TO DOWNLOAD A COPY OF THIS POSTER, VISIT WWW.WATERS.COM/POSTERS ©2014 Waters Corporation

INTRODUCTION

Sequence variants (SV) are unintended amino acid

substitution in the primary structure, and are

classified as product-related impurities. The presence of sequence variants may pose concerns

regarding bioactivity, stability, and immunogenicity.

Sequence variants are usually present at very low-

level in a therapeutic protein. From an analytical

stand point, the detection and characterizing SV in a complex digest mixture that is several orders of

magnitude more concentrated, remains a significant

challenge.

A comparative analytical strategy is presented to

identify sequence variant among multiple samples. The strategy was developed and tested by analyzing

monoclonal antibody samples which contain spiked

synthetic peptides with amino acid substitutions.

SEQUENCE VARIANT IDENTIFICATION STRATEGIES BASED ON LC/MS PEPTIDE MAPPING ANALYSIS.

Stephane Houel1; Scott Geromanos1, Andrew Tudor2, Barry Dyson2, Ying Qing Yu1; Weibin Chen1 1Waters Corp., Milford, MA 01757; 2Waters Corp., Manchester, UK.

EXPERIMENTAL

Sample Preparation: Three tryptic digests of Trastuzumab (Digest A, B or C) were prepared separately, each at a final

concentration of 2.4 picomole/µL. In digest C, two synthetic peptides of T14 containing substituted amino acid residues (see

Table 2) were spiked at 1.8 femtomole/µL.

LC-MS Analysis: Each sample was analyzed in replicate injection

(4µL inj.). A blank injection was performed between the samples to minimize the sample carryover. The samples were analyzed in

the order of Digest A, Digest C and finally Digest B.

LC/MS Conditions:

LC: Waters ACQUTIY UPLC I-Class Column: Acquity UPLC PST 2.1x150mm BEH C18 300Å, 1.7 µm

MS Conditions: Instrument Waters SynaptTM G2-Si HDMS

Mode: ESI positive mode Capillary Voltage: 3.0 kV

Cone Voltage: 10 V Source Temperature: 100 °C

Desolvation Temperature: 350 °C

Informatics:

Progenesis QI

Figure 1. Data were collected using Waters Synapt G2-Si

CONCLUSIONS

Progenesis can detect and quantify low abundance sequence variant peptides (0.1%)

in a monoclonal antibody digest.

HD-DDA spectra of sequence variant peptides loaded on a 2.1 mm column at 7

femtomoles provides enough information to pin-point the substitution.

Variability increases for features at low abundance and between digest samples.

Nevertheless, more than 95% of the detected features are common to all three

digests.

RESULTS & DISCUSSION

Figure 2: Proposed workflow to identify sequence variant peptides.

Workflow for Sequence Variant Identification Table 1: Native and mutant T14 feature characteristics.

Table 2: Quantification of native and mutant T14 peptides.

Quantification based on Progenesis QI

Progenesis Provides Intuitive Visualization Interface to Compare Feature Difference

Figure 7: 3D montage of features #11 and #11400 for each run of the three digests.

Feature #11 Feature #11400

Digest C

Digest B

Digest A

0

10

20

30

40

50

60

70

80

90

100

1.E-16 1.E-14 1.E-12 1.E-10 1.E-08 1.E-06 1.E-04 1.E-02 1.E+00

CV

(%

)

Anova P value

Digest A

Digest B

Digest C

0

50

100

150

200

250

300

0 5 10 15 20 25 30

Fre

qu

en

cy (

Co

un

ts)

CV (%)

1E4< Abundance <3E5

3E3< Abundance <1E4

1E3< Abundance <3E3

2E2< Abundance <1E3

1E1< Abundance <2E2

0

100

200

300

400

500

600

700

0 5 10 15 20 25 30

Fre

qu

en

cy (

Co

un

ts)

CV (%)

Digest A

Digest B

Digest C

All Digests

0

10

20

30

40

50

60

70

80

90

100

0 5 10 15 20 25 30

Cu

mu

lati

ve R

ela

tive

Fre

qu

en

cy (

%)

CV (%)

Digest A

Digest B

Digest C

All Digests

Figure 4: Frequency (a) and cumulative relative frequen-cy (b) diagrams of the coefficient of variation (CV) for each digest individually (intra-run variability) as well as

for all of the runs (inter-digest variability)

Figure 5: Venn diagram showing all the features detected in all 3 digests. A total of 8229 features (charge states 2+ to 13+) were detected, among which 95.15% are common in all three samples. Less than 2% of features are unique to any one digest, which can be automatically generated and exported to MassLynx to ac-

quire high quality HD-DDA spectra (see figure 8) for SV identification.

Digest B

Digest C

Digest A

7830(95.15%)

9(0.11%)

96(1.17%)

Figure 6: Relationship between the Anova p val-ue and the coefficient of variation (CV) for unique features to each digest.

Figure 3: Frequency (a) and cu-mulative relative frequency (b) diagrams of the coefficient of variation (CV) for digest A at different abundance ranges.

Raw Data

LC-HDMSE

U n d e r s t a n d ,

R e v i e w a n d

Process LC-HDMSE

Data Collected

from an Analytical

Setting for SV

Identification

Figure 8: HD-DDA spectra of T14 with one (N9S) or two substitutions (N3S and N9S).

VDSALQSGSSQESVTEQDSK

VDNALQSGSSQESVTEQDSK

b3 b4 b5 b6 b7 b9 b11

y4y3 y5 y7 y9 y10 y14

y8

y6

y6

y7

y8

y9

y5y4

y3

y10

b5

b6

b10

HD-DDA Spectra Aid in Sequence Variant Peptide Identification

Dat

a P

roce

ssin

g

Digests A and B Digest C

Reduction/Alkylation

Buffer exchange

Trypic digestion

TRASTUZUMAB

Sequence variant peptides spiked into an aliquot of digest C

Buffer A added toan aliquot of digests A and B

Sam

ple

Pre

par

atio

nD

ata

Acq

uis

itio

n

LC-HDMSe

Dat

a P

roce

ssin

g

Detection of unique features: Sequence Variant candidates

LC-DDA/HD-DDAwith inclusion ListD

ata

Acq

uis

itio

n

(Sequence Variant Search in UNIFI:Under Development)

a)

a)

b)

0

10

20

30

40

50

60

70

80

90

100

0 5 10 15 20 25 30

Cu

mu

lati

ve R

elat

ive

Freq

uen

cy (

%)

CV (%)

1E4< Abundance <3E5

3E3< Abundance <1E4

1E3< Abundance <3E3

2E2< Abundance <1E3

1E1< Abundance <2E2

b)