A Generic HPLC Method for Absolute Quantification of ...

4 downloads 0 Views 382KB Size Report
A Generic HPLC Method for Absolute Quantification of Oxidation in Monoclonal Antibodies and Fc-Fusion. Proteins Using UV and MS Detection. Oxidation of ...
A Generic HPLC Method for Absolute Quantification of Oxidation in Monoclonal Antibodies and Fc‐Fusion  Proteins Using UV and MS Detection

Link to Anal. Chem. 

Christof Regl1,2, Therese Wohlschlager1,2, Johann Holzmann2,3, and Christian G. Huber1,2 1Department of Molecular Biology, Division of Chemistry and Bioanalytics, University of Salzburg, Austria  2Christian Doppler Laboratory for Biosimilar Characterization, University of Salzburg, Austria  3Analytical Characterization Biopharmaceuticals, Sandoz GmbH, Kundl, Austria

Forced oxidation: 30 min  0.35% H2O2





2,3

40

5

4

1

6

20 (b) 0

5

10

15 20 Time [min]

25

30

Digestion with IdeS (FabRICATOR®) followed by reduction of disulfide bonds with TCEP 

100

2524.6979 +162 Da

50

+162 Da

2557.1146

0

100

Peak 2,3

2539.3011 2523.0970 +162 Da

50

+162 Da 2555.5039

0

Peak 4 2521.4972

Q ExactiveTM

UV AIF‐MS Absolute quantification of methionine oxidation in Fc/2

b34

4000 2000 0 8000

b34 + O

(b) y45

4000 0 4000

y45 + O

2000 0

5

10

2510

2520

20

(I) UV

b34 + O 1869.4478 b34 z=2 1861.4500 z=2

100

50

0 1000 b1 b26 b51 b76 b101 b126 b151 b176 b201

1400

m/z

2200

2600

3000 y186 y161 y136 y111 y86 y61 y36 y11 y1

Figure 4. Fragment ion spectrum of singly oxidized Fc/2 subunit (peak 2,3 in Figure 1b) obtained upon AIF at 96 eV in the HCD cell. The diagnostic fragments b34 (red) for methionine residue Met256 and y45 (blue) for Met432 are annotated in the raw‐spectrum (a). The positions of the b34 and y45 fragments used for quantification are indicated in the Fc/2 sequence (b).

(II) TICC (III) XICC

(IV) AIF‐ (IV) AIF‐ MS (b34) MS (y45)

LODa),b) [ng µL–1]

0.32

0.41

0.45

1.44

1.11

LOQa),c) [ng µL–1]

0.96

1.24

1.36

4.32

3.33

0.9975

0.9956

0.9950

0.9878

0.9863

7.8

8.5

12.9

9.0

9.1

R² (12–60 ng µL–1)

1800

!@ G !@ P !@ S !@ V !@ F !@ L !@ F !@ P !@ P !@ K !@ P !@ K !@ D !@ T !@ L !@ M !@ I !@ S !@ R !@ T !@ P !@ E !@ V !@ T !@ C !@ !@ V !@ V !@ V !@ D !@ V !@ S !@ H !@ E !@ D %@ P !@ E !@ V !@ K !@ F !@ N !@ W !@ Y !@ V !@ D-G !@ V !@ E !@ V !@ H !@ N !@ !@ A !@ K !@ T !@ K !@ P !@ R !@ E !@ E !@ Q !@ Y !@ N !@ S !@ T !@ Y !@ R !@ V !@ V !@ S !@ V !@ L !@ T !@ V !@ L !@ H !@ Q !@ !@ D !@ W !@ L !@ N !@ G !@ K !@ E !@ Y !@ K !@ C !@ K !@ V !@ S !@ N !@ K !@ A !@ L !@ P !@ A !@ P !@ I !@ E !@ K !@ T !@ I !@ !@ S !@ K !@ A !@ K !@ G !@ Q !@ P !@ R !@ E !@ P !@ Q !@ V !@ Y !@ T !@ L !@ P !@ P !@ S !@ R !@ D !@ E !@ L !@ T !@ K !@ N !@ !@ Q !@ V !@ S !@ L !@ T !@ C !@ L !@ V !@ K !@ G !@ F !@ Y !@ P !@ S !@ D !@ I !@ A !@ V !@ E !@ W !@ E !@ S !@ N !@ G !@ Q !@ !@ P !@ E !@ N !@ N !@ Y !@ K !@ T !@ T !@ P !@ P !@ V !@ L !@ D !@ S !@ D !^ G !@ S !@ F !@ F !@ L !@ Y !@ S !@ K !@ L !@ T !@ !@ V !@ D !@ K !@ S !@ R !@ W !@ Q !@ Q !@ G !@ N !@ V !@ F !@ S !@ C !@ S !@ V !@ M !@ H !@ E !@ A !@ L !@ H !@ N !@ H !@ Y !@ !@ T !@ Q !@ K !@ S !@ L !@ S !@ L !@ S !@ P !@ G !@

Table 1. Limits of detection (LODs), limits of quantification (LOQs) of the different quantification approaches for Fc/2 a)Mean values from 5 injections carried out within 3 days. b)LODs calculated according to the regression line method using the slope b and the residual standard deviation sy of the calibration curves c)LOQ=3 LOD; d)Relative standard deviation of peak areas at LOQ, N=5 over 3 days.

Conclusions

+162 Da 2553.9080

2530 m/z 2540

15 Time [min]

Figure 3. Oxidation site assignment of the separation shown in Figure 1b based on XICCs of diagnostic fragments b34 and y45, in their non‐oxidized or singly‐oxidized (+ O) form obtained by AIF at 96 eV. Positions in the Fc/2 sequence are shown in Figure 4b.

2537.7009

+162 Da

0

IP‐RP‐HPLC

6000 3000 0

RSD%d)

50

Full scan‐MS

TICC

Method

100

5

Peak 1

2540.9007

35

Signal intensity [counts]

(a)

y45 2557.7509 z=2 y + O 45 2565.7472 z=2

4

1

0

60

Figure 1. IP‐RP‐HPLC separation of H2O2 stressed Rituximab upon digestion with IdeS, with disulfide bonds intact (a) or reduced (b). Four different Fc/2 variants are obtained: doubly oxidized (1), singly oxidized (2 and 3), and nonoxidized (4) Fc/2; the respective variants with intact disulfides are marked as 1′‐4′. The subunits Fd` (5) and LC (6) are observed when the intermolecular disulfides of the F(ab’)2 subunit (7) are reduced.

The obtained mAb oxidation variants were subsequently separated by ion‐pair reversed‐ phase high‐performance liquid chromatography (IP‐RP‐HPLC) coupled to a Q Exactive™ hybrid quadrupole‐Orbitrap mass spectrometer from Thermo Fisher Scientific™ 4 . Separation was performed on a Thermo Fisher Scientific™ MAbPac RP column (4 µm, 2.10 i.d. x 150.0 mm) at 200 µL min‐1 at 80°C with a linear gradient of 28.9‐29.0% ACN + 0.10% TFA in 15 min; 29.0‐30.0% ACN + 0.10% TFA in 9 min; 30.0‐45.0% ACN + 0.10% TFA in 5 min. Absolute quantification of oxidation in Fc/2 was accomplished based on concentration‐ response curves from 0.60 ng µL–1 up to 120 ng µL–1 of Rituximab Fc/2 obtained by UV spectroscopy as well as full scan mass spectrometry (MS), and MS upon all‐ion fragmentation (AIF‐MS) 5 . Data acquisition and evaluation was carried out with the Chromeleon 7.2 software from Thermo Fisher Scientific™. Quantification of MS data was attempted based on total ion current chromatograms (TICC) as well as on extracted ion current chromatograms (XICC). 4

300000 150000



3

TCEP IgG1 mAb Rituximab (MabThera®)



80

(a)

2,3

7

0

Relative Abundance [%]

2

1

Signal Intensity [mAU]

Methods To develop an analytical workflow for the characterization of oxidized therapeutic antibodies, IgG1 antibody Rituximab (MabThera®) was used 1 . Oxidized mAb species were generated by exposure to 0.35% hydrogen peroxide (H2O2) for 30 minutes at 22°C 2 . The reaction was quenched by buffer exchange to 175 mM L‐1 ammonium acetate. Subsequently the mAb was digested with immobilized IdeS followed by reduction of disulfides 3 .

Results 100

Relative Abundance [%]

Introduction & Aim of the Study Oxidation of methionine induced by reactive oxygen due to inappropriate storage conditions or contamination is a major shelf‐life‐limiting factor of biopharmaceuticals [1]. Methionine oxidation within the Fc region may lead to decreased bioactivity, and may cause faster plasma clearance [2]. We aim at middle‐down analysis of monoclonal antibody (mAb) oxidation variants, enabling faster analysis and avoiding artifacts occurring during the extensive sample preparation of a peptide map [3].

2550

2560

Figure 2. Zoom of full‐scan mass spectra of Fc/2 subunit oxidation variants (charge state 10+) obtained by the separation shown in Figure 1b. Each chromatographic peak contains the three glycosylation variants G0F, G1F and G2F indicated by +162 Da mass shifts, corresponding to an additional hexose. Mass shifts of +16 Da indicate oxidation.

• Full chromatographic separation of Fc/2 oxidation variants by IP‐RP‐HPLC • Methionine oxidation site assignment upon HCD fragmentation • Absolute quantification of oxidation down to low percent‐range in biopharmaceuticals by UV‐ spectroscopy as well as full scan MS and AIF‐MS • Data acquisition and evaluation with a single software platform

References [1] Houde, D., et al., J.Chromatogr. A, 2006. 1123(2): p. 189‐198 [3] An, Y., et al., MAbs, 2014. 6(4): p. 879‐893

The financial support by the Austrian Federal Ministry of Science, Research and Economy, and by a start‐up grant of the State of Salzburg is gratefully acknowledged.

[2] Stracke, J., et al., MAbs, 2014. 6(5): p. 1229‐1242

and