigon 30 devices and freeze-dried. Amino terminal sequencing was determined by standard Edman degradation techniques.

Determination of Anti TAG-72 Activity

a) anti-FLAG ELISA

A. Tag Elisa

Binding activities of the purified monomer and multimer forms of H4L49HF as well as sonicates of E. coli expressing these species were determined by ELISA (Enzyme Linked Immunosorbant Assay) on 96 well PVC plates coated with a 1/300 dilution of the TAG-72 prep (040191). The plates were prepared and blocked for two hours with 1 percent bovine serum albumin (BSA) in PBS with 0.025 percent sodium azide at 37.degree. C. After blocking, the plates were washed three times with 0.025 percent Tween 20 solution and stored at 4.degree. C.

Aliquots of 50 .mu.L of sample at each concentration were added in duplicate to the 96 well plate. The starting concentrations for the purified monomer, dimer, and tetramer were 10 .mu.g/mL for each, and 1:2 dilutions were made to a final concentration of 0.005 .mu.g/mL. The samples added to the plate included the purified monomer, dimer, tetramer, and various crude sonicates.

After adding the samples, the plate went through a series of incubations at 37.degree. C. The first incubation was for 1 hour with sample. The plate was washed five times with 0.025 percent Tween 20. Binding of FLAG.TM. containing single chain species was detected by subsequent incubations with murine anti-FLAG MAb, M2 (International Biotechnologies cat: IB13010) at 1/250 dilution. The plate was incubated for 1 hour removed, and washed. Alkaline phosphatase conjugated goat anti-mouse polyclonal antibody (GAMIG-ALPH, Southern Biotechnology Associates) was added at 50 .mu.L per well. The plate was incubated for 1 hour, removed, and washed. The diluent for both antibodies was in 1 percent BSA in PBS with 0.025 percent sodium azide. The phosphatase substrate (Kirkegaard & Perry Laboratories, Inc.) was prepared by mixing 1 ml of 5.times.buffer (Kirkegaard & Perry Kit), 4 mL water, and one substrate tablet (Kirkegaard & Perry Kit). To each well, 50 .mu.L were added. Color development was terminated after 15 minutes with 50 .mu.L/well of 1 M sodium hydroxide. The ELISA plate was read with the Molecular Devices Kinetic Microplate Reader using the Delta Soft.COPYRGT. Computer program.

B. BSM ELISA

Bovine submaxillary mucin (BSM) was applied to a PVC plate at concentrations ranging from 5.0 mg/ml to 0.00008 mg/ml decreasing at 1:2 dilutions in water. A blank of 50 .mu.L of PBS with 0.025 percent azide was added as a control. The plate was allowed to dry at room temperature overnight in a biological safety cabinet. The plate was blocked using 200 .mu.L of 1 percent BSA in PBS. The plate was incubated at 37.degree. C. for 50 minutes. The plate was removed from incubation and washed three times with 0.025 percent Tween 20. Dilutions (1/3, 1/9, 1/27) of a crude sonicate of pSC49FLAG were applied to the BSM control wells.

After obtaining results from the above ELISA, a BSM concentration of 60 .mu.g/ml was selected for further plating. The diluent was water. The BSM at 60 .mu.g/ml was loaded at 50 .mu.L per well using a multi-channel pipet and dried and blocked as above. The plates were washed three times with 0.025 percent Tween 20 and stored at 4.degree. C. Incubation and detection of FLAG-containing species was performed as above, using the M2 anti-FLAG antibody.

The relative binding of H4L49HF single chain species by TAG of BSM ELISA are given in FIG. 26. FIG. 26A shows results from TAG ELISA for purified monomer, dimer and tetramer. FIG. 26B shows results from TAG ELISA for E. coli pSC49FLAG sonicate containing monomer and multimer species. FIG. 26C shows results from various concentrations of the single chain species reacting with BSM. FIG. 26D is ELISA on BSM with pSC49FLAG sonicate.

In FIG. 26A, starting with concentrations of 10 .mu.g/ml, 1:3 dilutions were performed and the relative binding to TAG-coated plates was determined. It can be seen that the tetramer species gave the greatest binding, followed by dimer, and eventually by monomer. Using DeltaSoft software (BioMetallics, Princeton, N.J.), and assuming the dimer as the standard curve, the tetramer activity was 1.3 to 5 times greater than the dimer activity. The monomer activity was only 0.5 to 0.8 percent of the dimer activity.

b) Competition ELISA

This assay was performed as described in Example 1. The formula used to calculate percent competition is as follows: ##EQU2## Absorbance values for the samples were measured at 405 nm-450 nm. The average of duplicate readings was used. Initially samples (25 .mu.L) were applied to the TAG-72 coated microliter plates at 2.0.times.10.sup.-10 /mL. Biotin CC49 (4 .mu.g/.mu.L diluted 1:20,000--used 25 .mu.L) diluted the samples by a factor of 2. Serial dilutions (1:2) were performed, resulting in a total of 10 concentrations for each sample. The "no competition" sample consisted of 25 .mu.L of antibody buffer (1 percent BSA in Tris buffered saline (Pierce Chemical Co.) + 0.025 percent Tween and 25 .mu.L of the biotinylated CC49 competitor. Background samples were formulated with either antibody buffer with no added CC49 biotin, or a 5 .mu.g/mL solution of CC49 IgG in antibody buffer. Both give essentially the background reading.

The results of the competion ELISA are given in FIGS. 27 and 28, FIG. 27 based on moles of antibody binding site and FIG. 28 based on moles of antibody. These results show that the extra FLAG.TM. binding sites on the dimer and tetramer did not skew the results in favor of the multimers.

Immunohistochemistry (IHC)

Specimens. A total of 7 specimens of colorectal adenocarcinoma, including four primary adenocarcinomas and three intra-abdominal metastases (two mesenteric metastases and one hepatic metastasis), were used in these studies. Areas of reactive or normal tissue were also present for assessment of immunoreactivity of non-neoplastic tissues. Sequential 5 .mu.m thick sections were prepared. One section was stained with hematoxylin and eosin (H and E) with other sections used for the various IHC procedures.

The purified H4L49HF Fv2 product was used at 0.33, 1.7 and 8.3 .mu.g/mL levels as the primary probing agent for the specimens. Sonication (as per Example 1) of a 2.5 mL overnight culture of E. coli pSC49FLAG provided a crude test probe material and was used for the specimens. These sonicates contained a mixture of H4L49HF scFv, Fv2, Fv4 etc. whose concentration and composition was not known.

Immunohistochemistry Procedures (IHC)

IHC was performed using a modification of the avidin-biotin-peroxidase complex (ABC) method (Hsu et al., J. Histochem. Cytochem. 29, 577-580, 1981). The method utilized a commercial anti-mouse IgG kit (Vectastain Peroxidase Mouse IgG, Vector Laboratories, Burlingame, Calif.). Briefly, the technique was as follows:

1) Sequential 5-6 .mu. sections of the test tumors were deparaffinized, rehydrated, and endogenous peroxidase activity was quenched.

The sections were then sequentially flooded with the following reagents:

2) Normal equine blocking serum (Vectastain kit) for 20 minutes.

3) Add primary antibody for 45 minutes, either 200 .mu.L of H4L49HF Fv2 or sonicate of E. coli pSC49FLAG (from 2.5 mL of overnight cell growth), and add to sequential sections of the tumor. Diluent (PBS containing 0.1 percent: BSA, Sigma Chemicals Co. St. Louis, Mo.) was used on one section from every tumor at this step to serve as a negative control.

4) M-2 (International Biotechnologies Inc., San Pedro, Calif.), murine anti-FLAG MAb, was then used at a dilution of 1:100 for 45 mintues.

5) Secondary antibody, biotinylated anti-mouse IgG (Vectastain kit), for 45 minutes.

6) ABC reagent for 45 minutes.

7) The chromagen, 0.05 percent diaminobenzidine, in 0.02 percent H.sub.2 O.sub.2 for 3.5 minutes.

8) The slides were counterstained with Meyer's hematoxylin (Sigma Diagnostics, St. Louis, Mo.), dehydrated, and coverslipped by routine methods.

All reactions were conducted at room temperature. The slides were rinsed between each step, except following step 2, in three washes of PBS for at least five mintues each; the final rinse contained 0.1 percent BSA. IHC reactivity was evaluated for both the acellular and cellular parts of the tumor as well as for adjacent non-neoplastic structures

The H4L49HF purified dimer stained both cellular and acellular areas of the test colon carcinomas. Mucinous acellular material from the neoplastic cells was extensively reactive. Viable neoplastic cells had more variable reactivity with some tumors having high numbers of reactive cells and other tumors having few reactive cells. The specific cellular location of the staining was also variable within and between the different specimens. The atypical cytoplasm and/or (atypical) surface adjacent stained acellular material was most frequently stained. Lesser members of cells had diffuse cytoplasmic staining. There was little reactivity of H4L49H Fv2 with limited numbers of non-neoplasmic tissues examined except reactive colon mucosa immediately adjacent CRC. In this mucosa, the mucin vacuoles of goblet cells and the free mucin on the surface was reactive. The staining diminished with the distance from the tumor/reactive mucosa junction. Crude E. coli pSC49FLAG sonicate, containing all the H4L49HF scFv species (monomer, dimer, multimer, etc.) also gave similar results, indicating that these single chain antibodies are specific for tumor tissues. A control sonicate from E. coli pATDFLAG gave no staining of the TAG-72 positive areas.

Example 5

Synthesis and Expression of Genes Encoding Single Chain Species of HLA-DR

DNA and oligonucleotides

Plasmids containing full length cDNAs encoding HLA alleles DRB1*0101, *0401, *0402 and *0408 and HLA.alpha.l are derived as follows. Poly A+ mRNA is isolated from the peripheral blood lymphocytes of a number of volunteers. First strand cDNA synthesis is performed using oigo (dT) as the primer.

.alpha. Chain Amplification

Using the known sequence of the HLA-DR .alpha. chain [see, Schamboeck et al., Nucl. Acids Res. 11, 8663-8675 (1983)] a 5' end primer called HLADA5 and a 3' end primer called HLADA3 are synthesized to amplify the a chain sequence. These primers contain an Eco RI restriction enzyme site. The sequences of the primers being:

HLADA5 (SEQ ID NO:72):

5'-(TGACGAATTCGCCCAAGAAGAAAATGGCCA-)3'

HLADA3 (SEQ ID NO:73):

5'-(TCAGGAATTCTGGAAAACGCTGAAGATGAC)-3'

Amplifying the a chain from the single-stranded cDNA from above results in an approximately 1 Kb fragment, which is gel purified (GeneClean.TM. kit, La Jolla, Calif.), digested with Eco RI and gel purified again. The product is then subcloned into an Eco RI digested pUC18 vector. A clone designated pKEK.alpha.4 is chosen as the .alpha. chain clone.

.beta. Chain Amplification

Using the known sequence of the HLA-DR .beta. chains, [see, Hurley et al., J. Immunology, 140, 4019-4023 (1988)] a 5' end primer called HLADB5, and a 3' end primer called HLADB3.3 are synthesized to amplify the .beta. chain sequence. Both of these primers contain an Eco RI restriction enzyme site, the sequences of the primers being:

HLADB5 (SEQ ID NO:74):

5'-(TGACGAATTCCTGTCCTGTTCTCCAGCATG)-3'

HLADB3.3 (SEQ ID NO:75):

5'-(GACTCTAGACCCGGGTACCAATGCTGGGACTTCAGGCC)-3

Amplification products are digested with Eco RI and Xba I, gel purified and subcloned into and Eco RI/Xba I digested puC18 vector. A number of clones are isolated, two of which contain rheumatoid arthritis (RA) associated alleles (DRB1*0401 and DRB1*1001) are isolated.

Two additional RA-associated alleles (DRB1*0101 and DRB1*0408) are generated via a semisynthetic approach using the pool of clones which were generated above. The DRB1*0101 clone is generated by removing a SacII/Bsu36I fragment from a clone containing DRB1*0102/DRB and inserting a small fragment generated by annealing the 0101F and 0101R oligonucleotides described below. The small inserted fragment changes amino acids 85 and 86 of DRB1*0102 to the amino acids encoded in the DRB1*0101 allele. The clone, designated pSYN.beta.60, contains a chimeric .beta. chain cDNA encoding HLA-DRB1*0101.

0101F (SEQ ID NO:76):

5'(GGTGGACACCTATTGCAGACACAACTACGGGGTTGGTGAGAGCTTCACAGTG CAGCGGCGAGTTGAGCC)-3'

0101R (SEQ ID NO:77):

5'-(TTAGGCTCAACTCGCCGCTGCACTGTGAAGCTCTCACCAACCCCGTAGTTGT GTCTGCAATAGGTGTCCACCGC)-3'

The DRB1*0408 clone is generated by removing a Eco109I/Bsu36I fragment from a clone containing DRB1*0401 and inserting a small fragment generated by annealing two oligionucleotides DR.beta.-3AHV5' and DR.beta.-3AHV3'. The annealed fragment exchanges amino acids 71 through 86 from the DRB1*0401 allele to the DRB1*0408 allele.

DR.beta.-3AHV5'(SEQ ID NO:78):

5'-(GACCTCCTCGAGCAGAGGCGGGCCGCGGTGGACACCTACTGCCGGCACAACT ACGGGGTTGGTGAAAGCTTCACAGTGC)-3'

DR.beta.-3AHV3'(SEQ ID NO:79):

5'-TCAGGATAGACTCGCCGCTGCACTGTGAAGCTTTCACCAACCCCGTAGTTGTG CCGGCAGTAGGTGTCCACCGCGGCCCGCCTCTGCTCGAGGAG-3'

The scDR.beta./.alpha.-FLAG expression vector (see FIG. 29) is based on the single chain antibody expression vector pATD-FLAG (Example 4).

The following oligonucleotides are used in PCR amplifications to generate cDNAs encoding the human .alpha.1 and .beta.1 domains.

SCDRA5.1 (SEQ ID NO:80) 5'-GGACCTCGAGATCAAAGAAGAACATGTG-3'

SCDRA3.1 (SEQ ID NO:81) 5'-AGTCGCTAGCATTGGTGATCGGAGTATA-3'

SCDRB5.1 (SEQ ID NO:82) 5'-CCAGCCATGGCCGGGGACACCCGACCACGT-3'

SCDRB3.1 (SEQ ID NO:83) 5'-CACTAAGCTTATACACAGTCACCTTAGG-3'

Sequencing reactions of the pSL301 subclones containing either sdDR.beta.1 or sdDR.alpha.1 are done using the commercially available biotinylated T7 and T3 primers are used (New England Biolabs). Alternatively these domains are sequenced directly in the psdDR.beta./-.alpha.FLAG expression vector using the following primers.

VLSEQ1 (SEQ ID NO:84) 5'-CTATTGCCTACGGCAGCC-3'

VLSEQ2 (SEQ ID NO:85) 5'-AGCGTCATCCTTCTTCGC-3'

VHSEQ1 (SEQ ID NO:86) 5'-GATGCTAAAAAGAC-3'

VHSEQ2 (SEQ ID NO:87) 5'-AGGTTTTTATTTGTCATCATC-3'

Primers used for sequencing the full length .beta.-chain and .alpha. chain cDNA clones include the pUC/M13 forward (-40) and pUC/M13 reverse primers (New England Biolabs) and primers Aint1, Aint2, Bint1, Bint2, bint3, bint4.

Aint1 (SEQ ID NO:88) 5'(GGACGATTTGCCAGCTT)-3'

Aint2 (SEQ ID NO:89) 5'(CACCAGACCCACAGTCA)-3'

Bint1 (SEQ ID NO:90) 5'(GCTTCGACAGCGACGTG)-3'

Bint2 (SEQ ID NO:91) 5'(GTACACAGTCACCGTAG)3'

Bint3 (SEQ ID NO:92) 5'(ACCAGGAGGTTGTGGTG)-3'

Bint4 (SEQ ID NO:93) 5'(GCTCAGGAATCCTGTTG)-3'

PCR Amplification

The reactions are performed in a Perkin-ElmerCetus PCR System 9600. The amplification program consists of 35 cycles of the following thermal profile: 94.degree. C. for 30 seconds, 55.degree. C. for 1 minute and 72.degree. C. for 1 minute.

Gel Purification Of DNA Fragments

The DNA bands (restriction fragments or PCR products) are excised from agarose gels and DNA purified using either the Geneclean.TM. II DNA Isolation (for fragments >1000 bp) or MerMaid DNA Isolation kits (for fragments <1000 bp) using the manufacturer's protocols (BIO101, La Jolla, Calif.). The DNA is eluted from the glass beads using water.

DNA Sequencing

The sequence of the double stranded plasmid DNAs is determined following either the Sequenase.TM. Sequencing System (United States Biochemical, Cleveland, Ohio) or the TaqTrack.TM. Sequencing System (ProMega, Madison, Wis.) protocols. Nonradioactive sequencing using the biotinylated T7 and T3 primers is done using the UniPlex.TM. Luminescent Kit (Millipore, Bedford, Mass.). DNA analysis is performed on a Macintosh SE/30 using the MacVector.TM. programs of International Biotechnologies, Inc. (New Haven, Conn.).

Anti-FLAG Colony Screening Assay

Colonies containing appropriate size inserts are selected from each transformation and streaked onto a hydrophilic PVDF (polyvinylidene difluroide) membrane. The membrane-bound colonies are grown overnight (approximately 16 hours) at 37.degree. C. on an agar plate containing Luria nutrient broth (LB) and chloramphenicol at 25.mu.g/mL. Interspersed between the hydrophilic membrane and the agar/LB plate is a hydrophobic PVDF membrane. This second membrane acts as a "protein trap" which can be used to examine the protein secretion profile of each of the colonies on the overlying membrane. In most cases, this hydrophobic filter is developed using the anti-FLAG antibody M2 (International Biotechnologies, New Haven, Conn.) and anti-mouse alkaline-phosphatase conjugate according to standard procedures.

Protein Expression and Purification

Single chain DR.beta.1/.alpha.1-FLAG proteins are expressed constitutively in recipient E. coli strains under the transcriptional control of the penicillinase P promoter from Bacillus licheniformis and contain the C-terminal octamer peptide referred to as FLAG.TM. (see Example 5). Lysates are analyzed for expression of FLAG-containing peptides directly by standard immunoblotting procedures using the anti-FLAG antibody M2. Lysates for either direct analysis or for purification are prepared as follows. Overnight cultures of plasmid containing colonies (approximately 2 mL for direct blotting; 100-500 mL for purification) are pelleted by centrifugation at approximately 5,000 rpm in an SS34 rotor) at 40.degree. C., resuspended in phosphate-buffered saline (PBS), sonicated through three 20 second cycles at a setting of 14 microns and pelleted by centrifugation at 10,000 rpm (SS34 rotor). For purification, the lysate is filtered through a 0.2 micron syringe-mounted filter. Affinity purification is accomplished through use of a 10 mL or 1 mL anti-FLAG.TM. affinity columns. Loading and washes are in PBS. Elution is with 0.10 M glycine pH 3.12. Fractions containing the eluted peak are neutralized and analyzed for the presence of FLAG-containing proteins by SDS-PAGE and immunoblotting with the anti-FLAG antibody.

Immune sera from mice immunized with mouse L cells expressing HLA-DRB*0101 is also used to probe membranes. The assay is conducted as described above for anti-FLAG except that anti-FLAG is replaced by a 1:80 dilution of serum isolated from an immunized mouse.

Enzyme-linked Immunosorbent Assays (ELISAs)

Commercially available anti-HLA-DR and anti-FLAG antibodies and supernatants from hybridomas derived from HLA-DR-immunized mice are assayed for the capacity to bind wells coated with the purified proteins according to standard procedures. Such as described in Caligan, S. E. et al., Current Protocals in Immunology, New York: John Wiley & Sons, Inc., (1991)

DR.alpha.1 and DR.beta.1 Amplification and Cloning

To isolate independently folding HLA-DR.beta.1 and -.alpha.1 domains and to intersperse them with a linker in such a way as to optimize their potential interaction in forming a class II MHC-like protein structure when expressed in microbial systems, an amino acid which is predicted to be in the region extending from the .beta.1 domain to the .beta.2 domain (or .alpha.1 to .alpha.2) as the C-terminal codon of the amplified product is selected [Brown et al., Nature 332, 845-850 (1988)]. For .beta.1, amino acid 102 is selected and for .alpha.1, amino acid 84 is selected (Kabat et al., Supra)). Oligonucleotide primers are synthesized such that the N-terminal and C-terminal codons are flanked by restriction enzyme sites appropriate for cloning into the expression vector pATD-FLAG. Specifically, amplification with the .beta.1-specific primers results in PCR products containing an Nco I site near the 5'-end and a Hind III site near the 3'-end. Amplification with the .alpha.1-specific primers (SCDRAS.1 and SCDRA3.1) gives products flanked by Xho I and Nhe I sites on the .sub.5 '- and 3'-ends, respectively.

The .beta.1 domain is amplified from plasmids containing DRB1 alleles *0101, *0401, *0402 and *0408, (pcD-SYN.beta.60, pcD-RAE.beta.60, pcD-DAK.beta.21 and pcD-SYN.beta.65, respectively. Bands of approximately 320 bp are gel isolated, purified using the MerMaid kit, digested with Nco I and Hind III and gel purified again. For subcloning, the PCR products and the plasmid pATD-FLAG' are digested with Nco I and Hind III, purified and ligation reactions setup either with or without digested .beta.-chain PCR DNA. Plasmid pATD-FLAG' differs from pATD-FLAG only in the Xho I-to-Nhe I spacer (see sequences below). In pATD-FLAG, this spacer contains of 8 bp of DNA, placing the FLAG peptide translationally out of frame with the linker, thereby preventing constructs containing only .beta.1-encoding cDNA (between the Nco I and Hind III sites) from expressing the corresponding FLAG-containing polypeptide. In pATD-FLAG', this spacer contains of 18bp of DNA which places the FLAG peptide in-frame with the DR.alpha..beta. domain and linker allowing the constructs containing only the DR.beta.1 insert to express a FLAG-containing protein.

Colonies containing appropriate size inserts (approximately 315 bp) are selected from each ligation and streaked on a hydrophilic PVDF membrane. The membrane-bound colonies are assayed for production of FLAG peptide and FLAG-positive colonies from each set selected for sequencing. The HLA-DR.beta.1 insert from each of these colonies is subcloned into pSL301 (Invitrogen Corporation, La Jolla, Calif.) and the nucleotide sequence determined nonradioactively using biotinylated T7- and T3-promoter primers and the Sequenase.TM. kit.

Subcloned Nco I/Hind III fragments are assigned names corresponding to their respective HLA-DR alleles; "sdDR.beta.11" for the (sd=single domain) .beta.1-encoding domain of DRB1*0101, "sdDR.beta.41" for allele DRB1*0401, "sdDR.beta.42" for allele DRB1*0402 and "sdDR.beta.48" for allele DRB1*0408. DNA sequencing (FIGS. 30 (SEQ ID NO:23), 31 (SEQ ID NO:25), 32 (SEQ ID NO:27) and 33 (SEQ ID NO:29) for plamids containing sdDR.beta.11, sdDR.beta.42, sdDR.beta.48 and sdDR.beta.42, respectively) reveal that both sdDR.beta.11 and sdDR.beta.42 contained nucleotide sequences identical to the parental HLA-DR cDNAs from which they had been amplified. The sequences extended from the first amino acid codon of the mature .beta.1 domain (GGG=Gly) to amino acid codon number 102 (TAT=Tyr).

In FIGS. 30 through 33, the transcriptional unit consists of the Bacillus licheniformis penicillinase P promoter (nucleotides 2-292) and terminator (nucleotides 1076-1292), the pel B signal peptide (nucleotides 292-352) and the corresponding scDR.beta./.alpha.-FLAG protein-coding region (nucleotides 358-1072). The corresponding scDR.beta./.alpha.-FLAG region is composed of the DR.beta. domain, encoded by nucleotides 358-664 (between the Nco I and Hind III sites), the linker regions (nucleotides 664-742; underlined), the corresponding DR.alpha. domain, encoded by nucleotides 743-1048 (between the Xho I and Nhe I sites) and the FLAG peptide-coding regions (nucleotides 1049-1072; dotted and underlined).

The psdDR.beta.-FLAG constructs lack the DR.alpha. coding domain and instead contain a linker formed by ligation of the annealed oligonucleotides XSN and NSX into the Xho I/Nhe I digested vector. The Xho I to Nhe I insertion is shown as follows:

The sdDR.beta.48 is originally intact only from codon 1 to codon 87, which is altered from GAA to GAG during semisynthesis of the HLA-DRB1*0408 cDNA). FIG. 32 shows the nucleotide and amino acid sequences after the missing codons 87-102 are replaced as follows. Briefly, the change in codon 87 results in the generation of a Hind III site at codons 87-89 such that digestion of the DR.beta.48 PCR product with Nco I and Hind III gives a Nco I/Hind III fragment with a 3'-overhang at codon 88 instead of codon 103. Thus, it is necessary to reinsert the codons corresponding to amino acid residues 87-102. Because of the similarity between HLA-DRB1*0101 and HLA-DRB1*0408 in this region and because both alleles contain a Sac II site at codons 73-75, this is done by replacing the Sac II/Hind III region of the original (truncated) sdDR.beta.48 construct with the Sac II to Hind III region of sdDR.beta.11 (sdDR.beta.41 or sdDR.beta.42 may also be used). In the example shown (FIG. 32), the Sac II/Hind III region of sdDR.beta.11 is used to replace the missing codons. The .beta.1 domain of the resulting construct differs in amino acid sequence from the native HLA-DRB1*0408 .beta.1 domain only at amino acid 96 (glutamic acid) which in the native *0408 allele is a tyrosine. This difference lies outside the region of the class II MHC molecule thought to be involved in antigen-binding and is not expected to alter the overall functional properties of the encoded protein.

The .alpha.1 domain is amplified from the plasmid pcD-KEK.alpha.4 (the pCD vector is described in Okayama, H. et al., Molecular and Cellular Biology, 3, 280-289 (1983). The pCD-1 vector is identical to the original vector except that the Pst I-Kpn I multiple cloning site is replaced by ligating the Pst I/Kpn I digested the following annealed oligonucleotides

pCD-1 (SEQ ID NO:99) 5'GAGAATTCAGGTAC-3'

pCD-2 (SEQ ID NO:100) ACGTCTCTTAAGTCC-3'

A band of approximately 270 bp is gel isolated, purified using the MerMaid kit, digested with Xho I and Nhe I and gel purified again. For subcloning, the PCR product and plasmids approximately 5-10 ng) encoding the .beta.1 domains (inserted between the Nco I and Hind III sites of pATD-FLAG') are digested with Xho I and Nhe I, purified and ligation reactions set up either with or without (.about.20 ng)digested .alpha.-chain PCR DNA. Competent E. coli cells are transformed with 1-2 .mu.l of ligation mix and plated on LB/chloramphenicol plates. When insert-containing ligation mixes give more colonies per plate than the negative control ligations, colonies are picked, grown overnight in 2 mL LB/chloramphenicol medium and plasmid DNA is isolated. Restriction digests (Xho I and Nco I or Nco I and Nhe I) confirm that a band of the right size has been inserted. The .alpha.-chain is excized using Xho I and Nhe I and subcloned into plasmid pSL301 and the nucleotide sequence determined.

sdDR.beta.11-FLAG Expression

Single-domain (sd) DR.beta.11-FLAG protein is purified by sonication and affinity chromatography as described above. The sdDR.beta.11-FLAG (and scDR.beta.42) protein elutes in a single peak. Fractions from the chromatographic elution are analyzed by SDS-PAGE and anti-FLAG immunoblotting. Coomassie-stained SDS-polyacrylamide gels show that the major Coomassie stained band in fractions in the purified preparation runs more quickly than the purified single-chain antibody sc49-FLAG (MW approximately 29.0 kD) and migrates at a MW of about 15,000. The immunoblot of the same gel shows that the purified band is recognized by anti-FLAG antibody. The Coomassie-stained gel also shows that the scDR.beta.11-FLAG protein constitutes a fairly minor component of the whole lysate (<1 percent) and the major protein (>50 percent) of the purified protein. Thus, sdDR.beta.11-FLAG (and sdDR.beta.42-FLAG; data not shown here) appears to be expressed as a stable protein with an apparent molecular weight of .about.15,000 as expected for a molecule consisting of 141 amino acids (DR.beta.1 domain=102 amino acids; linker=26 amino acids; FLAG' filler region=3 amino acids; and the FLAG-peptide coding region including the Nhe I site=10 amino acids). Although these constructs contain the DR.beta.1 domain, a linker region and a FLAG-peptide coding region, it is unlikely that the latter two components contribute significantly to the overall stability of the scDR.beta.1 proteins. These regions can be removed by digestion of the psdDR.beta.-plasmids with Hind II and Nhe I and ligation of a linker which introduces a stop codon immediately after the Hind III site. One such linker is generated by ligation of the oligonucleotides sdDRtruncl and sdDRtrunc-2 into the digested psdDR.beta.-plasmid. Sequences of two such oligonucleotides are as follows:

sdDRtrunc-1 (SEQ ID NO:101) 5'AGCTTTAAG-3'

sdDRtrunc-2 (SEQ ID NO:102) 3'AATTCGATC-5'

where the stop codon (TAA) is shown in bold letters.

To ascertain whether an immune response directed against HLA-DRB1*0101 might also recognize the sdDR.beta.11-FLAG protein, an immunoblot of the purified sdDR.beta.11 protein is developed using serum derived from a C3H mouse immunized with mouse L cells expressing the human HLA-DRB1*0101 protein. The immunoblot shows that the major band in this blot is the 15 kD sdDR.beta.11-FLAG. Thus, it appears the single domain DR.beta.1 proteins are stable and can be recognized by polyclonal immune serum raised against the native HLA-DRB1*0101 molecule. ELISA results (FIG. 34) shows that hybridomas generated by immunization of mice with cells expressing native HLA-DRB1*0101 can be found which are capable of binding the sdDR.beta.11-FLAG protein. These results suggest that the overall structure of the free sdDR11-FLAG protein is similar to that of the .beta.1 domain of the nature molecule. This is of considerable value in the discovery of antibodies and other molecules capable of binding discrete subsets of Class II MHC proteins. It also suggests considerable utility of the free .beta.1 domain as an immunogen as a potential therapeutic in suerantigen-dependent disorders.

Expression and Purification of Single-Chain DR.beta./.alpha.-FLAG Proteins

As described for the sdDR.beta.11-FLAG above, scDR.beta./.alpha.-FLAG proteins are also being expressed and purified from E. coli transformed with plasmids encoding these molecules. The scDR.beta./.alpha.-FLAG proteins are similar in size to a single chain antibody (such as sc49-FLAG) and migrate at a MW of about 25 to 30 kD. The scDR.beta./.alpha.-FLAG (single-chain DR) proteins can be purified by affinity chromatography in a manner similar to that described for the sdDR.beta.-FLAG proteins described above. Thus, a Coomassie-stained SDS-polyacrylamide gel will show that the purified protein migrates at a MW similar to that of purified sc49-FLAG. The affinity-purified protein represents the major protein component of the purified preparation. The similarity of the single-chain DR to the native HLA-DR molecule is being assessed in a number of ways. First, polyclonal immune sera raised in mice against cell surface HLA-DR is being used in immunoblotting experiments to determine whether antibodies raised against the native protein bind to the engineered single-chain molecule. Secondly, immunoblots and enzyme-linked immunosorbent assays (ELISA) are being used to assess the capacity of commercially available and public domain antibodies known to recognize HLA-DR to bind to the purified scDR.beta./.alpha.-FLAG proteins. Thirdly, hybridomas derived from HLA-DR-immunized mice are being screened for the capacity to bind to the purified single-chain proteins. Functional activity of the scDR.beta./.alpha.-FLAG molecules is being measured by competition-assays involving an influenza peptide (influenza hemagglutinin 307-319) known to bind a wide variety of class II MHC proteins. The data suggests that, like sdDR.beta.-FLAG proteins, the scDR.beta./.alpha.-FLAG proteins are expressed as stable proteins in E. coli , that they migrate at the appropriate MW and are recognized on immunoblots by at least one publicly available antibody (L227) known to bind native HLA-DR.

Characterization of Dimers and Multimers of scDR.beta./.alpha.-FLAG Proteins

The presence of dimers and multimers of single chain DR.beta./.alpha. molecules (scDR(.beta./.alpha.).sub.2 and scDR(.beta./.alpha.).sub.n respectively) is assessed using essentially the same techniques as described for single chain antibody molecules (Example 1). Briefly, the affinity purified scDR.beta./.alpha.-FLAG preparation is subjected to size-exclusion (gel filtration) chromatograph on a column containing an appropriate matrix (such as Superdex 75). The presence of peaks at molecular weights corresponding to approximately 48,000, approximately 96,000 and approximately 144,000 kilodaltons indicates the presence of dimers, tetramers, hexamers as is the case with the scFv molecules. The absence of peaks at approximately 72,000, approximately 120,000 and approximately 168,000 kilodaltons indicates the lack of trimers, pentamers and heptamers, respectively.

Given the present model of multimer formation, the latter peaks are usually absent or very minor components. Isoelectric focusing (of the purified preparation or crude lysate) is also used to assess the presence of distinct charged forms of the scDR.beta./.alpha.-FLAG indicating the influence of multimerization on the ionic properties of the single chain protein. Denaturing gel electrophoresis of the purified protein is used, like gel filtration, to confirm the presence of multimeric forms of scDR.beta./.alpha. and to estimate the sizes of such multimeric forms.

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

Claims

What is claimed is:

1. A method for in vivo diagnosis of TAG-72-expressing cancer which comprises administering to an animal a pharmaceutically effective amount of a composition comprising a protein comprising a dimer or multimer of single chain polypeptides (SCPs), each SCP consisting of (a) a first polypeptide comprising an antigen binding portion of a light or heavy chain variable domain of antibody CC49, (b) a second polypeptide comprising an antigen binding portion of a light or heavy chain variable domain of antibody CC49, and (c) a peptide linker linking the first and second polypeptides (a) and (b) into one SCP, wherein (I) each of the first and second polypeptides (a) and (b) retains the capability of being assembled into an Fv region having binding affinity for TAG-72; (II) each SCP is non-covalently linked to at least one other SCP of the protein, to form said dimer or multimer, by the direct, non-covalent association of their polypeptides (a), (b), or (a) and (b) into Fv regions having binding affinity for TAG-72; (III) the protein is conjugated to an imaging marker; and (IV) the composition comprises at least one pharmaceutically acceptable carrier.

2. The method of claim 1 wherein the imaging marker is .sup.125 I, .sup.131 I, .sup.111 In, .sup.105 Rh, .sup.153 Sm, .sup.67 Cu, .sup.67 Ga, .sup.166 Ho, .sup.177 Lu, .sup.186 Re, .sup.188 Re or .sup.99m Tc.

3. The method of claim 1 wherein the protein is a dimer.

4. The method of claim 1 wherein the protein is a multimer.

5. The method of claim 1 wherein the linker is from about 10 to about 50 amino acid residues.

Patent Citations (8)

Patent	Date	Inventor
US4474893	1984-10-01	Reading et al.
US4642334	1987-02-01	Moore et al.
US4714681	1987-12-01	Reading
US4946778	1990-08-01	Ladner et al.
US5001225	1991-03-01	Taylor
US5130297	1992-07-01	Sharma et al.
US5132405	1992-07-01	Huston et al.
WOX8809344	1988-12-01

Non-Patent Literature (35)

Filpula et al (Antibody Engineering, Eds McCafferty et al., IRL Press, Oxford, UK, pp. 253-268), 1996.
Hunkapiller et al., "Implications of the Diversity of the Immunoglobulin Gene Superfamily", Cold Spring Harbor Symposia on Quantitative Biology, LIV, 1989, pp. 15-29, Cold Spring Harbor Laboratory Press.
Wraith et al., "T-cell Recognition as the Target for Immune Intervention in Autoimmune Disease", Cell, vol. 57, pp. 709-715, Jun. 2, 1989.
Adorini, et al., "New Perspectives on Immunointervention in Autoimmune Diseases", Immunology Today, vol. 11, No. 11, pp. 383-386 (1990).
Bird et al., "Single-Chain Antigen-Binding Proteins", Science, vol. 242, pp. 423-426, Oct. 21, 1988.
Skerra et. al., "Assembly of a Functional Immunoglobulin, F.sub.v Fragment in Escherichia coli", Science, vol. 240, pp. 1038-1041, May 20, 1988.
Takkinen et al., "An Active Single-Chain Antibody Containing a Cellulase Linker Domain is Secreted by Escherichia coli", Protein Engineering, vol. 4, No. 7, pp. 837-841 (1991).
Traunecker et al., "Bispecific Single Chain Molecules (Janusins) Target Cytotoxic Lymphocytes on HIV Infected Cells", The EMBO Journal, vol. 10, No. 12, pp. 3655-3659 (1991).
Huston et al., "Protein Engineering of Antibody Binding Sites: Recovery of Specific Activity in an Anti-digoxin Single-Chain Fv Analogue Produced in Escherichia coli", Proc. Natl. Acad. Sci., Biochemistry, vol. 85, pp. 5879-5883, Aug. 1988.
Larson, "Improved Tumor Targeting With Radiolabeled Recombinant, Single-Chain, Antigen-Binding Protein", Journal of the National Cancer Institute, vol. 82, No. 14, pp. 1173-74, Jul. 18, 1990.
Colcher et al., In Vivo Tumor Targeting of a Recombinant Single-Chain Antigen-Binding Protein, Journal of the National Cancer Institute, vol. 82, No. 14, pp. 1191-1197, Jul. 18, 1990.
Bedzyk et al., "Immunological and Structural Characterization of a High Affinity Anti-fluorescein Single-chain Antibody", The Journal of Biological Chemistry, V. 265, No. 30, pp. 18615-18620, Oct. 25, 1990.
Winter et al., "Man-made Antibodies", Nature, vol. 349, pp. 293-299, Jan. 24, 1991.
Davis, "Single Chain Antibody (SCA) Encoding Genes: One Step Construction and Expression in Eukaryotic Cells", Biotechnology, vol. 9, pp. 165-169, Feb. 1991.
Whitlow et al., "Single-Chain Fv Proteins and Their Fusion Proteins", Methods: A Companion to Methods in Enzymology, vol. 2, No. 2, pp. 97-105, Apr. 1991.
Skerra et al., "The Functional Expression of Antibody F.sub.v Fragments in Escherichia coli: Improved Vectors and a Generally Applicable Purification Technique", Biotechnology, vol. 9, pp. 273-278, Mar. 1991.
Gibbs et al., "Construction and Characterization of a Single-Chain Catalytic Antibody", Proc. Natl. Acad. Sci., U.S.A., vol. 88, pp. 4001-4004, May 1991.
Hodgson, "Making Monoclonals in Microbes", Biotechnology, vol. 9, pp. 421-425, May 1991.
Pluckthun, "Antibody Engineering: Advances From the Use of Escherichia coli Expression Systems", Biotechnology, vol. 9, pp. 545-550, Jun. 1991.
Bird et al., "Single Chain Antibody Variable Regions", Trends in Biotechnology, vol. 9, pp. 132-137 (1991).
Glockshuber et al., "A Comparison of Strategies to Stabilize Immunoglobulin F.sub.v -Fragments", Biochemistry, vol. 29, pp. 1362-1367 (1990).
Denzin et al., "Single-chain Site-specific Mutations of Fluorescein-Amino Acid Contact Residues in High Affinity Monoclonal Antibody 4-4-20", Journal of Biological Chemistry, vol. 266, No. 21, pp. 14095-14103 (1991).
Pantoliano et al., "Conformational Stability, Folding, and Ligand-Binding Affinity of Single-Chain Fv Immunoglobulin Fragments Expressed in Escherichia coli", Biochemistry, vol. 30, pp. 10117-10125 (1991).
Gregoire et al., "Engineered Secreted T-Cell Receptor .alpha..beta. Heterodimers", Proc. Natl. Acad. Sci., U.S.A., vol. 88, pp. 8077-8081, Sep. 1991.
Pack et al., "Miniantibodies: Use of Amphipathic Helices to Produce Functional, Flexibly Linked Dimeric F.sub.v Fragments with High Avidity in Escherichia coli", Biochemistry, vol. 31, No. 6, pp. 1579-1584 (1992).
Soo Hoo et al., "Characterization of a Single-chain T-cell Receptor Expressed in Escherichia coli", Proc. Natl. Acad. Sci., U.S.A., vol. 89, pp. 4759-4763, May 1992.
Novotny et al., "A Soluble, Single-chain T-cell Receptor Fragment Endowed With Antigen-combining Properties", Proc. Natl. Acad. Sci., U.S.A., vol. 88, pp. 8646-8650, Oct. 1991.
Williams et al., "The Immunoglobulin Superfamily-Domains for Cell Surface Recognition", Ann. Rev. Immunol., vol. 6, pp. 381-405 (1988).
Williams et al., "Structural Diversity in Domains of the Immunoglobulin Superfamily", Cold Spring Harbor Symposia on Quantitative Biology, LIV, pp. 637-647, 1989, Cold Spring Harbor Lab. Press.
Bradwel et al (Monoclonal Antibodies for Cancer Detection & Therapy, 1985, Academic Press, London pp. 65-85.
Waldman, Science, 1991, 252:1657-1662.
Sheer et al, Cancer Res, 48: 6811-6818, 1988.
Rodwell et al, Biotechnology, 1985, 3: 889-894.
Hird et al (Genes and Cancer, John Wiley & Sons, 1990, Chichester, pp. 183-189.
Kramer (4.sup.th Annual Symposium, Jan. 23-26, 1992, Diagnostic & Therapeutic Applications in Benign & Malignant Disorders Key Biscayne FL, pp. 52-54, Meeting Abstract.

Abstract

Patent Drawings (73 sheets)