Samuel Lunenfeld Research Institute, Mount Sinai Hospital, Toronto, Ontario M5G 1X5, Canada
`
`Gene, 150(1994)409%410 0 1994 Elsevier Science B.V. All rights reserved. 0378-l 119/94/$07.00 GENE 08393 A rapid procedure for the humanization of monoclonal antibodies* (Antibody humanization; site-directed mutagenesis; polymerase chain reaction; efficiency) Jasbir Singh Sandhu
`
`Received by H. Zachau: 29 January 1993; Revised/Accepted: 25 April/28 April 1994; Received at publishers: 29 August 1994 409 SUMMARY An efficient and rapid procedure for the humanization of murine monoclonal antibodies (MumAb) is described. It consists of site-directed mutagenesis (SDM) to transfer the murine complementarity-determining regions (MuCDR) onto human framework regions (HuFR), followed by polymerase chain reaction (PCR) of the SDM product. Using SDM/PCR, rapid and correct humanization of MumAb heavy chains is clearly demonstrated. Compared to current protocols this method considerably reduces the time and labour required to generate humanized mAb. To test large numbers of HumAb for immunotherapy, a simple and efficient method for humanizing MumAb is required. Both SDM (Riechmann et al., 1988) and PCR (Lewis and Crowe, 1991) have been used separately to humanize MumAb. However, these procedures have either very low efficiency, or are time consuming. Described here is a new efficient humanization method, which consists of SDM followed by PCR of the SDM product (see Fig. 1). The effectiveness of the SDM/PCR humanization procedure is demonstrated by humanizing the Vn domain of MumAb (40-40) (Hudson et al., 1987) and DX48 (Lewis and Crowe, 1991). These two MumAb are specific for the hapten digoxin. The DNA fragment (HindIII-BamHI) coding for the heavy chain variable domain (HuVNp) from the human-
`Dr. J.S. Sandhu, Samuel Lunenfeld Research Institute, Mount Sinai Hospital, 600 University Avenue, Toronto, Ontario M5G 1X5, Canada. Tel. (1-416) 586-8251; Fax (1-416) 586-1570. *On request, the author will supply detailed experimental evidence for the conclusions reached in this Brief Note. Abbreviations: CDR, complementarity-determinining region(s); Ab, antibody(ies); FR, framework region; Hu, human; HuVNp, nitrophenyl- acetyl-specific Vu; Ig, immunoglobulin; mAb, monoclonal Ab; Mu, murine; oligo, oligodeoxyribonucleotide; PCR, polymerase chain reac- tion; ss, single strand(ed); SDM, site-directed mutagenesis (sed); V,, heavy-chain variable domain, SSDI 0378-1119(94)00604-O ized mAb HuNp (specific for the hapten nitrophenactyl) was removed from the plasmid pSV,,, HuV,,-HuIgG (provided by Dr. Neuberger, Cambridge, UK), and cloned into the corresponding sites in M13mp18, to produce the construct M13mpl8HuV,,. The M13mpl8HuV,r ss DNA was used as the template for the SDM. The oligos used for the SDM coded for MumAb DX48 or (40-40) V, CDRs, and HuVNp FRs. The SDM reactions were done using the modified method of Taylor et al. (1985). During each SDM either MuCDRl, or MuCDRl and -2, or MuCDRl, -2 and -3 were transferred onto HuVNp FRs. A portion (SO%, v/v) of the SDM DNA was used to transform competent E. coli cells (TGl). The resulting colonies were screened with [32P]oligos coding for (mAb (40-40) or DX48) Vn MuCDRl (Sambrook et al., 1989). The number of bacte- rial colonies positive for Vu MuCDRl for mAb (40-40) or DX48 are shown in Table I. Plasmid DNA from 10 MuCDRl-positive colonies was prepared, and the V, domains were sequenced (Sanger et al., 1977; the results are shown in Table I). PCR of the SDM DNA (20%, v/v) was done using primers designed to generate the Vu domains of the HumAb (40-40) or DX48. The PCR DNA was digested with PstI + BstEII and an aliquot (lo%, v/v) was inserted between the corresponding restriction sites in the expression plasmid pVHD1.3-VKD1.3tagl (pro- vided by Dr. G. Winter, Cambridge, UK). The ligated
`
`Correspondence
`
`to:
`
`IMMUNOGEN 2169, pg. 2
`Phigenic v. Immunogen
`IPR2014-00676
`
`

`

`E. cali
`DH5. The number of colonies positive for mAb (40-40) or DX48 V, MuCDRl obtained by
`SDM
`
`Ncif
`
`r IElllI I,I,,i Itit,, 1 ,,,,,I I rtnl ll,,,, 1 Fig. 1. The SDM/PCR humanization procedure. The oligos (1, 11 and III) coding for the MuCDRs and HuFRs were annealed to the HuV,, ss DNA. Step A: second ss synthesis was done with phage T4 DNA polymerase using [a-?S]dCTP in the presence of phage T4 DNA ligase to generate the mutant heteroduplex DNA. Step B: the first ss does not have the [a-3sS]dCTP and, therefore, can be nicked with
`and digested with exonuciease IfI. The first ss was repolyme~~d with E. coli DNA polymerase I, and the two DNA ends were ligated with phage T4 DNA ligase. Step C depicts the annealing of the PCR primers 1 and 2, to the double-stranded DNA generated by the SDM reactions. In
`D, DNA suitable for cloning into a plasmid expression vector was produced by PCR. DNA was transformed into
`
`The number of V, MuCDRl-positive colo- nies obtained for mAb (40-40) and DX48 is higher by the SDMjPCR protocol compared to SDM procedure alone. For example, using SDM to graft the three Vzl MuCDRs of mAb (40-40) onto HuFRs, no MuCDRl-positive colo- nies were detected. However, using SDM/PCR, 30 MuCDRl-positive colonies were detected, and 10 had the correct V, nt sequence. These data clearly demonstrates the advantage of PCR amplification of the low copy number of the SDM DNA coding for the humanised VH. The data in Table I show that as the number of MuCDRs transferred in a single SDM or SDMjPCR increases (from TABLE I E. cofi colonies generated by the SDM and SDMjPCR humanization procedure Number of MuCDRs transferred onto HuFRs in each SDM” I 1 and 2 1.2and3 Heavy chain (40-40)b SDM 100 (80%) 50 (60%) 0 (0%) SDM~PCR 210 (90%) 100 (70%) 30 (30%) Heavy chain (DX48p SDM 150 (70%) 65 (50%) 4 (25%)d SDMjPCR 360 (70%) 190 (60%) 55 (30%) Aliquot (SO%, v/v) of the DNA from each SDM was used to transform E. co/i TGI. The remaining 20% (v/v) of the SDM DNA was used for PCR. Aliquot (lo%, v/v) of the PCR DNA was used for ligation and transformation of
`and SDMjPCR are shown. The V, domain of 10 muCDRl-positive colonies from each SDM and SDM/PCR was sequenced and the percentage of colonies with the correct DNA sequence are shown in bracket. ‘In each SDM or SDMjPCR either MuCDRs 1, or 1, 2 or 1,2, 3 were transferred onto HuFRs. bHeavy chain of the antidigoxin mAb 40-40. “Heavy chain of the antidigoxin mAb DX48. dPlasmid DNA was available from only four colonies, and one colony had the correct V, nt sequence. 1 to 3) the number of colonies positive for mAb 40-40 or DX48 MuCDRl with the correct nt sequence decreases. After humanization of DX48 and (40-40) heavy chain by SDM and SDM/PCR, the VH of MuCDRl-positive colonies was sequenced. The sequence data showed that for both SDM and SDM/PCR humanization protocols, the number of sequence errors in the humanized (40-40) and DX48 VH DNA was similar. Therefore, the PCR step, does not significantly increase the number of sequence errors in the humanized V, DNA. REFERENCES Hudson, N.W., Medgett-Hunter, M., Panka, D.J. and Margolies, M.N.: Immunoglobulin chain recombination among antibodies by hybri- doma-hybridoma fusion. J. Immunol. 130 (1987) 2715-2723. Lewis, A.P. and Crowe, J.S.: Immunoglobulin complementarity- determining region grafting by recombinant polymerase chain reac- tion to generate humanised monoclonal antibodies. Gene 101 (1991) 297-302. Riechman, L., Clark, M., Waldmann, H. and Winter, G.: Reshaping human antibodies for therapy. Nature 332 (1988) 323-327. Sambrook, J., Fritsch, E.F. and Maniatis, T.: Molecular Cloning. A Laboratory Manual, 2nd ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, 1989. Sanger, F., Nicklen, S. and Coulson, A.R.: DNA sequencing with chain- terminating inhibitors. Proc. Natl. Acad. Sci. USA 74 (1977) 5463-5467. Taylor, J.W., Ott, J. and Eckstein, F.: The rapid generation of oligo- nucleotide directed mutations at high frequency using phosphoro- theioate-modi~ed DNA. Nucleic Acids Res. 13 (1985) 8764-8785.
`
`Step
`
`E. coli
`
`DH5, and the resulting colonies were screened using [32P]oligos coding for (mAb (40-40) or DX48) VH MuCDRl. The number of MuCDRl-positive clones for mAb (~-40) or DX48 are shown in Table I. Plasmid DNA from ten MuCDRl-positive colonies was made, and the VH domains were sequenced (see Table I). Most of SDM DNA (SO%, v/v), but only a small portion of PCR DNA (2% (v/v) of the original SDM DNA) was used to trans- form
`
`E. coli.
`
`IMMUNOGEN 2169, pg. 3
`Phigenic v. Immunogen
`IPR2014-00676
`
`