Modular DSP approaches for complex non-mab molecules

Transcription

1 Modular DSP approaches for complex non-mab molecules Dr. Stefan R. Schmidt MBA SVP Process Science & Production

2 Agenda 1 Introduction Process related impurities Product related impurities Virus inactivation Summary and conclusion 2

3 Expertise with a broad spectrum of proteins >100 therapeutic proteins produced % monoclonal antibodies and fusion proteins >40 % biosimilars and biobetters Produced substance classes API Cytokines & growth factors 10% Enzymes & blood factors 14% Other recombinant proteins 6% Biosimilars 27% Originator drugs vs. biosimilars and biobetters Fusion proteins 17% Antibodies 53% Biobetters 14% Originators 59% 3

4 General considerations for DSP development Assessment of physico-chemical product properties Explorative capture study pi Structure e 280nm pka titration Hydrophobicity ph 3.5 stability ph titration curve Temp. stability Mech. stability Basic requirements Protein concentration in intermediates 4 g L -1 Process steps at ambient temperature General process related considerations/requirements Stability data on Capture (& Harvest and Intermediates at -20 C, RT and 5 C, +/- stirring) Aggregation/precipitation propensity (SEC, T 580nm ) Contribution of steps for virus removal/inactivation (virus safety concept) Process step development Confirmatory purification sequence Quality targets GMP readyness Facility fit Virus risk Economics Definition of success criteria Overall process yield Removal factors of impurities: o process related (HCP, DNA, ProtA) o product related (fragments, aggregates, isoforms etc.) Final DS specification 4

5 Enz Vac Molecule Novel? Yes No Biosimilar Development Concept of modular toolbox FP mab Standard Process Yes Antibody? Yes Platform? No No Capture No Affinity available? Yes Low ph tolerant? No Yes IEX or MixMode Resin screen ph Titration Major impurities? Alternative Virus Inactivation Detergents? No Yes Stability assessement Polishing Org. Solvents? No Yes Aggregates? No Sequence of evaluation Capture ph-titration Stability assessment Technical solution UV, x-ray Yes Charged? No Hydrophobic? No Yes Yes IEX or MixMode Resin screen HIC: Resin screen Yes DNA/HCP? No Polishing Hydroxyapatite Virusfiltration 5

6 Variety of DSP processes for complex proteins Step 1 Step 2 Step 3 Step 4 Step 5 Step 6 Step 7 Step 8 Step 9 Step 10 g L -1 % FcF1 Affinity ph AIEX CIEX VF UF/DF FcF2 Affinity ph MIX C MIX A VF UF/DF FcF3 DepF Affinity ph DepF AIEX CIEX VF UF/DF FcF4 Affinity ph CHT VF UF/DF FcF5 Affinity DepF CIEX ph MIX A VF UF/DF FcF6 Affinity ph Precip DepF CIEX VF UF/DF FcF7 Affinity DepF ph CIEX MIXA VF UF/DF Vac1 UF/DF ph AIEX MIX C UF/DF AIEX VF UF/DF Vac2 UF/DF AIEX CIEX AIEX VF UF/DF Enz1 UF/DF DepF Affinity HIC AIEX ph CIEX VF UF/DF Enz2 MIX C HIC CHT Isopr. UF/DF AIEX VF UF/DF FP1 UF/DF DepF Triton AIEX HIC UF/DF CHT CIEX VF UF/DF FP2 AIEX MIX C ph UF/DF AIEX CHT UF/DF VF

7 Getting order and structure Ursus Wehrli

8 Modules of DSP unit operations and sequence of steps Conditioning Capture Virus Polishing Virus & Aliquot FcF1 FcF2 FcF3 FcF4 FcF5 FcF6 FcF7 Vac1 UF/DF Affinity ph AIEX CIEX VF UF/DF Affinity ph MIX C MIX A VF UF/DF DepF Affinity ph DepF AIEX CIEX VF UF/DF Affinity ph CHT VF UF/DF Affinity DepF CIEX ph MIX A VF UF/DF Affinity ph Precip DepF CIEX VF UF/DF Affinity DepF ph CIEX MIXA VF UF/DF ph AIEX MIX C UF/DF AIEX VF UF/DF Vac2 UF/DF AIEX CIEX AIEX VF UF/DF Enz1 UF/DF DepF Affinity HIC AIEX ph CIEX VF UF/DF Enz2 MIX C HIC CHT Isopr. UF/DF AIEX VF UF/DF FP1 UF/DF DepF Triton AIEX HIC UF/DF CHT CIEX VF UF/DF FP2 AIEX MIX C ph UF/DF AIEX CHT UF/DF VF 8

9 Agenda 1 Introduction Process related impurities Product related impurities Virus inactivation Summary and conclusion 9

10 FP2: Potential folding pathways leading to variants Tetramer formation depends on linker length and ph Linker can be target for fragmentation or truncation Some intermediates occur in folding pathway Monomeric byproducts can be suppressed by additives in media Aggregate formation still possible Closed monomer Domain swapped tetramer Kipriyanov et al. JMB, (1999), 293,

11 [ppm] Yield [%] FP2: HCP removal & yield HCP overall step AIEX MIX C ph UF/DF AIEX CHT UF/DF VF Harvest C10 C20 V10 I10 C30 C40 I20 I30 I

12 Activity after neutralization [%] Start 3h 24h 40h 48h 3h 24h 40h 48h FP2: Effect of low ph and incubation AIEX MIX C ph UF/DF AIEX CHT UF/DF VF Incubation kinetic 20 C 2-4 C 140% 120% 100% 80% 60% 2-8 C 20 C 40% 20% 0% [h] At least 24 h are required for complete refolding and maximal restoration of activity Incubation at 20 C leads to fragmentation (presence of acidic protease) In the process FP2 is incubated at 2-8 C for 48 h 12

13 HCP removal by precipitation with caprylic acid Caprylic acid Interaction Protein Precipitation Precipitate can be removed by depth filtration Concentrations >3 mmol L -1 sufficient for virus inactivation At high concentration yield is reduced Optimization process: o Titration of ph and CA concentration o Select concentration that keeps API still soluble CA impacts subsequent viral removal steps! Low ph eluate CA addition ph adjustment Homogenization Depth filtration 13

14 HMWS Yield Model HCP FcF6: HCP removal by precipitation with caprylic acid HCP reduction already at low conc. of CA ph only has influence on aggregation Predictable model 14

15 Caprylic acid-induced impurity precipitation HCP Clearance from a monoclonal antibody: Influence of ph and CA concentration HCP Red. 1.5 log Yield 95% Robust HCP clearance at high mab yield 15

16 Agenda 1 Introduction Process related impurities Virus inactivation Summary and conclusion Product related impurities 16

17 Example of product related impurities (mab) Aggregates 1. Full antibody 1 2. Incomplete antibody 2 3. Paired heavy chain 4. Half antibody 5. Free heavy chain 6. Free light chain (non-red SDS-PAGE, silverstain)

18 CE-SDS Content % Concentration in g/l CE-SDS Content % Concentration in g/l FcF7: Peak cut in C20 eliminates incomplete molecules Affinity DepF ph CIEX MIXA VF UF/DF Pool A7 A8 A9 A10 A11 A12 B1 B2 B3 B4 B5 B6 0 Monomer Dimer Half-trimer Trimer A7 A8 A9 A10 A11 A12 B1 B2 B3 B4 B5 B

19 Truncations Aggregates Product related impurities HMW Mono LMW FcF4 Affinity ph CHT VF UF/DF Vac1 UF/DF ph AIEX MIX C UF/DF AIEX VF UF/DF HMW Mono LMW FcF4 seems to be low ph sensitive Aggregates can be successfully removed by CHT chromatography Non mammalian Vac1 contains a lot of LMW species (clipped isoforms) from the beginning LMW have lower net charge elimination in subsequent ion exchange chromatography 19

20 FP1: Glyco-isoform separation Background: Molecule: highly glycosylated fusion protein Resin: DEAE Sepharose FF Purpose of the step: Removal of HCP and DNA Removal of low glycosylated isoforms Results: DNA reduction factor: 4 log Glycosylation: 100 % Yield: 45 % Load RP-HPLC L W E SDS-PAGE IEF M S L E M S L E Glc FP Glc Elute 20

21 Agenda 1 Introduction Process related impurities Product related impurities Summary and conclusion Virus inactivation 21

22 Virus inactivation alternatives and issues ph < 4 (key step for elution of mabs/fcf from Protein A resin) o Can lead to denaturation and proteolysis o Might induce aggregation during high protein concentration at elution Detergents (see next slide) Organic solvents (see next slide) Alternatives Arginine solutions >0.5 mol L -1 UV light (254 nm) o Might cause protein-dna aggregates Gamma irradiation (>25 kgy) o Access to source and equipment 22

23 LRF Virus inactivation with detergents (FP1) UF/DF DepF Triton AIEX HIC UF/DF CHT CIEX VF UF/DF Detergent (e.g. 1 % Triton X-100, 0.3 % TNBP, 1 % Tween 80) for more than 4 h, at RT o Tend to oxidation and byproducts o permitted residual level 3.25 ppm removal in subsequent steps 6 5 1% Triton X-100 at 20 C Comparison of 2 experiments Reproducible results [min] XMuLV XMuLV PRV PRV 23

24 LRF proper oligomer [%] [min] [min] activity [%] [min] [min] Virus inactivation with organic solvent (Enz2) Variables: Conc, T, t 15 C 27 C Some effect on activity Low impact on oligomer Higher T more effective 180 min are sufficient 14,5% 2-propanol, 22 C 8 XMuLV 7 PRV 6 XMuLV at 18 C propanol [%] 20 40,0-60,0 60,0-80,0 80,0-100, propanol [%] 20 70,0-75,0 75,0-80,0 80,0-85, [min] 10 2-propanol [%] propanol [%] 20 92,5-93,0 93,0-93,5 93,5-94,0 91,0-92,0 92,0-93,0 93,0-94,0 24

25 Agenda 1 Introduction Process related impurities Product related impurities Virus inactivation Summary and conclusion 25

26 Summary modular approach General principles o Identify major impurities in harvest select suitable analytic method o Assess molecular properties o Find specific capture with reasonable capacity Frontload stability studies (ph, mechanical, temperature, proteases) o Utilize that knowledge for definition of virus safety strategy HCP issues o Reduce HCP levels as early on otherwise reduction of column capacity o Think about alternative methods (e.g. caprylic acid precipitation) Aggregation o Identify underlying cause, eliminate according to properties (charge, hydrophobicity, size ) Product related impurities o Eliminate first mis-assembled variant, then aggregates and truncations 26

27 References Kipriyanov et al., JMB, (1999), 293, Data from our process development and production teams We are hiring! Scientist: Process Science DSP Scientist: Process Science USP Senior Scientist: Cell-Line Development Director: BioProcess Science Director: Process Development DSP Director: Process Unit DSP Production More than 50 open positions, see under: Or contact me: 27