Deamidation specific CZE methods for characterization and quality control of biopharmaceuticals

Transcription

1 Deamidation specific CZE methods for characterization and quality control of biopharmaceuticals Frank Moffatt, Alena Böhmova & Maria A. Schwarz Amazing where you can go

2 Well Characterized Biologics (?) Complex Heterogeneous Impure Unstable 2

3 Asparagine Post-translational modification by glycosylation Racemization D/L D-Asp, D-imide, D-isoAsp Deamidation Asn Asp Deamidation/Isomerization isoasp Succinimide formation dimerization 3

4 Deamidation High ph Asn Asp & isoasp & D- amino acids Low ph Asn Asp Sequence dependent (Standing & Robinson) Stability AsnVal & AsnIle >>> AsnGly Secondary structure (Topp) same sequence peptide protein Conserved regions of IgG (Chelius) SNG, ENN, LNG, & LNN GNT, TNY, YNP, WNS, SNF, CNV, SNT, WNS, FNW, HNA, FNS, SNK, GNV, HNH, SNY, LNW, SNL, NNF, DNA, GNS, & FNR 4

5 Monoclonal Antibody Structural Features & Stability Deamidation "Big 4 stability issues" 1. Oxidation 2. Deamidation 3. Fragmentation 4. Aggregation Asn (esp.) & Gln unstable under physiological conditions Natural human antibodies 23% deamidation in conserved (Fc) 384 (Liu) n Aggregation 5

6 IgG MAb Whole-Column-Imaged Capillary Isoelectric Focusing Whole Sialidase (- sialic acid) Carboxypeptidase (- C-terminal lysine) Both 7.5 pi 9.2 Resolution - good Site specific information - none Hsieh, How Useful is the Whole-Column-Imaging Capillary Isoelectric focusing (CIEF) in the Detection of Different Modifications of Monoclonal Antibodies, Poster CEPharm06 6

7 Deamidation changes pi & net charge Deamidation 7

8 Forced Deamidation mab C cief Deamidation Increase in proportion of acid variants Unknown Number of deamidation sites Location of deamidation sites stressed (forced deamidation) unstressed 8

9 Fragmented Antibody CE FAb-CE SM 9

10 mab A - cief versus FAb-CE SM Aggregates & Acidic Variants F c F ab Papain digestion higher resolution than to intact MAb no interferences due to aggregates localization of the charge variants av of F c bv of F c bv of Fab av of F ab digested aggregates intact acidic variants (av) pi basic variants (bv) 10

11 Advantages of FAb-CE SM Information about deamidation Stability indicating Avoids aggregation complications Fast Quantitative 11

12 FAb-CE SM Summary Useful for stability, comparability or biosimilarity Resolves more charge variants than cief Changes are located on Fc or Fab Rational evaluation of biological consequences Cell based bioassay for changes in Fc ELISA for changes in Fab Easily integrated into standard work-flow 12

13 Evidence for Deamidation in Papain digest Next steps Confirm & quantify deamidation 13

14 SDPM SM Specific deamidation peptide mapping 1. Peptide mapping using endoproteinase Asp-N 2. Designed to provide maximum information about deamidation 3. Standard work-flow to minimize timelines and risks 14

15 Ideal match between issue & enzyme Asp-N Asp-N R R' R R' or isoasp or Asp from deamidation AspN versus trypsin Lower optimum ph 6-8 fewer artifacts than trypsin (ph 8-9) (Simpson) Fewer peptides Natural abundance Lys 8 % Arg 5 % Asp 5 % 15

16 Influences upon Complexity number of aspartic acids (number of total digestion fragments) ratio or aspartic acid to asparagine number of deamidation sites probability of deamidation 16

17 Ribonuclease A - AspN Peptide Map P4 DVQAVCSQKNVACKNGQTNCYQSYSTMSIT [53-82] Asn Asn Asn Asn IsoAsp Asn 1 Site deamidation Theor. max. 3 new peptides 1 new isoasp peptide 2 new fragment peptides Asn IsoAsp IsoAsp 2 Site deamidation Theor. max. 8 new peptides 1 new isoaspx2 peptide 2 new isoasp peptides 4 new single cleavage peptides 1 new double cleavage peptides 17

18 Asp-N Peptides by CZE Predictable effect upon electrophoretic mobility Asn isoasp always reduces mobility of peptide Fragments depend upon size/charge 18

19 Asp-N digestion of mab D Stressed versus unstressed no additional peaks unstressed no reduced signals same CZE profiles stressed, 7d, 37 C Conclusions no deamidation no further investigation 19

20 CZE-separation of a therapeutic protein mab E Identification of all peptides CZE-MS Confirmed by spiking experiments calculation of apparent mobility (z/r H ) 20

21 CZE-separation of therapeutic protein mab E forced deamidation p3 Identification of peptides CZE-MS dp3-3 dp3-2 iso-p4 dp4-1,2 dp3-1 p(5+8) 21

22 MAb Fab fragment Deamidation facilitated by heat Unmodified peptides Modifications found P3 D N N' D D D' D N D iso 4/8 P4 D N D D D iso D 3/3 P8 D N D D D D iso 3/3 P10 D N D 1/3 22

23 Work in progress 1. Taming the work-flow - complexity often challenging Overlapping peaks Missing peptides Unidentified peptides 2. Improving the specificity of Asp-N 3. Analogous use of Glu-C for Gln deamidation 23

24 Calculated deamidation factor sum parameter for all deamidation sites Detection range of deamidation peptides 24

25 Deamidation factor Calculation (1) Calculate A% of peptides unaffected by deamidation (identification of the protein). Comparable results obtained for test sample and reference sample (SST). (2) Define one unaffected peak as internal standard (3) Calculate A% of the internal standard including the deamidation peaks (all peaks in red marked regions). (4) Calculate the A% of the internal standard in both test sample and reference sample. (5) Calculate the deamidation factor as the ratio of the A%(ISTD) of the reference sample and the test sample. 25

26 Summary of Asp-N digest Avoids aggregation complications High selectivity Simpler than tryptic digest Fast data processing State of potential deamidation sites Determination of Asp & isoasp Applicable for pegylated proteins 26

27 Work-flow 27

28 Integration Strategy BOTTOM-UP Characterization & stability indicating assays ISO9001 (2008) Perform SDPM SM along with tryptic peptide mapping Does it provide valuable information conveniently? Do site specific changes correlate with a whole protein method? TOP-DOWN QC Testing/Stability validate methods cgmp Is deamidation a critical quality attribute? Can we use a whole protein method? Can we use FAb-CE SM? If not use a peptide map (SDPM SM ) with UV absorbance detection 28

29 Acknowledgments - Novartis Biologics providing therapeutic proteins Aalen University (Germany) Prof. C. Neusüss CZE-MS

30 "You can always make it better" Sunset in Basel Dr. Frank Moffatt, FRSC, MRQA Business Development Tel