challenges in peptide mapping mass spectrometry of ... · lc04 lc05 lc06 lc07 lc08 lc10 lc11 lc13...
TRANSCRIPT
Challenges in peptide mapping mass spectrometry of biopharmaceuticals
07.MAR.2018
Will Burkitt, Ben Holmes, Johanna Paris, Nisha Patel, John O’Hara
Characterisation within UCB2
Context to the type of analysis performed
UCB is a global biopharma focused on severe diseases in two therapeutic areas: neurology and immunology
Characterisation is on the biology side of the business:
• Biopharmaceuticals rather than small molecules
• Antibody based molecules produced in cell culture (mammalian and e-coli)
Discovery
• Target identification• V-region discovery• Candidate selection
Development
• Toxicology• Phase I• Phase II• Phase III
Commercial
• Phase IV
Characterisation
Characterisation within development3
Who do we work with and what do we do?
Activities with in development:
Process scale-up and optimisation
• Upstream (cell-culture)
• Down stream purification
Formulation
Product understanding
Analytical
• Method development
• Stability
• Bioassay
Types of study:
• Comparability
• Reference standard characterisation
• Process understanding
• Force degradation study
• Stability
• Investigations
What types of analysis do we perform?4
Wide range of physical/biophysical
Gel separations
• SDS-PAGE
• CE
HPLC separations
• Cation exchange
• Size exclusion
Higher Order
• CD, FTIR, DSC, AUC, Biacore, DLS
Mass spectrometry
• Peptide mapping reduced and non-reduced
• Intact mass denatured and native
Gives information on the primary structure
What people want to know…
• Expected sequence confirmation
• PTM levels
• Sequence variants
Peptide mapping5
Method
Sample preparation (digestion)• Denature (guanidine hydrochloride)
• Reduce (dithiothreitol)
• Alkylate (iodoacetamide)
• Buffer exchange
• Digest (trypsin for 3 hours)
Separation• Reversed phase C18
• Water:acetonitrile with 0.1% formic acid
• 45 minute run time
Detection• Q-Exactive Plus Orbitrap
Data processing• Biopharama Finder
Ideal method:
• Reproducible
• Not induce PTMs during analysis
• Scalable
• Fast
• Specific cleavage
• Sensitive
Manufacture overview of a biopharmaceutical6
Modification levels
Peptide: DTLMISR
Non-modified selected ion chromatogram:
Oxidised selected ion chromatogram:
7
Example of how the level at which a PTM is present i s calculated (oxidation)
How are these calculated?
Easy example:
• One modification site
• Good chromatographic separation
• 16 Da mass shift
To calculate oxidation level:
Integrate peak area of non-modified and modified
Not quantitative
Can be used to compare levels across batches within the same sample set
0
10
20
30
40
50
60
70
80
90
100
Rel
ativ
e In
tens
ity
SIC: E:\Data\UCB\Innovate\A33\06JUL2017\11JUL2017_02_IgG4_Demo.raw15.81
16.2812.68 16.9513.89 17.36 18.06 18.5214.95 15.65 19.4013.5910.95
10 11 12 13 14 15 16 17 18 19RT (min)
NL: 3.04E8
0
10
20
30
40
50
60
70
80
90
100
Rel
ativ
e In
tens
ity
SIC: E:\Data\UCB\Innovate\A33\06JUL2017\11JUL2017_02_IgG4_Demo.raw13.92
15.8114.1313.1712.33 14.55 15.2313.52 16.36 19.5818.2316.7011.26
10 11 12 13 14 15 16 17 18 19RT (min)
NL: 4.35E7
Modification Level =���������������
���������(�����������������)
Example of measurement of modification level
Repeatability
Method can give linear and repeatable results
Linearity
8
Easy example norleucine misincorporation for methioninine
y = 0.04536x + 0.00189R² = 0.99996
0.0%
0.5%
1.0%
1.5%
2.0%
2.5%
3.0%
3.5%
4.0%
4.5%
5.0%
0 0.2 0.4 0.6 0.8 1
Mea
sure
d ra
tio o
f nor
leuc
ine
to
met
hion
ine
Ratio of low level batch to high level batch
Lower
concentration
at about 1.3%
Repeat
injections
at about 2.3%
Higher
concentrations
at about 4.4%
Std Dev CV Std Dev CV Std Dev CV
Norleucine
level0.06% 3.6% 0.05% 2.0% 0.13% 2.6%
Modifications searched for
Standard modifications:
• Glycosylation (glycoforms)
• C-terminal lysine clipping
• N-terminal pyroglutamic acid formation
• Methionine oxidation
• Deamidation
• Glycation
9
How many should be included?
Additional modifications:
• Tryptophan oxidation
• Histidine oxidation
• O-linked glycosylation
• Aspartic acid isomerisation
• Glycation degradation products (AGEs)
• Leader sequence
• Sequence variants
• …
• Non-hypothesis modifications (wild card search)
What are the issues with this approach?10
Levels reported are not quantitative
• Comparison between runs and instruments is difficult
• Comparison between occasions can be difficult
• How should levels be calculated?
When is a sample different
• Analytical difference Vs biochemically relevant difference
Detection of repeated sequences
• Would the presence of additional light chain be detected?
Too much data
• Large numbers of samples
• How to spot relevant differences quickly
• Which differences relate to the various parameters?
What improvements are we trying to make?
Isotopically labelled products
• For each product make a small amount that has been isotopically labelled so that this can be used as an internal standard.
• Focus on the amount of unmodified peptide present
• Any modification of a peptide will reduce the amount of unmodified peptide
• Doesn’t matter if you know what modifications to look for
11
Two separate approaches
Statistical methods
• Focus only on the PTMs and the levels that have been measured for each
• Takes some account of loading and digestion differences
• Include all modifications: hypothesis/non-hypothesis and sequence variants
Mass Spectrometry Analysis
Labelled internal standard methodology12
Principle
Trypsin Digestion
mAb X mAb XReferenceStandard
LabelledmAb X
LabelledmAb X
Equal amounts
Labelled internal standard methodology (2)13
Labelled peptide Labelled peptide
Peptide from mAb X Peptide from mAb X Ref Std
C DBA
IA = Intensity of peptide from mAb X IC = Intensity of peptide from Ref StdIB = Intensity of labelled peptide in mAb X sample ID = Intensity of labelled peptide in Ref Std sample
Measure intensity of peptide signals:
Calculate ratio of peptide and labelled peptide in each sample:
��
��
��
��
Calculate ratio of ratios to give the amount of the mAb X peptide relative to the mAb X Ref. Std. peptide:
��������
=��
��
��
��= ��� !!"#�$%&'�!#��()��)
Labelled internal standard methodology (3)14
Differences
Labelled peptide Labelled peptide
Peptide from mAb X Peptide from mAb X Ref. Std.
C DBA
Decrease in signal from peptide results in decrease measured level of peptide
��������
=��
��
��
��= ��� !!"#�$%&'�!#��()��)
Decrease in signal results from modification of the peptide e.g. deamidation, oxidation etc……even modifications that we don’t know to look for
Labelled internal standard results15
Com
parison of similar batches
0.9
0
0.9
5
1.0
0
LC01
LC02
LC04
LC05
LC06
LC07
LC08
LC10
LC11
LC13
LC14
LC15
LC16
LC18
LC19
HC01
HC02
HC03
HC04
HC06
HC08
HC09
HC10
HC11
HC12
HC13
Tw
o b
atch
es sa
me
pro
cess
LC02: 1.3+
0.9%H
C02:+
1.4+0.7%
0.4% less W
ox
0.9
0
0.9
5
1.0
0
LC01
LC02
LC04
LC05
LC06
LC07
LC08
LC10
LC11
LC13
LC14
LC15
LC16
LC18
LC19
HC01
HC02
HC03
HC04
HC06
HC08
HC09
HC10
HC11
HC12
HC13
Tw
ore
fere
nce
stan
da
rds d
iffere
nt
pro
cesse
s
0.9
5
1.0
0
LC01
LC02
LC04
LC05
LC06
LC07
LC08
LC10
LC11
LC13
LC14
LC15
LC16
LC18
LC19
HC01
HC02
HC03
HC04
HC06
HC08
HC09
HC10
HC11
HC12
HC13
Labelled internal standard results
Change of cell line and m
edia
Cell line 1 +
Media A
Cell line 2 +
Media A
Cell line 1 +
Media B
0.9
5
1.0
0
LC01
LC02
LC04
LC05
LC06
LC07
LC08
LC10
LC11
LC13
LC14
LC15
LC16
LC18
LC19
HC01
HC02
HC03
HC04
HC06
HC08
HC09
HC10
HC11
HC12
HC13
0.9
5
1.0
0
LC01
LC02
LC04
LC05
LC06
LC07
LC08
LC10
LC11
LC13
LC14
LC15
LC16
LC18
LC19
HC01
HC02
HC03
HC04
HC06
HC08
HC09
HC10
HC11
HC12
HC13
Labelled internal standard results17
Acidic species
Statistical methods18
Heatmaps
• Only data used is the levels of PTMs that have been measured for each sample
• Aids the data analysis as using the PTM levels provides some normalisation between the samples
• All modifications are included: hypothesis/non-hypothesis and sequence variants
• For each modification the levels are normalised to the mean and the number of standard deviations from the mean plotted
Values Std Dev from mean
Sample 1 0.59 0.71
Sample 2 0.36 -0.11
Sample 3 0.18 -0.75
Sample 4 0.31 -0.29
Sample 5 0.66 0.96
Sample 6 0.19 -0.71
Sample 7 0.06 -1.18
Sample 8 0.87 1.71
Sample 9 0.59 0.71
Sample 10 0.06 -1.18
Mean 0.39
Std Dev 0.28
Heatmaps19
As originally envisaged…
Heat maps20
What I got
Heatmaps21
Bioreactor conditions
ן Too much data
ן Combining datasets form multiple occasions
ן Processing data does not scale well
Issues still to be solved22
ן Automation
o Current demand requires improvements in automation of sample preparation and data processing
ן Host Cell Proteins (HCPs)
o Current levels of sensitivity allow detection some of these with our existing methodology
ן Higher Order Structure
o Can peptide mapping methods be extended to give information on higher-order structure?
Issues still to be solved23
ן Peptide mapping with mass spectrometry can provide a very detailed analysis of biopharmaceutical samples.
ן Large amounts of data are produced.
ן Plotting this data as a heat map can aid interpretat ion of this data and focus attention on potential issues.
ן Using isotopically labelled internal standards of the product of interest has the potential to give a rapid comparison of the similarity of batches.
ן There is still lots more work to do…
Conclusions24
Acknowledgements25
Characterisation Upstream
Questions?26
Thanks!