Multilevel Characterization of Biotherapeuticsusing CESI ... · HC LC-HC 219 SCDK 222 1 LC LC-HC...

Darmstadt, October 1th, 2015

Multilevel Characterization of Biotherapeutics using CESI-MS:

from Intact Protein to Peptide Mapping Approach

Yannis-Nicolas François

Laboratory of Mass Spectrometry of Interactions and System, University of Strasbourg, France

CESI User Meeting

Separation in capillary electrophoresis

electrophoretic mobility

electroosmotic mobility

• Analytes are separated depending on their charge and size

• CE provides fast separation

great efficiency

low sample consumption

Advantages of CE-MS

� Great efficiency

� Selectivity

� Sensitivity

� Structural information

� Ultra-low flow rate

CE-ESI-MS Coupling

Advantages of CE-MS

� Great efficiency

� Ultra-low flow rate

CE-ESI-MS Coupling

Minimize ion suppression2

Maximize ionization efficiency

Increase sensitivity1

1Wilm, Mann International Journal of Mass Spectrometry 1994, 136, 167–1802Schmidt, Karas, Dulcks , J Am Soc Mass Spectrom 2003, 14, 492–500

1. Monoclonal antibodies primary structure characterisation by CESI-MS

2. Biosimilarity assessment by CESI-MS

3. Antibody Drug Conjugate characterization by CESI-MS

4. Glycoform Separation and Characterisation of Cetuximab Variants by Middle-up Off-line CE-UV/ESI-MS

Content

CESI-MS coupling

• CESI-MS allows to be operated using nano flowrates

Favorable to ESI ionization

• CESI-MS showed improved sensitivity compared

to sheath liquid interface

� Faserl et al., Anal. Chem. 2011, 83, 7297-7305

� Busnel et al., Anal. Chem. 2010, 82, 9476-9483

Diagram and picture of the CESI interface

mAbs characterisation by CESI-MS/MS

• Monoclonal antibody (mAb) therapeutics attract the most interest due to their strong therapeutic potency.

� In 2015, over than 43 are marketed

� more than 30 mAbs in clinical trial phase III

• mAbs specificity for its antigen opens new avenues for therapeutic treatments

� oncology

� autoimmune diseases

� Transplant rejection prevention

• mAbs are complex and heterogeneous glycoproteins representing a challenge to analytical sciences

� Characterization on different level of the mAbs

� Necessity of precise and high throughput characterization

Zhang Z. et al., Mass Spec. Rev., 2009 (28), 147-176

A. Beck et al., Anal. Chem. 2012, 84, 4637-4646

Average mass: 148,057 Da (1,328 a.a.)

LC : -N30T – (D/isoD, +1 Da)

HC : -N55T – (D/isoD, +1 Da)

HC : -N387T – (D/isoD, +1 Da)

Monoclonal AntibodyTrastuzumab

mAbs characterisation workflow

In-solution tryptic digestionAnalysis by

t-ITP CESI-MS/MS

Amino acid sequence

characterisation

PTMs hot spots

characterisation

Glycosylations

(structure)

CESI8000 coupled to 5600 TripleTOF MS

• Primary structure characterisation workflow based on bottom-up proteomics strategy

Amino acid sequence characterisation

• MS/MS amino acid sequence characterisation (trastuzumab)

100% sequence coverage achieved in a single injection through

only purely tryptic unmodified peptides

Gahoual R. et al., Anal. Chem., 2014 (86), 9074-9081

variable domain

complementarity determining region

constant domain

identified peptides


MS/MS spectrum of digested peptides LT04

APK

(m/z 315.2039 ; 2+)

MS/MS spectrum of digested peptides HT15

DYFPEPVTVSWNSGALTSGVHTFPAVLQS

SGLYSLSSVVTVPSSSLGTQTYICNVNHKP

SNTK

(63 amino acids ; m/z 1119.898 ; 6+)

Implementation of CE allows separation and successful detection of a larger variety of peptides

A. Beck et al., Anal. Chem. 2012, 84, 4637-4646

Average mass: 148,057 Da (1,328 a.a.)

LC : -N30T – (D/isoD, +1 Da)

HC : -N55T – (D/isoD, +1 Da)

HC : -N387T – (D/isoD, +1 Da)

Trastuzumab (Herceptin)

Detection of 15

different glycosylations

(MS)

9 structures

characterized (MS/MS)

Glycosylations characterization

• mAbs glycosylations are characterized simultaneously using the same CESI-MS/MS data

Gahoual R. et al., mAbs, 2013 (5), 479-490

• Glycopeptides MS signal intensity used to estimate glycoforms relative abundances

15 different glycoforms identified in trastuzumab case

Possibility to detect weakly abundant glycosylation

Glycosylations characterization


PTMs hot spots characterization

N-terminal glutamic acid cyclization characterization

• CE mechanism separates of peptide with N-terminal glutamic acid cyclization from the

unmodified peptide

Results suggest partial

modification of sample

Favorable conditions to estimate

sample modification level

HT01

pyroglu - HT01

Extracted ion electropherograms of peptides HT01 and modified HT01



Methionine oxidation

• Methionine oxidation causes peptide mass shift (+15.99 Da) leading to the separation of the

modified peptide in CZE

confirmed by MS/MS spectra


Methionine (M)

methionine

sulfoxide (oxiM)

EIEs and MS/MS spectra of peptides HT21 (intact and modified)


Asparagine (N) aspartic acid (deaN)

• Deamidation (+ 0.98 Da) involves mobility change in CZE enabling the separation of the unmodified peptide

CE separation of deamidated peptides eases the

identification of the modification by MS


EIEs and MS/MS spectra of peptides LT04 (intact and modified)

Asparagine deamidation


Aspartic acid isomerization

CE separation prior to MS analysis allows in this particular case to include

aspartic acid isomerization in the overall characterization workflow

HT23 (-D283-)

HT23 (-isoD283-)

EIEs and MS/MS spectra of peptides HT23 (intact and modified)


Aspartic acid separation in CZE

• Aspartic acid isomers separation by CZE confirmed using a synthetic peptide

NH2-GLEWIGYISY D GTNNYKPSLK-OH

NH2-GLEWIGYISY isoD GTNNYKPSLK-OH

Synthetic peptides sequences :





Content

mAbs biosimilarity assessment

• As several mAbs patent are ending in the next few months/years, other companies should have the

possibility to commercialize « unprotected » mAbs

• mAbs complexity and production process (cell line selection) makes it nearly impossible to produce

strictly the same product as the innovator company

• FDA and EMA are introducing guidelines to help biopharma companies to determine the key features

needed for a biosimilarity between two products in term of structure, PK and PD => reducing clinical

trials

Guidelinesbiosimilars

approval

EMA, CHMP/437/04

EMA, CHMP/437/04 Rev 1

2nd case

cetuximab vs.candidatebiosimilar

Amino acid sequence similarity

RGAC

D. Ayoub et al., mAbs 2013, 5, 699-710

• A single analysis of each sample sufficient to

conclude on the complete similarity regarding

AA sequence

• Complete sequence coverage is obtained

through peptides without miscleavages or

PTMs

• CESI-MS/MS enabled to confirm an error,

recently reported in the litterature


Glycoforms characterisation

• Fc/2 glycosylation site characterisation

Heterogenous glycoforms

could be identified

Difference in glycoforms

distribution could be observed

Cetuximab possess two different

N-glycosylation sites

Significant number of glycans

could be characterized



• Fd glycosylation site characterisation

Glycoforms exhibited by the candidate biosimilar are significantly different from cetuximab

Rejected as biosimilar� 30 % of glycans contains N-acetylneuraminic acid






Content

ADC characterisation by CESI-MS/MS

• Antibody drug conjugates (ADCs): New class of biopharmaceutical drugs for cancer treatment

http://adcreview.com/knowledge-center/adcs-101/how-

do-antibody-drug-conjugates-work/


• Antibody drug conjugates (ADCs): New class of biopharmaceutical drugs for cancer treatment

• Cysteine-linked ADCs

http://adcreview.com/knowledge-center/adcs-101/how-

do-antibody-drug-conjugates-work/

Drug

Monoisotopic mass of drug: 1316.7869 Da

MMAE: 716,4962 (+ 2H)

Chain Disulfide interchain Séquences Nb of vcMMAE

HC LC-HC 219SCDK222 1

LC LC-HC 212GEC214 1

HC HC-HC 223THTCPPCPAPELLGGPSVFLFPPKPK248 1

HC HC-HC 223THTCPPCPAPELLGGPSVFLFPPKPK248 2

Objectives :

A. Characterization of intact ADCs: Determination of average DAR et DAR Distribution

In native conditionNanoESI infusion using CESI8000

B. Peptide mapping:

Identify

- Drug conjugate peptides

- Amino acid sequence

- PTMs: N-glycosylation, deamidation, oxidation, cyclization…


7.38

0

8

0 ==

∑

∑

=

=

ndrug

nDrug

A

nADAR

0-Drug

Deconvoluted Mass spectra

Characterisation of Intact ADCs

• DAR Distribution

Raw MS Signal

• Average DAR

1-Drug

2-Drugs

3-Drugs

4-Drugs

6-Drugs

5-Drugs 7-Drugs 8-Drugs

2637 2638 2636 2639

DIVLTQSPASLAVSLGQR ATISCK

ASQSVDFDGDSYMNWYQQK PGQPPK

VLIYAASNLESGIPAR

FSGSGSGTDFTLNIHPVEEEDAATYYCQQSNED

PWTFGGGTK LEIKR TVAAPSVFIFPPSDEQLK

SGTASVVCLLNNFYPR EAK VQWK

VDNALQSGNSQESVTEQDSK

DSTYSLSSTLTLSK ADYEK HK

VYACEVTHQGLSSPVTK SFNRGEC

QIQLQQSGPEVVK PGASVK ISCK ASGYTFTDYYITWVK

QK PGQGLEWIGWIYPGSGNTK YNEK FK GK

ATLTVDTSSSTAFMQLSSLTSEDTAVYFCANYGNYWFAYWG

QGTQVTVSAASTK GPSVFPLAPSSK STSGGTAALGCLVK

DYFPEPVTVSWNSGALTSGVHTFPAVLQSSGLYSLSSVVTVP

SSSLGTQTYICNVNHK PSNTK VDK KVEPK SCDK

THTCPPCPAPELLGGPSVFLFPPK PK DTLMISR

TPEVTCVVVDVSHEDPEVK FNWYVDGVEVHNAK TK PR

EEQYNSTYR VVSVLTVLHQDWLNGK EYK CK VSNK

ALPAPIEK TISK AK GQPR EPQVYTLPPSR DELTK

NQVSLTCLVK GFYPSDIAVEWESNGQPENNYK

TTPPVLDSDGSFFLYSK LTVDK SR WQQG

NVFSCSVMHEALHNHYTQK SLSLSPG


• MS/MS amino acid sequence characterisation

94% sequence coverage achieved in a single injection using tryptic digestion

N. Said et al. Anal. Chem. 2015, Submitted



VLIYAASNLESGIPAR





















• MS/MS amino acid sequence characterisation

98,8% sequence coverage achieved in a single injection using chymotryptic digestion




VLIYAASNLESGIPAR




















Drug characterization

Detection and characterization of the 4 modified peptides using MS/MS data


Conjugate peptide: LC-HC interchain 208SFNRGEC214

[M+3H]3+710,0489

709,8 709,9 710,0 710,1 710,2 710,3 710,4 710,5 710,6 710,7 710,8 710,9 711,0 711,1 711,2 711,3 711,4 711,5 711,6 711,7 711,8 711,9 712,0Mass/Charge, Da

0

710,3836

710,7173

711,0522

710 720 730 740 750 760 770 780 790 800 810 820 830 840 850 860 870 880 890 900 910 920 930 940 950 960 970 980 990 1000 1010 1020 1030 1040 1050 10600

100%

1065,0620710,0489

[M+2H]2+

1064,0 1064,2 1064,4 1064,6 1064,8 1065,0 1065,2 1065,4 1065,6 1065,8 1066,0 1066,2 1066,4 1066,6 1066,8 1067,0 1067,2 1067,4Mass/Charge, Da

0

1065,06201064,5597

1065,5624

1066,06281066,06281066,06281066,0628


150 200 250 300 350 400 450 500 550 600 650 700 750 800 850 900 950 1000 1050 1100 1150 1200 1250Mass/Charge, Da

0

100% 686,4849

106,0638

1104,4784

718,5126 1121,5014 1261,55721244,53421218,5530506,3556

152,1048538,7443

246,1219134,0941 687,4867321,2148 443,2894

730,4783691,31261005,4146605,4266 1105,4937

386,1779263,1473 1076,4844 1122,50471219,5557505,2475261,1572 1004,4272539,2417 636,4489668,4739

170,1160 788,3295502,3309

Drug fragmentation

Extracted ion electropherogram

MS spectra

MS/MS spectra: precursor 710.05 Da

Conjugate peptide: LC-HC interchain 219SCDK222


226,5 226,6 226,7 226,8 226,9 227,0 227,1 227,2 227,3 227,4 227,5 227,6 227,7 227,8 227,9 228,0 228,1

Mass/Charge, Da

0,0e0

2,0e3

4,0e3

6,0e3

8,0e3

1,0e4

1,2e4

1,4e4

1,6e4

1,8e4

2,0e4

2,2e4

2,4e4

2,6e4

2,8e4

3,0e4

3,2e4

3,4e4

3,6e4

3,8e4

4,0e4

4,2e4

4,4e4

4,6e4

Inte

nsity

226,5768

227,0786

227,5774

884,4 884,5 884,6 884,7 884,8 884,9 885,0 885,1 885,2 885,3 885,4 885,5 885,6 885,7 885,8 885,9 886,0 886,1 886,2 886,3 886,4 886,5 886,6 886,7 886,8 886,9 887,0 887,1 887,2

Mass/Charge, Da

0,0e0

5,0e3

1,0e4

1,5e4

2,0e4

2,5e4

3,0e4

3,5e4

4,0e4

4,5e4

5,0e4

5,5e4

6,0e4

6,5e4

Inte

nsity

884,9568

884,4556

885,4582

885,9588

886,4611

CE separation of drug peptides eases the

identification of the modification by MS

SCDK SCDK






Content

Introduction



Intact protein analysis

Introduction

� 400 mM ε-amino-caproic acid pH 5.7

� Triethylenetetramine as additive

Not compatible with ESI-MS detection

• Classical condition described in the literature1,2

1He et al, Anal. Chem. 2010, 82, 3222-32302Gassner et al, Electrophoresis 2013, 34, 2718-2724



Intact protein analysis

Off-line coupling strategy

On-line Off-line

Fraction collection

CE Separation

Minutes

2 4 6 8 10 12 14 16 18 20 22

AU

-0,005

0,000

0,005

0,010

0,015

0,020

0,025

AU

-0,005

0,000

0,005

0,010

0,015

0,020

0,025

UV - 200nm

Ins + Lyso + Myo + Ribo 20 uM

PACE MDQ

Beckman CoulterProteineer FC

Bruker

CESI8000 - MS

M. Biacchi et al. Electrophoresis 2014, 35 (20), 2986-2995

MS

Analysis by

CE-UV/Fraction collection

mAbs characterisation workflow

IdeS digestion

Fc/2 variant

characterization

F(ab’)2 characterization

Analysis by

NanoESI infusion

Characterisation of FC/2 variants

M. Biacchi et al. Anal. Chem. 2015, 87, 6240-6250

� Characterization of 7 Fc/2

glycoforms.

� Separation due to C-terminal

lysine truncation

Peak 1 Fc/2

Peak 3 Fc/2-K

Peak 2: Mixture Fc/2 and Fc/2-K

Characterisation of FC/2 variants

M. Biacchi et al. Anal. Chem. 2015, Submitted

� Characterization of 7 Fc/2

glycoforms.

� Separation due to C-terminal

lysine truncation

Peak 1 Fc/2

Peak 3 Fc/2-K Heterodimer

Fc/2/Fc/2-K

Homodimer

Fc/2-K

Homodimer

Fc/2

Separation of Fc/2 aggregates

Characterisation of F(ab’)2 variants

M. Biacchi et al. Anal. Chem. 2015, 87, 6240-6250

� Characterization of 8 F(ab’)2

glycoforms.

� Separation of F(ab’)2 glycoforms

Separation based on the

presence of sialic acid

Conclusion

• CESI-MS/MS allowed to conclude in each case on the biosimilarity assessment

• Charaterisation and Glycoform Separation of Cetuximab Variants by Middle-up Off-line CE-UV/ESI-MS

� Separation and characterisation of FC/2 variants (FC/2 and FC/2–K)

� Separation and characterisation of glycoform F(ab’)2 variants (presence of sialic acid)

• Monoclonal antibodies primary structure characterisation by CESI-MS

� 100% amino acid sequence characterisation

� 15 glycoforms characterisation

� All PTMs hot spots characterisation

• ADC characterisation by CESI-MS

� Average DAR measurement

� 98.8% amino acid sequence characterisation

� MS/MS characteirsation of the 4 modified peptides

Acknowledgments

Emmanuelle Leize-Wagner

Rabah Gahoual

Michael Biacchi

Nassur Said

Philippe Hammann

Lauriane Kuhn

Philippe Wolff

Johanna Chicher

Laboratory of Mass spectrometry of

Interactions and Systems (LSMIS)

Plateforme Protéomique

Strasbourg Esplanade

Centre d’Immunologie Pierre Fabre

Alain Beck

Elsa Wagner-Rousset

Marie-Claire Janin-Bussat

Daniel Ayoub

Olivier Colas

Jean-Marc Busnel

Acknowledgments

Steve Lock

Jim Thorn

Milla Neffling

Anton Heemskerk

Michel Anselme

Thank you for your attention“La petite France” - Strasbourg

mAbs characterisation by CESI-MS/MS

The t-ITP CESI-MS/MS method developed demonstrated its robustness

on different samples including technical replicates in each case

Results summary obtained with the t-ITP CESI-MS/MS method


59

Glycoforms abundances method

• Glycosylations relative quantification was performed using maximum intensities or peak area

Similar results with both

methodologies

Maximum intensity was selected for

the quantification

58

Glycoforms relative abundances

• Glycosylations relative abundances estimated from the CESI-MS/MS data were confronted to other

techniques if data were available

40.81

37.81

9.04

7.19

1.64 1.19 0.85 0.74 0.72

0

5

10

15

20

25

30

35

40

45

H3N4F1 (G0F) H4N4F1 (G1F) H5N2 (Man5) H5N4F1 (G2F) H4N4 (G1) H3N3F1 (G0F-

GlcNac)

H6N4F1 H3N4 (G0) H4N3F1 (G1F-

GlcNac)

D. Ayoub et al., mAbs 2013, 5, 699-710

44


position sequence

trastuzumab biosimilar

unmodif modif unmodif modif

N-term glutamic acid cyclization

1 -19 EVQLVESGGGLVQPGGSLR 98.2 1.8 97.3 2.7

Asparagine deamidation

51 - 59 IYPTNGYTR 89.4 10.6 92.3 7.7

374 - 395 GFYPSDIAVEWESNGQPENNYK 85.8 14.3 100.0 0.0

25 - 42 ASQDVNTAVAWYQQKPGK 96.1 3.9 96.0 4.0

Methionine oxidation

252 - 258 DTLMISR 95.3 4.7 94.5 5.5

420 - 442 WQQGNVFSCSVMHEALHNHYTQK 97.8 2.2 95.2 4.8

Aspartic acid isomerization

99 - 124 WGGDGFYAMDYWGQGTLVTVSSASTK 91.4 8.6 86.0 14.0

278 - 291 FNWYVDGVEVHNAK 92.0 8.0 96.5 3.5

396 - 412 TTPPVLDSDGSFFLYSK 75.3 24.7 93.9 6.1

Gahoual R. et al., mAbs 2014, in press

1st case

trastuzumab vs.candidatebiosimilar


• Complete sequence coverage obtained

for trastuzumab

• Biosimilar candidate sequence could be

successfully identified except HC K217

Suggesting an amino acid

substitution between the two samples

Gahoual R. et al., mAbs 2014, 6, 1464-1473


Interpretation of unidentified MS/MS spectra

Unambiguous characterisation of the amino

acid substitution of biosimilar candidate

CESI-MS/MS spectra of trastuzumab biosimilar candidate

V D K R217 V E P K

rejected candidate



• Glycosylation distribution evaluated for each sample using CESI-MS/MS data

Identification of a significant

number of glycoforms

Minor differences of glycoforms

could be distinguished


56

Poor crystallization = No MALDI-MS signal

CE-MALDI

Biacchi et al., Anal. Chem., 2015 submitted

Multilevel Characterization of Biotherapeuticsusing CESI ... · HC LC-HC 219 SCDK 222 1 LC LC-HC...

Documents

Transcript of Multilevel Characterization of Biotherapeuticsusing CESI ... · HC LC-HC 219 SCDK 222 1 LC LC-HC...