native and im-ms characterization of adcs - waters.com · native and im-ms characterization of adcs...
TRANSCRIPT
Native and IM-MS
characterization of ADCs
Alain VAN DORSSELAER
BioOrganic Mass Spectrometry Laboratory, IPHC
University of Strasbourg, CNRS
Strasbourg - France
1 ASMS 2015 Waters User’s meeting
mAbs as therapeutics
mAbs : a very important class of human therapeutics - > 50 mAbs & derivatives currently approved (EMA/FDA)
- indications: inflammation, autoimmunity, cancer, transplantation,…
- > 400 mAbs & derivatives in pre-clinical development and clinical trial
anti-cancer mAbs in clinical studies - 84 unmodified IgG (51%)
- 10 bispecific (6%)
- >35 ADC (15%)
- 17 engineered (10%)
- 16 fragments (10%)
mAbs are complex therapeutic proteins
(150 kDa glycoproteins)
Next generation empowered mAbs :
Bispecific mAbs (bsAbs) and Antibody-Drug-Conjugates (ADCs)
mAb + cytotoxic Double-barreled shotgun:
two antigens targetted
A single shot gun
A highly
cytotoxic drug IgG4
(natalizumab)
IgG4
(6F4-2)
+ GSH
Bispecific mAb (bsAb)
(natalizumab/6F4-2)
+
2 different half mAbs
Intact mAb
analysis
(150 kDa)
Fragment analysis
(25-100 kDa) Peptide mapping
(700-7000 Da)
Higher Order Structure
Aggregates
mAb/Ag complexes
(>150 kDa)
Size
mAb characterization is a nightmare
Native IM-MS Top-down MS
(FT-ICR + Orbital trap)
LC-MS
(nano)LC-MS/MS
enzymatic digestion Native MS
HDX-MS Crosslinking MS
The MS
toolbox
4
1991
Native
MS
2006
Ion Mobility MS
2012
High
resolution
native MS
Native MS: a brief history
2000
Automated
nanoESI
2002
Haemocyanins
2.3 MDa
Katta et al. JACS 1991
1991
Holo-myoglobine
17.5 kDa
McKay et al. JACS 2006
2006
Ribosomes
2.3 MDa
2000 4000 6000 8000 10000 12000 14000 16000 18000 20000 22000 24000m/z0
100
%
x20x1010474.19
10236.41
10008.84
4422.33
4173.91
3949.99
9791.424696.49
14088.06
13870.68
10723.23 13659.90
13455.13
10982.64 13255.21
14313.58
16702.72
16496.93
16298.87
16101.28
15914.78
16912.72
17126.64
17349.11
19404.00
18990.11
18796.87
19615.76
19830.56
20053.1921493.64
2000 4000 6000 8000 10000 12000 14000 16000 18000 20000 22000 24000m/z0
100
%
x20x1010474.19
10236.41
10008.84
4422.33
4173.91
3949.99
9791.424696.49
14088.06
13870.68
10723.23 13659.90
13455.13
10982.64 13255.21
14313.58
16702.72
16496.93
16298.87
16101.28
15914.78
16912.72
17126.64
17349.11
19404.00
18990.11
18796.87
19615.76
19830.56
20053.1921493.64
Sanglier et al. JASMS 2002
2013
Bacteriophage HK97
capsid
18 MDa
Increase in size,
nb of partners
and heterogeneity
Binding
Conformation
1996
nanoESI
Snijder et al. Angew. Chem 2013
Synapt
G1
Synapt
G2
Native MS analysis
Native MS gives information on assemblies maintained
by noncovalent interactions
ESI-MS
instrument
3000 3500 m/z
Gas phase Solution
Sample preparation step
= desalting step Instrumental setting
optimizations
- Buffer exhange
- Use of volatile aqueous
buffer at neutral pH
- Ammonium buffer
(0-1M) with controled pH
- Control of the energy
communicated to the
ions in the 1st pumping
stage region of the
instrument
(Vc, Pi adjustment)
Data interpretation
- Binding stoichiometries
- Binding specificity
- Solution affinities
- Dynamics of
assembly/disassembly
Be carefull !
Native Mass Spectrometry: a recursive question
6
We dream to get from Native MS and IM-MS experiments :
• Accurate mass measurements, stoichiometries,…
• Information on interactions forces
• Subunit spatial arrangement, structural information, shape, conformation changes,…
… but all this information concerns molecules in water solution
Toutankhamon
1- Electrospray mecamism
2- Evaporation in interface
Dried, in vacuo
Specific
non-covalent interactions
salts
In water solution
?
Native MS workflow for mAb analysis
For mAbs minimal and easy sample preparation are well descibed
Purified
mAb, bsAb or ADC
Deglycosylation Step
(PNGase F, IgZero, etc.)
Data Interpretation
144190
Mass
143700 144000 144300 144600
Mass
146700 147200 147700 148200
147239
147077 147401
147563
PNGase-FG0F/G0F
G1F/G0F
G1F/G1F
G2F/G0F
G2F/G1F
144190
Mass
143700 144000 144300 144600
Mass
146700 147200 147700 148200
147239
147077 147401
147563
PNGase-FG0F/G0F
G1F/G0F
G1F/G1F
G2F/G0F
G2F/G1F
If necessary
Buffer Exchange (spin columns, microconcentrators,
SEC etc.)
Native MS and IM-MS (with ESI, needles or Chip-based introduction)
Synapt G2 HDMS (Waters)
+ nanoMate Triversa (Advion)
How do we jumped into the mAb field?
8
- mAb native MS
- mAb/Ag binding stoichiometries
- Native IM-MS
2008-2012: Reference Monoclonal Antibodies
(mAb)
- Native MS for monitoring of bsAb formation
- IM-MS for bsAb
- Real-time IM-MS to monitor bsAb formation
2013: towards bispecific mAbs
(bsAbs)
- Native MS mandatory for ADCs’ analysis
- Native MS for drug load profiles, average
DAR
- IM-MS for ADC heterogeneity assessment
- IM-MS for drug load profiles, average DAR
2014: immunoconjugate analysis
(ADC)
1. Native MS and IM-MS
for reference mAb analysis
9
2. Native MS and IM-MS
for bsAb formation
3. Native MS and IM-MS
for ADC formation
Intact mAb analysis : benefits of native MS for intact mAb analysis
trastu 5uM ACNH4 150mM pH7.5 Viva6 120v 6mbar
m/z1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 6000 6500 7000 7500 8000 8500 9000 9500 10000
%
0
100
6000 5000 4000 3000 2000 m/z 9000 7000 8000
25+
dimer
3%
Trastu IgGZero denat 2uM
m/z600 800 1000 1200 1400 1600 1800 2000 2200 2400 2600 2800 3000 3200 3400 3600 3800 4000 4200 4400 4600 4800
%
0
100
3000 2500 2000 1500 1000 m/z 4500 3500 4000
45+
Denaturing Conditions mAb 2µM in H2O:ACN:FA (50:50:1)
Native Conditions mAb 5µM in AcNH4 150mM pH7.5
Full MS spectrum
Deconvoluted
mass spectrum
Deconvoluted
mass spectrum
Full MS spectrum
Beck et al. J Mass Spectrom. 2015 Feb;50(2):285-97.
In Native conditions
sensitivity and mass accuracy
are as good as
in Denaturing conditions.
But spectra are more simple,
and you see interactions !!!
Non covalent dimer
not detected
Native MS analysis of immune mAb/Antigen complexes
Direct determination of mAb/Ag binding stoichiometries
m/z5000 5200 5400 5600 5800 6000 6200 6400 6600 6800 7000 7200 7400 7600 7800 8000 8200 8400 8600 8800 9000 9200 9400 9600 9800
%
0
100
mass130000 140000 150000 160000 170000 180000 190000 200000 210000 220000 230000 240000 250000 260000
%
0
100
147099 5 Da
23+
6000 7000 8000 9000 m/z 0
%
100
kDa 150 200 250
0
%
100
mAb (5 µM)
m/z5000 5200 5400 5600 5800 6000 6200 6400 6600 6800 7000 7200 7400 7600 7800 8000 8200 8400 8600 8800 9000 9200 9400 9600 9800
%
0
100
mass130000 140000 150000 160000 170000 180000 190000 200000 210000 220000 230000 240000 250000 260000
%
0
100
220878 15 Da
245485 15 Da
+3 x 24.6 kDa
+4 x 24.6 kDa 30+
6000 7000 8000 9000 m/z 0
%
100
kDa 150 200 250
0
%
100
mAb (5 µM)
+
JAM-A (40 µM)
m/z5000 5200 5400 5600 5800 6000 6200 6400 6600 6800 7000 7200 7400 7600 7800 8000 8200 8400 8600 8800 9000 9200 9400 9600 9800
%
0
100
196368 8 Da
171783 15 Da
mass130000 140000 150000 160000 170000 180000 190000 200000 210000 220000 230000 240000 250000 260000
%
0
100
+2 x 24.7 kDa 27+
6000 7000 8000 9000 m/z 0
%
100
kDa 150 200 250
0
%
100
+1 x 24.7 kDa
mAb (5 µM)
+
JAM-A (10 µM)
Atmanene C. et al. Anal Chem. 2009, 81(15):6364-73.
Unexpected 1:4
mAb:Ag stoichiometry
Native Ion Mobility MS (IM-MS) for conformational studies
By IM-MS : simultaneous measurement of drift times and m/z ratios
ESI
source
IM
cell
MS
analyzer Detector
Instrumental setting
optimizations
- Wave height and velocity
- Pressure in the IM cell
Data interpretation
- Conformation of
biomolecules
- Conformational changes
Trap
T-Wave
Transfer
T-waveQuadrupoleIon
Guide
Pusher Detector
Primary
pumpingTurbomolecular pump
INTERFACE
TOF ANALYZER
Reflectron
Source
NanoESI
Ar He ArN2
IMS
Cell
Vanne
ArAr
Drift time
m/z
12
Native Ion Mobility MS (IM-MS) for conformational studies
- Ions separation takes place according to ion mobility
Ion Mobility cell
2+ 2+ 1+
Drift time +
1+ Ion separation according to
ion mobility
Size, shape Charge
Compactness
Drift time
z
Drift time
2+ 1+ 2+ 1+
Drift time
N2
z
x
y
z
x
y
z
x
y
z
x
y
x
y
zx
y
z
L
EtDz
MMTkP
T
N
e
iongasb
112760
15.27316
3
Drift times can be related to collisional cross sections (CCS)
Information on ion gas phase conformation (2D representation of 3D structure) 13
Ion Mobility Mass Spectrometry (IM-MS)
For mAb conformational characterization
SM classique IM-MS
SMSupra IM-MS
Denaturing MS
Native MS Native IM-MS
Denaturing IM-MS
Beck A, Sanglier-Cianférani S, Van Dorsselaer A. Anal Chem. 2012, 84(11):4637-46
IM-MS to probe mAb conformational heterogeneity
2010: the first
separation of
2 mAb isoforms
by IM-MS
reported
Lc214 S-S Hc127
Hc232 S-S Hc232
Hc233 S-S Hc233
Lc214 S-S Hc232
Hc127 S-S Hc233
1. Native MS and IM-MS
for reference mAb analysis
16
2. Native MS and IM-MS
for bsAb formation
3. Native MS and IM-MS
for ADC formation
17
Native MS analysis for analysis for bsAb formed by Fab Arm Exchange
IgG4
(natalizumab)
IgG4
(6F4-2)
+ GSH
Bispecific mAb (bsAb)
(natalizumab/6F4-2)
+
2 mAb features are important for the exchange
reaction of IgG4:
- CH3 domain
- Ser at position 228
In vivo, IgG4s can exchange half molecules by a
dynamic process called Fab-arm exchange (FAE).
In vitro, the FAE process can be mimicked by a
reaction with glutathione (GSH)
Serine 228 in IgG4 introduces flexibility in the core hinge region. Besides the usual disulfide
bonds connecting two heavy chains (covalent between HC), intrachain disulfide bonds may
form instead (only non-covalent bonds between HC).
covalent bonds
between HC
The 2 arms are
only maintained
by non-covalent bonds
between HCs Variable
rate
Native MS and IM-MS for bispecific mAb characterization
Debeane et al. Anal. Chem. 2013, Anal Chem. 2013, 85, 9785-92
IgG4 Fab Arm Exchange monitoring
Sample : Natalizumab and Hz6F4-2 IgG4
bsAb
6F4-2
natalizumab
Native MS
+ GSH
Time resolved
IM-MS
O14611FDE.raw : 1
bsAb
bsAb
O14605FDE.raw : 1
23+
24+
+ GSH
+
Fab arm Exchange
+ GSH
Native MS and IM-MS to monitor:
- Dynamics of Fab Arm Exchange
- Bispecific antibody formation
Debaene F. et al. Anal Chem. 2013 85(20):9785-92.
1. Native MS and IM-MS
for reference mAb analysis
19
2. Native MS and IM-MS
for bsAb formation
3. Native MS and IM-MS
for ADC formation
ADCs are next generation empowered therapeutic mAbs
ADCs are tripartite molecules, made by the chemical conjugation of cytotoxins to mAbs
Cytotoxic drug Linker
*Beck A. and Reichert JM., mAbs 2014
cell-killing
capabilities
+
targeting delivery
of the conjugated
drug to tumor cells
Highly
cytotoxic drug
2 already approved ADCs’ on the market:
- brentuximab vedotin (Adcetris®, Seattle Genetics)
- ado-trastuzumab emtansine (Kadcyla®, Genentech)
- > 30 in clinical trials*
3 main types of conjugation:
- at cysteine -SH groups after reduction of the interchain disulfide bonds (brentuximab vedotin)
- at lysine side chains amine (ado-trastuzumab emtansine)
- at engineered cysteine residues at specific sites without reduction (thiomabs)
20
Drug load profile
Unconjugated D0
Average
DAR
Higher order
structures Ion mobility
MS?
Native
MS ?
- Cristallography
- HDX-MS
- SEC-MS
- Bottom up
(nano)LC-MS/MS
- Hydrophobic Interaction (HIC)
- Ion Exchange (IEC)
- Size exdclusion chromatography
- Reversed phase (rpHPLC)
- CE-SDS
Minutes3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 11.00
D0
D2 D4
D6
D8
- Middle- and Top-down MS
(MS/MS on intact ADCs )
The analytical toolbox for ADC analysis
- Middle-up
LC-MS
8
0
8
0
n
n
DAR
DAR
A
nA
DAR = 4.0
Site of
conjugation
21
Determining the
Drug Antibody Ratio (DAR)
with MS ?
ADCs’ analytical characterization is more challenging
than unconjugated mAbs
Cysteine-conjugation generates heterogeneous population of drug-loads
22
S-S
Reduction
step
Drug
Conjugation
step
Covalent
species
Non-covalent
species
Need for translational complementary analytical techniques
Compared to “unconjugated” mAbs, ADCs have increased level of complexity:
- a mixtures of species with different nb of drug molecules (0, 2, 4, 6, 8) at different positions
- a mixture of covalent and non-covalently associated populations
At least
one inter-chain
S-S bond missing
Cysteines
Our reference Cys-linked ADC: brentuximab vedotin
mAb
Brentuximab
anti-CD30
(chIgG1)
-Cys-SH (Hinge)
Linker
Peptide protease-clivable linker
-Cit-Val-
Drug
antimicrotubule drug
monomethylauristatin E
indicated for treatment of hematological malignancies (Hodgkin lymphoma, systemic
anaplastic large cell lymphoma)
Cysteines
HIC : the gold standard for Cys-linked ADC
Minutes 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 11.00
D4 D2
D6
D0
8
0
8
0
n
n
DAR
DAR
A
nA
DAR = 4.0
3 main ADC quality attributes (QA) can be determined by Hydrophobic Interactions (HIC)
- average DAR
- drug load distribution
- Relative amount of unconjugated mAb (D0)
D8
Cysteines
05-Dec-2013Brentuximab IgZero Viva5 replicat 4 5uM AcNH4 150mM pH7.4 80v 6mbar
m/z500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 6000 6500 7000 7500
%
0
100
m/z500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 6000 6500 7000 7500
%
0
100
6000 7000 m/z1000 3000 500040002000
37+
15+
37+
Full scan mass spectrum
Native MS: mandatory for cysteine-linked ADC analysis
- Very minor ion signals correspond to intact ADC
- Non-covalent interactions are disrupted
in denatured conditions
25
100 80 40 60 20 120 140 Mass (kDa)
145,910 Da
123,500 Da
101,072 Da
75,586 Da
25,042 Da
53,177 Da
MaxEnt deconvoluted mass spectrum
Denatured conditions
IgGZero treated ADC @ 2 µM in H2O:ACN:FA (50:50:1)
Cysteines
Full scan mass spectrum
26
Native conditions
IgGZero treated ADC @ 5 µM in 150 mM AcONH4 pH 7.5
05-Dec-2013Brentuximab IgZero Viva5 replicat 4 5uM AcNH4 150mM pH7.4 80v 6mbar
m/z500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 6000 6500 7000 7500
%
0
100
m/z500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 6000 6500 7000 7500
%
0
100
6000 7000 m/z 1000 3000 5000 4000 2000
Contamination (PEG)
free LC + 1
payload
Clarification/simplification of native mass spectra due
to fewer charges
(increased sensitivity)
Native MS: mandatory for cysteine-linked ADC analysis Cysteines
Intact ADC signals:
D0, D2, D4, D6, D8 ?
MaxEnt deconvoluted mass spectrum
27
Native conditions
IgGZero treated ADC @ 5 µM in 150 mM AcONH4 pH 7.5
mass144000 145000 146000 147000 148000 149000 150000 151000 152000 153000 154000 155000 156000 157000 158000 159000
%
0
100
O19456FDE 24 (1.627) M1 [Ev-62181,It48] (Gs,10.000,5175:6905,10.00,L10,R10); Sm (Mn, 2x2.00); Sb (25,5.00 ); Cm (2:29) TOF MS ES+ 5.05e5
146 Mass (kDa)
150 148 154 152 158 156
D4
D2
D6
D8 D0
MWexp
(Da) mass
(Da)
Error
(ppm)
D0 145903 7.3 50.2
D2 148539 8.1 54.2
D4 151172 5.8 38.3
D6 153814 12.5 81.4
D8 156446 9.2 59.1
- ADC is almost detected as an intact molecule
- Mass measurement affords drug binding stoichiometries to be assessed
- Native MS maintains non-covalent interactions in the gas phase
Native MS: mandatory for cysteine-linked ADC analysis Cysteines
Native MS
profile
HIC
profile
Native MS analysis to assess average DAR
- Native MS data are in good correlation with HIC data for all QA
- Native MS affords in one single run
1) drug load profile determination and 2) average DAR calculation 28
= 4.0 DAR
Comparison between HIC and Native MS for brentuximab vedotin
= 3.9 DAR
Cysteines
Off-line HIC native MS coupling
Native MS allows to confirm mass homogeneity of each HIC fraction :
no mixtures of different drug binding stoichiometries are observed within one HIC fraction
Min
ute
s
3.0
0
4.0
0
5.0
0
6.0
0
7.0
0
8.0
0
9.0
0
10
.00
1
1.0
0
D0
1
3
D4
D2
2
D6
4
D8
5
2014-09-19 Bretux IgZero HIC Fraction1 3uM AcNH4 150mM pH7.4 6mbar 180v
mass140000 141000 142000 143000 144000 145000 146000 147000 148000 149000 150000 151000 152000 153000 154000 155000 156000 157000 158000 159000
%
0
100
O23505FDE 16 (1.090) M1 [Ev0,It27] (Gs,10.000,5261:7072,7.00,L30,R30); Sm (Mn, 2x5.00); Sb (2,40.00 ); Cm (1:29) TOF MS ES+ 4.22e5
A: 145939.97±0.85
145904 ± 2 Da
D0
2014-09-23 Bretux IgZero HIC Fraction5 3uM AcNH4 150mM pH7.4 6mbar 80v
mass140000 141000 142000 143000 144000 145000 146000 147000 148000 149000 150000 151000 152000 153000 154000 155000 156000 157000 158000 159000
%
0
100
O23515FDE 11 (0.755) M1 [Ev-68567,It25] (Gs,10.000,5168:7016,7.00,L30,R30); Sm (Mn, 2x5.00); Sb (2,40.00 ); Cm (1:29) TOF MS ES+ 8.24e4
151173 ± 1 Da
D4
2014-09-21 Bretux IgZero HIC Fraction3 0.5uM AcNH4 150mM pH7.4 6mbar 180v
mass140000 141000 142000 143000 144000 145000 146000 147000 148000 149000 150000 151000 152000 153000 154000 155000 156000 157000 158000 159000
%
0
100
O23509FDE 24 (1.627) M1 [Ev-81894,It25] (Gs,10.000,5383:7158,7.00,L30,R30); Sm (Mn, 2x5.00); Cm (1:29) TOF MS ES+ 3.27e5
A: 148558.11±0.99
148539 ± 1 Da
D2
2014-09-25 Bretux IgZero HIC Fraction7 3uM AcNH4 150mM pH7.4 6mbar 180v
mass140000 141000 142000 143000 144000 145000 146000 147000 148000 149000 150000 151000 152000 153000 154000 155000 156000 157000 158000 159000
%
0
100
O23521FDE 6 (0.419) M1 [Ev0,It76] (Gs,10.000,5168:7016,7.00,L30,R30); Sm (Mn, 2x5.00); Cm (1:29) TOF MS ES+ 2.50e5
A: 156515.52±3.37
156457 ± 3 Da D8
150 155 145 Mass (kDa)
1
3
2
5
2014-09-24 Bretux IgZero HIC Fraction6 3uM AcNH4 150mM pH7.4 6mbar 180v
mass140000 141000 142000 143000 144000 145000 146000 147000 148000 149000 150000 151000 152000 153000 154000 155000 156000 157000 158000 159000
%
0
100
O23518FDE 17 (1.157) M1 [Ev-97372,It21] (Gs,10.000,5168:7016,7.00,L30,R30); Sb (2,40.00 ); Cm (1:29) TOF MS ES+ 2.61e5
153815 ± 2 Da
D6 4
29
Cysteines
Advantages of native MS compared to HIC
2 3 4 5 6 7
8 9 10 11 min.
D0
D2?
D4? D1?
Hz1613F12 SG3249DAR2 IgZero 5uM AcNH4 150mM pH7.5 NAP2 6mbar 120v
m/z145500 146000 146500 147000 147500 148000 148500 149000 149500 150000 150500 151000 151500 152000 152500 153000 153500
%
0
100146 152 mass (kDa) 145 150 148 147 151 149 153
D0
D1
D3
D2
D4
= 2.0 DAR
Native MS profile
HIC profile
- Broad HIC peaks
- Difficulty to unambiguously assess
drug binding
- No odd drug load expected
= 1.8 DAR - Native MS mass accuracues allows
unambiguous drug binding
stoichiometry assessment
- odd drug load are clearly detected
- No peak broadening as on HIC
Native MS allowed to precise unclear HIC results 30
D1’?
Cysteines
Native IM-MS of intact cysteine-linked ADC
Deglycosylated Parent mAb O25596FDE.raw : 1
24+
12 16 20 22 tD (ms)
7500
7000
6000
5500
m/z
181410
6500
5000
23+
21+
IgGZero ADC
O23564FDE.raw : 1
12 16 20 22 tD (ms)
7500
7000
6000
5500
m/z
181410
6500
5000
Deglycosylated ADC
SynaptG2, 3µM, AcNH4 150mM pH7,4 Vc=80v, Pi=6mbar, Bias25, WH=40V, WV=923m/s, He=130mL/min, N2= 45mL/min
31
Cysteines
Native IM-MS of intact cysteine-linked ADC
Deglycosylated Parent mAb O25596FDE.raw : 1
24+
12 16 20 22 tD (ms)
7500
7000
6000
5500
m/z
181410
6500
5000
23+
21+
IgGZero ADC
O23564FDE.raw : 1
12 16 20 22 tD (ms)
7500
7000
6000
5500
m/z
181410
6500
5000
Deglycosylated ADC
32
SynaptG2, 3µM, AcNH4 150mM pH7,4 Vc=80v, Pi=6mbar, Bias25, WH=40V, WV=923m/s, He=130mL/min, N2= 45mL/min
IM-MS enlighten and characterize the heterogeneity of ADCs
Cysteines
Native IM-MS of intact cysteine-linked ADC
ADC heterogeneity in drug binding is observed on native IM-MS plots
Each individual even-drug load can be separated in IM-MS
No positional isomers were detected (to low IM cell resolution on intact ADC)
Deglycosylated Parent mAb Deglycosylated ADC O25596FDE.raw : 1
16 17 18 19 tD (ms)
6700
6600
6500
6400
m/z
23+
22+
O23564FDE.raw : 1
16 17 18 19 tD (ms)
6700
6600
6500
6400
m/z
DO 23+
D0 22+
D2 23+
D6 24+
D4 23+
D8 24+
33
SynaptG2, 3µM, AcNH4 150mM pH7,4 Vc=80v, Pi=6mbar, Bias25, WH=40V, WV=923m/s, He=130mL/min, N2= 45mL/min
Cysteines
Native IM-MS of intact cysteine-linked ADC
resolving power of IM cell
tD/tD at FWHM : 16.4 ± 0.8
14.2
D014.2
D214.9
D415.6
D616.3
D817.0
13 14 15 16 17 18 19 tD (ms)
34
Each individual even-drug load can
be separated in IM-MS
Parent
mAb (24+
charge state)
ADC (24+ charge state)
(brentuximab vedotin)
- D0 D2 D4 D6 D8
tD (ms) 14.2 ± 0.0 14.2 ± 0.1 14.9 ± 0.1 15.6 ± 0.1 16.3 ± 0.1 17.0 ± 0.1
(tD)
(ms) - - + 0.7 + 0.7 + 0.7 +0.7
CCS from IM-MS (nm²) 68.0 ± 0.0 68.1 ± 0.1 68.8 ± 0.1 69.5 ± 0.1 70.3 ± 0.1 71.1 ± 0.1
CCS
(nm²) - - + 0.7 + 0.7 + 0.8 +0.8
Predicted*
CCS (nm2) 68.2 68.2 69.0 69.8 70.6 71.4
Predicted CCS
(nm²) - - + 0.8 + 0.8 + 0.8 +0.8
Drug binding induces constant and reproducible tD and CCS differences
Cysteines
Semi-quantitative analysis of native IM-MS
For each charge state, intensities of the extracted drift peaks were plotted across a series
for each drug binding stoichiometry (D0, D2, D4, D6 and D8) and a Gaussian curve fitted.
To give a relative intensity of each conformer, the area underneath each curve was
calculated as a percentage of the total observed signal, here (D0+D2+D4+D6+D8).
= 3.7 ± 0.1 DAR
(triplicates)
35
Norm
aliz
ed
Inte
nsity
(%)
Charge States22 24 26 28
0
20
40
60
80
100 D0: 14%D2: 27%D4: 30%
D8: 10%D6: 19%
Cysteines
Minutes 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 11.00
D0
D2 D4
D6
D8
Average DAR determination
from semi quantitative native MS and IM-MS
MaxEnt1 deconvolution
Quantitation based on ion
intensities
n=8
D0
D2 D4
D6
D8
DAR
= 3.9 ± 0.1
DAR
= 3.7 ± 0.1 DAR
= 4.0
Number of
drug load
Calculation based on area
measured under fited gaussian
of charge states extracted ions
n =3
Native MS
HIC
Average DAR determination and drug-load profiles obtained
from native MS or native IM-MS are in good agreement with HIC data
mass144000 145000 146000 147000 148000 149000 150000 151000 152000 153000 154000 155000 156000 157000 158000 159000
%
0
100
O19456FDE 24 (1.627) M1 [Ev-62181,It48] (Gs,10.000,5175:6905,10.00,L10,R10); Sm (Mn, 2x2.00); Sb (25,5.00 ); Cm (2:29) TOF MS ES+ 5.05e5
146 Mass (kDa) 150 148 154 152 158 156
D4
D2
D6
D8 D0
Native IM-MS
36 Debaene F. et al. Anal Chem. 2014, 86(21):10674-83.
Cysteines
Benefits of native MS of Lysyne-conjugates
mAbs have more than 60 surface exposed Lys residues (compared to 32 Cys) available for reactions but
only 32 cysteines, among which 8 only are involved in the interchain disulfide bridges of chimeric, humanized
and human IgG1.
Lysine conjugation generates even more heterogeneous samples even if all species are covalent
Lysines
Benefits of native MS for lysine conjugate analysis
trastuzumab emtansine + IG0 native dil5X 5uM
m/z1000 2000 3000 4000 5000 6000 7000
%
0
100
m/z1000 2000 3000 4000 5000 6000 7000
%
0
100
6000 7000 m/z 1000 3000 5000 4000 2000
trastuzumab emtansine + IG0 native dil5X 5uM
m/z1000 2000 3000 4000 5000 6000 7000
%
0
100
m/z1000 2000 3000 4000 5000 6000 7000
%
0
100
6000 7000 m/z 1000 3000 5000 4000 2000
Native Conditions IgGZero treated ADC @ 5 µM in 150 mM AcONH4 pH7.5
Denatured conditions IgGZero treated ADC @ 2 µM in H2O:ACN:FA (50:50:1)
25+
16+
45+
Full MS spectrum Deconvoluted
mass spectrum
Full MS spectrum Deconvoluted
mass spectrum
Very complex mass spectrum :
overlapping charge state distribution
hamper appropriate Dn quantification
Deconvolution fails in detecting D8 species
Decomplexifictaion of the mass
spectrum
Deconvolution fails in detecting D8
species
* + 220 Da linker adducts
Lysines
Benefits of charge state reduction for lysine ADC analysis
trastuzumab emtansine + IG0 native dil10X 2.8uM + 10 mM Imid
m/z5000 6000 7000 8000 9000 10000 11000
%
0
100
m/z5000 6000 7000 8000 9000 10000 11000
%
0
100
trastuzumab emtansine + IG0 native dil10X 2.8uM + 10 mM Imid
m/z5000 6000 7000 8000 9000 10000 11000
%
0
100
m/z5000 6000 7000 8000 9000 10000 11000
%
0
100
25+
23+
16+
T-DM1 (2µM)
+ 10 mM imidazole
T-DM1 (2µM)
Detection of
D8 species
MaxEnt1 deconvolution
n=8
Deconvoluted
mass spectrum
= 3.0 DAR
= 3.5 DAR
* + 220 Da linker adducts
Lysines
Native MS and IM-MS for Lysine-linked ADCs
Marcoux et al. Protein Science. 2015, Protein Sci. 2015 Feb 18. doi: 10.1002/pro.2666
Sample : trastuzumab emtansine (Kadcyla)
Average DAR and drug load profile
from IM-MS
Average DAR and drug load profile
from native MS
DAR = 3.5 ± 0.1
Number of drug loads
Lysines
Special feature:
Head comparison of trastuzumab and
cetuximab with corresponding biosimilar and
biobetter candidates
Beck A, Debaene F, Diemer H, Wagner-Rousset E, Colas O,
Van Dorsselaer A and Cianférani A, JMS; 2015 in press
What’s coming next ?
41
originator
Trastuzumab
Trastuzumab-B
originator biosimilar
Take home message
42 From May C. et al., Biochemical Pharmacology 2012.
Native MS
HDX-MS Ion Mobility -MS
Native MS is a rapid, robust and reliable method for first-
line semi-quantitative mAb characterization
Native IM-MS for semi-quantitative
global conformational
characterization of mAbs
HDX for structural mAb characterizartion
(epitope mapping, comparability studies, etc.)
Many Thanks for your attention !
The Antibody Physico-Chemistry group
Elsa WAGNER-ROUSSET
Olivier COLAS
Nathalie CORVAÏA
Amandine BŒUF (past)
Daniel AYOUB (past)
Alain BECK
The native MS group of LSMBO
Julien MARCOUX
Guillaume TERRAL
Johann STOJKO
François DEBAENE (past)
Cédric ATMANENE (past)
Alain VAN DORSSELAER
Sarah CIANFERANI
43