United States Patent [19]
`Moore et al.
`
`[54] HYBRID DNA PREPARED BINDING
`COMPOSITION
`
`[75] Inventors: Kevin W. Moore, San Bruno;
`Alejandro Zaffaroni, Atherton, both of
`Calif.
`
`[73] Assignee: Schering Corporation
`
`(cid:9) (cid:9)
`
`[21] Appl. No.: 461,071
`
`[22] Filed: (cid:9)
`
`Jun. 5, 1995
`
`Related U.S. Application Data
`
`[62] Division of Ser. No. 394,923, Feb. 23, 1995, abandoned,
`which is a continuation of Ser. No. 210,540, Mar. 17, 1994,
`abandoned, which is a continuation of Ser. No. 61,760, May
`13, 1993, abandoned, which is a continuation of Ser. No.
`928,526, Aug. 11, 1992, abandoned, which is a continuation
`of Ser. No. 740,862, Jul. 31, 1991, abandoned, which is a
`continuation of Ser. No. 235,835, Aug. 18, 1988, abandoned,
`which is a continuation of Ser. No. 558,551, Dec. 5, 1983,
`Pat. No. 4,642,334, which is a continuation of Ser. No.
`358,414, Mar. 15, 1982, abandoned.
`[51] Int. C1.6
` C12P 21/02; C12P 21/08;
`C12N 1/21; C12N 15/13
` 435/696; 435/172.3; 435/252.33
`[52] U.S. Cl. (cid:9)
` 735/240.2, 172.3,
`[58] Field of Search (cid:9)
`735/69.1, 69.6; 514/2; 530/387.1, 387.3;
`435/69.1, 69.6, 172.1, 172.3, 320.1, 252.33
`
`[56]
`
`References Cited
`
`U.S. PATENT DOCUMENTS
`
`4,036,945 7/1977 Haber et al. (cid:9)
`4,642,334 2/1987 Moore et al. (cid:9)
`
` 424/1.49
` 530/388
`
`FOREIGN PATENT DOCUMENTS
`
`A 0 035 265 9/1981 European Pat. Off. .
`
`OTHER PUBLICATIONS
`
`Bodey et al., "Human Cancer Detection and Immunotherapy
`with Conjugated and Non—conjugated Monoclonal Antibod-
`ies", Anticancer Res., vol. 16, pp. 661-674 (1996).
`Jain, "Vascular and Interstitral Barners to Delivery of Thera-
`peutic Agents in Tumors", Cancer Metastasis Rev. vol. 9(3),
`pp. 253-266, 1990 (Abstract only).
`Baumgartner et al., "Immunotherapy of Endothoxemia and
`Septicemia", Immunobiol., vol. 187, pp. 464-477, (1993).
`Givol, 1974, Essays in Biochemistry 10:73-104.
`Gough, Aug., 1981, TIBS 203-205.
`Early et al., 1981, Genetic Engineering: Principles and
`Methods 3:157-183.
`Leder, 1982, Scientific American 19:2702-2710.
`
`licomilollomilpillpoommoulmou
`
`[11] Patent Number: (cid:9)
`[45] Date of Patent: (cid:9)
`
`5,840,545
`Nov. 24, 1998
`
`Gough et al., 1980, Biochemistry 19: 2702-2710.
`Adams et al., 1980, Biochemistry 19: 2711-2719.
`Inbar et al., Proc. Nat. Acad. Sci. USA 69(9):2659-2662.
`Hochman et al., 1973, Biochemistry 12(6):1130-1135.
`Sharon et al., 1976, Biochemistry 15(7):1591-1594.
`Kooistra et al., 1978, Biochemistry 17(2):345-351.
`Accolla et al., 1980, Proc. Nat. Acad. Sci., USA
`77(1):563-566.
`O'Sullivan et al., 1979, Annals of Clinical Biochemistry
`16:221-240.
`Miozzari et al., 1978, J. Bacteriol 133(3):1457-1466.
`Pliickthun, 1991, Bioltechnology 9:545-551.
`Kakimoto et al., 1974, The Journal of Immunology
`112(4): 1373-1382.
`Lin et al., 1978, Proc. Natl. Acad. Sci. USA
`75(6):2649-2653.
`Rosemblatt et al., 1978, Biochemistry 17:3877-3882.
`Cunningham, Understanding Immunology, pp. 38-39.
`Alberts, 1989, Molecular biology of the cell, pp. 1025-1026.
`Francis et al., 1974, Proc. Nat. Acad. Sci., 71:1123-1127.
`Raso et al., 1980, Journal of Immunology
`125(6):2610-2616.
`Hales et al., 1980, Methods in enzymology 70:334-355.
`O'Sullivan et al., 1981, Methods in enzymology 73:147-166.
`Boss et al., 1984, Nucleic Acids Research 12:3791-3806.
`Wood et al., 1985, Nature 314:446-448.
`Cabilly et al., 1984, Proc. Natl. Acad. Sci. USA
`81:3273-3277.
`Kenton et al., 1984, Proc. Natl. Acad. Sci. USA
`81:2955-2599.
`Liu et al., 1984, Proc. Natl. Acad. Sci. USA 81:5369-5373.
`Gherna et al., 1985, American Type Culture Collection
`Sixteenth edition, p. 243.
`Gubler et al., 1986, J. Immunol. 136:2492-2497.
`Saxena et al., 1970, Biochemistry 9:5015-5023.
`
`Primary Examiner—David Guzo
`Attorney, Agent, or Firm—Cynthia L. Foulke; Edwin P.
`Ching
`
`[57] (cid:9)
`
`ABSTRACT
`
`Proteinaceous binding compositions are prepared employing
`hybrid DNA technology, where the variable region polypep-
`tides of immunoglobulins are substantially reproduced to
`provide relatively small protein molecules having binding
`specificity and lacking the undesirable aspects of the heavy
`regions of immunoglobulins. The compositions find a wide
`range of use, particularly for physiological purposes for
`diagnosis and therapy. The binding compositions may be
`modified by labeling with radioisotopes, fluorescers, and
`toxins for specific applications in diagnosis or therapy.
`
`2 Claims, No Drawings
`
`Genzyme Ex. 1019, pg 580
`
`(cid:9)
`

`
`5,840,545
`
`2
`host antibodies against the constant region of the immuno-
`globulin or against any other part of the molecule.
`It is therefore important that methods be developed which
`permit the preparation of homogeneous compositions hav-
`ing high specificity for a particular ligand, while avoiding
`the shortcomings of complete immunoglobulins, and pro-
`viding the many advantages of lower molecular weight.
`2. Description of the Prior Art
`Discussions concerning variable regions of heavy and
`light chains of immunoglobulins may be found in Sharon
`and Givol, Biochem. (1976) 15:1591-1594; Rosemblatt and
`Haber, Biochem. (1978) 17:3877-3882; and Early and
`Hood, Genetic Engineering (1981) 3:157-188. Synthesis of
`part of a mouse immunoglobulin light chain in a bacterial
`clone is described by Amster et al., Nucleic Acids Res.
`(1980) 8:2055-2065. See also the references cited through-
`out the specification concerning particular methodologies
`and compositions.
`
`SUMMARY OF THE INVENTION
`
`5
`
`10
`
`15
`
`20 (cid:9)
`
`Novel protein complexes are provided by producing
`homogeneous compositions defining the variable regions of
`the light and heavy chains of an immunoglobulin, which
`25 individually or together form a specific binding complex to
`a predetermined haptenic or determinant site. Employing
`hybrid DNA technology, cDNA is tailored to remove nucle-
`otides extraneous to the variable regions of the light and
`heavy chains. The resulting tailored ds cDNA is inserted into
`30 an appropriate expression vector which is then introduced
`into a host for transcription and translation. The resulting
`truncated light and heavy chains define at least a major
`portion of the variable regions and are combined to form a
`complex capable of specifically binding to a predetermined
`35 haptenic site with high affinity
`
`DESCRIPTION OF THE SPECIFIC
`EMBODIMENTS
`
`1
`HYBRID DNA PREPARED BINDING
`COMPOSITION
`
`This application is a division of application Ser. No.
`08/394,923, filed Feb. 23, 1995, now abandoned, which is a
`continuation of application Ser. No. 08/210,540, filed Mar.
`17, 1994, now abandoned, which is a continuation of appli-
`cation Ser. No. 08/061,760, filed May 13, 1993, now
`abandoned, which is a continuation of application Ser. No.
`07/928,526, filed Aug. 11, 1992, now abandoned, which is
`a continuation of application Ser. No. 07/740,862, filed Jul.
`31, 1991, now abandoned, which is a continuation of appli-
`cation Ser. No. 07/235,835, filed Aug. 18, 1988, now
`abandoned, which is a continuation of application Ser. No.
`06/558,551, filed Dec. 5, 1983, now U.S. Pat. No. 4,642,334,
`which is a continuation of application Ser. No. 06/358,414,
`filed Mar. 15, 1982, now abandoned.
`
`BACKGROUND OF THE INVENTION
`
`1. Field of the Invention
`The mammalian immunological system is unique in its
`broad ability to produce protein compounds having
`extremely high specificity for a particular molecular struc-
`ture. That is, the proteins or immunoglobulins which are
`produced have a conformation which is specifically able to
`complement a particular structure, so that binding occurs
`with high affinity. In this manner, the mammalian immune
`system is able to respond to invasions of foreign molecules,
`particularly proteins in surface membranes of
`microorganisms, and toxins, resulting in detoxification or
`destruction of the invader, without adverse effects on the
`host.
`The primary immunoglobulin involved in the defensive
`mechanism is gamma-globulin (IgG). This immunoglobulin,
`which is a glycoprotein of about 150,000 daltons, has four
`chains, two heavy chains and two light chains. Each of the
`chains has a variable region and a constant region. The
`variable regions are concerned with the binding specificity
`of the immunoglobulin, while the constant regions have a
`number of other functions which do not directly relate to the
`antibody affinity.
`In many situations it would be desirable to have mol-
`ecules which are substantially smaller than the
`immunoglobulins, while still providing the specificity and
`affinity which the immunoglobulins afford. Smaller mol-
`ecules can provide for shorter residence times in a mamma-
`lian host. In addition, where the immunoglobulin has to be
`bound to another molecule, it will be frequently desirable to
`minimize the size of the final product. Also there are many
`economies in being able to produce a smaller molecule
`which fulfills the function of a larger molecule.
`There are situations where it will be desirable to be able
`to have a large number of molecules compactly held
`together. By having smaller molecules, a greater number can
`be brought together into a smaller space. Furthermore,
`where the binding molecule can be prepared by hybrid DNA
`technology, one has the opportunity to bind the binding
`portion of the molecule to a wide variety of other
`polypeptides, so that one can have the binding molecule
`covalently bonded at one or both ends to a polypeptide
`chain.
`Where immunoglobulins are used in in vivo diagnosis or
`therapy, antisera from an allogenic host or from a mono-
`clonal antibody may be immunogenic. Furthermore, when
`conjugates of other molecules to the antibody are employed,
`the resulting conjugate may become immunogenic and elicit
`
`50
`
`40 (cid:9)
`
`45
`
`The subject invention concerns a hybrid DNA strategy for
`the preparation of specific binding polypeptides, normally
`comprised of two different polypeptide chains, which
`together assume a conformation having high binding affinity
`to a predetermined ligand or haptenic site thereof. The
`polypeptide chains form binding sites which specifically
`bind to a predetermined ligand to form a complex having
`strong binding between the ligand and the binding site. The
`binding constant or avidity will generally be greater than
`105, more usually greater than 106, and preferably greater
`than 108. The haptenic binding site or determinant binding
`site of the polypeptide chain may be associated with a hapten
`or antigen.
`One or both of the different polypeptide chains derived
`from the variable region of the light and heavy chains of an
`55 immunoglobulin may be used to provide specific binding to
`a ligand. For the most part each of the polypeptide chains of
`the light and heavy variable regions would be employed
`together for binding to the ligand. In describing this
`invention, it will be understood that while the two different
`60 chains are indicated as forming a complex, either of the
`chains could be used individually, where feasible due to
`sufficient binding affinity of the particular chain to the
`reciprocal ligand.
`The two polypeptide chains which, individually or
`65 together, provide the compositions of this invention will
`form a receptor site, analogous to the binding site of an
`immunoglobulin. The composition will be referred to as an
`
`Genzyme Ex. 1019, pg 581
`
`

`
`1 5
`
`2
`
`30 (cid:9)
`
`3
`rFv with the individual chains referred to as L-rFv or H-rFv.
`The L- and H- designations will normally mean light and
`heavy respectively, but in some instances the two chains
`may be the same and derived from either the light or heavy
`chain sequences. The polypeptide chains of the rFv will
`generally have fewer than 125 amino acids, more usually
`fewer than about 120 amino acids, while normally having
`greater than 60 amino acids, usually greater than about 95
`amino acids, more usually greater than about 100 amino
`acids. Desirably, the H-rFv will be from about 110 to 125
`amino acids while the L-rFv will be from about 95 to 115
`amino acids.
`The amino acid compositions will vary widely, depending
`upon the particular idiotype involved. Usually there will be
`at least two cysteines separated by from about 60 to 75
`amino acids and joined by a disulfide bond to form cystine.
`The two chains will normally be substantial copies of
`idiotypes of the variable regions of the light and heavy
`chains of immunoglobulins, but in some situations it may be
`sufficient to have combinations of either the light or the
`heavy variable region chains.
`In many instances, it will be desirable to have one or both
`of the rFv chains labeled or bound to a support. Various
`labels may be employed, such as radioisotopes, fluorescers,
`or toxins. In other situations, one or both of the chains may
`be bound to an inert physiologically acceptable support,
`such as synthetic organic polymers, polysaccharides, natu-
`rally occurring proteins, or other non-immunogenic sub-
`stances.
`In some situations, it may be desirable to provide for
`covalent crosslinking of the two chains, which could involve
`providing for cysteine residues at the carboxyl termini. The
`chains will normally be prepared free of the constant
`regions, including or being free of all or a portion of the J
`region. The D region will normally be included in the
`transcript of the H-rFv.
`For the most part only a relatively small percent of the
`total amino acids will vary from idiotype to idiotype in the
`rFv. Therefore, there will be areas providing a relatively
`constant framework and areas that will vary, namely, the
`hypervariable regions.
`The C-terminus region of the rFv will have a greater
`variety of sequences than the N-terminus and, based on the
`present strategy, can be further modified to permit variation
`from the naturally occurring heavy and light chains. A
`synthetic oligonucleotide can be employed to vary one or
`more amino acids in a hypervariable region.
`The preparation of the rFv employing hybrid DNA tech-
`nology will now be described in greater detail.
`The preparation of the rFv will be divided into three parts:
`(1) isolation of appropriate DNA sequences; (2) introduction
`of the DNA sequences coding for the members of the rFv
`into an appropriate expression vector; and (3) expression
`and isolation of the mimetic variable regions of the light
`(L-rFv) and heavy (H-rFv) chains to provide the rFv.
`
`I. Isolation of Appropriate DNA Sequences.
`In preparing the DNA sequences, a source of the genes
`encoding the variable region will be required. The variable
`regions may be derived from IgA, IgD, IgE, IgG or IgM,
`most commonly, from IgM and IgG. This can be achieved by
`immunizing an appropriate vertebrate, normally a domestic
`animal, and most conveniently a mouse. The immunization
`may be carried out conventionally with one or more repeated
`injections of the immunogen into the host mammal, nor-
`mally at two to three week intervals. Usually three days after
`
`5,840,545
`
`5 (cid:9)
`
`4
`the last challenge, the spleen is removed and dissociated into
`single cells to be used for cell fusion to provide hybridomas.
`The immunogen will be the antigen of interest, or where
`a hapten, an antigenic conjugate of the hapten to an antigen.
`In order to prepare the hybridomas, the spleen cells are
`fused under conventional conditions employing a fusing
`agent, e.g. PEG6000, to a variety of inter- or intra- species
`myeloma cells, particularly mouse cells such as SP-2/0,
`NS-1, etc. and then suspended in HAT selective media. The
`1° surviving cells are then grown in microtiter wells and
`immunologically assayed for production of antibodies to the
`determinant site(s) of interest.
`Assays for antibodies are well known in the art and may
`employ a variety of labeled antigens or haptens, where the
`labels are conveniently radioisotopes, fluorescers, enzymes,
`or the like. Other techniques may also be employed, such as
`sandwich techniques involving two antibodies, one bound to
`a support and the other being labeled. The cells from
`microtiter wells scored as positive are cloned either by
`limiting dilution or cloning in soft agar. The resulting cloned
`cell lines are then propagated in an appropriate nutrient
`medium and, if necessary, may be stored frozen in liquid
`nitrogen.
`25 After selection of a particular cell line providing a mono-
`clonal antibody of interest, the cells are expanded.
`Conveniently, the cells may be grown to a density of about
`1x106 cells/ml in a 1 L culture. The cells are then harvested
`by centrifugation and lysed.
`In order to obtain the desired DNA sequence, one can look
`to either the gene expressing the variable region or the
`messenger RNA, which expresses the variable region. The
`difficulty with employing genomic DNA is in juxtaposing
`the sequences coding for the variable region, where the
`35 sequences are separated by introns. One must isolate the
`DNA fragment(s) containing the proper exons, excise the
`introns and then splice the exons in the proper order and
`orientation. For the most part, this will be difficult, so that
`the alternative technique employing the messenger RNA
`40 will be the method of choice.
`Where the messenger RNA is to be employed, the cells
`will be lysed under RNase inhibiting conditions. The mes-
`senger RNA has the advantage that the mature messenger is
`free of introns, so that the sequence is continuous for the
`45 entire variable region. Difficulties with messenger RNA
`have been encountered, due to incomplete reverse transcrip-
`tion but these difficulties can be minimized The first step is
`to isolate the messenger RNA. Conveniently, messenger
`RNA can be separated from other RNA because of its
`so polyadenylation, employing an oligo-(dT) cellulose column.
`The mixture of messenger RNAs will be obtained free of
`other RNA. The presence of messenger RNAs coding for the
`heavy and light chain polypeptides of the immunoglobulins
`may then be assayed by hybridization with DNA single
`55 strands of the appropriate genes. Conveniently, the
`sequences coding for the constant portion of the light and
`heavy chains may be used as probes, which sequences may
`be obtained from available sources (see, for example, Early
`and Hood, Genetic Engineering, Setlow and Hollaender eds.
`60 Vol. 3, Plenum Publishing Corp., New York (1981), pages
`157-188.)
`Whether the messenger RNA codes for the correct immu-
`noglobulin may be determined by in vitro translation
`employing a rabbit reticulocyte cell-free extract (Pelham and
`65 Jackson, Eurp. J. Biochem. (1976) 66:247-256). The result-
`ing translation product may then be isolated by employing
`antibodies specific for one or more of the regions of the
`
`Genzyme Ex. 1019, pg 582
`
`

`
`5,840,545
`
`5
`chain of interest, for example, using rabbit anti(mouse IgG)
`where the chains are derived from mouse immunoglobulin.
`The immunoprecipitate may be further analyzed by poly-
`acrylamide gel electrophoresis, and the presence of com-
`plexes determined by using radiotagged receptors for
`antigen-antibody complexes, such as S. aureus protein A, Rf
`factor, or the like. In addition, RNA blot hybridization can
`be employed to further insure that the correct messenger
`RNA is present.
`The crude mixture of mRNA sequences containing the
`desired mRNA sequences will be treated as follows. In order
`to enhance the probability that full length cDNA is obtained,
`the method of Okayama and Berg, Mol. Cell. Biol. (1982)
`may be employed. Alternatively, the methods described by
`Efstradiadis and Villa-Komaroff (1979) in Genetic Engi-
`neering: Principles and Methods 1, Setlow and Hollaender,
`eds., New York, Plenum Press, pages 15-36, or Steinmetz et
`al. (1981) Cell 24:125-134, may be employed. The first
`strand of cDNA is prepared employing a virus reverse
`transcriptase in the presence of primer. A second strand may
`then be prepared employing reverse transcriptase, the Kle-
`now fragment of DNA polymerase I or T4 polymerase. If
`necessary, the resulting ds cDNA may then be treated with
`a single-strand-specific nuclease, such as Si nuclease for
`removal of single stranded portions to result in ds cDNA,
`which may then be cloned.
`
`II. Preparation of Genes Coding For L-rFv and H-
`rFv and Introduction into an Expression Vector For
`Amplification
`
`A wide variety of vectors may be employed for amplifi-
`cation or expression of the ds cDNA to produce the light and
`heavy chains of the immunoglobulin. A vector having an
`appropriate restriction site is digested with the appropriate
`endonuclease. The ds cDNA obtained from the reverse
`transcription of the mRNA may be modified by ligating
`linkers, treatment with terminal transferase or other tech-
`niques to provide staggered (complementary) or blunt ended
`termini. The vectors may have one, two or more markers for
`selection of transformants. Desirably, the vector will have a
`unique restriction site in one of multiple markers, so that the
`transformants may be selected by the expression of one
`marker and the absence of expression of the other marker.
`Various markers may be employed, such as biocide
`resistance, complementation of an auxotroph, viral
`immunity, or the like.
`After transforming an appropriate host with the ds cDNA
`prepared from the mRNA, e.g. E. coli, B. subtilis, S.
`cerevisiae, etc., in accordance with conventional ways, the
`transformants are plated and selected in accordance with the
`particular markers. The resulting colonies are screened, by
`restriction electrophoretic pattern, hybridization to a labeled
`probe or by any other conventional means. See, for example,
`Hanahan and Meselson (1980), Gene 10:63-67. One proce-
`dure employs colony hybridization, where the transformants
`are grown on a solid medium to produce colonies. Cells
`from the colonies are transferred to a nitrocellulose replica
`filter, the transferred cells incubated for further growth,
`lysed, dried and baked. The replica filter is then hybridized
`with appropriate radio-isotope labeled probes. conveniently,
`there are readily available probes for the determinant sites
`present in the constant regions of a variety of mammalian
`immunoglobulins. The colonies may be probed based on the
`nature of the particular immunoglobulin, as well as the
`different determinant sites, which may be present with the
`particular immunoglobulin.
`
`10
`
`6
`The host colonies, usually bacterial, which have DNA
`which hybridizes to either the light or heavy chain probes are
`picked and then grown in culture under selective pressure. In
`order to maintain selective pressure, it is desirable that the
`5 vector which is employed have biocidal, particularly
`antibiotic, resistance. After sufficient time for expansion of
`the host, the host cells are harvested, conveniently by
`centrifugation. The hybrid plasmid DNA may then be iso-
`lated by known procedures. (Gunsalus et al., J. Bacteriol.
`(1979) 140:106-133).
`The isolated plasmid DNA is then characterized by
`restriction enzyme digestion and DNA sequence analysis.
`These analyses insure that the isolated cDNA clones com-
`pletely encode the variable region and, optionally, the leader
`15 sequences for the light or heavy chain of the desired immu-
`noglobulin. Furthermore, by having a restriction map of the
`variable regions and leader sequences, as well as the flank-
`ing sequences, one can determine the appropriate restriction
`sites for excising a DNA fragment which will allow for
`20 appropriate modification of the DNA sequence for insertion
`into a vector and expression of the polypeptide of interest.
`Where no unique restriction site is available at an appropri-
`ate position in the flanking regions, partial digestion may be
`employed, with selection of fragments having the variable
`25 region and, optionally, the leader sequence intact. Where the
`5' and 3' flanking regions are too extended, these can be
`chewed back using Bal 31 to varying degrees by varying the
`period of digestion.
`Furthermore, by knowing the DNA sequence of the cod-
`30 ing strand in the region of the C-terminus of the heavy and
`light chain variable regions, a stop codon may be introduced
`at the C-terminus by the following procedure of in vitro
`mutagenesis. The cDNA is restricted with the appropriate
`enzyme(s) to provide a variable region coding segment with
`35 additional 5' and 3' flanking sequences. This segment is
`purified, for example, by gel electrophoresis, gradient den-
`sity centrifugation, etc. After isolating the desired segment,
`the two strands of the segment are dissociated, conveniently
`by boiling. Alternatively, the undesired strand of the intact
`40 cDNA-plasmid clone may be nicked and digested.
`A synthetic, single-stranded DNA oligomer is prepared,
`conveniently by synthesis, which will have at least about 12
`nucleotides, more usually about 15 nucleotides, and will
`generally have fewer than about 50 nucleotides, usually
`45 fewer than 30 nucleotides, since a more extended oligomer
`is not required.
`Where heteroduplexing is involved, the non-
`complementary nucleotides will usually be flanked by at
`least about three, more usually at least about six nucleotides
`so complementary to the hybridized strand. The heteroduplex-
`ing oligonudleotide will be complementary to the sequence
`at or about a significant juncture i.e. between the leader
`sequence and the variable region or the variable region and
`the constant region. The synthetic DNA oligomer will be
`55 complementary to the coding ("sense") strand of the
`variable-region sequence, but altered to encode a termina-
`tion codon at the C-terminus of the variable region. That is,
`the oligomer will be complementary to the coding strand
`except at or about the amino acid which is involved at the
`60 juncture of the variable region and the D-, J- or C-regions of
`the light and heavy chains, particularly at or intermediate the
`D- or J-regions or intermediate the J-region, or at the J-
`region and C-region juncture. It is intended that there will be
`some variation in the polypeptides which are prepared, so far
`65 as extending beyond the variable domains or not including
`all of the amino acids at the C-terminus of the variable
`region.
`
`Genzyme Ex. 1019, pg 583
`
`

`
`5,840,545
`
`7
`An excess amount of the oligomer is combined with the
`denatured strands of the restriction fragment under suffi-
`ciently stringent hybridization conditions. Thus, the oligo-
`mer specifically heteroduplexes to the complementary por-
`tions of the coding strand, while providing one or more stop
`and/or nonsense codons to insure the termination of expres-
`sion at the desired amino acid at the C-terminus.
`After sufficient time for hybridization at the desired level
`of stringency, sufficient amounts of the four deoxynucle-
`otides are added in conjunction with the Klenow fragment of
`DNA polymerase I. A strand complementary to the coding
`sequence of the variable-region and any 5'-flanking
`sequence is synthesized by enzymatic elongation of the
`primer resulting in a sequence complementary to the strand
`to which the oligonucleotide is bound. The single-stranded
`DNA sequence on the coding strand located 3' to the region
`hybridized to the synthetic oligonucleotide is degraded by
`the 3'-5' exonuclease activity of the DNA polymerase. In
`this manner, ds cDNA is obtained which specifically codes
`for the variable-region and upstream flanking regions asso-
`ciated with the light and heavy chains. Each of the heavy and
`light chains is encoded to terminate expression at a prede-
`termined codon in the V, D or J region.
`The resulting heteroduplexed blunt-ended ds CDNA frag-
`ments are then employed for preparation of homoduplexed
`ds CDNA coding for the light and heavy variable regions
`with the stop codons at the desired sites. Conveniently, the
`blunt ended fragments are modified as described previously,
`e.g. joined to linkers which code for restriction sites which
`are absent in the variable region sequences, or may be tailed
`e.g. polyG or polyC, or used directly for insertion. With
`restriction site linkers, after insertion of the fragment into an
`appropriate vector having complementary termini, the frag-
`ment can be recovered by restriction at the linker sites. The
`linkers are joined to the coding sequences with an appro-
`priate ligase, e.g. T4 ligase, the resulting fragment restricted
`to provide cohesive ends, and the product annealed to the
`complementary ends of a vector.
`At this stage, the vector which is employed provides for
`amplification and convenient isolation of transformants hav-
`ing the variable region coding sequence insert. Numerous
`vectors for amplification in bacteria or other hosts exist such
`as pBR322, pSC101, pRK290, 2 kt-plasmid, etc. The hybrid
`plasmid containing the mismatched sequences will replicate
`in the host to generate two different plasmid molecules, one
`with the original sequence and one with the "tailored" or
`"site mutated" sequence derived from the synthetic oligo-
`nucleotide. Therefore, each transformant colony is grown in
`small (approximately 2 ml) culture for plasmid isolation.
`The transformants are grown, the plasmid DNA isolated
`in accordance with known procedures, and used for a second
`cycle of transformation to provide individual clones repli-
`cating the tailored sequence. The clones may be screened by
`filter blot hybridization, probing with a labeled synthetic
`oligonucleotide which will include the synthetic oligonucle-
`otide employed in tailoring the variable region sequence, or
`other convenient technique. Thus, plasmids are obtained
`having ds cDNA flanked by appropriate restriction sites and
`having a stop codon at a predetermined site.
`Having now defined the 3'-terminus of the coding strand
`or, alternatively, the C-terminus amino acid, the 5'-region or
`N-terminus of the polypeptide is now defined. Of course, the
`particular order in which the two termini are modified is
`primarily one of convenience, and can even be done
`simultaneously, where primer repair is used at the 5'-end of
`the coding strand in conjunction with site mutation at the
`3'-end.
`
`2 5
`
`30
`
`8
`Different strategies may be evolved, depending upon the
`nature of the host in which expression is to be obtained, and
`whether such host recognizes the leader sequence as a
`secretory signal for secretion of the polypeptide with con-
`5 (cid:9) comitant removal of the leader sequence polypeptide. Where
`this opportunity is not available, the strategy will involve
`removal of the leader sequence to provide a start codon at
`the 5'-terminus of the sequence of the coding strand coding
`for the variable region, which sequence can be inserted into
`10 an expression vector, so as to be under the control of a
`predetermined promoter and ribosomal start site.
`Based on the sequence of the leader region or the
`sequence coding for the N-terminus of the variable region,
`different oligonucleotides for homo- or heteroduplexing can
`15 be prepared.
`Where the leader sequence is retained, primer repair is
`employed to remove the 5'-flanking sequence of the coding
`strand. When the primer repair of the N-terminus is per-
`formed simultaneously with the C-terminus mutagenesis,
`20 after treatment with the DNA polymerase, the resulting
`partial double stranded DNA will be treated with a 5'-3'-
`single strand exonuclease to remove the 5'-flanking region as
`well as a ligase to provide for covalent linking of the
`replicated strand to the N-terminus oligonucleotide.
`Where the leader sequence is to be removed, in vitro
`mutagenesis is employed to introduce an f-met codon at the
`N-terminus of the DNA sequence coding for the variable
`region.
`Alternative strategies may be employed for recovering the
`desired ds cDNA and performing the in vitro mutagenesis.
`If useful restriction sites are distant from the coding regions,
`the plasmid may be digested with the appropriate restriction
`endonuclease, followed by digestion with a double-strand
`35 exonuclease e.g. Bal 31. The resulting ds cDNA may be
`cloned and the proper sequence selected and modified, as
`appropriate, as described above. If the non-coding flanking
`region at the 5'-terminus of the coding strand is too long, it
`may be digested with an endonuclease, where a convenient
`40 restriction site is available or by digestion with an exonu-
`clease e.g. Bal 31.
`By repeating the above described procedure for modify-
`ing the 3'-terminus, except that the oligonucleotide is now
`complementary to the non-coding (nonsense) strand, and
`45 includes an initiation codon at the 5'-end (primer repair) or
`within the oligonucleotide (in vitro muta-genesis), the
`5'-terminus