`
`[19]
`
`nu Patent Number:
`
`4,816,397
`
`Boss et al.
`
`[45] Date of Patent: Mar. 28, 1989
`
`[54] MULTICHAIN POLYPEPTIDES OR
`PROTEINS AND PROCESSES FOR THEIR
`PRODUCTION
`
`[75]
`
`Inventors: Michael A. Boss, Slough; John H.
`Kenton; John S. Emtage, both of
`High Wycombe; Clive R. Wood,
`Near Fordingbridge, all of United
`Kingdom
`
`[73] Assignee: Celltech, Limited, Slough, United
`Kingdom
`
`[21] Appl. No.:
`
`672,265
`
`[22] PCT Filed:
`
`Mar. 23, 1984
`
`[86] PCT No.:
`
`PCI‘/GB84/00094
`
`§ 371 Date:
`
`Nov. 14, 1984
`
`§ 102(e) Date:
`
`Nov. 14, 1984
`
`[87] PCT Pub. No.: W084/03712
`
`PCT Pub. Date: Sep. 27, 1984
`
`Foreign Application Priority Data
`[30]
`Mar. 25, 1983 [GB] United Kingdom ................. 8308235
`
`[51]
`
`Int. c1.4 ...................... C12P 21/on; C12N 15/00;
`C12N 1/00; C12N 1/20
`[52] U.S.Cl. ................................... 435/68;435/172.3;
`435/243; 435/255; 435/320, 435/252.31,
`435/252.33
`[58] Field of Search .................... .. 435/68, 172.3, 243,
`435/320, 253, 255
`
`[56]
`
`References Cited
`U.S. PATENT DOCUMENTS
`
`......................... 435/68
`
`9/1983 Hartley et al.
`4,403,036
`4,642,334 2/1987 Moore et al.
`
`FOREIGN PATENT DOCUMENTS
`
`041313 12/1981 European Pat. Off.
`041767 12/1981 European Pat. Off.
`055945 7/1982 European Pat. Off.
`075444
`3/1983 European Pat. Off.
`088994 9/1983 European Pat. Off.
`0125023 ll/1984 European Pat. Off.
`
`.
`.
`.
`.
`.
`.
`
`OTHER PUBLICATIONS
`
`Adams et al., Biochemistry, vol. 19, pp. 2711-2719,
`1980.
`
`Haley et al., DNA, vol. 1, pp. 155-162, 1982.
`Gough et al., Biochemistry, vol. 19, pp. 2702-2710,
`1980.
`
`Iserentant et al., Gene, vol. 9, pp. 1-12, 1980.
`Seidman et al: “Immunoglobulin light—chain structural
`gene sequences cloned in a bacterial plasmid,” Nature,
`vol. 271, pp. 582-585, 1978.
`
`Primary Examiner-Blondel I-Iazel
`Attorney, Agent, or Firm-_-Cushman, Darby & Cushman
`[57]
`ABSTRACI‘
`
`Multichain polypeptides or proteins and processes for
`their production in cells of host organisms which have
`been transformed by recombinant DNA techniques.
`According to a first aspect of the present invention,
`there is provided a process for producing a heterolo-
`gous multichain polypeptide or protein in a single host
`cell, which comprises transforming the host cell with
`DNA sequences coding for each of the polypeptide
`chains and expressing said polypeptide chains in said
`transformed host cell. According to another aspect of
`the present invention there is provided as a product of
`recombinant DNA technology an Ig heavy or light
`chain or fragment thereof having an intact variable
`domain. The invention also provides a process for in-
`creasing the level of protein expression in a transformed
`host cell and vectors and transformed host cells for use
`in the processes.
`
`037723 10/1981 European Pat. Off.
`
`.
`
`18 Claims, 13 Drawing Sheets
`
`Genzyme Ex. 1007, pg 126
`
`
`
`Genzyme Ex. 1007, pg 126
`
`
`
`U.S. Patent Mar. 28, 1989»
`
`Sheet 1 of 13
`
`4,816,397
`
`C terminus
`
`H6. 1
`
`Genzyme Ex. 1007, pg 127
`
`Genzyme Ex. 1007, pg 127
`
`
`
`U.S. Patent Mar. 28, 1989
`
`Sheet 2 of 13
`
`4,816,397
`
`H3 HH Sacsuc H3 pATM_15
`
`2ISOLATE1 1 RloS
`
`1_|-“sq;
`
`1.H3 UP
`
`ATEANG
`
`4-" T
`..P_/_933AcrAGG
`pCTS4
`
`CTGTTGTG
`CAFCGACAACAUGAGTCCTTA
`Rah
`
`.
`pCT S419-1
`3- pcTsa19-1 AAQEGTATCGATTGATCAAIE
`pNP3
`%
`AAGGGTA
`TTGATCAA_[6_
`p NPA.
`AAGGGT
`TTGATCAEIQ
`
`Legend.
`
`E - EcorRI
`
`H =
`
`Hinfl
`
`H3=
`
`Hind III
`
`FIG. 2
`
`
`
`G——
`
`PCT 54
`
`Genzyme Ex. 1007, pg 128
`
`Genzyme Ex. 1007, pg 128
`
`
`
`US. Patent Mar. 28, 1989
`
`Sheet 3 of 13
`
`4,816,397
`
`P?
`
`r
`
`A
`
`BumHl
`
`Pr
`
`1*
`
`‘A’
`pCDNA
`
`EQQ :,‘_"“1:DT)"(bD)
`EDIIED If 29 III
`1.l+kb Psf fragment
`
`(8)
`
`'
`
`
`
`'rr-
`-(TA GATCA
`ACTAG ll-_—‘:lW—d T
`«
`9°75‘:
`°°?>"?§
`+1J»kb}1Psf frugnszen/
`\3comptete ;u com
`
`um HI
`
`
`
`Genzyme Ex. 1007, pg 129
`
`Genzyme Ex. 1007, pg 129
`
`
`
`US. Patent Mar. 28, 1989
`
`Sheet 4 91' 13
`
`4,816,397
`
`Possible 2° sfrucfures of p m RNAs
`
`A 5’-3’
`
`pNP9.
`
`C
`
`A
`U U
`'35: C5
`A=U
`
`A G = 7-6 K.:ul.
`(1-0 ru)
`'
`
`GGGUALEAU
`
`Qt‘?
`
`=AGCAG C‘
`
`pNp11_
`
`A
`5;
`C'G
`G =C
`U =A
`-AAGGUAU =ACUGCAGCAGC-
`
`AG:7-8 K ml.
`(6-9ru )
`
`‘
`
`NP12
`
`P
`
`- QUAUUGCACA
`
`UU CA
`G
`5
`G = C
`A=U _
`$39
`
`'
`
`AG: 7-6 Kcul.
`(18 ru)
`
`=AGCAGC-
`
`PN.P1‘+-
`
`(101-2 ru)
`
`-AAQQGUAUGAU CAEJECAAGUGCAACUGCAG
`
`FIG. 4
`
`Genzyme Ex. 1007, pg 130
`
`Genzyme Ex. 1007, pg 130
`
`
`
`U.S. Patent Mar. 28, 1989
`
`Sheet 5 of 13
`
`4,816,397
`
`Bgl II
`
`§)§*’
`
` Psfl
`
`Olionucleolides
`
`ligufe
`
`EcoRY.'
`
`
`Genzyme Ex. 1007, pg 131
`
`Genzyme Ex. 1007, pg 131
`
`
`
`U.S. Patent Mar. 28, 1989
`
`Sheet 6 of 13
`
`4,816,397
`
`Genzyme Ex. 1007, pg 132
`
`Genzyme Ex. 1007, pg 132
`
`
`
`U S. Patent
`
`Mar. 28, 1939
`
`Genzyme Ex. 1007, pg 133
`
`Genzyme Ex. 1007, pg 133
`
`
`
`U.S. Patent Mar. 28, 1989
`
`Sheet 8 of 13
`
`4,816,397
`
`>\-2»
`
`12345
`
`FIG. 8
`
`Genzyme Ex. 1007, pg 134
`
`Genzyme Ex. 1007, pg 134
`
`
`
`U.S. Patent Mar. 28, 1989
`
`Sheet 9 of 13
`
`4,816,397
`
`
`
`123l.567_8910
`
`FIG. 9
`
`Genzyme Ex. 1007, pg 135
`
`Genzyme Ex. 1007, pg 135
`
`
`
`U.S. Patent Mar. 28, 1989
`
`Sheet 10 of 13
`
`A 4,816,397
`
`Genzyme Ex. 1007, pg 136
`
`Genzyme Ex. 1007, pg 136
`
`
`
`U.S. Patent Mar. 28, 1989
`
`Sheet 11 of 13
`
`4,816,397
`
`24
`
`I
`‘—g
`3,
`E
`c:
`1E
`9
`3'
`3
`3
`
`10
`
`51
`::
`E
`60!
`3‘
`.2
`1.3
`.3
`3
`3'
`
`
`
`0-2
`
`3
`3
`< DJ
`3 .5
`A 504
`3
`‘
`-‘,!_<3
`....U'l~<f
`E91
`
`II
`
`O
`
`
`20
`22
`24
`
`26
`
`
`30
`32
`31.
`36
`28
`FRACTIONS
`
`2’E‘
`
`FlG.11
`
`Genzyme Ex. 1007, pg 137
`
`Genzyme Ex. 1007, pg 137
`
`
`
`U.S. Patent Mar. 28, 1989
`
`Sheet 12 of 13
`
`4,816,397
`
`0.3
`
`_
`
`'
`
`NIP-cap—BSA binding activity
`from:
`
`fraction 26( A ). purified
`Igpund>.(CI)andB1-8(0).
`Binding in the presence of
`30uM NIP-cup( A , I, 0,
`respectively)
`
` A5l.OnminELISA92
`
`Serial dilution 1:1
`
`FIG. 12
`
`Genzyme Ex. 1007, pg 138
`
`Genzyme Ex. 1007, pg 138
`
`
`
`US. Patent Mar. 28, 1989
`
`Sheet 13 of 13
`
`4,816,397
`
`
`
`hcpten concentration (M )
`
`Binding of antibodies to NIP-cap BSA
`B1-8 IgM(I), fraction 26(4)}
`purified lg }.x or).( 0 ). in the presence
`of free NIP-cup(----) or NIP-cap(
`).
`
`FIG. 13
`
`Genzyme Ex. 1007, pg 139
`
`Genzyme Ex. 1007, pg 139
`
`
`
`1
`
`4,816,397
`
`MULTICI-IAIN POLYPEPTIDES OR PROTEINS
`AND PROCESSES FOR THEIR PRODUCTION
`
`This invention relates to multichain polypeptides or
`proteins and processes for their production in cells of
`host organisms which have been transformed by recom-
`binant DNA techniques.
`In recent years advances in molecular biology based
`on recombinant DNA techniques have provided pro-
`cesses for the production of heterologous (foreign)
`polypeptides or proteins in host cells which have been
`transformed with heterologous DNA sequences which
`code for the production of these products.
`Theoretically, the recombinant DNA approach may
`be applied to the production of any heterologous poly-
`peptide or protein in a suitable host cell, provided that
`appropriate DNA coding sequences can be identified
`and used to transform the host cell. In practice, when
`the recombinant DNA approach was first applied to the
`production of commercially useful products, its applica-
`tion for the production of any specified polypeptide or
`protein presented particular problems and difficulties,
`and the success of applying this approach to the produc-
`tion of any particular polypeptide or product was not
`readily predictable.
`However, a large number of heterologous single
`chain polypeptides or proteins have now been produced
`by host cells transformed by recombinant DNA tech-
`niques. Examples of such heterologous single chain
`polypeptides or proteins include human interferons, the
`A and B chains of human insulin, human and bovine
`growth hormone, somatostatin, calf prochymosin and
`urokinase. Such transformed host cells provide a repro-
`ducible supply of authentic heterologous polypeptiede
`or protein which may be produced on an industrial scale
`using industrial fermentation technology.
`It should be pointed out that some of these polypep-
`tides, for instance uroldnase, after secretion by a host
`cell appear as two chain molecules. However, in such
`cases, the molecule is synthesised by the host cell as a
`single chain polypeptide, coded for by a single DNA
`sequence, which is cleaved in the host cell subsequent to
`synthesis to form the two chain structure.
`It is known that in both human and animal systems
`there are a number of polypeptides or proteins which
`have multichain structure in which the chains are not
`derived from the cleavage of a single chain polypeptide
`coded for by a single DNA sequence. In such cases, the
`gene for each of the chains may be located at different
`points on the same chromosome or even on different
`chromosomes. In these cases, the polypeptide chains are
`synthesised separately and then assembled into the com-
`plete molecule subsequent to synthesis. Heretofore, no
`such multichain polypeptide or protein has been pro-
`duced by recombinant DNA techniques from a single
`host cell.
`A particular example of a class of such multichain
`polypeptides or proteins is the immunoglobulins.
`Immunoglobulins, commonly referred to as antibod-
`ies, are protein molecules produced in animals by B-
`lymphocyte cells in response to challenge with foreign
`antigenic agents, such as bacteria, viruses and foreign
`proteins. The immunoglobulins comprise a crucial part
`of the immune systems of humans and animals. The
`immunoglobulins recognise specific parts of the foreign
`agents and bind onto them. The specific parts are usu-
`ally known as antigenic determinants or antibody bind-
`
`5
`
`10
`
`15
`
`20
`
`25
`
`30
`
`35
`
`45
`
`50
`
`55
`
`60
`
`65
`
`'
`
`2
`ing sites. A given foreign agent is likely to have a num-
`ber of different antigenic determinants.
`A typical immunoglobulin (lg) molecule is shown in
`FIG. 1 of the accompanying drawings, to which refer-
`ence is now made. The Ig molecule comprises two
`identical polypeptide chains of about 600 amino acid
`residues (usually referred to as the heavy chains H),
`disulphide bonded to each other, and two identical
`shorter polypeptide chains of about 220 amino acid
`residues (usually referred to as the light chains L), each
`light chain being disulphide bonded to one end of each
`heavy chain as shown.
`When the Ig molecule is correctly folded, each chain
`is formed into a number of distinct globular areas, usu-
`ally known as domains, joined by a more linear poly-
`peptide chain. The light chains have two such domains,
`one of which is of variable sequence V1,, and the other
`of which is of constant sequence CL. The heavy chains
`have a single variable domain VH adjacent the variable
`domain V; of the light chain, and three or four constant
`domains CH1-3 9, 4 .
`The variable domains on both the heavy and the light
`chains
`each contain three hypervariable
`regions
`(HV1-3) which, when the Ig molecule is correctly
`folded, are located adjacent one another and form the
`antigen binding site. It is this area which recognises and
`binds to the antigenic determinant for which the Ig
`molecule is specific.
`The constant domains of the Ig molecule do not take
`part in the binding to the antigenic determinant, but
`mediate the actions triggered by the binding of the Ig
`molecule to the antigenic determinant. It is believed
`that this triggering is caused by an allosteric effect in-
`duced by the binding of the Ig molecule to the antigenic
`determinant. The constant domain may enable the Ig
`molecule to fix complement or may cause mast cells to
`release histamine.
`Ig’s may be categorised by class or subclass, depend-
`ing on which of a number of possible heavy chain con-
`stant domains they contain, there being eight such possi-
`ble heavy" chains in mice. Thus, for instance, and lg
`molecule with a p. heavy chain belongs to the class IgM,
`and one with a 7; heavy chain to the class IgG1.
`'
`Ig’s may also contain one of two light chains, desig-
`nated as K and 7t light chains, which have different
`constant domains and different sets of variable domains.
`The structure of the Ig molecule and the location of
`the genes coding for the various domains thereof are
`discussed more fully by Earlyand Hood, in Genetic
`Engineering, Principles and Methods, Vol. 3, pages
`153-158 (edited by Setlow and Hollaender, Plenum
`Press).
`.
`It is known that Ig molecules on digestion with se-
`lected enzymes can produce a number of immunologi-
`cally functional fragments. Two such fragments are
`known as the Fab and (Fay); fragments. The Fab frag-
`ment comprises one light chain linked to the V1; and
`CH1 domains of a heavy chain as shown in FIG. 1. The
`(Fa/,')2 fragment consists essentially of two F41, frag-
`ments linked together by a small additional portion of
`the heavy chains as shown in FIG. 1. These fragments
`and other similar fragments can be of use in various tests
`and diagnostic and medical methods.
`The principle method employed for the production of
`Ig’s involves the immunisation of susceptible animals
`with the antigenic agent to provide an immune reaction.
`Generally the animal is immunised a second time to
`
`Genzyme Ex. 1007, pg 140
`
`Genzyme Ex. 1007, pg 140
`
`
`
`3
`improve the yield of lg. The animal is then bled and the
`Ig is recovered from the serum.
`However, the product of this method is not a homo-
`geneous protein. The animal will produce Ig of differ-
`ent classes and also Ig specific for each of the different
`antigenic determinants on the antigenic agent, and its
`blood will therefore contain a heterogeneous mixture of
`Ig’s. Obtaining a specific Ig of particular class and de-
`sired specificity from such a mixture requires very diffi-
`cult and tedious purification procedures.
`Recently, it has become possible to produce a homo-
`geneous 1g of a single class and a single specificity by a
`technique first described by Kohler and Milstein (Na-
`ture, 256, 495-479, 1975). The technique involves the
`fusion of single lg-producing parent cells with cancer
`cells to produce a monoclonal hybridoma cell which
`produces the lg. lg produced by this technique is usu-
`ally known as monoclonal antibody. The nature of the
`monoclonal lg is determined by the class and specificity
`of the lg produced by the parent cell.
`Recently attempts have been made to use recombi-
`nant DNA techniques to produce fragments of Ig mole-
`cules. For instance, Amster et al. (Nucleic Acid Re-
`search, 8, No. 9, 1980, pp 2055 to 2065) disclose the
`cloning of double stranded cDNA sequences encoding
`for a mouse Ig light chain into a plasmid. An E. Coli
`strain transformed by the plasmid synthesised a protein
`thought to comprise the complete constant domain of
`the light chain and about 40 amino acid residues of its
`variable region.
`Kemp and Cowman (Proc. Natl. Acad. Sci. USA, 78,
`1981 , pp 4520 to 4524) disclose the cloning of cDNA
`sequences encoding for mouse heavy chain fragments
`and the transforming of an E. Coli strain which then
`synthesised heavy chain polypeptide fragments.
`In both these cases, the polypeptides were produced
`as fusion proteins, in which the fragments of the Ig
`polypeptides were fused with additional non-Ig poly-
`peptide sequences, and the incomplete variable do-
`mains. Thus, the polypeptide chains produced in these
`studies were not immunologically functional polypep-
`tides as they were incapable of combining with comple-
`mentary heavy or light chains to provide Ig molecules
`having intact antigen binding sites and immunological
`function.
`Research studies have also been carried out in mam-
`malian systems. For instance, Falkner and Zachau (Na-
`ture, 298, 1982, pp 286 to 288) report the cloning of
`cDNA sequences encoding for mouse light chains into
`a plasmid which was used to transfect genomic eukary-
`otic cells which could then transiently synthesise light
`chains.
`
`Rice and Baltimore (Proc. Natl. Acad. Sci. USA, 79,
`1982, pp 7862 to 7865) report on the transfection of a
`functionally rearranged K light chain Ig gene into a
`murine leukemia virus-transformed lymphoid cell line.
`The cell line is then able to express the gene continu-
`ously. In both these cases, the K genes used to transfect
`the mammalian cells were obtained from myeloma cells
`and the K polypeptides produced were of indeterminate
`immunological function.
`A further approach is exemplified in a series of papers
`by Valle et al. (Nature, 291, 1981, 338-340; Nature, 300,
`1982, 71-74and J. Mol. Biol., 160, 1982, 459-474),
`which describe the microinjection of mRNAs encoding
`for heavy or light chains of lg isolated from a mouse
`myeloma line into oocytes of Xenopus laevis. Under
`certain conditions complete Ig molecules were formed.
`
`10
`
`15
`
`20
`
`25
`
`30
`
`35
`
`40
`
`45
`
`50
`
`55
`
`60
`
`65
`
`4,816,397
`
`4
`However, the mRNAs were obtained from myeloma
`cells and the Ig molecules were of indeterminate immu-
`nological function.
`It can thus be seen that hitherto it has not been possi-
`ble to produce functional lg by recombinant DNA
`technology.
`According to a first aspect of the present invention,
`there is provided a process for producing a heterolo-
`gous multichain polypeptide or protein in a single host
`cell, which comprises transforming the host cell with
`DNA sequences coding for each of the polypeptide
`chains and expressing said polypeptide chains in said
`transformed host cell.
`
`According to a second aspect of the present inven-
`tion, there is provided a heterologous multichain poly-
`peptide or protein produced by recombinant DNA
`technology from a single host cell.
`The present invention is of particular, but not exclu-
`sive, application in the production of lg molecules and
`immunologically functional Ig fragments of the type
`referred to above. However, it will be appreciated that
`the present invention can be applied to the production
`of other multichain polypeptides or proteins.
`In relation to the product of Ig molecules according
`to the invention it will be appreciated that, in order to
`produce a functional molecule, the DNA sequences
`used to transform the host cell will need to encode for
`at least the V; and V1; domains of an Ig molecule.
`Moreover, these domains will need to be complemen-
`tary so that when the two polypeptide chains fold to-
`gether they forrn an antigen binding site of predeter-
`mined specificity.
`Preferably, the Ig molecule or fragment include a
`complete light chain and at least the CH1 domain in
`addition to the V1; domain of the heavy chain. Most
`preferably the Ig molecule is intact.
`It has also been shown by the present applicants that
`it is now possible to produce individual heavy and light
`chains having intact variable domains. This has not
`previously been possible. Therefore, according to a
`third aspect of the present invention there is provided as
`a product of recombinant DNA technology an Ig heavy
`or light chain or fragment thereof having an intact vari-
`able domain.
`
`Advantageously, the Ig molecule or functional frag-
`ment thereof according to the present invention has a
`variable region (formed by the V1, and V3 domains)
`which defines a binding site for an antigenic determi-
`nant of clinical or industrial importance. The DNA
`coding sequences necessary to produce such a molecule
`may be derived from naturally occuring or' hybridoma
`(monoclonal)Ig-producing cells with the desired speci-
`ficity.
`The constant domains of the Ig molecule or fragment,
`if present, may be derived from the same cell line as the
`variable region. However, the constant domains may be
`specifically altered, partially or completely omitted, or
`derived from a cell line producing a different class of lg
`to provide Ig molecules or fragments having desired
`properties.
`For example, an lg molecule may be produced hav-
`ing variable domains (V3 and VL) idential with those
`from a monoclonal antibody having a desired specific-
`ity, and constant domain(s) from a different monoclonal
`antibody having desired properties, for instance to pro-
`vide human compatibility or to provide a complement
`binding site.
`
`Genzyme Ex. 1007, pg 141
`
`Genzyme Ex. 1007, pg 141
`
`
`
`4,816,397
`
`5
`Such alterations in the amino acid sequence of the
`constant domains may be achieved by suitable mutation
`or partial synthesis and replacement or partial or com-
`plete substitution of appropriate regions of the corre-
`sponding DNA coding sequences. Substitute constant
`domain portions may be obtained from compatible re-
`combinant DNA sequences.
`,
`The invention may be utilised for the production of
`Ig molecules or fragments useful for immunopurif1ca-
`tion, immunoassays, cytochemical labelling and target-
`ting methods, and methods of diagnosis or therapy. For
`example, the Ig molecule or fragment may bind to a
`therapeutically active protein such as interferon or a
`blood clotting factor, for example Factor VIII, and may
`therefore be used to produce an affinity chromatorgra-
`phy medium for use in the immunopurification or assay
`of the protein.
`It is also envisaged that the Ig molecule may be syn-
`thesised by a host cell with another peptide moiety
`attached to one of its constant domains. Such a further
`peptide moiety may be cytotoxic or enzymatic. Alterna-
`tively, the moiety may be useful in attaching the Ig
`molecule to a biological substrate, such a cell or tissue,
`or to a non-biological substrate, such as a chromatogra-
`phy medium. Such a peptide moiety is herein referred to
`as a structural peptide moiety.
`It is further envisaged that cytotoxic, enzymic or
`structural peptide moieties could be attached to the Ig
`molecule by normal peptide chemical methods, as are
`already known in the art, rather than by being synthe-
`sised with the Ig molecule.
`The Ig molecule or fragment may also comprise a
`therapeutic agent in its own right. For instance, an Ig
`molecule or fragment specific for D blood group anti-
`gen may be useful for the prevention of haemolytic
`disease of the new born.
`Any suitable recombinant DNA technique may be
`used in the production of the multichain polypeptides or
`proteins of the present invention. Typical expression
`vectors such as plasmids are constructed comprising
`DNA sequences coding for each of the chains of the
`polypeptide or protein.
`It will be appreciated that a single vector may be
`constructed which contains the DNA sequences coding
`for more than one of the chains. For instance, the DNA
`sequences coding for Ig heavy and light chains may be
`inserted at different positions on the same plasmid.
`Alternatively, the DNA sequence coding for each
`chain may be inserted individually into a plasmid, thus
`producing a number of constructed plasmids, each cod-
`ing for a particular chain. Preferably the plasmids into
`which the sequences are inserted are compatible.
`The or each constructed plasmid is used to transform
`a host cell so that each host cell contains DNA sequen-
`ces coding for each of the chains in the polypeptide or
`protein.
`Suitable expression vectors which may be use for
`cloning in bacterial systems include plasmids, such as
`Col El, pcR1, pBR322, pACYC 184 and RP4, phage
`DNA or derivatives of any of these.
`For use in cloning in yeast systems, suitable expres-
`sion vectors include plasmids based on a 2 micron ori-
`gm.
`Any plasmid containing an appropriate mammalian
`gene promoter sequence may be used for cloning in
`mammalian systems. Such vectors include plasmids
`derived from, for instance, pBR322, bovine papilloma
`virus, retroviruses, DNA viruses and vaccinia viruses.
`
`l0
`
`15
`
`20
`
`25
`
`30
`
`35
`
`45
`
`50
`
`55
`
`65
`
`6
`Suitable host cells which may be used for expression
`of the heterologous multichain polypeptide or protein
`include bacteria, such as E. coli and B. subtilis, Strepto-
`myces, yeasts, such as S. cervisiae, and eukaryotic cells,
`such as insect or mammalian cell lines. Examples of
`suitable bacterial host cells include E. Coli HB 101, E.
`Coli X1776,E. Coli X2882, E. Cali PS 4l0,E.- Coli MRC
`1, E. Cali RV308, E. Coli E1038 and E Coli B.
`The present invention also includes constructed ex-
`pression vectors and transformed host cells for use in
`producing the multichain polypeptides or proteins of
`the present invention.
`After expression of the individual chains in the same
`host cell, they may be recovered to provide the com-
`plete multichain polypeptide or protein in active form,
`for instance to provide an Ig molecule of predetermined
`immunological function.
`It is envisaged that in preferred forms of the inven-
`tion, the individual chains will be processed by the host
`cell to form the complete polypeptide or protein which
`advantageously is secreted therefrom.
`However, it may be that the individual chains may be
`produced in insoluble or membrane-bound form. It may
`therefore be necessary to solubilise the individual chains
`and allow the chains to refold in solution to form the
`
`active multichain polypeptide or protein. A suitable
`procedure for solubilising polypeptide chains expressed
`in insoluble or membrane-bound form is disclosed in our
`copending application No. (Protein Recovery, Agent’s
`Ref. GF 402120 and 402121).
`It will be appreciated that the present application
`shows for the first time that it is possible to transform a
`host cell so that it can express two or more separate
`polypeptides which may be assembled to form a com-
`plete multichain polypeptide or protein. There is no
`disclosure or suggestion of the present invention in the
`prior art, which relates solely to the production of a
`single chain heterologous polypeptide or protein from
`each host cell.
`The present invention will now be described, by way
`of example only, with reference to the accompanying
`drawings, in which:
`FIG. 1. shows a diagrammatic representation of a
`typical intact lg molecule;
`FIG. 2 shows the construction of plasmids for the
`direct synthesis of a A light chain in E. Coli;
`FIG. 3 shows the construction of plasmids for the
`direct synthesis of a p. heavy chain in E. Coli;
`FIG. 4 is a diagrammatic representation of p.mRNA '
`sequences around the initiation codon;
`FIG. 5 shows the construction of plasmids having
`altered secondary structure around the initiationcodon;
`FIG. 6 is a polyacrylamide gel showing expression
`and distribution of 1.1. protein from E. Coli B;
`FIG. 7 is a polyacrylamide gel showing pulse chase
`autoradiograms of 1:. protein in E. Coli B and in E. Coli
`HBl0l;
`FIG. 8 is a polyacrylamide gel showing the results of
`A gene expression in E. Coli;
`FIG. 9 is a polyacrylamide gel showing the distribu-
`tion of recombinant it light chain polypeptide between
`the soluble and insoluble cell fractions;
`FIG. 10 is a polyacrylamide gel showing expression
`and distribution of A protein from E. Coli El03S-,
`FIG. 11 shows the results of the fracitonation of p.
`and A protein expressed by E. Coli B on DEAE Sepha-
`cel;
`
`Genzyme Ex. .1007, pg 142
`
`Genzyme Ex. 1007, pg 142
`
`
`
`4,816,397
`
`8
`well as Bell and Hinfl sticky ends. The two chemically
`synthesised oligonucleotides made to facilitate assembly
`of the gene had the sequences:
`
`5
`
`R45 5'-pGATCAATGCAGGCTG'I'I'GTG 3’
`
`R44 3’ CCGACAACACTGAGTCC'I'I‘Ap- 5'
`
`10
`
`pCT54 was cut with both Bell and HindIII and the
`resulting linear molecules isolated, mixed together with
`the two oligodeoxyribonucleotide linkers R44 and R45
`and both fragments 1 and 2, and ligated using T4 ligase
`(FIG. 2). The mixture was used to transform E. coli
`DHl to ampicillin resistance. Recombinant clones in
`pCT54 were identified by hybridisation of DNA from
`replica plated colonies on nitrocellulose to a nick-tran-
`slated probe derived from the pAT A 1-15 insert.
`A clone was identified which hybridised to lambda
`cDNA and also showed the predicted restriction frag-
`ment pattern. This plasmid (designated pCT54 19-1)
`was sequenced from the Clal site and shown to have the
`anticipated sequence except that there was a mutation
`of the fourth codon from CTG to ATG, changing the
`amino acid at this point from valine the methionine.
`The sequence in this area was:
`
`7
`FIG. 12 shows the specific hapten binding of recon-
`stituted Ig molecules; and
`FIG. 13 shows the heteroclitic nature of the hapten
`binding of the reconstituted Ig molecules.
`In the following examples,
`there is described the
`production of Ig light and heavy chain polypeptides
`derived from monoclonal antibodies which recognise
`and bind to the antigenic determinant 4-hydroxy-3-
`nitrophenyl acetyl (NP), using E. coli and S. cerevisiae as
`the host cells Recombinant DNA techniques were used
`to enable the host cells to express both the polypeptide
`chains.
`
`It will be appreciated that the invention is not limited
`to the specific methods and construction described
`hereafter.
`
`15
`
`Construction of Lambda Light Chain Expression
`Plasmid
`
`FIG. 2, to which reference is now made, shows sche-
`matically the method used to construct a A light chain
`expression plasmid.
`It was decided to express the lambda gene in E. coli
`by direct expression of the gene lacking the eucaryotic
`signal peptide but containing a methionine initiator
`residue at the amino-terminus (met-lambda). The ap-
`.
`.
`. GATTGATCA.ATG.CAG.GC’I'.G'I‘T.ATG.ACT.CAG.GAA.TCT.GCA.C’l'C.ACC.ACA.TCA
`met
`gln
`ala
`val met
`thr
`gln
`glu
`sex‘
`ala
`leu
`thr
`thr
`ser
`
`20
`
`proach used for bacterial synthesis of met-lambda was
`to reconstruct the gene in vitro from restriction frag-
`ments of a cDNA clone and to utilise synthetic DNA
`fragments for insertion into the bacterial plasmid pCT54
`(Emtage et a1., proc. Natl. Acad. Sci. USA., 80, 3671 to
`3675, 1933). This vector contains the E. coil‘ trp pro-
`moter, operator and leader ribosome binding site; in
`addition 14 nucleotides downstream of the ribosome
`binding site is an initiator ATG followed immediately
`by EcoR1 and HindIII sites and the terminator for E.
`coli RNA polymerase from bacteriophage T7.
`As a source of light chain we used a plasmid pAB}.-
`l-15 which contains a full-length A; light chain cDNA
`cloned into the Pstl site of pBR322. This M light chain
`is derived from a monoclonal antibody, designated S43,
`which binds to 4-hydroxy-3-nitrophenylacetyl
`(NP)
`haptens.
`In order to create a HindIII site 3’ to the end of the
`lambda gene for insertion into the HindIII site of
`pCT54, the cDNA was excised from pAB A 1-15 using
`Pstl. The cohesive ends were blunt ended using the
`Klenow fragment of DNA polymerase and synthetic
`HindIII
`linker molecules
`of
`sequence
`5’-
`CCAAGCTTGG-3' ligated. The DNA was digested
`with HindIII and the 850bp lambda gene isolated by gel
`electrophoresis and cloned into HindIII cut pATl53 to
`yield plasmid pAT A 1-15. The 3’ end of the lambda
`gene was isolated from pAT A 1-15 by HindIII plus
`partial Sacl digestion as a 630bp SacI~HindIII fragment
`(2 in FIG. 2). The HindIII cohesive end was dephos-
`phorylated by calf intestinal alkaline phosphatase dur-
`ing isolation of the fragment to prevent unwanted liga-
`tions at this end in subsequent reactions.
`A Hinfl restriction site is located between codons 7
`and 8 and the lambda sequence. The 5’ end of the
`lambda gene was isolated as a l48bp Hinfl to sacI frag-
`ment (1 in FIG. 2).
`Two’ oligodeoxyribonucleotides were designed to
`restore condons 1-8, and to provide an initiator ATG as
`
`30
`
`35
`
`40
`
`45
`
`50
`
`55
`
`60
`
`65
`
`The restriction enzyme sites in pCT54 between the
`Shine and Dalgamo sequence (AAGG), which is im-
`portant for ribosome binding, and the ATG allow for
`the adjustment of the SD-ATG distance, an important
`parameter in determining expression rates. The SD-
`ATG distance was reduced by cutting the plasmid with
`Clal or Bell and creating blunt ended species by diges-
`tion with S1 nuclease. 2 pg of ClaI cut DNA was di-
`gested with 200 units of S1 nuclease for 30 minutes at 30°
`using standard buffer conditions. The solution was de-
`proteinised with phenol and the DNA recovered by
`ethanol precipitation. This DNA on religation with T4
`DNA ligase and transformation into E. coli strain
`HB101 gave rise to a number of plasmids which had lost
`the Clal or BclI site.
`
`The plasmids which had lost their Clal site were
`sequenced in the region surrounding the initiator ATG.
`
`.
`
`.
`
`.
`
`.
`
`. AAGGGTATTGAICAATG CAG .
`SD
`met glu
`
`.
`
`. plasmid pNP3
`
`. AAGGG I'l'I GATCAATG CAG
`SD
`met glu
`
`plasmid pNP4
`
`In order to achieve high level expression a number of
`other approaches were followed. Firstly, a series of
`constructs were obtained which had increasing amounts
`of the 3’ untranslated region of the cDNA removed by
`Bal 31 exonuclease. Secondly, a high copy number
`plasmid containing ACDNA was constructed. This plas-
`mid contained a par function (Meacock, P. A. and Co-
`hen, S. N., Cell, 20, 529-542, 1980) as well as being
`present in high copy number. Thirdly, the pNP3 plas-
`mid was transformed into a number of protease-defb
`cient strains or into HB10l in conjunction with a prote-
`ase deficient dominant acting plasmid (Grossman, A. D.
`et al, Cell, 32, 151-159, 1983).
`
`Genzyme Ex. 1007, pg 143
`
`Genzyme Ex. 1007, pg 143
`
`
`
`9
`
`4,816,397
`
`10
`with PstI cut pCT54 Pst (see above) using T4 DNA
`ligase under standard conditions, followed by transfor-
`mation into HB101. A plasmid designated pNPl was
`isolated which was shown by restriction endonuclease
`pattern analysis to contain the 1.4 Kb p.cDNA fragment
`in an appropriate orientation (FIG. 3). pNPl was a
`plasmid which consisted of an appropriate 5’ end for
`expression, whilst pCT54 contained an appropriate 3’
`end. The full length gene was reconstructed -into pCT54
`by cutting