Case 1:18-cv-01363-CFC Document 79-3 Filed 03/22/19 Page 1 of 3 PageID #: 9456
`
`Mullan et al. BMC Proceedings 2011, 5(Suppl 8):P110
`http://www.biomedcentral.com/1753-6561/5/S8/P110
`
`ME ET I N G A BS T R A CT
`Open Access
`Disulphide bond reduction of a therapeutic
`monoclonal antibody during cell culture
`manufacturing operations
`Brian Mullan1*, Bryan Dravis2, Amareth Lim3, Ambrose Clarke4, Susan Janes3, Pete Lambooy2, Don Olson2,
`Tomas O’Riordan1, Bruce Ricart3, Alexander G Tulloch2
`
`From 22nd European Society for Animal Cell Technology (ESACT) Meeting on Cell Based Technologies
`Vienna, Austria. 15-18 May 2011
`
`Background
`Disulphide bonding is critical to maintaining immuno-
`globulin (IgG) tertiary and quaternary structure for ther-
`apeutic monoclonal antibodies (MAb). Both inter- and
`intra-chain disulphide bonds are formed intracellularly
`in the expression host prior to secretion and purification
`during MAb production processes. Disulphide bond
`shuffling has previously been reported for IgG2[1,2] and
`disulphide-mediated arm-exchange for IgG4[3,4], reflect-
`ing innate behaviour of these IgG classes. However, aty-
`pical and significant reduction of disulphide bonds has
`been recently observed in IgG1[5,6] that present signifi-
`cant issues for manufacturing of therapeutic MAbs.
`During manufacturing of preliminary lots of a recently
`transferred MAb manufacturing process (IgG1), gross
`disulphide bond reduction following affinity capture
`chromatography of clarified production bioreactor mate-
`rial was observed. Investigations leading to the identifi-
`cation of the nature of this reduction process, and
`process steps to mitigate against its future occurrence,
`are described here. The MAb was co-developed with
`MacroGenics, Rockville, MD.
`
`Methods
`Production Bioreactor material for downstream proces-
`sing was supplied from a 16 day, fed-batch, GS-CHO
`culture [7]. The Production Bioreactor was a single-use
`(Wave System) with a 100L (full scale) or 10L (lab
`model) working volume. Clarified harvest intermediate
`(CHI) hold studies were performed in either 560L Lev-
`Mix units (full scale) or 5L Braun benchtop bioreactors
`
`* Correspondence: mullan_brian@lilly.com
`1Manufacturing Science and Technology, Eli Lilly & Co, Kinsale, Cork, Ireland
`Full list of author information is available at the end of the article
`
`(lab model). Purification to produce Affinity Capture
`chromatography eluted mainstream was performed
`using a 20cm x 20cm MAbselect Protein A resin (GE
`Healthcare) column, and an AKTA Process skid (GE
`Healthcare).
`LC-MS analysis was performed on a Polymer Labora-
`tories PLRP-S HPLC column and analyzed using an Agi-
`lent 1100 HPLC system coupled to an Applied
`Biosystems QSTAR XL mass spectrometer, following
`sample preparation. CE-SDS analysis was performed
`using a Beckman Coulter PA800 capillary electrophor-
`esis instrument fitted with bare-fused silica capillary and
`UV detection at 220 nm, following sample preparation.
`Microchip CE-SDS analysis was performed using a Lab-
`on-chip microanalyser (Agilent). Free thiols were quanti-
`fied using Ellman’s reagent. Metabolic analysis was con-
`ducted by Metabolon (Durham, NC).
`
`Results
`Identification of disulphide reduction of IgG during
`Primary Recovery
`The IgG manufacturing process as transferred from the
`co-developing partner included a cell culture settling
`step following the Production Bioreactor and prior to
`Primary Recovery.
`Disulphide bond reduction was first detected during
`initial development runs by routine Non-reduced (NR)
`CE-SDS in-process analysis after Affinity capture chro-
`matography (data not shown). NR-CE-SDS analysis
`identified elevated levels of free light chain and half anti-
`body molecules, when compared to Reference Standard.
`Additional analysis, employing microchip-based NR-
`CE-SDS methods indicated that the antibody reduction
`occurred during the primary recovery cell settling step
`
`© 2011 Mullan et al; licensee BioMed Central Ltd. This is an open access article distributed under the terms of the Creative Commons
`Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in
`any medium, provided the original work is properly cited.
`
`

`

`Case 1:18-cv-01363-CFC Document 79-3 Filed 03/22/19 Page 2 of 3 PageID #: 9457
`
`Mullan et al. BMC Proceedings 2011, 5(Suppl 8):P110
`http://www.biomedcentral.com/1753-6561/5/S8/P110
`
`Page 2 of 3
`
`(results not shown). This was confirmed by LC-MC ana-
`lysis (results not shown). Assessment of disulphide
`bonding pattern and intactness by LC-MS peptide map-
`ping identified both inter- and intra-chain disulphide
`scrambling (results not shown).
`
`disulphide bonded) IgG1 in CHI. When O2 was present,
`IgG1 remained intact under all conditions evaluated.
`Only when O2 was deliberately absent, or stripped away,
`would the harvest material or CHI demonstrate poten-
`tial for reduction (Figure 1b).
`
`Delineation of events leading to IgG reduction
`Initial investigations to understand process behaviour
`during primary recovery identified that reducing species,
`including free thiols (which increase over the course of
`the Production Bioreactor, up to 1mM), were present at
`the end of the Production Bioreactor (Figure 1a). Dis-
`solved oxygen was also shown to deplete during the cell
`settling phase following harvest (data not shown). From
`this, an initial working hypothesis was formed that redu-
`cing species, including free thiols, became reactive at
`low dissolved oxygen concentrations and led to IgG1
`disulphide bond reduction.
`A revised process control strategy was implemented
`(see below) to prevent oxygen depletion and maintain
`dissolved oxygen levels above a minimum level. This
`involved including an aerated and agitated hold for Clar-
`ified Harvest Intermediate (CHI) in the process.
`Further studies identified that O2 is critical to main-
`taining a stable environment for oxidised (i.e., normally
`
`Metabolic behaviour of reducing intermediates
`The working hypothesis was that by maintaining suffi-
`cient levels of dissolved oxygen in the CFM, the thiol
`species could be reacted out (oxidised) and a stable
`environment for oxidised IgG1 created (Figure 1a). How-
`ever, the relationship between IgG reduction and thiol
`redox state is not first order (Figure 1c), and the rate of
`thiol oxidation was found to be dependent on the
`source of Production Bioreactor material (i.e., varied
`with different harvest lots). This indicated the involve-
`ment of an additional component, potentially catalytic,
`which has not yet been identified in our studies. Thiore-
`doxins have been identified as such a catalytic compo-
`nent by others [5,6], and these need to be recycled after
`one redox cycle via Thioredoxin Reductase / NADP(H).
`Metabolic analysis of cell and media material from
`Production Bioreactors indicated high levels of oxidised
`homocysteine and cysteine (both reactive redox mole-
`cules), which correlated with decreasing levels of folate
`
`Figure 1 IgG disulphide bond reduction under various conditions for cell-settled and immediately clarified harvest material. CHI, Clarified harvest
`intermediate; DiS, Disulphide; LoC, Lab-on-a-Chip (Agilent).
`
`

`

`Case 1:18-cv-01363-CFC Document 79-3 Filed 03/22/19 Page 3 of 3 PageID #: 9458
`
`Mullan et al. BMC Proceedings 2011, 5(Suppl 8):P110
`http://www.biomedcentral.com/1753-6561/5/S8/P110
`
`Page 3 of 3
`
`doi:10.1186/1753-6561-5-S8-P110
`Cite this article as: Mullan et al.: Disulphide bond reduction of a
`therapeutic monoclonal antibody during cell culture manufacturing
`operations. BMC Proceedings 2011 5(Suppl 8):P110.
`
`(B6) and cobalamine (B12), both of which are involved
`in recycling homocysteine. Overall, this analysis identi-
`fied numerous options for media optimisation to miti-
`gate against IgG reduction. However, given the success
`of process controls (described below), and the late stage
`of process development (pre-validation) these media
`optimisation options were not pursued.
`
`Process controls to mitigate disulphide reduction of IgG
`A process control strategy was implemented including:
`• Establishing a minimal dissolved oxygen level in the
`Production Bioreactor prior to harvesting
`• Immediately clarifying the Production Bioreactor
`material (i.e., eliminating the cell settling step)
`• Holding the CHI in a hold vessel (LevMix container,
`agitated hold) that had been partially pre-filled with pro-
`cess air.
`
`Conclusions
`• Gross disulphide bond reduction was observed during
`late stage development of an IgG1 monoclonal antibody
`being commercialised for a therapeutic indication;
`• Disulphide bond reduction had a second, or higher,
`order link to low dissolved oxygen levels in process
`intermediates, and the involvement of a catalytic factor
`was also indicated;
`• Implementation of an appropriate control strategy
`(and associated process analytics) informed by process
`development has ensured no recurrence of this issue
`(for n=15 full scale lots).
`
`Author details
`1Manufacturing Science and Technology, Eli Lilly & Co, Kinsale, Cork, Ireland.
`2Bioprocess R&D, Eli Lilly & Co, Indianapolis, Indiana, USA. 3Bioproduct
`Analytical Chemistry, Eli Lilly & Co, Indianapolis, Indiana, USA. 4Analytical
`Technical Operations, Eli Lilly & Co, Kinsale, Cork, Ireland.
`
`Published: 22 November 2011
`
`2.
`
`3.
`
`4.
`
`References
`1. Wypych J, et al: Human IgG2 antibodies display disulfide-mediated
`structural isoforms. J Biol Chem 2008, 283:16194-16205.
`Liu YD, et al: Human IgG2 antibody disulfide rearrangement in vivo. J
`Biol Chem 2008, 283:29266-29272.
`van der Neut Kolfschoten M: Anti-inflammatory activity of human IgG4
`antibodies by dynamic Fab arm exchange. Science 2007, 317:1554-1557.
`Labrijn AF, et al: Therapeutic IgG4 antibodies engage in Fab-arm
`exchange with endogenous human IgG4 in vivo. Nature Biotechnology
`2009, 27:767-773.
`Trexler-Schmidt M, et al: Identification and Prevention of Antibody
`Disulfide Bond Reduction During Cell Culture Manufacturing.
`Biotechnology and Bioengineering 2010, 106:452- 461.
`Kao Y-H, et al: Mechanism of Antibody Reduction in Cell Culture
`Production Processes. Biotechnology and Bioengineering 2010, 107:622-632.
`7. Mullan B, et al: Transfer, Implementation and Late Stage Development of
`an End-To-End Single-Use Process for Monoclonal Antibody
`Manufacture. American Pharmaceutical Review 2011, 58-64.
`
`5.
`
`6.
`
`Submit your next manuscript to BioMed Central
`and take full advantage of:
`
`• Convenient online submission
`
`• Thorough peer review
`
`• No space constraints or color figure charges
`
`• Immediate publication on acceptance
`
`• Inclusion in PubMed, CAS, Scopus and Google Scholar
`
`• Research which is freely available for redistribution
`
`Submit your manuscript at
`www.biomedcentral.com/submit
`
`