Proc. Nati. Acad. Sci. USA
`Vol. 86, pp. 10029-10033, December 1989
`Immunology
`
`A humanized antibody that binds to the interleukin 2 receptor
`(chimeric antibody/antibody affinity/autoimmune disease)
`CARY QUEEN*, WILLIAM P. SCHNEIDER*, HAROLD E. SELICK*t, PHILIP W. PAYNE*,
`NICHOLAS F. LANDOLFI*, JAMES F. DUNCAN*t, NEVENKA M. AVDALOVIC*, MICHAEL LEVITT§,
`RICHARD P. JUNGHANS¶, AND THOMAS A. WALDMANN¶
`*Protein Design Labs, 3181 Porter Drive, Palo Alto, CA 94304; §Department of Cell Biology, Stanford University, Stanford, CA 94305; and IMetabolism
`Branch, National Cancer Institute, National Institutes of Health, Bethesda, MD 20892
`Contributed by Thomas A. Waldmann, August 30, 1989
`
`ABSTRACT
`The anti-Tac monoclonal antibody is known
`to bind to the p55 chain ofthe human interleukin 2 receptor and
`to inhibit proliferation of T cells by blocking interleukin 2
`binding. However, use of anti-Tac as an immunosuppressant
`drug would be impaired by the human immune response
`against this murine antibody. We have therefore constructed a
`"humanized" antibody by combining the complementarity-
`determining regions (CDRs) of the anti-Tac antibody with
`human framework and constant regions. The human frame-
`work regions were chosen to maximize homology with the
`anti-Tac antibody sequence. In addition, a computer model of
`murine anti-Tac was used to identify several amino acids
`which, while outside the CDRs, are likely to interact with the
`CDRs or antigen. These mouse amino acids were also retained
`in the humanized antibody. The humanized anti-Tac antibody
`has an affinity for p55 of 3 x 109 M-1, about 1/3 that ofmurine
`anti-Tac.
`
`The cellular receptor for the lymphokine interleukin 2 (IL-2)
`plays an important role in regulation of the immune response
`(reviewed in ref. 1). The complete IL-2 receptor (IL-2R)
`consists of at least two IL-2-binding peptide chains: the p55
`or Tac peptide (2, 3), and the recently discovered p75 peptide
`(4, 5). Identification and characterization of the p55 peptide
`were facilitated by the development of a monoclonal anti-
`body, anti-Tac, which binds to human p55 (2). The p55
`peptide was found to be expressed on the surface of T cells
`activated by an antigen or mitogen but not on resting T cells.
`Treatment of human T cells with anti-Tac antibody strongly
`inhibits their proliferative response to antigen or to IL-2 by
`preventing IL-2 binding (3, 6).
`These results suggested that anti-IL-2R antibodies would
`be immunosuppressive when administered in vivo. Indeed,
`injection of an anti-IL-2R antibody into mice and rats greatly
`prolonged survival of heart allografts (7, 8). Anti-IL-2R was
`also effective in rats against experimental graft-versus-host
`disease (9). In animal models of autoimmune disease, an
`anti-IL-2R antibody alleviated insulitis in nonobese diabetic
`mice and lupus nephritis in NZB x NZW mice (10). Anti-Tac
`itself was highly effective in prolonging survival of kidney
`allografts in cynomolgus monkeys (11).
`In human patients, the specificity of anti-Tac for activated
`T cells might give it an advantage as an immunosuppressive
`agent over OKT3 (monoclonal anti-CD3), which is effective
`in treating kidney transplant rejection (12), but which sup-
`presses the entire peripheral T-cell population. In fact, in
`phase I clinical trials for kidney transplantation, prophylactic
`administration of anti-Tac significantly reduced the incidence
`of early rejection episodes, without associated toxicity (13).
`Furthermore, treatment with anti-Tac induced temporary
`
`The publication costs of this article were defrayed in part by page charge
`payment. This article must therefore be hereby marked "advertisement"
`in accordance with 18 U.S.C. §1734 solely to indicate this fact.
`
`partial or complete remission in three of nine patients with
`Tac-expressing adult T-cell leukemia (14). However, as a
`murine monoclonal antibody, anti-Tac elicits a strong human
`antibody response against itself, as does OKT3 (15). This
`response would prevent its long-term use in treating autoim-
`mune conditions or suppressing organ transplant rejection.
`The immune response against a murine monoclonal anti-
`body may potentially be reduced by transforming it into a
`chimeric antibody. Such antibodies, produced by methods of
`genetic engineering, combine the variable (V) region binding
`domain of a mouse (or rat) antibody with human antibody
`constant (C) regions (16-18). Hence, a chimeric antibody
`retains the binding specificity of the original mouse antibody
`but contains less amino acid sequence foreign to the human
`immune system. Chimeric antibodies have been produced
`against a number of tumor-associated antigens (19-21). In
`some but not all cases, the chimeric antibodies have mediated
`human complement-dependent cytotoxicity (CDC) or anti-
`body-dependent cellular cytotoxicity (ADCC) more efficient-
`ly than the mouse antibodies (21).
`When the murine antibody OKT3 is used in human pa-
`tients, much of the resulting antibody response is directed
`against the V region of OKT3 rather than the C region (15).
`Hence, chimeric antibodies in which the V region is still
`nonhuman may not have sufficient therapeutic advantages
`over mouse antibodies. To further reduce the immunogenic-
`ity of murine antibodies, Winter and colleagues constructed
`"humanized" antibodies in which only the minimum neces-
`sary parts of the mouse antibody, the complementarity-
`determining regions (CDRs), were combined with human V
`region frameworks and human C regions (22-25). We report
`here the construction of chimeric and humanized anti-Tac
`antibodies. 11 For the humanized antibody, sequence homol-
`ogy and molecular modeling were used to select a combina-
`tion of mouse and human sequence elements that would
`reduce immunogenicity while retaining high binding affinity.
`
`MATERIALS AND METHODS
`Construction of Plasmids. cDNA cloning was by the
`method of Gubler and Hoffman (26), and sequencing was by
`the dideoxy method (27). The plasmid pVK1 (Fig. 1A) was
`constructed from the following fragments: an approximately
`4550-base-pair (bp) BamHI-EcoRI fragment from the plas-
`
`Abbreviations: IL-2R, interleukin 2 receptor; CDR, complementar-
`ity-determining region; CDC, complement-dependent cytotoxicity;
`ADCC, antibody-dependent cellular cytotoxicity; V, variable; J,
`joining; C, constant.
`tPresent address: Biospan, 440 Chesapeake Drive, Redwood City,
`CA 94063.
`tPresent address: Beckman Instruments, 1050 Page Mill Road, Palo
`Alto, CA 94304.
`"The sequences reported in this paper have been deposited in the
`GenBank data base (accession nos. M28250 and M28251).
`
`10029
`
`PETITIONER'S EXHIBITS
`
`Exhibit 1034 Page 1 of 5
`
`

`
`10030
`
`Immunology: Queen et al.
`
`Proc. Natl. Acad. Sci. USA 86 (1989)
`
`E H
`
`Eco RI
`
`Eco RI
`
`Gpt
`
`Hybridize
`primer
`
`Ago\
`
`V
`Extend
`
`I
`
`J
`
`C
`
`A
`
`Amp
`
`B
`
`Xba
`
`|14+I
`
`Denature.
`Hybridize
`rev. primer
`
`_
`
`Extend
`and cut
`
`|
`4
`
`D5
`
`,$
`
`Xba I
`
`V
`
`J
`
`(A) Schematic diagram of the plasmids PVK1 and pLTac.
`FIG. 1.
`Light chain exons are shown as boxes. An arrow indicates the
`direction of transcription from the K promoter. EH, heavy chain
`enhancer. Not drawn to scale. (B) Schematic diagram of the method
`used to excise the V-J region. SD, splice donor sequence; rev.
`primer, reverse primer.
`mid pSV2gpt (28) containing the amp and gpt genes; an
`1800-bp EcoRI-Bgl II fragment from pKcatH (29) containing
`the heavy chain enhancer and K promoter; and a 1500-bp
`EcoRI-Xba I fragment containing the human CK region (30).
`Similarly, pVyl was constructed starting from a 4850-bp
`BamHI-EcoRI fragment of the plasmid pSV2hph (a gift of A.
`Smith, A. Miyajima, and D. Strehlow, Stanford University),
`which is analogous to pSV2gpt except that the gpt gene is
`replaced by the hyg gene (31). This fragment was combined
`with the EcoRI-Bgl II fragment from pKcatH and a 2800-bp
`HindIII-Pvu II fragment containing the human yl constant
`region, isolated from a phage kindly provided by L. Hood
`(32). In each case, the fragments were combined by standard
`methods (ref. 33, pp. 390-401), with an Xba I linker inserted
`between the K promoter fragment and the 5' end of the C
`region fragment.
`Construction of Chimeric Genes. EcoRI fragments contain-
`ing the anti-Tac light and heavy chain cDNAs were sepa-
`rately inserted into the EcoRI site of the phage M13mpllD,
`a variant of M13mpll (34) in which the EcoRI and Xba I sites
`of the polylinker were filled in and joined. The resulting
`phage, in which the 5' ends of the cDNAs abutted the Xba I
`site, were respectively denoted M13L and M13H. The V-J (J,
`joining) segments of the cDNAs, followed by splice donor
`signals, were precisely excised from these phage, using a
`double-priming scheme (Fig. 1B). For the light chain, the
`following primer was synthesized (Applied Biosystems
`model 380B DNA synthesizer) and purified by gel electro-
`phoresis: 5 '-CCAGAATTCTAGAAAAGTGTACTTAC-
`GTTTCAGCTCCAGCTTGGTCCC-3'. From the 3' end, the
`first 22 residues of the primer are the same as the last 22
`residues of the JK5 segment (noncoding strand). The next 16
`nucleotides are the same as the sequence that follows JK5 in
`
`mouse genomic DNA and therefore includes a splice donor
`signal. The final 10 nucleotides of the oligonucleotide include
`an Xba I site.
`We hybridized this oligonucleotide to M13L and extended
`it with the Klenow fragment of DNA polymerase. The DNA
`was heat-denatured, hybridized with an excess of the "re-
`verse primer" 5'-AACAGCTATGACCATG-3', again ex-
`tended with Klenow DNA polymerase, and digested with
`Xba I. The digested DNA was run on a gel, and an approx-
`imately 400-bp fragment was excised and inserted into the
`Xba I site of pVK1. Sequencing showed that the fragment
`consisted of the V-J region of the light chain cDNA followed
`by the splice donor "tail," as expected (Fig. 1B), and pLTac,
`a clone with the appropriate orientation, was chosen. In an
`analogous fashion, the heavy chain V-J segment, followed by
`the mouse JH2 splice donor sequence, was excised from
`M13H and inserted into the Xba I site of pVyl to yield pGTac.
`Computer Analysis. Sequences were manipulated and ho-
`mology searches were performed with the MicroGenie Se-
`quence Analysis Software (Beckman). The molecular model
`of the anti-Tac V region was constructed with the ENCAD
`program (35) and examined with the MIDAS program (36) on
`an IRIS 4D-120 graphics workstation (Silicon Graphics).
`Construction of Genes for Humanized Antibody. Nucleotide
`sequences were selected that encoded the protein sequences
`of the humanized light and heavy chain V regions including
`signal peptides (Results), generally utilizing codons found in
`the mouse anti-Tac sequence. These nucleotide sequences
`also included the same splice donor signals used in the
`chimeric genes and an Xba I site at each end. For the heavy
`chain V region, four overlapping 120- to 130-nucleotide-long
`oligonucleotides were synthesized that encompassed the
`entire sequence on alternating strands. The oligonucleotides
`were phosphorylated with polynucleotide kinase, annealed,
`extended with T4 DNA polymerase, cut with Xba I, and
`ligated into the Xba I site of pUC19 (34), using standard
`reaction conditions. An insert with the correct sequence was
`recloned in pVyl. The humanized light chain V region was
`constructed similarly..
`Transfections. For each antibody constructed, the light
`chain plasmid was first transfected into Sp2/0 mouse my-
`eloma cells (ATTC CRL 1581) by electroporation (Bio-Rad
`Gene Pulser) and cells were selected for gpt expression (28).
`Clones secreting a maximal amount of light chain, as deter-
`mined by ELISA, were transfected with the heavy chain
`plasmid and cells were selected for hygromycin B resistance
`(31). Clones secreting a maximal amount of complete anti-
`body were detected by ELISA. The clones were used for
`preparation of chimeric and humanized antibodies.
`Antibody Purification. Medium from confluent cells was
`passed over a column of staphylococcal protein A-Sepharose
`CL-4B (Pharmacia), and antibody was eluted with 3 M
`MgCl2. Antibody was further purified by ion-exchange chro-
`matography on BakerBond ABx (J. T. Baker). Final anti-
`body concentration was determined, assuming that 1 mg/ml
`has an A280 of 1.4. Anti-Tac antibody itself was purified as
`described (2).
`Affiiity Measurements. Affinities were determined by com-
`petition binding. HuT-102 human T-lymphoma cells (ATTC
`TIB 162) were used as source of p55 Tac antigen. Increasing
`amounts of competitor antibody (anti-Tac, chimeric, or hu-
`manized) were added to 1.5 ng of radioiodinated (Pierce
`lodo-Beads) tracer anti-Tac antibody (2 uCi/,g; 1 Ci = 37
`GBq) and incubated with 4 x 105 HuT cells in 0.2 ml of
`binding buffer (RPMI 1040 medium with 10% fetal calf serum,
`human IgG at 100 ,g/ml, 0.1% sodium azide) for 3 hr at room
`temperature. Cells were washed and pelleted, and their
`radioactivities were measured, and the concentrations of
`bound and free tracer antibody were calculated. The affinity
`of mouse anti-Tac was determined by Scatchard plot analy-
`
`PETITIONER'S EXHIBITS
`
`Exhibit 1034 Page 2 of 5
`
`

`
`Immunology: Queen et al.
`sis, using anti-Tac itself as the competitor. Then the affinities
`of the chimeric and humanized antibodies were each calcu-
`lated according to the formula [XI - [anti-Tac] = (1/K,) -
`(1/Ka), where Ka is the affinity of anti-Tac (9 x 109 M-1), K.
`is the affinity of the competitor X, [ ] indicates the concen-
`tration of competitor antibody at which bound/free tracer
`binding is RO/2, and Ro is maximal bound/free tracer binding
`(37).
`
`RESULTS
`Cloning of Light and Heavy Chain cDNA. A cDNA library
`in Agt1O was prepared from anti-Tac hybridoma cells and
`screened with oligonucleotide probes for the mouse K and y2a
`constant regions. The cDNA inserts from four K-positive and
`four y2a-positive phage were subcloned in M13mpl9. Partial
`sequencing showed that two of the K isolates had one
`sequence, and the other two had another sequence. In one
`pair, a VK gene segment was joined to the J,,2 segment out of
`its reading frame. In addition, the conserved cysteine at
`position 23 was absent from this V segment, and the se-
`quences of the two isolates differed slightly. Presumably,
`these clones were the result of an aberrant joining event in
`one K allele, which continued to undergo somatic mutation
`after the formation of the hybridoma.
`The V-J segments of the other pair of K clones were
`sequenced completely and were identical. This light chain
`uses the J,,5 segment. Partial sequencing of the four y2a
`clones showed they were all from the same gene. The V-J
`segments of two were sequenced completely and were iden-
`tical. This heavy chain uses the JH2 segment and is of
`subgroup II (38). The DNA sequences have been deposited
`with GenBank; 11 the deduced protein sequences are shown in
`Fig. 2. As both alleles of the K light chain were accounted for
`and only one heavy chain sequence was detected, we tenta-
`tively assigned these sequences to the anti-Tac antibody
`genes.
`Construction of Chimeric Genes. Plasmid vectors were
`prepared for the construction and expression ofchimeric light
`and heavy chain genes. The plasmid pVK1 (Fig. 1A) contains
`the human genomic C, segment, including 336 bp of the
`preceding intron and the poly(A) signal. It also contains the
`promoter sequence from the MOPC 41 K gene and the heavy
`chain enhancer sequence, which synergize to form a very
`strong transcriptional unit (29). There is a unique Xba I site
`between the promoter and the intron. A similar plasmid,
`pVyl, was prepared by using the human CG1 region in place
`of the C,, region. In that case, the region inserted between the
`Xba I and BamHI sites extended from about 210 bp 5' of the
`CH1 exon to beyond the CH3 exon.
`Our strategy was to insert the V-J region from the anti-Tac
`K cDNA, followed by a splice donor signal, at the Xba I site
`B
`A
`
`Proc. Natl. Acad. Sci. USA 86 (1989)
`
`10031
`
`of pVK1 to construct the plasmid pLTac. Doing so created a
`chimeric K gene with a short synthetic intron between the
`mouse V-J and human CK segments (Fig. LA). For this
`purpose, we used a form of double primer-directed mutagen-
`esis (Materials and Methods; Fig. 1B). Similarly, the V-J
`region from the anti-Tac y2a heavy chain cDNA, followed by
`a splice donor signal, was inserted into the Xba I site of pVyl.
`The resulting plasmid, pGTac, contained a chimeric heavy
`chain gene, with a synthetic intron between the mouse V-J
`and human C,,1 segments.
`Construction of a Humanized Anti-Tac Antibody. In select-
`ing a human antibody to provide the variable region frame-
`work for the humanized anti-Tac antibody, we reasoned that
`the more homologous the human antibody was to the original
`anti-Tac antibody, the less likely would combining the anti-
`Tac CDRs with the human framework be to introduce dis-
`tortions into the CDRs. The anti-Tac heavy chain sequence
`was therefore compared by computer with all the human
`heavy chain sequences in the National Biomedical Research
`Foundation Protein Identification Resource (release 15). The
`heavy chain V region of the Eu antibody (of human heavy
`chain subgroup I; ref. 38) was 57% identical to the anti-Tac
`heavy chain V region (Fig. 2B); all other complete VH regions
`in the data bank were 30-52% identical. However, no one
`human light chain V region was especially homologous to the
`anti-Tac light chain. We therefore chose to use the Eu light
`chain (of human light chain subgroup I; ref. 38) together with
`the Eu heavy chain to supply the framework sequences for
`the humanized antibody. The CDRs in the humanized anti-
`body were of course chosen to be identical to the anti-Tac
`CDRs (Fig. 2).
`A computer program was used to construct a plausible
`molecular model of the anti-Tac V domain (Fig. 3), based on
`homology to other antibody V domains with known crystal
`structure and on energy minimization. Graphic manipulation
`shows that a number of amino acid residues outside of the
`CDRs are in fact close enough to them to either influence
`their conformation or interact directly with antigen. When
`these residues differ between the anti-Tac and Eu antibodies,
`the residue in the humanized antibody was chosen to be the
`anti-Tac residue rather than the Eu residue. This choice was
`made for residues 27, 30, 48, 67, 68, 98, and 106 in the
`humanized heavy chain, and for 47 and 59 in the humanized
`light chain (Figs. 2 and 3; amino acids shown in blue in Fig.
`3), although we now consider the light chain residue 59,
`which was chosen on the basis of an earlier model, to be
`doubtful. In this way, we hoped to better preserve the precise
`structure of the CDRs at the cost of possibly making the
`humanized antibody slightly less "human."
`Different human light or heavy chain V regions exhibit
`strong amino acid homology outside of the CDRs, within the
`framework regions. However, a given V region will usually
`
`D I Q M T Q S P S T L S A S V G D R V T
`
`1
`
`Q V Q L V Q S G A E V K K P G S S V K V
`
`Q I V L T Q S P A I M S A S P G E K V T
`
`1 Q V Q L Q Q S G A E L A K P G A S V K M
`
`1
`
`1
`
`21
`
`21
`
`41
`
`40
`
`61
`
`60
`
`81
`
`80
`
`I T C R A S Q S I N T W L A W Y Q Q K P
`
`I T C S A S S S I
`
`S Y M H W F Q Q K P
`
`G K A P K L L M Y K A S S L E S G V P S
`liiil
`II
`G T S P K L W I Y T T S N L A S G V P A
`
`R F I G S G S G T E F T L T I S S L Q P
`liiil
`II
`II
`1 1
`R F S G S G S G T S Y S L T I S R M E A
`
`D D F A T Y Y C Q Q Y N S D S K M F G Q
`
`E D A A T Y Y C H Q R S T Y P L T F G S
`
`101
`
`100
`
`G T K V E V K
`
`G T K L E L K
`
`21
`
`21
`
`41
`
`41
`
`61
`
`61
`
`81
`
`81
`
`S C K A S G G T F S R S A I I W V R Q A
`II
`II
`1111
`S C K A S G Y T F T S Y R M H W V K Q R
`
`P G
`
`G L E W M G G I V P M F G P P N Y
`
`P G Q G L E W I G Y I N P S T G Y T E
`
`A Q K F Q G R V T I T A D E S T N T A Y
`
`N 0 K F K D K A T L T A D K S S S T A Y
`
`M E L S S L R S E D T A F Y F C A G G Y
`
`M Q L S S L T F E D S A V Y Y C A R G
`
`101 G I Y S P E E Y N G G L V T V S S
`
`100 GGVFD Y W GQ TTLTVSS
`
`FIG. 2. Amino acid sequences of the humanized
`anti-Tac light (A) and heavy (B) chains. The se-
`quences of the Eu antibody light and heavy chains
`(upper lines) are shown aligned above the mouse
`anti-Tac light and heavy chain sequences (lower
`indicating identity of amino acids.
`lines), with a
`The three CDRs in each chain are underlined, and
`the other mouse amino acids used in the humanized
`antibody are double underlined. Hence, the human-
`ized sequences are the same as the upper (Eu)
`sequences, except where the amino acid is under-
`lined or double underlined.
`
`PETITIONER'S EXHIBITS
`
`Exhibit 1034 Page 3 of 5
`
`

`
`10032
`
`Immunology: Queen et al.
`
`Proc. Natl. Acad. Sci. USA 86 (1989)
`
`Model of the mouse anti-Tac antibody V region, generated with the ENCAD program and displayed with the MIDAS program. Amino
`FIG. 3.
`acids in the CDRs are shown in red; amino acids potentially interacting with the CDRs are shown in blue; other mouse amino acids used in the
`humanized antibody are shown in yellow, as described in the text. Thus, all amino acids transferred from the anti-Tac sequence to the humanized
`antibody are shown in red, blue, or yellow. Residue 1 is the first amino acid of VK; residue 301 is the first amino acid of VH.
`culture medium of cells producing the chimeric or humanized
`contain exceptional amino acids, atypical of other human V
`antibody. When analyzed by reducing SDS/polyacrylamide
`regions, at several framework positions. The Eu antibody
`gel ejectrophoresis, the antibodies showed only two bands,
`contains such unusual residues at positions corresponding to
`having the expected molecular weights 50,000 and 25,000.
`93, 95, 98, 106, 107, 108, and 110 of the humanized heavy
`Flow cytometry showed that the chimeric and humanized
`chain and 47 and 62 of the light chain (Fig. 2), as determined
`antibodies bound to Hut-102 and CRII.2 cells, two human
`by visual comparison of the Eu heavy and light chain V
`T-cell lines that express the p55 chain of the IL-2R, but not
`regions with other human V regions of subgroup 1 (38). The
`to CEM and other cell lines that do not express the IL-2R. To
`Eu antibody contains several other unusual residues, but at
`determine the binding affinity of the chimeric and humanized
`the listed positions, the murine anti-Tac antibody actually has
`antibodies, their ability to compete with labeled mouse
`a residue much more typical of human sequences than does
`anti-Tac for binding to Hut-102 cells was determined. The
`Eu. At these positions, we therefore chose to use the anti-Tac
`affinity of chimeric anti-Tac was indistinguishable from that
`residue rather than the Eu residue in the humanized antibody,
`of anti-Tac (data not shown), as expected from the fact that
`to make the antibody more generically human. Some of these
`their entire V regions are identical. The affinity of humanized
`residues had already been selected because oftheir proximity
`anti-Tac for membrane-bound p55 was 3 x 109 M-1, about
`to the CDRs, as described above (the remaining ones are
`1/3 the measured affinity of 9 x 109 M-1 of anti-Tac itself
`shown in yellow in Fig. 3).
`(Fig. 4).
`These criteria allowed the selection of all amino acids in the
`humanized antibody V regions as coming from either anti-Tac
`or Eu (Fig. 2). DNA segments encoding the desired heavy
`and light chain amino acid sequences were synthesized.
`These DNA segments also encoded typical immunoglobulin
`signal sequences for processing and secretion, and they
`contained splice donor signals at their 3' end. The light and
`heavy chain segments were cloned, respectively, in pVKl and
`pVyl to form the plasmids pHuLTac and pHuGTac.
`Properties of Chimeric and Humanized Antibodies. Sp2/0
`cells, a nonproducing mouse myeloma line, were transfected
`sequentially with pLTac and pGTac (chimeric genes) or with
`pHuLTac and pHuGTac (humanized genes). Cell clones
`were selected first for antibiotic resistance and then for
`maximal antibody secretion, which reached 3 ,lg/106 cells per
`24 hr. S1 nuclease mapping of RNA extracted from the cells
`transfected with pLTac and pGTac showed that the synthetic
`introns between the V and C regions (Fig. LA) were correctly
`spliced (data not shown). Antibody was purified from the
`
`DISCUSSION
`Because monoclonal antibodies can be produced that are
`highly specific for a wide variety of cellular targets, antibody
`therapy holds great promise for the treatment of cancer,
`autoimmune conditions, and other diseases. However, this
`promise has not been widely realized, largely because most
`monoclonal antibodies, which are of mouse origin, are im-
`munogenic when used in human patients and are ineffective
`at recruiting human immune effector functions such as CDC
`and ADCC. A partial solution to this problem is the use of
`chimeric antibodies (16), which combine the V region binding
`domains of mouse antibodies with human antibody C regions.
`Initially, chimeric antibodies were constructed by combining
`genomic clones of the V and C region genes. However, this
`method is very time consuming because of the difficulty of
`genomic cloning, especially from tetraploid hybridomas.
`
`PETITIONER'S EXHIBITS
`
`Exhibit 1034 Page 4 of 5
`
`

`
`Immunology: Queen et al.
`
`Proc. Natl. Acad. Sci. USA 86 (1989)
`
`10033
`
`W 2.0L RO P
`
`1 1.5
`
`D 1.0
`0~~~~~~
`
`z
`
`Ro ...
`
`0.5
`0.0
`4.0
`3.5
`3.0
`2.5
`2.0
`1.5
`1.0
`LOG CONCENTRATION OF COMPETITOR, pM
`
`Competitive binding of labeled anti-Tac tracer to Hut-102
`FIG. 4.
`cells. Duplicate samples are shown. e, Mouse anti-Tac competitor;
`v, humanized anti-Tac competitor.
`
`More recently, cDNA clones of the V and C regions have
`been combined, but this method is also tedious because of the
`need to join the V and C regions precisely (20, 21). Here we
`show that the V region from a readily obtainable cDNA clone
`can be easily joined to a human genomic C region, which need
`only be cloned once, by leaving a synthetic intron between
`the V and C regions. When linked to suitable transcriptional
`regulatory elements and transfected into an appropriate host
`cell, such chimeric genes produce antibody at a high level.
`Chimeric antibodies represent an improvement over
`mouse antibodies for use in human patients, because they are
`presumably less immunogenic and sometimes mediate CDC
`or ADCC more effectively (21). For example, chimeric
`anti-Tac mediates ADCC with activated human effector cells,
`whereas murine anti-Tac does not (unpublished data). How-
`ever, the mouse V region can itself be highly immunogenic
`(15). Winter and colleagues therefore took the further, inno-
`vative, step of combining the CDRs from a mouse (or rat)
`antibody with the framework region from a human antibody
`(22-25), thus reducing the xenogeneic elements in the hu-
`manized antibody to a minimum. Unfortunately, in some
`cases the humanized antibody had significantly less binding
`affinity for antigen than did the original mouse antibody. This
`is not surprising, because transferring the mouse CDRs from
`the mouse framework to the human framework could easily
`deform them.
`In humanizing the anti-Tac antibody, which binds to the
`p55 chain of the human IL-2R, we have introduced two ideas
`that may have wider applicability. First, the human frame-
`work was chosen to be as homologous as possible to the
`original mouse antibody to reduce any deformation of the
`mouse CDRs. Second, computer modeling was used to
`identify several framework amino acids in the mouse anti-
`body that might interact with the CDRs or directly with
`antigen, and these amino acids were transferred to the human
`framework along with the CDRs. The resulting humanized
`antibody has a high affinity, 3 x 109 M-1, for its antigen.
`Further work is needed to determine to what extent the
`choice of human framework and the preservation of partic-
`ular mouse amino acids in fact contributed to the affinity of
`the humanized antibody. The extent to which humanization
`eliminates immunogenicity will need to be addressed in
`clinical trials, where humanized anti-Tac will be administered
`to patients with Tac-expressing lymphomas or selected au-
`toimmune diseases or to patients receiving organ transplants.
`
`1.
`2.
`
`Waldmann, T. A. (1989) Annu. Rev. Biochem. 58, 875-911.
`Uchiyama, T., Broder, S. & Waldmann, T. A. (1981) J. Immunol.
`126, 1393-1397.
`
`3.
`
`4.
`
`5.
`
`6.
`
`7.
`
`8.
`
`9.
`
`10.
`
`11.
`
`12.
`
`13.
`
`Leonard, W. J., Depper, J. M., Uchiyama, T., Smith, K. A., Wald-
`mann, T. A. & Greene, W. C. (1982) Nature (London) 300, 267-269.
`Tsudo, M., Kozak, R. W., Goldman, C. K. & Waldmann, T. A.
`(1986) Proc. Natl. Acad. Sci. USA 83, 9694-9698.
`Sharon, M., Klausner, R. D., Cullen, B. R., Chizzonite, R. &
`Leonard, W. J. (1986) Science 234, 859-863.
`Depper, J. M., Leonard, W. J., Robb, R. J., Waldmann, T. A. &
`Greene, W. C. (1983) J. Immunol. 131, 690-696.
`Kirkman, R. L., Barrett, L. V., Gaulton, G. N., Kelley, V. E.,
`Ythier, A. & Strom, T. B. (1985) J. Exp. Med. 162, 358-362.
`Kupiec-Weglinski, J. W., Diamantstein, T., Tilney, N. L. & Strom,
`T. B. (1986) Proc. Nat!. Acad. Sci. USA 83, 2624-2627.
`Volk, H.-D., Brocke, S., Osawa, H. & Diamantstein, T. (1986) Clin.
`Exp. Immunol. 66, 126-131.
`Kelley, V. E., Gaulton, G. N., Hattori, M., Ikegami, H., Eisen-
`barth, G. & Strom, T. B. (1988) J. Immunol. 140, 59-61.
`Reed, M. H., Shapiro, M. E., Strom, T. B., Milford, E. L., Car-
`penter, C. B., Weinberg, D. S., Reimann, K. A., Letvin, N. L.,
`Waldmann, T. A. & Kirkman, R. L. (1989) Transplantation 47,
`55-59.
`Ortho Multicenter Transplant Study Group (1985) N. Engl. J. Med.
`313, 337-342.
`Kirkman, R. L., Shapiro, M. E., Carpenter, C. B., Milford, E. L.,
`Ramos, E. L., Tilney, N. L., Waldmann, T. A., Zimmerman, C. E.
`& Strom, T. B. (1989) Transplant. Proc. 21, 1766-1768.
`14. Waldmann, T. A., Goldman, C. K., Bongiovanni, K. F., Sharrow,
`S. O., Davey, M. P., Cease, K. B., Greenberg, S. J. & Longo,
`D. L. (1988) Blood 72, 1805-1816.
`Jaffers, G. J., Fuller, T. C., Cosimi, A. B., Russell, P. S., Winn,
`H. J. & Colvin, R. B. (1986) Transplantation 41, 572-578.
`Morrison, S. L., Johnson, M. J., Herzenberg, L. A. & Oi, V. T.
`(1984) Proc. Natl. Acad. Sci. USA 81, 6851-6855.
`Boulianne, G. L., Hozumi, N. & Shulman, M. J. (1984) Nature
`(London) 312, 643-646.
`Neuberger, M. S., Williams, G. T., Mitchell, E. B., Jouhal, S. S.,
`Flanagan, J. G. & Rabbitts, T. H. (1985) Nature (London) 314,
`268-270.
`Sun, L. K., Curtis, P., Rakowicz-Szulczynska, E., Ghrayeb, J.,
`Chang, N., Morrison, S. L. & Koprowski, H. (1987) Proc. Natl.
`Acad. Sci. USA 84, 214-218.
`Whittle, N., Adair, J., Lloyd, C., Jenkins, L., Devine, J., Schlom,
`J., Raubitschek, A., Colcher, D. & Bodmer, M. (1987) Protein Eng.
`1, 499-505.
`Liu, A. Y., Robinson, R. R., Hellstrom, K. E., Murray, E. D., Jr.,
`- C-hang, C. P. & Hellstrom, I. (1987) Proc. Natl. Acad. Sci. USA 84,
`3439-3443.
`Jones, P. T., Dear, P. H., Foote, J., Neuberger, M. S. & Winter, G.
`(1986) Nature (London) 321, 522-525.
`Verhoeyen, M., Milstein, C. & Winter, G. (1988) Science 239,
`1534-1536.
`Reichmann, L., Clark, M., Waldmann, H. & Winter, G. (1988)
`Nature (London) 332, 323-327.
`Hale, G., Dyer, M. J. S., Clark, M. R., Phillips, J. M., Marcus, R.,
`Riechmann, L., Winter, G. & Waldmann, H. (1988) Lancet i,
`1394-1399.
`Gubler, U. & Hoffman, B. J. (1983) Gene 25, 263-269.
`Sanger, F., Nicklen, S. & Coulson, A. R. (1977) Proc. Nat!. Acad.
`Sci. USA 74, 5463-5467.
`Mulligan, R. C. & Berg, P. (1981) Proc. Natl. Acad. Sci. USA 78,
`2072-2076.
`Garcia, J. V., Bich-Thuy, L. T., Stafford, J. & Queen, C. (1986)
`Nature (London) 322, 383-385.
`Hieter, P. A., Max, E. E., Seidman, J. G., Maizel, J. V., Jr., &
`Leder, P. (1980) Cell 22, 197-207.
`Sugden, B., Marsh, K. & Yates, J. (1985) Mol. Cell. Biol. 5,
`410-413.
`Ellison, J. W., Berson, B. J. & Hood, L. E. (1982) Nucleic Acids
`Res. 10, 4071-4079.
`Maniatis, T., Fritsch, E. F. & Sambrook, J. (1982) Molecular
`Cloning: A Laboratory Manual (Cold Spring Harbor Lab., Cold
`Spring Harbor, NY).
`Yanisch-Perron, C., Vieira, J. & Messing, J. (1985) Gene 33,
`103-119.
`Levitt, M. (1983) J. Mol. Biol. 168, 595-617.
`Ferrin, T. E., Huang, C. C., Jarvis, L. E. & Langridge, R. (1988) J.
`Mol. Graphics 6, 13-27.
`Berzofsky, J. A. & Berkower, I. J. (1984) in Fundamental Immu-
`nology, ed. Paul W. E. (Raven, New York), pp. 595-644.
`Kabat, E. A., Wu, T. T., Reid-Miller, M., Perry, H. M. & Gottes-
`man, K. S. (1987) Sequences ofProteins ofImmunological Interest
`(National Institutes of Health, Bethesda, MD).
`
`15.
`
`16.
`
`17.
`
`18.
`
`19.
`
`20.
`
`21.
`
`22.
`
`23.
`
`24.
`
`25.
`
`26.
`27.
`
`28.
`
`29.
`
`30.
`
`31.
`
`32.
`
`33.
`
`34.
`
`35.
`36.
`
`37.
`
`38.
`
`PETITIONER'S EXHIBITS
`
`Exhibit 1034 Page 5 of 5