vectors for the expression of pcr-amplified immunoglobulin

© 1993 Oxford University Press Nucleic Acids Research, 1993, Vol. 21, No. 12 2921-2929

Vectors for the expression of PCR-amplifiedimmunoglobulin variable domains with human constantregions

Michael A.Walls, Ku-chuan Hsiao and Linda J.HarrisBristol-Myers Squibb Pharmaceutical Research Institute—Seattle, Department of MolecularImmunology, 3005 First Avenue, Seattle, WA 98121, USA

Received November 6, 1992; Revised and Accepted April 15, 1993 EMBL accession nos*

ABSTRACT

Cassette vectors have been constructed for mammalianexpression of complete Immunoglobulin heavy andlight chain genes whose variable regions are producedby the polymerase chain reaction (PCR). The light andheavy chain vectors have promoter, leader, partialIntron, enhancer and constant region segments withinmodified pSV2-gpt and pSV2-neo plasmlds,respectively. Variable (V) regions are obtained by PCRusing a two step process: 1) the V gene Is amplifiedfrom genomlc or cDNA, cloned Into an intermediatevector and sequenced; 2) the first PCR product servesas the template for a second amplification In whichrestriction enzyme recognition sites and limitedflanking Intron sequence are added. The second PCRproduct Is Inserted into the expression vector, whichis then transfected Into mouse myeloma cells. Thesevectors contain human constant regions and may beused to express chlmerlc, humanized or human Iggenes. This report describes the design of thesevectors and their application for the expression ofchlmerlc 60.3, an anti-CD18 antibody.

INTRODUCTION

Recombinant monoclonal antibodies are currently being clinicallytested as therapeutic agents. However, repeated clinical treatmentwith non-human antibodies induces an adverse anti-immunoglobulin response (e.g., human anti-mouse antibody orHAMA [1,2]). To reduce this type of response, it becomesnecessary to express potential therapeutic antibodies as eitherchimeric, humanized, or human monoclonals [3,4]. This usuallyrequires the cloning of variable (V) region genes. To ensure theproper folding of recombinant immunoglobulin (Ig) moleculesfor applications where whole antibody is preferred, it is beneficialto utilize vectors suitable for eucaryotic expression over thosethat employ procaryotic cells [5]. Although there are severalexamples of transient expression vectors available for use ineucaryotic cells [6 -8] , stable expression systems are mostappropriate for manufacturing applications.

Previously, most vectors for stable expression of Ig genesrequired cloning relatively large genomic DNA fragments [9,10].We have designed new vectors that accept V genes obtainedyby a rapid PCR based approach, thus avoiding the timeconsuming preparation and screening of a cDNA or genomicDNA library. When V genes are amplified from cDNA, theyare deficient in promoter and intron sequences, which can beimportant for efficient expression. Depending on the primersused, amplified V genes also may lack all or part of leader (signalpeptide) sequence. Our new cassette vectors are designed toprovide all of these essential elements: promoter, leader, intron,enhancer and constant ( Q region gene. To use these vectors,V genes are amplified from cDNA, the PCR product is clonedinto an intermediate vector and sequenced. Prior to cloning intothe expression vectors, V genes are modified by PCR to includerestriction sites and splice recognition sequences.

This methodology was used to clone and express 60.3 as achimeric Ab. 60.3 is a mouse monoclonal antibody (IgG2a, x)which recognizes a functional epitope on the CD 18 componentof leukocyte integrins [11,12]. It prevents adherence andaggregation of polymorphonuclear leukocytes, thus blockingneutrophil mediated damage during shock and reperfusion injury.Because of its clinical value, it would be advantageous to express60.3 in chimeric (chi60.3) and humanized forms. This reportdescribes the cloning of 60.3 heavy (H) and light (L) chain Vregions into our cassette vectors and their stable expression bymurine myeloma cells.

MATERIALS AND METHODSReagentsUnless specifically noted in the text, the chemical and biologicalreagents used in this study were purchased from BoehringerMannheim (Indianapolis, IN).

Cells

P3X63Ag8.653 (Ag8.653) is an established mouse myeloma cellline (ATCC CRL 1580) that does not produce detectable H chainIg. Ag8.653 cells produce a mRNA encoding an aberrant L chain.

* Z1733O, Z17336 and Z17370

Downloaded from https://academic.oup.com/nar/article-abstract/21/12/2921/1045823by gueston 12 April 2018

2922 Nucleic Acids Research, 1993, Vol. 21, No. 12

Human Ck

pICMHEXX

I

ori -i pGk.5 <*Hcs.HHE.Ck. W Human Ck

pG.10NINi.ASV40E

Human Ck

leader MCS for insertionof V genes

Figure 1. Construction of light chain vector pGx. 11. See Materials and Methods for details of construction. Certain intermediates are represented only by name.Below the name of each intennediate are the distinguishing characteristics of that vector. The direction of transcription is indicated by arrows. Not all restrictionsites are designated at each step. Restriction she abbreviations are: B, BamH I; Bg, Bgl II; C, Cla I; E, EcoR I; H, Hind HI; H/Bg, fusion of filled in Hind IDand Bgl H sites; N/Nd, overlapping Not I and Nde I sites; Na, Nar I; Nd, Nde I; Nr, Nru I; Pv, Pvu II; Sa, Sal I; Sau/B, fusion of Sau3A I and BamH I sites;Sp, Sph I; Ss, Sse83871; Xh, Xho I. Regions of the vectors are abbreviated as follows: amp, ampicillin resistance gene; Ecogpt, E.coli xanthine-guanine phosphoribosyltransferase gene; SV40 E, SV40 enhancer; A SV40 E, SV40 E with 140 bp deleted; MCS, multiple cloning site; Cx, exon encoding human C* gene, residingin a 2.8 kb EcoR I fragment; MHE, mouse heavy chain enhancer, 1 kb Xba I fragment; ori, bacterial origin of replication; Pro/L, Sau 3A I fragment containingpromoter, leader and a portion of the L-V intron from the 4B9 light chain.

When using an Ag8.653 derived hybridoma as the source ofmRNA, the deduced sequence should be compared with that ofthe Ag8.653 L chain.

Construction of pGx. l l (Figure 1)

Hind HI and Bgl n sites were deleted from the 5' end of theiJcogpt gene in pSV2-gpt [13] by restriction with Hind IH andBgl n , filling in with Klenow polymerase, and religation (this121 bp fragment is not required for expression of Ecogpt gene).The EcoR I to Pvu II fragment containing pBR322 sequenceswas substituted with the analagous portion of pUC18 (PvuII—Pvu II fragment; the EcoR I overhang was filled in withKlenow polymerase prior to ligation with a Pvu II site, thusregenerating the EcoR I site) to form pG.3. This was done toincrease the yields of plasmid DNA [14].

pG.5 was constructed from pG.3 by replacement of a 750 bp

EcoR I-BamH I fragment with the 64 bp EcoR I-BamH Imultiple cloning site (MCS) from pIC20R [ 15]. To create pG. 12,a Not I site was inserted at the Nde I site of pG.5 usingoligonucleotide linkers:

5'-T AGCGGCCGC A-3'3'-CGCCGGCGT AT-5'

The Not I site was used for linearization of plasmid DNA priorto transfection. A 2.75 kb EcoR I fragment containing the humanCx gene was placed into the EcoR I site of the MCS of pG. 12to form pGx.3.

pG.9 was assembled from pG.3 by removal of a 140 bp portionof the SV40 enhancer from pG.3. This was achieved by fill inand ligation of Sph I and Pvu II sites. (pG.9 was constructedfor unrelated experiments; as is apparent from Figure 1, the final


Nucleic Acids Research, 1993, Vol. 21, No. 12 2923

amp

ori SV40 ori

promoter leader

I EcoR V-Sac I EcoR V-Sac I| fragmanl fragment

promoter leader

Hind tll-BaniH Ifragment

Human Cyl

amp

Hrnd IH-BamH Ifragment

Figure 2. Construction of heavy chain vector pNyl.12. See Materials and Methods for details of construction. Below the name of each construct are the distinguishingcharacteristics of that vector. The direction of transcription is indicated by arrows. Restriction site abbreviations are: A, Asp718 I; B, BamH I; C, Cla I; E, EcoRI; Ev, EcoR V; H, Hind HI; S, Sac I; Sa, Sal I; Xh, Xho I. Regions of the vectors are abbreviated as follows: amp, ampknllin resistance gene; neo, G418 andneomycin resistance gene; SV40 E, SV40 enhancer, MCS, multiple cloning site; Cyl, exons and introns encoding human C7I gene, residing in a 2.8 kb HindIQ-BamH I fragment; MHE, mouse heavy chain enhancer, 0.7 kb Xba I-EcoR I fragment; ori, bacterial origin of replication; Pro/L, PCR fragment containingpromoter, leader and a portion of the L-V intron from L6 heavy chain.

vector, pGx.ll , could have been more simply constructedwithout removal of the SV40 enhancer.) Subsequently, a Not Isite was inserted at the Nde I site of pG.9 to form pG.10.

The 195 bp Not I to BamH I fragment from pG.12 was putinto the same sites of pG. 10 to make pG. 11. This replaced an879 bp fragment from pG. 10 with a 195 bp fragment consistingof the pIC20R MCS. The 3 kb Nar I-Cla I fragment of pGx.3,containing human Cx, was directionally subcloned into the samesites of pG.ll to create pGx.4. A 1 kb fragment containing themouse heavy chain enhancer (MHE) was transferred frompICMHEXX to pGx.4 as a Cla I to Hind m fragment, thusproducing pGx.5. pICMHEXX was made by insertion of the1 kb Xba I fragment (filled in with Klenow polymerase) fromRBL 216 [16] into the filled in Bgl H site of pIC19R [15]. A579 bp Sau3A I fragment that included the 4B9 (human GroupB Streptococcus Ab) promoter [17] from pGxA1.9 [18] wasplaced into the BamH I site of the MCS in pGx.5 to constructpGx. 11. The following unique restriction sites between the leadersequence and the MHE are available for addition of V, genes:Sal I, Sse8387 I, Hind m, Nru I, and Xho I.

Construction of PN7I.I2 and pNyl.16 (Figure 2)The Hind HI site in pSV2-neo [19] was destroyed by digestionwith Hind HI, filling in with Klenow polymerase and religation.The 750 bp EcoR I—BamH I fragment was replaced by the 64bp EcoR I to BamH I polylinker from pIC20R to form pN.5.

The 695 bp Xba I to EcoR I region of the MHE was amplifiedby PCR using primers which deleted EcoR I and Xba I sites,but provided restriction sites for EcoR V, Sac I, Asp 718 I, and

Xho I at the 5' end and sites for Hind HI, Sal I and BamH Iat the 3' end. This PCR product was cloned into pIC20R toproduce pMHE.pcr. The 723 bp EcoR V to BamH I fragmentfrom pMHE.pcr was ligated into the same sites in the MCS regionof pN.5 creating pN.8.

A 699 bp fragment comprised of promoter, leader, and intronsequences [20] from the L6 (a murine anti-tumor mAb [21]) Hchain locus was amplified by PCR using primers which addedrecognition sequences for EcoR V and Sac I at the ends. AfterDNA sequence verification in an intermediate vector, thepromoter/leader fragment was introduced into EcoR V and SacI sites of pN.8 to produce pN.ll .

A 2.8 kb Hind in to BamH I fragment containing the human7I gene [22] was inserted into the same sites of pN. 11 to constructpNyl. 12. Vector PN71.16 is identical to PN71.12 except severalmutations were introduced into the promoter (see Results andDiscusssion). The H chain vectors have the following uniquerestriction sites for insertion of VH genes: Asp718 I/Kpn I,Eel 136 I/Sac I, and Xho I. Potential sites for linearization ofplasmids are EcoR I, Cla I, and EcoR V.

Amplification of V genes by PCRTotal cellular RNA was extracted from the 60.3 hybridoma [11]according to the method of Davis et al [23]. First strand cDNAwas synthesized according to the manufacturer's instructions byusing an oligo(dT) primer (Invitrogen, San Diego, CA). cDNAsof 60.3 H and L chains were amplified by PCR using degenerateprimers complimentary to the framework 1 and constant regions[24,25], which were modified to facilitate cloning. Xho I sites



were used at the 5' ends of the sense heavy (MH-SP-ALT. 1 orMH-SP-ALT.2) and light (EcoRI/FRl-ML(k)) chain primers.Pst I sites were utilized for antisense primers (MH-gamma-CONST and Hindm/ML(k)CONST).

For both H and L chains, PCR products were cloned intopUC18 using either the Xho I and Pst I sites or by blunt endedligation into the Sma I site. Ligation products were used totransform competent E.coli DH5a. Clones containing 0.5 kbEcoR I—Sal I fragments were subjected to double stranded DNAsequencing [26] with Sequenase (U.S. Biochemicals, Cleveland,OH). Products obtained by blunt ended ligation were used in thenext step.

Insertion of V genes into expression vectorsIn order to clone die amplified VL and VH genes into pGx. 11and PN7I.I6, respectively, they became templates for a secondPCR using the following primers (60.3 coding sequences areunderlined, the second part of the leader sequence (L2) is in boldprint, restriction sites are enclosed by brackets and the remaindercorresponds to intron sequences):

Heavy chain sense primer (EcoR I/Sac I)5 '-AJGAATTCGAGCTC]TTTTCTGATAACGTTGTCCTT-CTGTTTCTTGCAGGTGTCCAGTGTCAGGTCCAACTTC-AGCAGCCTGGG-3'Heavy chain anti-sense primer (Sal I/Xho I)5' -AJGTCGACCTCGAG1TGTGAGGACTCACCTGAGG-AGACGGTGACTGAGGTGCCT-3'

Light chain sense primer (EcoR I/Hind HI)5'-AJGAATTCAAGGTT]TCCTGACTACATGAGTGCAT-TTCTGTTTTATTTCCAATTTCAGATACCACCGGAQA-C ATTGTGCT AAC AC AATCTCC A-3'

(The penultimate nucleotide of the light chain sense primer (C,double underlined) corrected for an inappropriate base (A)introduced by the original sense primer).Light chain antisense primer (Sal I/Xho I)5'-AJGTCGACCTCGAGlATCACTTACGTTTGATTTC-C AGCTTGGTGCCTCC AC-3'

Amplified V genes were first placed into intermediate vectorsfor sequencing and then directionally cloned into expressionvectors. Alternatively, V genes can be sequenced directly fromthe expression vectors. For the heavy chain, the Sac I and SalI digested PCR product was cloned into the Sac I and Xho I sitesof PN71.16 (the PCR product could not be digested with XhoI due to internal Xho I sites; Sal I and Xho I have compatiblesticky ends). For the light chain, the Hind HI and Xho I siteswere used for cloning.

Transfection, limiting dilution, and soft agarose cloningHeavy and light chain expression vectors were digested with ClaI and Not I, respectively. Ag8.653 cells (107) were transfectedby electroporation wim 10 /ig each of H and L chain as describedby Raff et al [18], except the PBS was not supplemented withCa + + /Mg + + . After incubation on ice for 10 min, the contentsof the cuvettes were transferred into a flask containing 20 misof culture medium (Iscove's Modified Dulbecco's Medium, 10%fetal calf serum) and incubated at 37°C for 48 hrs in a CO2incubator. Cells were plated in 96-well microtiter plates atdensities between 2 x 104 and 1 x 103 cells per well in culturemedium containing G418 (0.6 mg/ml, GIBCO, Gaidiersburg,

MD). These cells were fed every third day for 2 weeks beforescreening for whole antibody by ELISA.

Master wells secreting the most Ig underwent limiting dilutioncloning. Cells were suspended in culture medium at 1 cell/400 id (1 cell/2 wells) before being distributed in 96 well plates(200 /tl per well). Plates were checked approximately every3 days for growth. Supernatants from wells with positive growthwere tested by ELISA for presence of Ab; culture supernatantsfrom individual wells were used in the Western Blots describedin this report.

Soft agarose cloning was performed essentially according otthe procedure of Raff et al [ 18], except that the cells in agarosewere layered over a 5 ml agarose base, and 5 % fetal bovine serumwas used.

ELISAThe presence of chimeric 60.3 was detected by a conventionalELISA using goat anti-human IgG antiserum (AntibodiesIncorporated, Davis, CA) as a capture reagent and horseradishperoxidase (HRPO) conjugated goat anti-human kappa (SouthernBiotech, Birmingham, AL) for detection. Purified chimeric L6[21] was used as a standard.

SDS-PAGE and Western blot analysisCulture supernatant containing 20 ng of reduced Ab was analysedby SDS-PAGE on a 4-20% gradient gel using the Novex(Encinitas, CA) gel system, which is based on die discontinousbuffer system of Laemmli [27]. The Western blot was performedaccording to the manufacturer's instructions using the enhancedchemiluminescent (ECL) system (Amersham, Chicago, PL).HRPO conjugated antibodies were used at a dilution of 1:2,000in blotto (0.5 % nonfat milk, 0.01 % thimerosol, 0.01 % antiform-A in PBS). X-ray film (Hyperfilm™-ECL, Amersham) wasexposed for <30 seconds.

RESULTS AND DISCUSSION

Design of vectors

High level expression of Ig from genomic DNA constructstransfected into murine myeloma cell lines has been welldocumented [5], while Liu et al [28] and others [29] have reportedweak expression of antibody using cDNA constructs. It is knownthat an Ig promoter/enhancer pair can elicit efficient expressionin mouse myeloma cells. Furthermore, there is considerableevidence that introns are essential for the efficient expression ofIg genes [30,31]. Recently, one group described a kappa vectorthat omitted both the leader-V and J-C introns [32]; expressionwas not observed until the J-C intron was incorporated in theirvector. Since non-coding intragenic sequences may have apositive influence on expression, our vectors were designed tomimic the genomic structure of rearranged Ig genes. Thearrangement of these genes in a 5' to 3' direction is: promoter,leader exon 1, leader-V (L-V) intron, V(D)J exon, J-C intron,eidier a C gene exon (x, and X genes) or exons and introns (ji,8, y, e, and a genes), and 3' untranslated region. Thesecomponents are transcribed as a unit, with die precursor mRNAbeing processed and translated into protein.

As shown schematically in Figure 3, genomic Ig DNAfragments have been assembled into modified versions ofpSV2-gpt (containing £cogpt gene [ 13]) and pSV2-neo (encodingneomycin and G418 resistance [ 19]) for our light and heavy chain



L2-V(D)J

L1

y \, Amplified V region

amplified V gene is inserted into MCS of cassette vector

C

Cassettevector

Proa. fragm*nt MCS MHE Ctregmvm

Final=»uv expression

vector

Figure 3. Components of the cassette vectors and amplified V genes. Schematic repesentation of the amplified V gene, the cassette vector, and the expression vector.The parts of the cassette vector contained within modified pSV2-gpt and pSV2-neo plasmids are: the Pro/L section [ • ] ; a MCS [ HIE ] for insertion of the V gene;the MHE [E2]; and the C region [ • ] . The component which includes the amplified L2-V(D)J exon [ • ] is inserted into the MCS of the cassette vector to formthe final expression vector.

r t t v e r a *initiator

lpninip W f l . 1 2 GATATCfcGJUUlATGAAXACCCACATTTTJpWTl-16 I . . . . . •[ j .AAA.T I

J tTTAaakOCCCAGTCAlttTGCAAAtoAaAGAAWCrrTAGAGfrAGAAAllCTGAG ["... „ - , « „ , < H \» l . c . T . T T- • • A • -I- T TT nuclaotldoa

NAG/GTRAGT

Figure 4. Comparison of the Pro/L sections of pNyl. 12 and pNyl. 16. Schematic representations of the Pro/L regions of pN71.12 (containing the wild type promoter)and PN7I.I6 (mutated promoter). A base in the pN-yl.16 sequence that is identical to one found in PN7I.12 is depicted by a single dot (•)• The names of significantregions are given above the pN-yl. 12 sequence. Restriction sites and sequence motifs are enclosed within boxes. Consensus sequences of the motifs are presentedbelow the boxes. Abbreviations for the nucleotides: A: adenine; C: cytosine; G: guanine; T: thymine; W: A or T; S: C or G; M: A or C; R: A or G. The leadersequence (LI) is underlined.

vectors, respectively. The components grouped 5' to 3' withinthe modified pSV2 vectors are: 1) an Ig promoter/leader sequenceregion (Pro/L; which includes partial intron sequence); 2) a MCSfor subcloning in the V gene of interest; 3) the MHE [33] and4) a human C region gene (Cx, C. or C7i).

The component farthest upstream is an Ig promoter. It has beenshown that the combination of octamer motif (ATTTGCAT andATGCAAAT for the light and heavy chain, respectively) andTATA box element in the promoter are minimal requirementsfor the expression of Ig genes in lymphoid cells [34,35]. TheH chain promoter frequendy contains a heptamer motifimmediately 5' of the octamer [36]. The heptamer has beenshown to enhance Ig H promoter activity in combination wtihthe octamer and TATA box motifs [37]. All of these elementsare within 100 nt of the RNA start site; this value was chosenas the minimal Ig promoter length necessary for our expressionvectors.

The Ig promoter is followed by the leader sequence thatencodes the signal peptide. Although signal sequences lacksequence homology between each other, they do share structuralsimilarity and are capable of facilitating the export of heterologous

proteins [38]. In Ig genes, the leader sequence is composed oftwo exons. Leader exon 1 (LI) encompasses all but the last 11nucleotides that encode the signal peptide. These 11 nucleotides(L2) follow the L-V intron and form the initial part of theL2-V(D)J exon. In our design, LI is incorporated as a componentof the vectors, whereas L2 is added as part of of the V(D)J gene.

In our cassette vectors the promoter, LI, and the 5' portionof the L-V intron are integral parts of the Pro/L region (Figure 3).Sequences for Pro/L sections were taken from genomic constructswhose expression resulted in high levels of Ab production (> 100/ig/ml after cloning). The Pro/L segments from L6 [20,21] and4B9 [17,18] were used in H and L chain cassette vectors,respectively.

Restriction sites were added by PCR to both ends of the L6Pro/L region, which included 607 nt upstream of the initiatormethionine. In reviewing die sequence of rhis region, we noticedseveral mismatches to consensus sequences that may influenceexpression levels. We chose to 'correct' these mismatches anddetermine whether the mutated promoter (in vector PN7I.I6)performed differendy than the wild type L6 promoter (in vectorPN7I.I2). Bases were altered by the PCR based SOEing (splicing



by overlap extension) technique [39]. The mutations are shownschematically in Figure 4 and include addition of a matrixassociation region (MAR), alterations to the TATA box, splicedonor signal and heptamer motifs, and removal of an invertedTATA box and its associated initiator methionine. MARs (oftenlocated within or near a topoisomerase II site) have been shownto stimulate several fold the transcription of transfected Ig genes[40]. The role of the TATA box [34,35,41] in transcription hasbeen well documented. Furthermore, the presence of an invertedTATA box upstream of the octamer may interfere withtranscription in the forward direction [42]. A heptamer motif isfrequently found upstream of the octamer and may positivelyinfluence expression [36,37]. Modifications to the splice donorsignal [43] could alter both the rate of RNA splicing and the rateof signal peptide cleavage [44] since the mutated sequenceencodes a Gly instead of Ser at the —4 position.

The 4B9 Pro/L includes 402 nt upstream of the initiatormethionine and is in a Sau3A I fragment containing LI and someintron sequences. Interestingly, there is a single nt mismatchwithin the octamer consensus motif (CTTTGCAT vsATTTGCAT). The CTTTGCAT is characteristic of the humanVxIJJ subgroup. The significance of this disparity is not known,but it does not appear to interfere with transcription of transfectedIg genes [45]. As with the heavy chain promoter, a mutatedversion of this promoter has been prepared; however, it has notyet been tested.

In the L-V intron there are three important sequence motifsinvolved in splicing: a splice donor site, branch point lariat motifand a splice acceptor site. There is evidence that a minimum of50 bp of intron sequence between the splice donor site and thebranch point lariat signal is necessary for efficient splicing [46].More than 50 bp of L-V intron sequence has been included inboth of our expression vectors. In the L and H chain cassettevectors, following LI are the 5' region of the L-V intron andthe 3' segment of the J-C intron. The two partial introns are joinedby a polylinker that accepts the amplified V region; providedon this PCR product are die L2-V(D)J exon and the remainingportions of flanking introns.

In both of our vectors, the MHE follows a MCS and is includedin the 3 ' part of the partial J-C intron component. The 1 kb XbaI fragment is used for the L chain vector, and the 0.7 kb XbaI to EcoR I fragment for H chain cassette vectors; this truncatesunnecessary intronic sequence flanking the enhancer [47]. Thesefragments are functionally equivalent in myeloma cells and werechosen for compatible restriction sites.

In our cassette vectors, the enhancer is followed by 3' intronsequence that includes a branch point consensus sequence andan acceptor site for splicing to the C region. This intron sequenceis provided by the DNA fragment containing either the x or 7Igenomic C region gene. This fragment also has 3' untranslatedsequence including the polyadenylation signal [48]. The C genein H chain vectors can be switched to g2, -y3, or -y4 by substitutingthe Hind HI—BamH I fragments, whereas the C regions of otherheavy chain isotypes can be modified to have flanking Hind HIand BamH I sites, and then inserted into this location.

Preparation of V genes for cloning into expression vectors

For insertion into our expression vectors, the V gene must beflanked by L-V and J-C intron sequences on the 5' and 3' sides,respectively. In an ideal case, this would be accomplished witha single round of PCR using a cDNA template and degenerateprimers. The sense primer would contain a restriction enzyme

site, intron sequence, and degenerate sequence for annealing tothe 5' part of the L2-V(D)J region. The antisense primer wouldinclude a restriction site, intron sequence, and degeneratesequence for the 3' end of the J regions. However, the lowefficiency in the synthesis of long degenerate oligonucleotidesmakes this approach impractical. Moreover, a second PCR isnecessitated by the use of either signal peptide or constant regionprimers, since intron sequences must be provided 5' and 3' ofthe L2-V(D)J exon.

With our method, two PCR steps are used to obtain the V geneinsert. In the first step, the V(D)J region is amplified usingdegenerate primers from either the leader or framework regions(sense orientation) and from the constant region (anti-senseorientation). The use of sense primers from the leader regionis preferable becasue oligonucleotides homologous to frameworkregion 1 (FR1) may cause mutations in the V gene. Followingsequence determination, a set of non-degenerate primers can bedesigned to provide intron and coding sequences, as describedabove. Either the original cDNA or the product of the first PCRcan be used as template for the second PCR.

The cDNAs of 60.3 V genes were made from mRNA, andthen amplified by PCR with degenerate framework primers asdescribed in Materials and Methods. Amplified V genes werecloned into an intermediate vector for sequencing. Initially, thisinformation was used to select clones containing V-gene like openreading frames. The sequences of several independent cloneswere then compared to each other to eliminate those withmisincorporations caused by reverse transcriptase [49] and/or Taqpolymerase I [50]. After obtaining the DNA sequences, thededuced amino acid sequences of 60.3 V genes were comparedto known murine Ig sequences [51] and to the N-terminal aminoacid sequence obtained by Edman degradation. The 60.3 VH andWx domains belong to subgroups JJB and UJ, respectively.Sequences will be reported in another publication (Hsiao et al,manuscript in preparation).

After the initial sequence determination, the V genes wereamplified again, with the addition of flanking intron sequencesthat are required by the expression vectors. The senseoligonucleotide includes in a 5' to 3' direction: (1) an enzymesite for cloning into the MCS of the cassette vector; (2) intronsequence containing the branch point lariat motif and apolypyrimidine tract [52]; preceding (3) a splice acceptor site[43]; (4) the last 11 nucleotides that encode the signal peptide(L2); and (5) the 5' portion of the V(D)J gene. The 11 nucleotidesin (4) are derived from 4B9 and L6 (for L and H chains,respectively) and encode amino acids which conform to the rulesfor signal sequence cleavage sites (e.g., - 3 , — 1 rule [44]). Onlysection (5) of the primer is homologous to and anneals with thetemplate.

In a 5' to 3' direction, the anti-sense primer supplies: (1) asecond cloning site; (2) intron sequence that has a splice donorsite for joining to the C region gene; and {3) a portion of the3' end of the V(D)J gene. Only region (3) anneals to the template.Partial introns included in the primer sequences are derived fromeither genomic 4B9 DNA or L6 DNA for the L and H chainconstructs, respectively. A stretch of 4 to 10 extra nucleotidescan be found at the 5' end of each primer to allow the enzymesto cut more efficiently. The sequences of PCR products fromthe second step were verified prior to their insertion intoexpression vectors. Alternatively, these PCR products can becloned directly into the expression vectors and sequence can beconfirmed using primers from the expression vector.



1 2 3 4

974-

69.0-46 0 -

3 0 0 - —

21 5 -

14.3- £.

C , , , a D

2 3 4

3 4 1 2 3 4

97.4-

69.0-46 0-

3 0 0 -

21.5-

14.3-

Flgure 5. Western blot analysis of murine and chimeric 60.3 antibodies. Culturesupematants containing 20 ng of reduced Ab were analysed by SDS-PAGE ona 4-20% gradient gel. Proteins were transferred to nitrocellulose membranesand reacted witfi specific HRPO-conjugated goat antibodies as follows: A) anti-murine IgG2a heavy chain; B) anti-human IgG heavy chain; C) ano-murine xlight chain; and D) anti-human x light chain. Lanes 1: culture supernatant fromAg8.653 cell line; lanes 2; culture supernatant from murine 60.3 hybridoma; lanes3: culture supernatant from transfectomas produced with light chain vector(pGx. 11) and mutated (pN-yl. 16) heavy chain vector, lanes 4: culture supernatant,from transfectomas produced with light chain vector (pGx.ll) and wild type(PN7I.I2) heavy chain vector.

ExpressionPlasmid constructs with chimeric H (with either the wild typeor mutated promoter) and L chain were cotransfected byelectroporation into Ag8.653 mouse myeloma cells, and thenselected for expression of the neomycin resistence gene (asdescribed in Materials and Methods). Two weeks posttransfection, the presence of the chimeric Ab in culturesupematants was determined by ELISA. As seen in Table I, theheavy chain vector with the mutated promoter (PN7I.I6) gavesignificantly better results (more master well supematantscontaining >300 ng/ml Ig) than the vector with the wild typepromoter (pNyl.12). When seeded at a low density ( lxlO 3

cells/well), the outgrowth with the two vectors was equivalent.Cells from the best producing master wells were subjected to

limiting dilution cloning and the Abs from the resultant cloneswere analyzed by Western blot (discussed below) and bindingassays. The best producing master wells were expanded into 24well plates. After one round of soft agarose cloning, the spentculture supematants contained 1 to 3 /tg/ml of chimeric 60.3 Ab.In a separate experiment, the same light chain vector and themutated version of the heavy chain vector were used for theexpression of humanized 60.3 (Hsiao et al, manuscript inpreparation). After two rounds of limiting dilution cloning andone round of soft agarose cloning, spent culture supematants heldup to 10 /tg/ml of humanized 60.3 Ab.

As stated above under 'Design of Vectors', there were severalmutations to the heavy chain promoter in pNyl.16. Thesemutations could have enabled the transfectomas to express moreIg by altering any one or a combination of the following:integration of the vector, rates of RNA transcription, RNAsplicing, or cleavage of the signal peptide. Futher experiments

Table I.

Promoter

wild typemutated

Comparison of

<50

924832

vectors with mutated and wild

Antibody Concentration50 to 300

3087

type promoters

(ng/ml)>300

641

Ag8.653 cells were transfected as described in Materials and Methods with 10MgpGx.ll and either 10 jigpN?!. 12 (heavy chain vector with wild-type promoter)or 10 it%pHyl. 16 (heavy chain vector with mutated promoter). Cells were platedat 2 x 10*/well in media containing G418. Two weeks post transfection, culturesupematants (diluted 1 to 50) were screened for Ig by ELISA. Numbers indicatethe quantity of master wells (out of 960) with the indicated antibody concentration.In similar experiments where the cells were plated at lower density, outgrowthwith the two heavy chain vectors was equivalent. Control, Ag8.653 cellselectroporated in the absence of DNA did not survive the selection with G418.These experiments were repeated three times with similar results.

will be done to distinguish between these possibilities and todetermine which mutations are responsible for the increasedexpression. Similar experiments will also be done with the lightchain vector.

The size and immunoreactivity of the chimeric antibody inculture supematants was determined by Western Blot analysis(Figure 5). Reduced chimeric proteins were detected by specificreagents (HRPO labelled goat anti-human x and HRPO labelledgoat anti-human >). Only one band each was detected with theH and L chain reagents. These bands were of similar size to themurine controls (detected with reagents specific for murineantibodies). In addition, FACS assays indicated mat chimeric 60.3Ab binds as well as the murine 60.3 Ab to CD 18 positive cells(Hsiao et al, manuscript in preparation).

Comparison with other vectorsMany vectors have been developed for the expression of Ig geneswhere the V region was obtained in the absence of a promoter(i.e., from cDNA or by PCR). These vectors fall into two generalclassifications: 1) transient expression vectors [6-8] that cangenerate in a short time sufficient antibody for experimentation;and 2) stable expression vectors which are more appropriate formanufacturing purposes. Most of the vectors currently availablefor stable expression require the investigator to provide a leadersequence along with the V region [53—58]. This may be achievedwhen degenerate leader region primers are employed to amplifythe V gene. However, in cases where the V gene can not beobtained with leader sense primers, it may be necessary to takeadvantage of degenerate primers homologous to FR1. To usethese primers with certain vectors, the signal sequence, with orwithout L-V intron, must be added during amplification of theV gene. Two groups [59,60] describe vectors containing leadersequences, but they clone V genes into their vectors usingrestriction sites in FR1. This causes several amino acids to beencoded by the vector and/or primer and may result inmodifications of the original sequence. Since our vectors includethe leader sequence plus a partial L-V intron, the V gene isinserted without altertion, by using restriction sites in the intron.Equally important, our vectors permit the flexibility of usingdegenerate sense primers from either the leader sequence or fromFR1 of the V gene. If FR1 primers are required, the deducedamino acid code can be compared with the actual sequenceinformation obtained by N terminal analysis. When necessary,the nucleotide sequence is modified to correct any discrepancies.



SUMMARY

The vectors presented in this paper represent a rapid method forthe construction of chimeric, humanized, or human antibodies.Our vectors can be used for expressing Ig without the loss ofantibody affinity or specificity (Hsiao et al, manuscript inpreparation). The L and H chain expression vectors preserve thegenomic organization of rearranged Ig genes by utilizing an Igpromoter plus enhancer and by including intron sequences.cDNAs were prepared from mRNA templates derived from the60.3 hybridoma cell line. Resultant cDNAs were amplified byPCR, their products sequenced, amplified again with the additionof introns and sequenced a second time, men subcloned into ourmammalian expression cassette vectors for subsequenttransfection into a mouse myeloma cell line. The rate limitingstep of this methodology is the requirement to sequence severalindependent clones after each PCR. This ensures that the fidelityof the V gene sequence has not been compromised due to possibleerrors introduced either from single round copying of cDNA frommRNA with reverse transcriptase, or from amplification roundswith Taq polymerase I. However, this sequencing effort isminimal relative to the classical preparation of chimeric antibodieswhich involved first generating then screening a genomic orcDNA library.

Without changing the overall structure of the heavy chainvector, we have introduced several mutations into thepromoter/leader region: addition of a matrix associated region,removal of a reverse TATA box and associated initiatormethionine, and mutations resulting in sequences which betterconform to the consensus for TATA box, heptamer motif, andsplice recognition signals. One or more of these mutations ledto an increase in the number of transformants secreting higherlevels of Ig (>300 ng/ml in master well culture supernatants).We are investigating the mechansim and specific mutation(s)which allowed this increase.

ACKNOWLEDGEMENTS

We thank Scott Austin for preparing oligonucleotides, CherylBradley and Anne Newell for cloning transfectomas, AlisonMalacko and Hans Marquardt for amino acid sequencing,Michael Neubauer for DNA sequencing and H.Perry Fell, KimFolger, Stephen McAndrew, Monica Van der Vieren, ReithaWeeks and Dale Yelton for critical review of the manuscript.

REFERENCES

1. Levy.R.L., and Miller.R.A. (1983)/*. Rev. Med, 34, 107-127.2. Shawler.D.L., Bartholomew,R., Smith.L., and Dilliman,R. (1985) J.

Immunol, 135, 1530-1535.3. Benjamin.RJ., CobboM.S.P., OarkJvl.R., and Wakhnarm,H. (1986)/ Exp.

Med., 163, 1539-1552.4. BrOggemann.M., Williams.G.T., Caskey.H.M., Teale.C, Spicer.M.A.,

Suraai.M.A., and Nueberger.M.S. (1990)BehringInst. Mitt., 87, 21-24.5. Morrison.S.L., and Schom,J. (1990) Important Adv. Oncol, 3 -18 .6. Daugherty.B.L., DeMartino.J.A., Law,M.-F., Kawka.D.W., Singer.I.I.,

and Mark,G.E. (1991) Nucleic Adds Res., 19, 2471-2476.7. Kettleborough.C.A., SaldanhaJ., Heath.V.J., Morrison.C.J., and

Bendig.M.M. (1991) Protein Eng., 4, 773-783.8. Carter.P., Presta.L., Gorman.C.M., Ridway.J.B.B., Henner.D.,

Wong.W.T.L., RowiandAM., Kotts.C, Carver.M.E., and Shepard.H.M.(1992) Proc. NmL Acad Set USA. 89, 4285-4289.

9. Bouliannc.G.L., Hozumi.N., and Schuhnan.MJ. (1984) Nature, 312,643-646.

10. Monison.S.L., Johnson.M.J., Herzenberg.L.A., and Oi.V.T. (1984) Proc.Nad. Acad. Set. USA, 81, 6851-6855.

11. Beatty.P., LedbetterJ.A., Martin.P.J., Price.T.H., and HansenJ.A. (1983)/ Immunol., 131, 2913-2918.

12. Beatty.P., Ochs.H.D., HarlanJ.M., Price.T.H., Rosen.H., Taylor.R.F.,HansenJ.A., and Klebanoff.S.J. (1984) Lancet, 1, 535-536.

13. Mulligan,R.C, and Berg.P. (1980) Science, 209, 1422-1427.14. Yanisch-Perron.C, Vierra.J., and Messing^. (1985) Gene, 33, 103-119.15. MarshJ.M., Erfle.M., and Wykes.E.J. (1984) Gene, 32, 481-485.16. Lang,R., Stanton,L., and Marcujf. (1982) Nucleic Adds Res., 10,611-620.17. The nucleocide sequence data for the 4B9 L chain promoter will appear in

the EMBL, GenBank and DDBJ Nucleotide Sequence Databases under theaccession number Z17330; nucleotides 47 through 623.

18. Raff,H.V., Bracfley.C, Brady.W., DonaWson,K., Ljpsich.L., Maloney.G.,Shuford.W., Walls.M., Ward,P., Wolff.E. and Harris.L. (1991) / Infect.Dis., 163, 346-354.

19. Southern.P.J., and Berg.P. (1982) /. Mol Appl. Genet., 1, 327-341.20. The nucleotide sequence data for the L6 H chain promoter will appear in

the EMBL, GenBank and DDBJ Nucleotide Sequence Databases under theaccession number Z17336; nucleotides 26 through 724.

21. Fell.H.P., Gayk.M.A., Ydton,D., Lipsich,L., Schieven.G.L., MarkaU.S.,Aruffo.A., Heustrom.K.E., Hellstrom,I. and Bajorath.J. (1992) J. BiolChem., 267, 15552-15558.

22. The nucleotide sequence data for the 2.8 Kb Hind m - P v u II fragmentencoding the C7I gene will appear in the EMBL, GenBank and DDBJNucleotide Sequence Databases under the accession number Z17370.

23. Davis,L.G., Dibner.M.D. and BatteyJ.F. (1986) Basic Methods in MolecularBiology. FJsevier, N.Y.

24. Larrick,J.W., Coloma,M.J., Del Valle.J., Fernandez.M.E., Fry.K.E. andGavilonodo-CowleyJ.V. (1990) Scand J. Immunol., 32, 121-128.

25. ColomaJ.J., Larrick.W., Ayala.M. and GavUondo-CowleyJ.V. (1991)BioTechniques, 11, 152-156.

26. Hsiao,K.-C. (1991) Nucleic Acids Res., 19, 2787.27. Laemmli.U.K. (1970) Nature, 227, 680.28. Lhi.A.Y., Robinson,R.R., Hellstrom.K.E., Murray Jr.,E.D., Chang.C.P.,

and Hellstrom,I. (1987) Proc. Nail Acad Sd. USA, 84, 3439-3443.29. Sahagan,B.G., Dorai.H., Sahzgaber-MuUerJ., ToneguzzoJ1., Guindoo.C.A.,

Lilly.S.P., McDonald.K.W., Morrissey.D.V., Stone.B., Davis.G.,McIntrosh,P., and Moore.G. (1986) J. Immunol, 137, 1066-1074.

30. Neuberger.M.S., and Williams.G.T. (1988) Nucleic Adds Res., 16,6714-6723.

31. GOlies.S.D., Dorai.H., WesolowskU., Majeau.G., Delano.Y., BoydJ.,Gardener.J., and James K. (1989) BioTechnology, 7, 799-804.

32. Coloma.MJ., Hastings^., Wims,L.A., and Morrison.S.L. (1992) J.Immunol Meth., 152, 89-104.

33. Gillies.S.D., Morrison.S.L., Oi.V.T., and Tonegawa.S. (1983) Cell, 33,717-728.

34. Parslow.T.G., Blair.D.L., Murphy, W.J., and Granner.D.K. (1984) Proc.Natl Acad Sd. USA, 81, 2650-2654.

35. Falkner.F.G. and Zachau.H.G. (1984) Nature, 310, 71-74.36. Eaton.S. and CalameJC. (1987) Proc. Natl Acad Sd. USA, 84, 7634-7638.37. Landolfi.N.F., Capra.J.D., and Tucker.P.W. (1988) Ataum, 323, 548-551.38. Gierasch.L.M., Biochem. 28, 923-930.39. Ho.S.N., Hunt,H.D., Horton,R.M., PuDen^.K. and Pease.L.R. (1989) Gene,

77, 51-59.40. Cockerill.P.N., Yuen.M.H. and Garrad.W.T. (1987)7. Biol Otem.,262,

5394-5397.41. BOcher.P. and Trifonov.E.N. (1986) Nucleic Adds Res., 14, 10009-10026.42. Doyen,N., Dreyfus, M., and Rougeon.F. (1989) Nucleic Adds Res., 14,

4683-4690.43. Gelfand.M.S. (1989) Nucleic Acids Res., 17, 6369-6382.44. von Heinje.G. (1986) Nucleic Adds Res., 14, 4683-4690.45. Goldrick.M.M., and Gottlieb.P.D. (1988) J. Immunol, 140, 646-653.46. Smith.C.W., and Nidal-Ginard.B. (1989) Cell, 56, 749-758.47. Tsao.B.P., Wang, X.-F., Peterson, C.L., and Calame.K. (1988) Nucleic

Acids Res., 16, 3239-3253.48. Bimstiel.M.L., Busslinger.M., and Strub.K. (1985) Cell, 41, 349-359.49. Gopinathan.K.P., Weymouth.L.A., Kunkel.T.A., and Loeb.L.A. (1979)

Nature, 278, 857-859.50. Saiki,R.K., Gdfand.D.H., Stoffel.S., Scharf.SJ., Higuchi.R., Hom.G.T.,

MuUis.K.B., and Erlich.H.A. (1987) Science, 239, 487-491.51. Kabat, E.A., Wu.T.T., Peny^I.M., Gottesman.K.S., and Foeller,C. (1991)

Sequences of Proteins of Immunological Interest, Fifth Edition, U.S.Department of Health and Human Services, Washington, D.C.

52. Kdler.E.B., and Noon.W.A. (1984) Proc. Natl Acad Sd. USA, 81,7417-7420.

53. Verhoeyen,M., Milstein.C., and Winter.G. (1988) Science, 239, 1534-1536.



54. Queen.C., Schneider.W.P., Selick.H.E., Payne,P.W., Landolfi.N.F.,Duncan,J.F., Avdalovic.N.M., Levitt,M., Junghans.R.P. andWaldmann.T.A. (1989) Proc. NatL Acad. So. USA, 86, 10029-10033.

55. Co.M.S., Deschamps.M., Whidey.R.J., and Queen.C. (1991) Proc. NatLAcad. Sd. V.SA-, 88, 2869-2873.

56. FooteJ., and Winter.G. (1992) /. Mol. BioL, 224, 487-499.57. Hastings^., Morrison.S.L., Kanda.S., SaxtonJl.E., and Irie.R.F. (1992)

Cancer Res., 52, 1681-1686.58. XiangJ., and Chen.Z., (1992) Immunol., 75, 209-216.59. Jones.P.T., Dear.P.H., FooteJ., Neubojer.M.S. and Winter.G. (1986)

Nature, 321, 522-525.60. Oriandi.R., Gflssow.D.H., Jones.P.T., and Winter.G. (1989) Proc. NatL

Acad. USA, 86, 3833-3837.


vectors for the expression of pcr-amplified immunoglobulin

Documents