r
`
`NOVA ACTA LEOPOLDINA
`I Abhandlungen der Deut8chen Akademie der Naturforscher Leopoldina
`un Anftrage des Praidiuma hernuagegeben von
`JOACHIMHERMANN SCHARF
`Director Ephemeriduin der Akndemie
`
`NEUE FOLGE
`
`NUMMER 269
`
`BAND 61
`
`Leopoldina-Symposium
`Functional and Regulatory
`Aspects of Enzyme Action
`May 25 to 28, 1988
`in Halle (Saale)
`
`Organized and edited by:
`Ernst J. M. HELMREICH (Würzburg)
`Mitglied der Akadernie
`
`Helmut HOLZER (Freiburg)
`Mitglied der Akadenue
`Alfred SCHELLENBERGER (Halle)
`Mitglied des Senates der Akademie
`
`With 148 Figures and 37 Tables
`
`Deutsche Akademie der Naturforscher Leopoldina. Halle (Siialeb 1999
`EXHIBIT 1193 Jefferson Foote, Ph.D.
`2/4/18 Planet Depos-T. Rosate, RDR, CRR, CSR 10891
`
`
`
`1 of 13
`
`Celltrion, Inc. 1193
`Celltrion v. Genentech
`IPR2017-01374
`
`

`

`
`
`Redaktion: Prof .Dr .nd.Dr.rer.nat .Dr h. c. Joachim-Hernnn SiARF ale
`Director 4phemeridtim,
`Prof. Dr. ac .nat. Alfred SCHELLENBERM ur
`Dip1.-Phya. Rainer-M. JNXBI,
`wissenach. Assistent der Akademie
`
`Die Zeitschrift erschaint im Kcimssionsverlag Johann Antroaius Barth,
`-70 10 Leipzig, Postschlietach 109, Rut 70131.
`Jedes Heft 1st einzeln käuflich!
`
`Lizenztrr: Deutsche Akadeinie der Naturforsctr Leopoldina, Halle (Saale)
`Cbefredalcteur urid Herausber: Prof .Dr.Dr.Dr.h.c. J.-H. SC{ARF,
`MR-4010 Halle (Saale), Postschliefach 302
`
`Veröffenti.icht unter der Lizen7z1unhier 1393 des Presseamtes bairn Vorsitzen-
`den des !'tinisterrates der 1
`Printed in the German Democratic Republic
`Gesanitherstellung: Drucithaus "FreTheit", Halle (Saale), ET Merseburg
`
`
`
`2 of 13
`
`Celltrion, Inc. 1193
`Celltrion v. Genentech
`IPR2017-01374
`
`

`

`
`
`CONTENTS
`
`IM
`
`Qapter 1: Enzyme Catalysis - Basic Principles
`
`SIN, R.L.: TheiBition-state Structure and Its Role in Enzyme
`Catalysis ani Enzyme Regulation
`
`HU&R, R.: Flexibility and Rigidity in Proteins and Protein Pig-
`ment Qxp].exes
`
`FISaIER, G.: Slow Co]1foxtjona1 thar€es and Their Enzyn].ogy
`
`Chapter 2: Enzyme Catalysis - Theoretical Aspects
`
`BUJNIELL, T.L..: TheThree-dimensional Suctizes of Aspartic
`Proteirses and Their Inhibitors - Lessons for 1g Design
`
`KEIETI, T.: Kinetic Power, a Key Parameter of Metabolic Cortho].
`in HczlcgerEous and Heterogeneous Systers
`
`HESS, B ., MARK.E, M., !1)UER, S.C., and PIESSER, T.: Nonlirear.
`Dynamics in Chemistry and Biology
`
`EDY1E, J.: Humanized Antibodies
`
`Chapter 3: Enzyme Catalysis - Special Mechanisms
`
`1HAJER, R.K.: Structure and Furtion of Methyl-0:*1 Reductase,
`a Nickel-Porphimid Containing Enzyme in Methamgenic
`Archaebacteria
`
`CHRISTEN, P., (1RING, H., KOa*IAR, S., 1)BLER, H.P., rIEBTA, P.,
`and HALE, T.: Enzymic Vitamin B6-Catalysis: Mechanistic and
`Evolutionary Aspects
`
`9
`
`13
`
`31
`
`35
`
`57
`
`59
`
`79
`
`103
`
`11.3
`
`123
`
`
`
`3 of 13
`
`Celltrion, Inc. 1193
`Celltrion v. Genentech
`IPR2017-01374
`
`

`

`
`
`
`
`PALM, D., and SGWIJ, R.: The )thanism of a-G]noan PIxsrIorylasee
`Studied by Substrate Analogs and Site-directed ?4itagenesis
`
`1113
`
`Chapter
`
`: Enzyme Catalysis and Regulation - Role of Domain
`Structure and Subunit Interactions
`
`NADOVA, N.K..: Pumtional Aspects of Protein-Protein Interactions
`in 0].igneric Enzymes. NAD-deident Dehydrogenases
`1'ZLEIM(AIJF, H., and v. D1REN, H.: Interacting ?tiltienzymes
`
`LANE, AN., SASSER, A., and ICRSC}JNER, K.: Catalysis of Coupled
`Reactions. Camunicafl.on Between the Two Active Sites of
`Tryptophare Synthase
`
`Chapter 5: Enzyme Regulation - PuixtLona]. and Kinetic Aspects
`
`SC1QMAN, H.K.: Effects of Amino, Acid Substitutions on the
`Catalytic and Regulatory Properties of E ..Cli Aspartate Trana-
`oaxtaisjlase (Aase)
`
`HERVE, G.: Mechanism of Allosteric HetexopiC Interactions in
`Eschichia Coil Aspartate Trenscarbaurlase
`
`SCL1ENR(R, A., HUBNER, G., KtNIG, S., FIATAIJ, S., and
`NEEF, H.: Substrate Activation of Pyruvate Decartoxylase -
`chanistic Aspects
`
`HOBNER, G., and WOLNA, P.: Oscillations in Regu].atorJ Enzyme
`Systems at Constant Substrate Input
`
`Chapter 6: Enzyme Regulation - Special Systems
`
`HOMER, H.: Regulatory Protein Phosphorylation in Yeast
`
`W.:
`HOIWJNN, E., KRZ1'S(€R, H., FRENZEL, J., and S
`Regulation of the Fructose 6-Phosphate/Fructose 2,6-Bisphos-
`phate Cycle
`
`HEK46N, H., PIEUFEER, T., and HELZ4(EIai, E .J .N.: Protein-Protein
`Interactions in Horticna]. Signal Transmission
`
`169
`
`189
`
`209
`
`213
`
`217
`
`225
`
`2113
`
`253
`
`269
`
`289
`
`
`
`4 of 13
`
`Celltrion, Inc. 1193
`Celltrion v. Genentech
`IPR2017-01374
`
`

`

`
`
`
`
`
`
`U., BOILER, W., !EINEQ(E, N., NIERDING, N., WALJNN, R.,
`and LL1K4, SM.; Second
`sserger-regulated Protein Kineses
`and Their Role in Hepatocyte and Platelet Furxtion
`
`SEVERIN, E.S., and ALAKHOV, V.Yu.: Role of
`Regulation of InteceUular Sys
`
`cAMP in The
`
`Chapter 7; Protein Tw'xxver and Intracellular Proteolysis
`
`LEVINE, R.L.., RIVETr, A.J., and (RVEM, J.: Ooddative )tdifica-
`tion of Proteins: Potential Physiologic and Pathologic Roles
`
`SWITR, R.L.: Regulation of Selective Intracellill Proteolysis
`in BaøilZua eubtiUe
`
`KAIUMJM, N., and 1ffNAMI, E.: Mechaniane and Regulation of
`Protein Degradation in Iysosane
`
`Chapter 8: Enzymes - InUel]ii].ar Traffic
`
`BLOL, G.: Inteoellular Protein flpogenesis
`
`RAPOPORT, T.A., WE314e 9 N., KZGIAt1A, T.V., and HARIMANN,. E.:
`Signal Recognition in Protein Iens1ocation Across the Enb-
`plasnic Reticulum Merbrere
`
`VESTR, D., and SCIWFZ, C.: Blockage and Quantification of
`Icondrial Protein Import Sites by an Internally Cross-
`]irked Hybrid Precursor Protein
`
`311
`
`329
`
`355
`
`367
`
`369
`
`385
`
`399
`
`409
`
`
`
`5 of 13
`
`Celltrion, Inc. 1193
`Celltrion v. Genentech
`IPR2017-01374
`
`

`

`Nova acta Leopoldina NF 61 Nr. 269, 103-110 (1989)
`
`-
`
`Humanized Antibodies
`
`by Jefferson FOOTE(Cambddge)
`
`With 2 Figures and 2 Tables
`
`Summary "Humanized" antibodies are created by trnspiantIng, via recombinant
`DNA techniques, the antigen binding site of a rodent monoclonal antibody onto a
`human immunoglobulln. This process has been applied three times In our laboratory,
`starting with antibodies to the hapten 3-nitro, 4-hydroxy-phenacetyl, to the protein
`antigen tysozyme, and with a therapeutic antibody, CaJTpath-1, recognizing a human
`lymphocyte marker. Findings obtained with each of the three examples are discussed.
`
`Physicians have long faced the dilemma that whereas antibodies of specificity
`against any pathogen can be raised In animals, the use of such antibodies in human
`patients is accompanied by potentially lethal allergic reactions; conversely, human
`antisera are more safely administered, but very few specificities are obtainable from
`human volunteers. Monoclonal human antibodies made by in vitro operations would
`seem to present a way out of this dilemma, but currently, the human-based analog of
`the murine hybridoma technology Is beset with methodological problems (CARSON
`and FREIMARK 1986).
`As an alternative to a frontal assault on the activation, selection, and propagation
`of human lymphocytes, several groups at the Laboratory of Molecular Biology have
`attempted to convert mouse monoclonal antibodies to human form by the methods of
`protein engineering. Michael Neuberger developed a system for stably re-introducing,
`and expressing, cloned Immunoglobulin genes In cultured mouse plasmacytoma cells
`(NEUBERGER 1983). The gene structure of immunoglobulins Is such that the variable
`domains, which determine antigen specificity, and constant domains are encoded on
`separate exons. This makes It very easy to create a "chimeric" molecule with human
`constant doThains replacing their mouse counterparts. This was achieved without
`
`
`
`6 of 13
`
`Celltrion, Inc. 1193
`Celltrion v. Genentech
`IPR2017-01374
`
`

`

`apparent effect on antigenic specificity (NEUBERGER et al. 1985). However, even in a
`chimeric antibody, retaining the variable domains dictates that fully a third of the
`molecule will remain 'mouse.
`Homology studies (WU and KABAT 1970, KABAT et al. 1987) demonstrated that
`the variable domain of an immunoglobulin was composed of four "framework" regions
`of highly conserved sequence, sandwiching three regions of extreme sequence
`variablihity. Biochemical studies and X-ray crystallography confirmed that residues in
`these positions were almost exclusively the ones which interacted directly with antigen,
`hence the name, complementarity determining regions (CDR's). Several years ago,
`Greg Winter proposed that the CDR's of a mouse monoclonal antibody could be
`combined, via gene synthesis, with human framework sequences, to yield a
`"humanizedu variable domain. Such constructs in turn could be joined with human
`constant regions, as in a chimeric, to constitute a molecule indistinguishable from a
`human antibody.
`Two questions may be asked of a humanized antibody. Does it continue to bind
`antigen? Does it function in vivo, finding its target while escaping the surveillance of
`the patient's immune system? The former question, a problem in structural chemistry,
`has been easier to answer.
`
`A genomic clone was obtained of the heavy chain of a mouse hybridoma line
`specific for this hapten (NEUBERGER 1983). (The hybridoma was of the lambda type.
`A peculiarity of this system is that there is very little heterogeneity in mouse lambda
`sequences, hence antigen specificity is largely a function of the heavy chain
`sequence.) A gene was synthesized from oligonucleotides, with a corresponding
`protein sequence Identical to the mouse sequence in the region of the CDR's, and a
`framework sequence identical to that of the human myeloma protein NEWM (SAUL et
`at. 1978). The mouse and humanized sequences are compared in Fig. 1. Shifting to
`the NEWM frameworks has the effect of putting 37 point mutations in the mouse gene.
`This construction, and the unmodified mouse variable domain, were each joined with
`human IgE constant regions, and transfected into a cell line which ordinarily secretes
`only a mouse lambda light chain (JONES et a). 1986). Clones were selected which
`
`104
`
`
`
`7 of 13
`
`Celltrion, Inc. 1193
`Celltrion v. Genentech
`IPR2017-01374
`
`

`

`secreted complete antibodies Into the culture medium, consisting of the mouse lambda
`chain in combination with the humanized heavy chain or the parallel mouse control.
`
`Mouse
`Humanized
`
`
`
`frsmswork 1
`
`,0
`I
`QUQLQQPSAELUKPeRSUKLSCKASGYTFT
`SYUflH
`QUQLQESBPGLURPSQTLSLTCTUSBSTFS
`• *sees S••
`S
`S...
`S
`
`Ta
`miwork 2
`
`49
`36
`UUKQRPGR6LEIiIG
`UURQPPGRGLEU 16
`
`RIDPHSB6TKYPIEKFKS
`
`framework 3
`
`95
`67
`KRTLTUDEPSSTRYISQLSSLTSED$RUVYCAR
`RUTnLUDTSHQFSLHL$$UTRROTRUYYCAR
`S •• S
`.. SS
`•S•$S•S••
`
`YDYYGSSYFDY
`
`frarniwork 4
`lID
`120
`UGQGTTLTUSS
`156 QGSLLJTUSS
`...
`
`Fig. 1 CDR gifting in the heavy chain of anti-NP
`
`The antigen binding properties of the two antibodies were tested by the
`fluorescence quench method (EISEN 1964), using the 5-amino-caproic amide of NP.
`As seen in Table I, the humanized molecule bound the hapten with an affinity less than
`a factor of 2 weaker than the mouse control. A second peculiarity of this system is that
`the anti-NP hybridoma binds the 5-iodo derivative of NP, NIP, more tightly than NP
`itself, even though the latter had been the original immunogen. Accordingly, a second
`measurement was made, of the affinity for NIP. Again the humanized molecule showed
`an affinity just slightly weaker than the mouse construct (Table 1). Thus not only was
`antigen binding retained through humanization, but also specificity In distinguishing
`between an Iodine and a Hydrogen atom at the same sing position. In structural terms,
`the hapten binding site on the humanized construct must be virtually identical to that of
`the mouse antibody, despite the 37 point mutations in the frameworks.
`
`105
`
`
`
`8 of 13
`
`Celltrion, Inc. 1193
`Celltrion v. Genentech
`IPR2017-01374
`
`

`

`Table 1 Hapten Dissociation Constants from Mouse and Humanized anti-NP
`
`Mouse
`Humanized
`
`NP
`
`1200 n
`1900
`
`NIP
`
`20 n
`70
`
`Although grafting the CDR's onto the human frameworks was dearly a success in
`the anti-NP case, It must be kept In mind that only the heavy chain was humanized.
`Furthermore, the haptens were small molecules, and bound to a small pocket on the
`heavy chain, whereas true antigens would be macromolecular, and would bind to a
`much larger surface on the variable domain.
`A more realistic target for humanization was the mouse anti-lysOzYme whose
`crystal structure has been solved In Roberto Poljak's laboratory (AMif at al. 1986).
`cONA clones were obtained for both chains. A humanized anti-lysozyme heavy chain
`was made directly from the anti-NP construct, using three synthetic ollgonudeotides
`which spanned the CDR's A human lgG2 constant region was added to complete the
`heavy chain. The heavy chain was expressed as before, and then assembled In vitro
`with the mouse antWysozyme light chain (VERHOEVEN at al. 1988). This
`half-humanized molecule also bound lysozyme, again with somewhat reduced affinity
`relative to the original hybfldoma, thus extending the findings In the humanized
`anti-NP experiment to a macromolecular antigen.
`A completely humanized antl-lyeozyme was made. The CDR sequences from the
`kappa light chain were combined with consensus human kappa frameworks (The
`frameworks were very similar to those of the human kappa light chain dlmer REI (EPP
`at al. 1974), of known three-dimensional structure.) A human kappa constant region
`was added, and the light and heavy chain constructs were co-transfacted Into a
`previously non-secreting mouse myeloma cell tine. Complete lmmunoglobulln was
`subsequently purified In suffident quantity for physical studies comparing the mouse
`and humanized antibodies In their Interaction with tysozyme.
`The fluorescence emission spectrum of the complex of anti-Iyeozyme and
`
`4
`
`106
`
`
`
`9 of 13
`
`Celltrion, Inc. 1193
`Celltrion v. Genentech
`IPR2017-01374
`
`

`

`
`
`lysozyme is considerably quenched relative to the sum of the independently obtained
`emission spectra of the two molecules. In the case of the mouse antibody, the
`quenching is so extreme that the addition of lysozyme, though Itself fluorescent,
`actually causes the overall fluorescence of an antibody sample to decline. This is
`shown In Figure 2, in which the mouse and-lysozyme and a control antibody directed
`against an entirely different antigen are titrated with lysozyme.
`
`E
`C
`
`600
`
`550
`
`500
`
`1!
`
`450
`
`400
`
`gMjl
`0
`
`anti-lysozyne
`
`1
`
`2
`lysozyme/antibody
`
`3
`
`4
`
`Fig. 2 Stoichlometry of tysozymeantl-lysozyme spectral change.
`The curvature in Figure 2 In the vicinity of the titration breakpoint Is a function of
`the equilibrium constant for the formation of the antibody-antigen complex. The value
`of this equilibrium constant can be calculated by performing titratlons at several
`antibody concentrations and fitting the resulting data to a simple binding equation. This
`method indicates a dissociation constant of 2 nM for the mouse, and 70 nM for the
`humanized antibody.
`Rapid kinetics measurements, using a stopped-flow apparatus, and exploiting the
`same spectral change, show very similar association rates for the mouse and
`humanized antibodies. The difference In the equilibrium constants is due to a half-fife
`on the order of 10 minutes for the lysozyme complex with the mouse antibody, and on
`
`
`
`10 of 13
`
`Celltrion, Inc. 1193
`Celltrion v. Genentech
`IPR2017-01374
`
`

`

`the order of 10 seconds In the case of the humanized molecule.
`The anti-tysozyme system has been developed for answering questions of
`structural chemistry concerning the interaction of humanized antibodies with antigen.
`Given the known three-dimensional structure and the physical techniques which have
`been applied, the effects of mutational alteration on the thermodynamic and kinetic
`parameters of this system can be Interpreted in a meaningful way.
`
`4. Humanized Campath-1
`In our approach to second question posed of a humanized antibody, whether it
`functions as Intended in the human body, we have attempted to develop a
`therapeutically useful molecule.
`Several years ago, Herman Waldmann and his colleagues developed a rat
`monoclonal antibody, Campath.10 (HALE at al. 1983). Uttle is known about the
`function of the human antigen it recognizes, but what Is significant Is that this antigen
`is found on the surfaces of only a few differentiated cells of the immune system, B and
`T lymphocytes, and monocytes, not on the undifferentiated stem cells, or on any other
`cell type. The rat antibody has been used clinically for treating graft-versus-host
`disease following bone marrow transplants, suppressing tissue rejection after organ
`transplants, and for the direct elimination of leukemic cells. Its drawbacks are the
`occurence of severe side effects, including anaphylactic shock, and that its
`effectiveness is compromised after approximately ten days by the induction of an
`anti-rat response.
`Asks limitations seemed attributable entirely to Its rat origin, Campath-1 seemed
`a good target for humanization. CDNA clones were made of the heavy and light chains.
`A humanized light chain gene was converted from the anti-lysozyme construct. A
`humanized heavy chain gene was converted from the anti-NP. In the latter
`construction, the change Ser-27 -> Phe was introduced. Modeling had suggested that
`this would Improve affinity, many human sequences occur with Phe in this position, so
`this change would not be expected to create an immunological inconsistency. Human
`kappa and IgGi constant regions were used to complete the construction
`(RIECHMANN at al. 1988). (A number of other rat and human heavy chain isotypes
`were employed in constructions used as controls.)
`
`108
`
`
`
`11 of 13
`
`Celltrion, Inc. 1193
`Celltrion v. Genentech
`IPR2017-01374
`
`

`

`The humanized Campath-1 obtained after transfection was subjected to a
`number of tests to determine its suitability for use in viva, and the results of these are
`compiled in Table 2. In an "enzyme linked immunosorbent assays (EUSA), comparing
`the relative affinity for the Campath-1 antigen, a humanized construction (with a rat
`light chain and heavy chain constant region, necessary for a valid comparison) was
`only slightly less active than the rat antibody. in the complement lysis assay, human
`lymphocytes were isolated, and the separate antibodies bound. Human serum was
`added, and complement reactions initiated by the presence of the bound antibodies
`caused the lysis of the cells. In this test, the humanized and rat molecules were almost
`identically effective. In antibody-dependent cell mediated cytotoxichy (ADCC) tests,
`human lymphocytes are again coated with antibody, and their lysis Is achieved by the
`addition of human effector K cells. In this case the humanized antibody proved more
`effective than the rat.
`
`Table 2 Properties of uCampathu Antibodies
`
`ELISA
`Complement Lysis .
`ADCC
`
`Rat
`
`0.7
`2
`0.008
`
`Humanized
`
`2.1
`3
`0.001
`
`The complement lysis and ADCC reactions are realistic mimics of the type of
`reactions expected in viva to eliminate the Campath-1 target cells. The evidence from
`these in vitro tests indicates that the humanized antibody should work as well as the rat
`molecule as a therapeutic agent. Clinical trials starting now should show whether the
`immunogenicity of the humanized Campath-1 has been reduced to the basal level
`expected of a truly human antibody.
`
`4cknowledgement: J. F. is a Fellow of the Jane Coffin Childs Memorial Fund
`for Medical Research.
`
`109
`
`
`
`12 of 13
`
`Celltrion, Inc. 1193
`Celltrion v. Genentech
`IPR2017-01374
`
`

`

`References
`
`AMIT, A. C., MARIUZZA, A. A., PHILLIPS, S. E. V., and POLJAK, R. J. (1986) Science
`233,747-753
`CARSON, D. A., and FREIMARK, B. D. (1986) Adv. Immunol. 38,275-311
`EISEN, H. N. (1964) Methods Med. Res 10, 115-121
`EPP, 0., COLMAN, P. ,FEHLHAMMER, H., BODE, W., SCHIFFER, M., HUBER, R.., and
`PALM, W. (1974) Eur. J. Biochem. 45,513-524
`HALE, G., CLARK, M., and WALDMANN, H. (1985) J. ImmunoL 134,3056-3061
`HALE, G., BRIGHT, S., CHUMBLEY, G., HOANG, T., METCALF, D., MUNRO, A. J., and
`WALDMANN, H. (1983) Blood 62,873-882
`JONES, P. T., DEAR, P. H., FOOTE, J., NEUBERGER, M. S., and WINTER, G. (1986)
`Nature 321,522-525
`KABAT, E. A., WU, T. T., REID-MILLER, M., PERRY, H. M., and GOUESMAN, K S.
`(1987) Sequences of proteins of Immunological interest, 2nd ed. (Bethesda:
`Dept. of Health and Human Services)
`NEUBERGER, M. S. (1983) EMBOJ. 2,1373-1378
`NEUBERGER, M. S., WILLIAMS, G. T., MITCHELL, E. B., JOUHAL, S. S., FLANAGAN,
`J. C., and RABBITS, T. H. (1985) Nature 314,268-270
`RIECHMANN, L, CLARK, M., WALDMANN, H., and WINTER, G. (1988) Nature 332,
`323-327
`SAUL, F. A., AMZEL, M., and POLJAK, R. J. (1978) J. Biol. Chem. 253,585-597
`VERHOEVEN, M., MILSTEIN, C., and WINTER, G. (1988) Science 239,1534-1536
`WU, T. T., and KABAT, E. A. (1970) J. Exp. Med. 132,211-250
`
`Address correspondence to:
`
`Jefferson Foote
`MAC Laboratory of Molecular Biology
`Hills Road
`Cambridge CB2 20H
`England
`
`110
`
`
`
`13 of 13
`
`Celltrion, Inc. 1193
`Celltrion v. Genentech
`IPR2017-01374
`
`