2
Brentuximab vedotin was deglycosylated using EndoS
endoglycosidase (IgGZERO
™
, Genovis). Titration
experiments involving J10.4 mAb and JAM-A were
monitored by native MS in order to determine the binding
stoichiometry. The fixed amount of J10.4 (5 μM) was
incubated with increasing amounts (1:1, 1:2, 1:4, 1:8) of
JAM-A up to 40 μM. The mixture of eleven distinct
deglycosylated humanized IgG antibodies included two
marketed therapeutic mAbs (rituximab and trastuzumab)
and nine point mutation variants of the Hz6F4-2 mAb
[4, 5]. They were mixed together prior to PNGase-F
deglycosylation.
Finally, all the samples were buffer exchanged against
150 mM ammonium acetate (AcONH4) pH 7.5.
Trastuzumab, deglycosylated Brentuximab vedotin,
and the mAb/antigen complexes of J10.4 mAb/JAM-A
were injected at 5 μM, and the deglycosylated IgG
mixture was injected at 1 μM on the Exactive Plus EMR
Orbitrap mass spectrometer.
Direct-Infusion Native MS Conditions
Chip-based infusion conditions
Instrumentation
TriVersa NanoMate
®
(Advion, USA) system
Ionization voltage (kV)
1.6–1.8
Gas pressure (psi)
0.3–0.6
The ESI Chip
®
consists of an array of 400 nanoelectro-
spray emitters with 5 µm inner diameters.
MS conditions
Instrumentation
Exactive Plus EMR Orbitrap MS system
(Figure 1)
EMR mode
ON
Mass range
(m/z)
350–20,000
Resolution
17,500 to 140,000, depending on spectral
complexity
Target value
3 x 10
6
Microscans
10
Max injection time (ms)
300
Insource CID energy (eV) 60 to 150, manually tuned for optimized
desolvation
S-lens level (%)
100 to 200, manually tuned for optimized
transmission and avoiding in-source
fragmentation
Injection flatapole DC (V) 8
Inter flatapole lens (V)
7
Bent flatapole DC (V)
6
C-Trap entrance lens
0
tune offset (V) EMR
Trapping gas pressure 4
setting factor
Spectra average
Enabled (10 to 50 scans are averaged to
achieve S/N ratio of >100)
Data Processing
Software
Thermo Scientific
™
Protein
Deconvolution software version 2.0 SP2
and version 3.0
Deconvolution parameters
Number of iterations
4
Noise compensation
On
Minimum adjacent charges
1 to 3
Results and Discussion
High-Resolution Native MS Analysis of Intact
Monoclonal Antibody Trastuzumab
Trastuzumab (Herceptin
®
) is a humanized IgG1 mAb,
approved for HER2-overexpressing breast cancer
treatment since 1998. Several mechanisms of action are
thought to contribute to trigger the tumor-inhibitory effect
of this protein therapeutic. Among them, trastuzumab can
mediate the effector functions of immune cells through its
constant region (Fc) by binding to the Fc gamma receptor III
(Fc
γ
RIII) and triggering antibody-dependent, cell-
mediated cytotoxicity (ADCC).
Based on the published amino acid sequence of both the
light and heavy chain of trastuzumab, the calculated mass
of this protein is C
6560
H
10132
O
2090
N
1728
S
44
= 148,057 Da.
This calculation includes 16 disulfide bridges (−32 Da),
two main glycoforms (G0F; +1445 Da), and near 99%
cleavage of two heavy chain C-terminal lysines
(−128 × 2 Da). Partial cyclization of one or two
N-terminal glutamic acids (−18 Da) may also occur as
well as methionine oxidations (+16 Da). Three Asn
deamidation/Asp isomerization hot spots have also been
described in the CDRs and shown to negatively impact
HER2 antigen binding when degraded (Figure 2A).
Trastuzumab was analyzed on the Exactive Plus EMR MS
with resolution set at both 17,500 and 35,000. The
deconvoluted mass spectrum calculated using Protein
Deconvolution software version 2.0 SP2 represents the
classical glycosylation pattern of a mAb with baseline-
resolved glycan peaks. Figure 2B shows the complete mass
spectrum at resolution of 35,000 and a zoom of the
corresponding 23
+
charge state of trastuzumab acquired
with the resolution set at both 17,500 and 35,000 in
native conditions. Compared to the raw spectrum
acquired at 17,500 resolution, an interference peak can
be resolved by using a resolution of 35,000 or higher.
The high resolution can resolve the analyte from the
interferences, therefore, ensuring the low ppm mass
accuracy. Molecular weights of each trastuzumab
glycoform were measured with good mass accuracy in
the low ppm range, as shown in Figure 2C. The mass
differences between species are +146 Da and +162 Da,
corresponding to a fucose or to the addition of multiple
hexose units, respectively.
