`
`Developments
`in
`Biological Standardization
`Vol. 74
`
`Edited by
`the
`International Association of Biological Standardization
`
`Acting Editors
`Joan C. May (Bethesda, MD) — F. Brown (Guildford)
`
`@ MARGE
`
`Basel ¢ Miinchen ® Paris © London ¢ New York ® New Delhi © Singapore ® Tokyo ¢ Sydney
`
`Mylan Exhibit 1126
`Mylan v. Regeneron, IPR2021-00881
`Page 1
`
`Mylan Exhibit 1126
`Mylan v. Regeneron, IPR2021-00881
`Page 1
`
`
`
`IV
`
`Bibliographic Indices
`This publication is listed in bibliographic services, including Current Contents® and Index Medicus.
`
`Drug Dosage
`The authors and the publisher have exerted every effort to ensure that drug selection and dosageset
`forth in this text are in accord with current recommendationsand practice at the time of publication.
`However, in view of ongoing research, changes in government regulations, and the constant flow of
`informationrelating to drug therapy and drugreactions, the readeris urged to check the packageinsert
`for each drug for any changein indications and dosage and for added warningsandprecautions. This
`is particularly important when the recommendedagentis a new and/orinfrequently employed drug.
`
`All rights reserved
`No partofthis publication may betranslated into other languages, reproducedorutilized in any form
`or by any means,electronic or mechanical, including photocopying, recording, microcopying, or by
`any information storage and retrieval system, without permission in writing from the publisher.
`
`© Copyright 1992 by S. Karger AG, P.O. Box, CH-4009 Basel (Switzerland).
`Printed in Switzerland by Médecine et Hygiéne, Genéve.
`
`ISBN 3-8055-5466-4
`
`Mylan Exhibit 1126
`Mylan v. Regeneron, IPR2021-00881
`Page 2
`
`Mylan Exhibit 1126
`Mylan v. Regeneron, IPR2021-00881
`Page 2
`
`
`
`This material may be protected by Copyright law (Title 17 U.S. Code)
`
`Mylan Exhibit 1126
`Mylan v. Regeneron, IPR2021-00881
`Page 3
`
`
`
`74
`
`D.B. Volkin and A.M. Klibanov
`
`Reversible thermal denaturation
`
`The native, catalytically active conformation of a protein is maintained by a
`delicate balance of noncovalent forces including hydrogen bonds, hydrophobic,
`ionic and van der Waals interactions. An increase in temperature affects the
`strength of these interactions to different extents, thereby distorting this delicate
`balance and causing protein molecules to unfold. Upon cooling, the non-covalent
`interactions return to theirinitial state, and the enzymeregainsits native, catalyti-
`cally active conformation. This reversibility stems from the «thermodynamic
`hypothesis» which states that the native conformation of a protein in a given
`environmentcorresponds to the minimum free energy of the entire system. This
`conformation is determined by the amino acid sequence. The thermodynamic
`hypothesis is based on the classic experiments of Anfinsen and coworkers who
`demonstrated that ribonuclease, once reduced and unfolded in urea, can refold
`into the native, catalytically active structure by removal of urea and re-oxidation
`of sulfhydryl groups (2).
`The reversible unfolding of proteins has been examined extensively andits
`mechanisms are well understood (3). The design of protein formulations fre-
`quently utilizes additives and solvents whichstabilize proteins against reversible
`denaturation (4). The reversible partial unfolding of a protein is usually the first
`step, subsequently followed by conformational or covalent processesin irreversi-
`ble thermo-inactivation.
`
`Irreversible thermal inactivation of proteins
`
`If a protein solution is heated and then rapidly cooled, but full catalytic
`activity is not recovered within a reasonable time, then the protein has undergone
`irreversible thermalinactivation. In this case, conformational or chemical changes
`(see below) have occurred whicheither prevent refolding or destroy the integrity
`of the polypeptide molecule.
`
`1. Aggregation
`
`Heat-induced protein aggregation is most commonly described as a two-step
`process. First, as the temperature rises, the conformational flexibility (« breath-
`ing») of a protein intensifies until eventually conformational changes (local
`changes in secondary andtertiary structure) or reversible protein denaturation
`(cooperative loss of higher ordered structure) take place. This partially exposes the
`buried, interior hydrophobic aminoacid residues to the aqueoussolvent. In the
`second stage of this process, the thermally altered protein molecules associate
`intermolecularly (primarily via hydrophobicinteractions) in order to minimize the
`unfavorable exposure of hydrophobic amino acid residues to water (5). Subse-
`quent chemical reactions mayalso occur,especially intermolecular disulfide cross-
`links (6).
`
`2. Mechanisms of protein thermo-inactivation
`
`Intermolecular aggregation can be circumvented by employingdilute protein
`solutions or by using immobilized enzymes (7). As a result,
`the actual.
`monomolecular processes that cause irreversible thermal inactivation of enzymes
`can be determined. In order to elucidate their mechanisms, the relative contribu-
`
`Mylan Exhibit 1126
`Mylan v. Regeneron, IPR2021-00881
`Page 4
`
`Mylan Exhibit 1126
`Mylan v. Regeneron, IPR2021-00881
`Page 4
`
`
`
`Protein structure changes and inactivation
`
`75
`
`tions of conformational and covalent reactionsto the overall thermo-inactivation
`pathway must be determined.
`A monomolecular model for irreversible conformational thermal inactivation
`was proposed based on studies with immobilized trypsin (8). At high tempera-
`tures, an enzymelosesits native non-covalent (intramolecular) interactions; sub-
`sequently, non-native interactions may form which, although thermodynamically
`unfavorable, remain for purely kinetic reasons so that the protein cannot spon-
`taneously refold to the native conformation at ambient temperatures.
`Twoseparate criteria were established to identify monomolecular conforma-
`tional inactivation. First, enzymes can be reactivated after thermo-inactivation
`has occurred. For example, immobilized trypsin was heated and rapidly cooled
`(losing virtually all enzymatic activity), then completely unfolded and reducedin
`urea, and subsequently re-oxidized in the absence of denaturant. Recovery of
`enzymatic activity was nearly complete, implying that immobilized trypsin inacti-
`vated almostentirely via incorrect structure formation (8). A second methodology
`to measureirreversible conformational inactivation is based on the difference
`between the rates of thermo-inactivation in the presence and absenceofreversible
`denaturants (9).
`;
`11) quantitatively
`Employing these two strategies,
`some authors (10,
`accounted for the processes causingirreversible thermal inactivation (pH range of
`4-8) of hen egg-white lysozyme at 100°C and bovine pancreatic ribonuclease A at
`90°C,respectively. First, the overall monomolecularrate constantsofirreversible
`thermal inactivation were determined for both enzymes (protein concentrations
`wereselected so that no aggregation occurred), as shownin thefirst line of Table
`I. Second, a comparison of the rate constants of thermo-inactivation in the
`presence and absence of denaturants, along with reactivation experiments via the
`method described above for immobilized trypsin, provided the rate constant of
`incorrect structure formation, as shown in Table I. In the case of ribonuclease,
`these incorrect structures were foundto stem from thethiol-catalysed interchange
`of disulfide bonds (see next section).
`Samples of thermally inactivated lysozyme and ribonuclease were also ana-
`lysed for peptide chain integrity and aminoacid destruction to identify deleterious
`covalent processes that contribute to inactivation. As can be seen in Table I, the
`degradative covalent changes in the primary structure of lysozyme and
`ribonuclease are pH-dependentreactions whoserelative contribution to irreversi-
`ble thermo-inactivation consequently varies with pH. The general nature of these
`covalentreactionsis discussed in the next section. These degradative reactions not
`only pointed to « weak links» in the primarystructure of proteins, but also helped
`to define the upper limit of protein thermostability.
`
`Hydrolysis ofpeptide bonds at aspartic acid residues
`
`In the case of both ribonuclease and lysozyme,the hydrolysis of the polypep-
`tide chain at the carboxyl terminus of aspartic acid at acidic pHs was shown to
`proceed at rates comparable to thermo-inactivation. A combination of SDSpoly-
`acrylamide gel electrophoresis and gel scanning densitometry was used,along with
`measurements of the appearance of new carboxy and aminotermini. By examin-
`ing a series of Asp-X peptides, it has been shown(12) that the Asp-Pro bondis
`particularly labile at high temperatures under acidic conditions. Replacement of
`
`Mylan Exhibit 1126
`Mylan v. Regeneron, IPR2021-00881
`Page 5
`
`Mylan Exhibit 1126
`Mylan v. Regeneron, IPR2021-00881
`Page 5
`
`
`
`76
`
`D.B. Volkin and A.M. Klibanov
`
`Table I. The rate constantsofirreversible thermo-inactivation of lysozyme and ribonuclease: the over-
`all process and contributions of individual mechanisms to thermo-inactivation (14).
`
`Irreversible thermo-inactivation
`of two model proteins
`
`
`
`Rate constants (hr~!)
`
`H6
`ae +
`
`
`
`
`
`Hen egg white lysozyme, 100°C (10)
`Directly measured overall process
`Dueto individual mechanisms:
`Deamidation of Asn/Gln residues
`Hydrolysis of Asp-X peptide bonds
`Destruction of cystine residues
`Formation ofincorrect structures
`
`H
`
`4.1
`
`Bovine pancreatic ribonuclease 90°C (11)
`Directly measured overall process
`Due to individual mechanisms:
`Deamidation of Asn/Gln residues
`Hydrolysis of Asp-x peptide bonds
`Destruction of cystine residues
`Formation ofincorrect structures'
`
`'
`
`Shownto be dueto thiol-catalysed disulfide interchange.
`
`such labile Asp residues by meansofsite-directed mutagenesiscan greatly stabilize
`proteins (13).
`
`Deamidation of asparagine and glutamine residues
`
`Deamidation of asparagine and/or glutamineresidues was found to contribute
`to the irreversible thermo-inactivation of both ribonuclease and lysozyme. The
`rate of ammonia release from thermo-inactivated enzyme was determined by
`either chemical or enzymatic methods. Moreover, since deamidation converts Asn
`to Asp and Gln to Glu, the extent of deamidation was further characterized by
`isoelectric focusing (IEF) coupled with quantitative gel scanning. By extraction of
`deamidated species from the IEFgel, it was found that each random deamidation,
`on average, lowers the specifc activity by 30-50% compared to the native enzyme
`(14).
`An in-depth investigation into the nature of the deamidation reaction in three
`model hexapeptides containing asparagine residues confirmed that the mechanism
`of deamidation underneutral to basic conditions involves an intramolecularcycli-
`zation (15). The main chain amide nitrogen acts as a nucleophile attacking the
`electrophilic asparagine amide causing ring closure to the imide with concomitant
`release of ammonia. The subsequent hydrolysis leads to a mixture of « and B
`aspartyl residues. A temperature and sequence dependence of deamidation in”
`model hexapeptides at 100°C at pH 7.4 was observed. It was found that Asn-Gly
`is 30-50 times morelabile than Asn-Pro and Asn-Leu duetosteric factors.
`
`Mylan Exhibit 1126
`Mylan v. Regeneron, IPR2021-00881
`Page 6
`
`Mylan Exhibit 1126
`Mylan v. Regeneron, IPR2021-00881
`Page 6
`
`
`
`Protein structure changes andinactivation
`
`77
`
`Recent studies have shownthatthe tertiary structure ofa protein is a principal
`‘determinant to deamidation under physiological conditions. By using neutron
`diffraction techniques, the time-aged, deamidatedprotein crystals of trypsin were
`examinedandthesites of deamidation did not correlate with the previously men-
`tioned peptide studies (16). Moreover,kinetic studies on the deamidation ofa par-
`ticularly labile Asn-Gly sequence in ribonuclease showedsignificantly (30-fold)
`increased lability in the unfolded versus folded enzyme(17).
`Asexpected from the cyclic imide mechanism,the rate of heat-induced deami-
`dation in ribonuclease and lysozymeincreased as the pH wasraised from 4 to 8
`(see Table I). Heat-induced deamidation seems to be a general phenomenon as
`seen by the continuous release of ammonia (levelling off at the total number of
`Asn and Gin residues) in 10 different globular proteins heated at pH 4 and 100°C
`(Volkin and Klibanov, unpublished results). In a practical application of this
`knowledge,site-directed mutagenesis has been used to removespecific asparagine
`residues from the dimer interface of triose phosphate isomerase. The resulting
`mutant enzyme showed a two fold increasein stability against irreversible thermo-
`inactivation (18). A similar approach has been Seinen applied to recom-
`binant interleukin-lo (19).
`
`Cystine destruction and thiol-catalysed disulfide interchange
`Amino acid analysis revealed the destruction of cystine residues during the
`thermo-inactivation of ribonuclease and lysozymeat neutralto slightly alkaline
`pH. The loss of disulfide bonds occurred at rates comparable to thermo-
`inactivation. By identifying the by-products of cystine destruction, this process
`was shown to proceed via a {-elimination reaction to yield dehydroalanine,
`thiocysteine and lysinoalanine residues (20).
`Other workers (11) demonstrated that the disulfide bonds of ribonuclease are
`reshuffled during irreversible thermo-inactivation. Not only was the enzymestabi-
`lized against irreversible thermo-inactivation by adding thiol scavengers such as
`copperchloride, but heating in the presenceoffree-labelled cystine resulted in the
`incorporation of its fragments into the enzymevia a mixed disulfide. In a thiol-
`catalysed disulfide interchange reaction, free thiols carry out nucleophilic attack
`on the sulfur atom of a disulfide (21).
`The generality of the heat lability of disulfide bonds in over a dozen proteins
`(containing between 1 and 19 cystine residues) at 100°C wasinvestigated (22). The
`B-elimination reaction was shownto generatefree thiols which in turn catalyse dis-
`ulfide interchange.
`
`Oxidation of cysteine residues
`
`By first quantitatively determining the mechanisms of irreversible thermo-
`inactivation for two model enzymes(ribonuclease and lysozyme), it was then pos-
`sible to study thermo-inactivation of more complicated enzyme systems. Some
`researchers (23, 24) examined the thermostability of several homologousbacterial
`a-amylases. The thermal
`inactivation of both mesophilic and thermophilic
`enzymes at 90°C and pH 6.5 was due to a monomolecular conformational
`process. The differences in thermostability between mesophilic and thermophilic
`enzyme were shownto be dueto additional salt bridges in the latter. When the
`conformationalinactivation process was suppressed (by additives or pH changes),
`
`Mylan Exhibit 1126
`Mylan v. Regeneron, IPR2021-00881
`Page 7
`
`Mylan Exhibit 1126
`Mylan v. Regeneron, IPR2021-00881
`Page 7
`
`
`
`78
`
`D.B. Volkin and A.M. Klibanov
`
`a-amylase underwentirreversible thermo-inactivation (at 90°C) via deamidation
`of its asparagine and/or glutamine residues and oxidation of cysteine residues.
`Similarly, in the case of immobilized glucose isomerase, conformational scram-
`bling (in the absence of substrate), deamidation of Asn and/or Gln residues
`(90°C) and cysteine oxidation (70°C) were shown to contribute to thermo-
`inactivation (25).
`Cysteine residues undergo auto-oxidation to form intra- or intermoleculardis-
`ulfide bondsor oxidation products such as sulfenic acid (21). These oxidative reac-
`tions are catalysed by divalent metal ions (especially copper) with enhancedrates
`at elevated pH. The productsof cysteine oxidation during the thermo-inactivation
`of «-amylase were determined to be 30%sulfenic acid and 70% new disulfides
`(23).
`
`CONCLUSIONS
`
`Recombinant DNAtechnology allows us to manufacture humanproteins for
`use as therapeutic agents. Consequently, the pharmaceutical scientist must now
`develop rational strategies to stabilize these proteinsin vitro. It is necessary to pre-
`pare formulations of proteins that will be stable under defined storage conditions
`for periods exceeding one year. A critical step toward developing such stabilized
`preparations is an understanding of the causes and mechanisms of temperature
`induced-inactivation, which can be subsequently extrapolated to ambient condi-
`tions.
`Upto this point, we have discussed the degradative pathways, both conforma-
`tional and covalent, that contribute to the loss of enzymatic activity during heat-
`ing in aqueoussolution. Environmental conditions can bealtered in a variety of
`ways (temperature, pH, salts, solvents, etc.) to cause protein inactivation (26).
`Similar to thermo-inactivation, these environmental changes usually induce
`partial unfolding or disruption of the tertiary structure,
`thereby exposing
`hydrophobicor chemically labile residues which leads to conformationalor cova-
`lent inactivation. Although the pathwaysare similar, additional «weak links»in
`the primary structure have been identified: oxidation of methionine residues,
`racemization of aspartic acid and serine residues, photodegradation of trypto-
`phan, and nucleophilic addition reactions involving lysine residues (26).
`The freezing and lyophilization of a protein solution mayalso lead to inactiva-
`tion, since both processes involve the concentration of solutes as wateris crystal-
`lized and sublimated, respectively. As water moleculescrystallize during freezing,
`solutes such as salts and protein molecules are concentrated, causing potentially
`significant shifts in pH and ionic strength. These shifts may cause conformational
`changesin a protein leading to aggregation. For example, the freezing of an iso-
`tonic saline solution (0.15M) to — 21°C causes a 24-fold increasein salt concentra-
`tion. Moreover, the pH of a sodium phosphate buffer can shift several pH units
`upon freezing (27). Another potential mechanism of protein inactivation during
`freezing is oxidation; the oxygen concentrationin a partially frozen system at -3°C
`is 1150-fold higher than that solution at 0°C (28).
`The subsequentstorage of lyophilized proteins can be accompaniedbyinacti--
`vation. Recent workin our laboratory has shownthat as lyophilized protein pow-
`ders are hydrated during storage, the protein molecules may aggregate in the solid
`
`Mylan Exhibit 1126
`Mylan v. Regeneron, IPR2021-00881
`Page 8
`
`Mylan Exhibit 1126
`Mylan v. Regeneron, IPR2021-00881
`Page 8
`
`
`
`Protein structure changes and inactivation
`
`79
`
`state, thereby losing the ability to dissolve in aqueous solution upon addition of
`water. In the case of lyophilized BSA, as well as several other proteins, this loss
`of solubility was shownto be dueto a thiol-disulfide interchange reaction. When
`the free cysteine residue of BSA was alkylated, and then freeze-dried and stored
`in a humid atmosphereas before,the solubility of the protein powder upon recon-
`stitution was dramatically increased (29).
`In summary, an understanding of causes and mechanismsofprotein inactiva-
`tion allows for the developmentof rational strategies (both protein engineering
`and formulation design) to minimize, or perhaps even eliminate, its occurrence.
`
`REFERENCES
`
`. Lumry, R. and Eyring, H. (1954). Conformational changes of proteins. Journal of Physical
`Chemistry 58, 110-120.
`. Anfinsen, C.B. and Scheraga, H.A. (1975). Experimental and theoretical aspects of protein fold-
`ing. Advances in Protein Chemistry 29, 205-301.
`. Lapanje, S. (1978). Physicochemical Aspects of Protein Denaturation. Wiley, New York.
`4. Timasheff, S.N. and Arakawa, T. (1989). Stabilization of protein structure by solvents. In:
`Protein structure: a practical approach (T.E. Creighton, ed.) pp. 283-289, IRL Press, Oxford.
`. Jaenicke, R. (1967). Intermolecular forces in the process of heat aggregation of globular proteins
`and the problem of correlation between aggregation and denaturation phenomena. Journal of
`Polymer Science Part C 16, 2143-2160.
`. Mozhaev, V.V. and Martinek, K. (1982). Inactivation and reactivation of proteins (enzymes).
`Enzyme and Microbial Technology 4, 299-309.
`. Klibanov, A.M. (1983). Stabilization of enzymes against thermal inactivation. Advances in
`Applied Microbiology 29, 1-28.
`(1978). On the mechanism of irreversible thermo-
`. Klibanov, A.M. and Mozhaev, V.V.
`inactivation of enzymes andpossibilities for reactivation of «irreversibly» inactivated enzymes.
`Biochemical and Biophysical Research Communications 83, 1012-1014.
`. Ahern, T.J. and Klibanov, A.M. (1986). Why do enzymesirreversibly inactivate at high tempera-
`tures? In: Protein structure. Folding and design (D.L. Oxender, ed.) pp. 283-289, Alan R. Liss,
`New York.
`Ahern, T.J. and Klibanov, A.M. (1985). The mechanism ofirreversible enzymeinactivation at
`100°C. Science 228, 1280-1284.
`Zale, S.E. and Klibanov, A.M. (1986). Why does ribonucleaseirreversibly inactivate at high tem-
`peratures? Biochemistry 25, 5432-5444.
`Marcus, F. (1985). Preferential cleavage at Asp-Pro peptide bondsin dilute acid. /nternational
`Journal of Peptide and Protein Research 25, 542-546.
`George-Nascimento, C., Lowenson, J., Borissenko, M., Calderon, M., Medina-Selby, A., Kuo,
`J., Clarke, S. and Randolph, A. (1990). Replacementoflabile aspartyl residue increasesthe stabil-
`ity of human epidermal growth factor. Biochemistry 29, 9584-9591.
`Ahern, T.J. and Klibanov, A.M. (1988). Analysis of processes causing thermalinactivation of
`enzymes. Methods in Biochemical Analysis 33, 91-127.
`Geiger, T. and Clarke, S. (1987). Deamidation, isomerization, and racemization at asparaginyl
`and aspartyl residues in proteins. Journal of Biological Chemistry 262, 785-794.
`Kossiakoff, A.A. (1988). Tertiary structure is a principal determinant to protein deamidation.
`Science 240, 191-194.
`Wearne,S.J. and Creighton, T.E. (1989). Effect of protein conformation on rate of deamidation:
`ribonuclease. A. Proteins, Structure, Function and Genetics 5, 8-12.
`Ahern, T.J., Casal, J.I., Petsko, G.A. and Klibanov, A.M. (1987). Control of oligomeric enzyme
`thermostability by protein engineering. Proceedings of the National Academy of Sciences USA
`84, 675-679.
`
`10.
`
`11.
`
`12.
`
`13.
`
`14,
`
`15.
`
`16.
`
`17.
`
`18.
`
`Mylan Exhibit 1126
`Mylan v. Regeneron, IPR2021-00881
`Page 9
`
`Mylan Exhibit 1126
`Mylan v. Regeneron, IPR2021-00881
`Page 9
`
`
`
`80
`
`D.B. Volkin and A.M. Klibanov
`
`19. Wingfield, P.T., Mattaliano, R.J., MacDonald, H.R., Craig, S., Clore, G.M., Gronenborn,
`A.M. and Schmeissner, U.
`(1987). Recombinant-derived interleukin-l« stabilized against a
`specific deamidation. Protein Engineering 1, 413-417.
`20. Whitaker, J.R. and Feeney, R.E. (1983). Chemical and physical modification of proteins by the
`hydroxide ion. CRC Critical Reviews in Food Science and Nutrition 19, 173-212.
`21. Torchinsky, Y.M. (1981). Sulfur in Proteins. Pergamon Press, Oxford.
`22. Volkin, D.B. and Klibanov, A.M. (1987). Thermal destruction processesin proteins involving cys-
`tine residues. Journal of Biological Chemistry 262, 2945-2950.
`23. Tomazic, S.J. and Klibanov, A.M. (1988a). Mechanismsofirreversible thermal inactivation of
`Bacillus «-amylases. Journal of Biological Chemistry 263, 3086-3091.
`24. Tomazic, S.J. and Klibanov, A.M. (1988b). Whyis one Bacillus «-amylase moreresistant against
`irreversible thermo-inactivation than another? Journal of Biological Chemistry 263, 3092-3096.
`25. Volkin, D.B. and Klibanov, A.M. (1989a). Mechanism of thermo-inactivation of immobilized
`glucose isomerase. Biotechnology and Bioengineering 33, 1104-1111.
`26. Volkin, D.B. and Klibanov, A.M. (1989b). Minimizing protein inactivation. In: Protein function:
`a practical approach (T.E. Creighton, ed.) pp. 1-124, Oxford Press, Oxford.
`27. Franks, F. (1990). Freeze-drying: from empiricism to predictability. Cryoletters 11, 93-110.
`28. Schwimmer,S. (1981). Source Book of Food Enzymology. Avi Publishing, Westport.
`29. Liu, W.R., Langer, R. and Klibanov, A.M. (1991). Moisture-induced aggregation of lyophilized
`proteins in the solid state. Biotechnology and Bioengineering 37, in press.
`
`GENERAL DISCUSSION
`
`PARTICIPANT:You showedin yourlast slide that you could, by alkylating
`and blocking the SH group explain the three microliters. But without blocking the
`SH group youdid get non-aggregation with seven microliters. How do youlink
`those two observations?
`S
`VOLKIN:I haveto confess that wasn’t my work,butI think that they carried
`it out at 7.5 microliters and it also blocked. This was just one example.
`YIM:I have two questions. You seem to indicate that the thermal inactivation
`whichis due to unfolding and a covalent and non-covalent inactivation seem to
`be additive with respect to inactivation. I think that might just be a very fortui-
`tious situation because percent of inactivation is not necessarily correlated with
`percentage of deamidation or anything like that.
`VOLKIN:Thereare manypossible chemical reactions that can cause inactiva-
`tion. These wereactually shownto causeinactivation. In the case of lysozyme and
`deamidation isoelectric focusing was carried out. The protein, the mono-, di-,
`trideaminated species were extracted from the gel, refolded, and the specific
`activity of each one was determined, and the decrease in activity was measured.
`YIM: This necessitates a linear correlation with percentage of inactivation
`versus percentage of deamination or oxidation.
`VOLKIN:Youare saying what if there is a deamidation on the surface, for
`instance, and that we wouldn’t expect to cause inactivation as much as deamida-
`tion of the interior residue. That certainly could be the case. These experiments
`were carried out at high temperatures where the protein is unfolded. So those
`residues are exposed.
`YIM: Doyouhave any data at lower temperatures so that one can extrapolate
`to more reasonable temperatures?
`
`v
`
`Mylan Exhibit 1126
`Mylan v. Regeneron, IPR2021-00881
`Page 10
`
`Mylan Exhibit 1126
`Mylan v. Regeneron, IPR2021-00881
`Page 10
`
`
`
`Protein structure changes and inactivation
`
`81
`
`VOLKIN:As wehave already heard, extrapolation is a very dangerous game.
`It‘is especially dangerousin these kinds of processes becauseit is not a linear rela-
`tionship because there are conformational effects. For instance, deamidation,
`whereat high temperatures we cansee certain patterns in peptide studies, we no
`longer see those patterns at low temperatures where the protein is folded. Soitis
`a very dangerous game to extrapolate lower. We werecarrying out these studies
`to look at the upper limit of protein thermostability to try to determine the weak
`links that limit protein stability.
`
`Mylan Exhibit 1126
`Mylan v. Regeneron, IPR2021-00881
`Page 11
`
`Mylan Exhibit 1126
`Mylan v. Regeneron, IPR2021-00881
`Page 11
`
`