Combination of Bottom-Up and Top-Down Characterization of … · 2017. 9. 20. · Combination of Bottom-Up and Top-Down Characterization of Biologics Using a High Throughput Capable

Combination of Bottom-Up and Top-Down Characterization of Biologics Using a High Throughput Capable Workflow in Proteome Discoverer Software Kai Scheffler,1 Torsten Ueckert,2 Carmen Paschke,2 Bernard Delanghe2

Thermo Fisher Scientific, 1Dreieich, 2Bremen, Germany

Po

ster No

te 64

48

2

C bi ti f B tt U d T D Ch t i ti f Bi l i U i Hi h Th h t C bl W kfl iCombination of Bottom-Up and Top-Down Characterization of Biologics Using a High Throughput Capable Workflow in p p g g g g p pP t Di S ftProteome Discoverer Software

S ff 1 2 C 2 2Kai Scheffler1, Torsten Ueckert2, Carmen Paschke2, Bernard Delanghe2, , , gTh Fi h S i tifi 1D i i h 2B GThermo Fisher Scientific, 1Dreieich, 2Bremen, Germany, , , y

O i Results continued FIGURE 6: Sequence coverage of the antibodies' light chain based on the topFIGURE 3. Top-down data of the reduced monoclonal antibody sample showing R lt ti dOverview Results continued FIGURE 6: Sequence coverage of the antibodies' light chain based on the topdown spectrum. Blue lines with kinks towards the left/right indicate theFIGURE 3. Top down data of the reduced monoclonal antibody sample showingthe Full MS spectrum (A), the MS/MS HCD spectrum of a single charge state of the Results continued

Purpose: Demonstrate the benefits of Thermo ScientificTM Proteome DiscovererTM 2.0 For the complete characterization of biologics it is essential to confirm the expectedi i t 100% d i dditi th it i f l ti

do spect u ue es t s to a ds t e e t/ g t d cate t eassignment of corresponding b/y-type fragment ions.antibodies' light chain (B). Trace C shows a zoom into the top-down spectrum

highlighting a series of fragments of charge state z=6 and a number of low Figure 8 lists the identified modifications found in the bottom up approach: oxidation,software for combining a top-down and bottom up protein analysis workflow for in-depth protein characteri ation

sequence aiming at 100% sequence coverage and in addition the monitoring of relativeabundances of glycoforms as well as the levels of modifications such as oxidation

highlighting a series of fragments of charge state z=6 and a number of lowabundant higher charge states z=10,12. Trace D highlights the deconvoluted

gdeamidation and the conversion of the N-termini of both the light and heavy chain from Glnto p ro Gl Fig re 9 highlights the relati e q antitation of the C terminal peptide of thedepth protein characterization.

M th d A S t HT d t b h ith l l ti

abundances of glycoforms as well as the levels of modifications such as oxidation,deamidation and truncation of the heavy chains’ C-terminal lysine residue. Figure 2A

g g g gportion of the top-down spectrum shown in trace C highlighting the partialseq ence aa81 83 DAA of the light chain

to pyro-Glu. Figure 9 highlights the relative quantitation of the C-terminal peptide of theheavy chain with and without the terminal lysine. For the full length peptide both chargeMethods: A Sequest HT database search with precursor area calculation was

combined with a ProSightPD and Byonic searchshows the base peak chromatogram and the separation achieved from the antibodydigest sample with a separation time of 20min Figure 2B is showing the XIC of the

sequence aa81-83 DAA of the light chain. heavy chain with and without the terminal lysine. For the full length peptide both chargestates +1 and +2 were considered, the truncated peptide was detected only in charge state 1combined with a ProSightPD and Byonic search.

Results: A 100% sequence coverage was obtained by combining bottom up and topdigest sample with a separation time of 20min. Figure 2B is showing the XIC of theoxonium ion indicative for the presence of glycopeptides that are shown in Figures 2C

due to the lack of the proton-capturing lysine. The areas obtained for the two versions of thepeptide allow the determination of the ratio truncated:untruncated peptide to be ~23100 1536.74AResults: A 100% sequence coverage was obtained by combining bottom-up and top-down experiments.

p g y p p gand 2D on the Full MS level highlighting the presence of the glycopeptides in its variousi f Th b i d f S HT d B i h i

peptide allow the determination of the ratio truncated:untruncated peptide to be ~23,corresponding to ~4.3% abundance of the untruncated form.Full MS Spectrum80

100

ance

1646.591280 881152 83 1440 83

Ap

I d iisoforms. The obtained sequence coverage from Sequest HT and Byonic are shown inFigures 4 and 5 for the light and heavy chain respectively Figure 7 lists the identified

p g

Figure 3 shows the Full MS spectrum of the reduced antibody with the overlapping charge60Abun

da 1280.881152.83 1440.831048.13 1773.181920 821690 20Introduction Figures 4 and 5 for the light and heavy chain respectively. Figure 7 lists the identifiedglycoforms of the heavy chains’ glycopeptide in two forms, identified with and without a

g p y pp g genvelopes for the light and heavy chains. For top-down analysis and fragmentation a singleh i l d i h 10D i l i i d i d f i20

40

ativ

e A 1920.821690.20 2095.35 2304.77

1950.112327 70

2112.48

In the development as well as the production phase of biologics such as monoclonaltib di th ifi ti f th t i id ll th FIGURE 1 P t Di 2 0 ft kfl f th d t l i f

missed cleavage. charge state was isolated with a 10Da isolation window in order to get a fragment ionspectrum of either light chain (Figure 3B) or the heavy chain (not shown) was obtained0

20

Rel

a 2327.702204.21

antibodies the verification of the correct amino acid sequence as well as theassessment of the presence of the correct type and relative amount of glycosylation

FIGURE 1. Proteome Discoverer 2.0 software workflows for the data analysis ofbottom-up (A) and top-down (B) spectra. The bottom-up workflow consist of

spectrum of either light chain (Figure 3B) or the heavy chain (not shown) was obtained.When top-down analyses are performed involving chromatographic separation the isolation

1000 1200 1400 1600 1800 2000 2200 2400m/zassessment of the presence of the correct type and relative amount of glycosylation

are essential steps and only a few out of many analytical methods applied. Mostbottom up (A) and top down (B) spectra. The bottom up workflow consist ofnodes for protein ID using the Sequest HT search algorithm, the precursor ion window can be widened up to several hundreds of Da in order to co-isolate and fragment

several charge states of either species in parallel Figure 3C highlights the number and100

e

1322.901176 02

commonly, enzymatic digests of the proteins are analyzed in a peptide mapping typeexperiment In addition top down experiments provide additional information regarding

area detector for relative quantitation and the new Byonic node. The top-downworkflow contains the new ProSight TopDown High/High Crawler for spectra

several charge states of either species in parallel. Figure 3C highlights the number anddensity of fragment ions observed, and in particular fragment ions in higher charges statesTop Down MS/MS Spectrum80nda

nce 1176.02 1310.06

1358.981161.60Bexperiment. In addition, top-down experiments provide additional information regarding

mass, amino acid and further modifications. Whereas previously data analysis had toworkflow contains the new ProSight TopDown High/High Crawler for spectradeconvolution combined with the Prosight absolute mass search node.

density of fragment ions observed, and in particular fragment ions in higher charges statesdue to the use of the protein mode available on the Q Exactive mass spectrometer. Figure

p p40

60

e A

bun

1058.52 1511.601235.53 1378 82

1161.60

1011 40, p y y

be run through different software packages, Proteome Discoverer software nowid d th t bl th l i f b th b tt d t d

g3D shows the deconvoluted spectrum of the zoomed MS/MS spectrum highlighting thefragment ion series b -b supporting the amino acid sequence “DAA” The assignment of all20

40

Rel

ativ

e 1378.82 1744.021526.141073.23974.97 1723.01 1781.33

1643 84

1011.40

provides new nodes that enable the analysis of both bottom-up and top-downexperiments in one workflow on one platform A B

fragment ion series b80-b83 supporting the amino acid sequence DAA . The assignment of alldetected fragment ions to the sequence of the light chain is highlighted in Figure 6. The1000 1100 1200 1300 1400 1500 1600 1700 1800 1900

0

R 1643.84 1807.64 1911.63

experiments in one workflow on one platform. A B g q g g g gisoform with Gln->pyro-Glu (D mass -17.0264 Da) was selected so this modification wasdd d t th O ll 48% f ll id l fi d ith th t

000 00 00 300 00 500 600 00 800 900m/z

1403 71 Zoom into MS/MS SpectrumMethods added to the sequence. Overall 48% of all residue cleavages were confirmed with the top-down approach This resulted in a combined sequence coverage of 100 %.100ce

1403.71z=6 1446.40

z=61434.56z=61422 72

Zoom into MS/MS Spectrum

ConclusionSample Preparation FIGURE 7: List of the identified glycoforms of the antibodies’ glycopeptidedown approach This resulted in a combined sequence coverage of 100 %.

60

80

unda

nc z=61422.72z=61411.20

z 10

1447.23z=6C ConclusionSample PreparationFor bottom-up experiments, the monoclonal antibody Rituximab was denatured for 30

g y g y p pEEQYNSTYR and TKPREEQYNSTYR using the Byonic search node.40

60

ve A

bu z=101437.71

z=121417.12

z=12 1427.71101444.72

61406.71 1431.721400 19

Here we demonstrate the advantages of the possibilities to implement nodes for Byoincy

min in 7 M Urea and 50 mM Tris HCL at pH 8.00. The sample was reduced with 5 mMDTT for 30 min at 37°C and alkylated by addition of 10 mM IAA for 30 min at room 0

20

Rel

ati z=10 z=6z=6

1431.72z=6

1400.19z=?

and ProSightPC into Proteome Discoverer software for a high-throughput capableanalysis of biologics In the example detailed on this poster results were obtained for:

DTT for 30 min at 37°C, and alkylated by addition of 10 mM IAA for 30 min at roomtemperature. The reaction was quenched by addition of 10 mM DTT. Thermo

1400 1405 1410 1415 1420 1425 1430 1435 1440 1445 1450m/z

0

analysis of biologics. In the example detailed on this poster results were obtained for:

§ The sequence coverage for the light (100%) and heavy chain (100%)

temperature. The reaction was quenched by addition of 10 mM DTT. ThermoScientificTM Pierce™ Trypsin Protease (MS Grade) was added and digestion allowed Second Level Head Deconvolution

m/z

8667 31b80 bb82 § The sequence coverage for the light (100%) and heavy chain (100%).

The identification of various glycoforms of the heavy chains’ glycopeptideto proceed over night at 37°C. Digests were stopped by addition of TFA toapproximately pH 3 00 Body text. AAD80

100

ance

8667.318596.328410.188525.25

b80 b81 b8382

§ The identification of various glycoforms of the heavy chains’ glycopeptide.

V i difi ti f li ht d h h i i ti l id ti d id ti

approximately pH 3.00.The sample for the top-down characterization of the reduced antibody was produced

A83A82D816080

Abu

nda 8525.25

D§ Various modifications of light and heavy chain, in particular oxidation, deamidation

and Gln->pyroGluc conversion

p p y pby offline buffer exchange against 50mM ammonium acetate followed by incubation in10 M TCEP f 1 h t 60°C di tl i t l i 20

40

lativ

e A

8477 158392 08 8428 20 and Gln >pyroGluc conversion.

§ The assessment of relative ratio of the truncation of the terminal lysine residue on the10 mM TCEP for 1 h at 60°C directly prior to analysis.

8400 8450 8500 8550 8600 86500

20

Rel 8477.158392.08 8428.20 8580.26 8688.31

§ The assessment of relative ratio of the truncation of the terminal lysine residue on the heavy chain.Liquid Chromatography FIGURE 8: This table shows the list of identified modifications based on the

FIGURE 2. Base Peak Chromatogram of the monoclonal antibody digest (toptrace) The peptide mixture was separated over a 20min gradient The XIC of m/z

8400 8450 8500 8550 8600 8650m/z y

§ The light chain top-down data confirmed the sequence and supported the bottom-up Thermo Scientific™ Vanquish™ UHPLC system with a 2.1 x 250 mm i.d. ThermoFIGURE 8: This table shows the list of identified modifications based on theincluded search criteria for the bottom-up database search:Gln->pyroGlu,

trace). The peptide mixture was separated over a 20min gradient. The XIC of m/z204.0685 shows where the glycopeptides elute (middle trace). A zoom into the g p q pp p

analysis with 48% obtained residue cleavages.Scientific™ Acclaim™ 120 C18 column and gradients of water and ACN with 0.1% formicacid each were used to separate the peptide mixture For the gradient 2 45% ACN

p pydeamidation and oxidation.

204.0685 shows where the glycopeptides elute (middle trace). A zoom into theaveraged Full MS spectra shows the variety and relative abundances of the various FIGURE 4: The sequence coverage obtained from the bottom-up approach for the

tib di li ht h i i 98 6% O l th t i tid EAK i d d t th

Referencesacid each were used to separate the peptide mixture. For the gradient 2-45% ACNwere ramped for 30min at a flow rate of 0.4 mL/min.

glycoforms and different charge states of the glycopeptide EEQYNSTYR. antibodies light chain is 98.6%. Only the tripeptide EAK was missed due to thesearch algorithms’ capabilities to identify peptides with at least 4 amino acids Referencesp

Mass Spectrometry Base Peak Chromatogramsearch algorithms capabilities to identify peptides with at least 4 amino acids.

A 1. LeDuc, R. D.; Taylor, G. K.; Kim, Y. B.; Januszyk, T. E.; Bynum, L. H.; Sola, J. V.;Mass Spectrometry

Data acquisition was performed on the Thermo ScientificTM Q ExactiveTM Plus and 100 8.42 NL: 2.11E9Base Peak ChromatogramA

Garavelli, J. S.; Kelleher, N. L. ProSight PTM: an integrated environment for proteinidentification and characterization by top down mass spectrometry Nucleic Acids

Data acquisition was performed on the Thermo Scientific Q Exactive Plus andThermo ScientificTM Q ExactiveTM HF mass spectrometers, both of which were 80

100

danc

e

10.739.93

10 77 14 51

Base Peak F: ms MS QE HF Vanquish identification and characterization by top-down mass spectrometry. Nucleic Acids

Res. 2004, 32: W340–W345.equipped with the protein mode feature. Bottom-up experiments were performedusing a HESI source; top down experiments were performed by direct infusion with a 4060

Abu

nd 10.77 14.512.707.99

4 50 5 60

_ _ q_Rituximab_20min

,

2. Durbin KR, Tran JC, Zamdborg L, Sweet SM, Catherman AD, Lee JE, Li M, Kellie JF,using a HESI source; top-down experiments were performed by direct infusion with aThermo Scientific™ Nanospray Flex™ ion source with the static nanospray needle 20

40

elat

ive 4.50 5.602.22 15.10

4.00 6.58 14.029.371.581.48 12 26 16 417 35 18 72 2. Durbin KR, Tran JC, Zamdborg L, Sweet SM, Catherman AD, Lee JE, Li M, Kellie JF,Kelleher NL. Intact mass detection, interpretation, and visualization to automate Top-

p y p ysetup.

FIGURE 9 The truncation of the C terminal lysine residue of the heavy chain is1000R

e 12.26 16.417.35 18.723.93 NL: 1.11E7

m/z= Down proteomics on a large scale. Proteomics. 2010 Oct;10(20):3589-97.Data AnalysisFIGURE 9 The truncation of the C-terminal lysine residue of the heavy chain isquite commonly observed. Using the precursor ion quantification in the bottom-80

100

3.904 58

m/z= 204.08568-204.08772 MS QE HF V i h

XIC m/z 204.0685B3. Leduc RD, Kelleher NL. Using ProSight PTM and related tools for targeted protein

identification and characterization with high mass accuracy tandem MS data CurrData analysis was performed using Proteome Discoverer 2.0 software with the quite commonly observed. Using the precursor ion quantification in the bottomup workflow an assessment of relative quantitation of the peptide with and

40

60 4.58 QE_HF_Vanquish_Rituximab_20min

identification and characterization with high mass accuracy tandem MS data. CurrProtoc Bioinformatics. 2007 Sep;Chapter 13:Unit 13.6.

y p gSequest HT search algorithm and ProsightPD [1-4] and Byonic [5] nodes enabled. without the C-terminal lysine can be performed including different charge states.

The ratio of the truncated to untruncated peptide based on the precursor are is2040

4.595 553 85 12 362 11 12 22 14 486 14 7 78 8 701 39 18 4317 10 p; p

4. Zamdborg L, LeDuc RD, Glowacz KJ, Kim YB, Viswanathan V, Spaulding IT, EarlyBP,The ratio of the truncated to untruncated peptide based on the precursor are is6.2e8:(2.6e6+2.5e7)=23 corresponding to ~4.3% of the untruncated form.0 2 4 6 8 10 12 14 16 18 20

05.553.85 12.362.11 12.22 14.486.14 7.78 8.701.39 18.4317.10 FIGURE 5. The sequence coverage obtained from the bottom-up approach for the

antibodies heavy chain is 100%Results 4. Zamdborg L, LeDuc RD, Glowacz KJ, Kim YB, Viswanathan V, Spaulding IT, EarlyBP,Bluhm EJ, Babai S, Kelleher NL. ProSight PTM 2.0: improved protein identification( ) p g

Time (min)QE_HF_Vanquish_Rituximab_20min #1838-2017 RT: 3.68-4.03 QE_HF_Vanquish_Rituximab_20min #2212-2260 RT: 4.47-4.55

antibodies heavy chain is 100%.Resultsand characterization for top down mass spectrometry. Nucleic Acids Res. 2007 Jul;35:W701 6

_ _ q _ _T: FTMS + p ESI Full ms [200.00-2000.00]

820.60664

_ _ q _ _T: FTMS + p ESI Full ms [200.00-2000.00]

933.03943

In this study we have evaluated the new options in the Proteome Discoverer 2.0ft t i l d kfl d f B i d P Si htPC i bi d b tt :W701-6.

5 Byonic: advanced peptide and protein identification software Curr Protoc Bioinformatics90100 z=4

90

100 z=3

zoom into Full MSzoom into Full MSsoftware to include workflow nodes for Byonic and ProSightPC in a combined bottom-up and top-down analysis of a monoclonal antibody. The bottom-up workflow (Figure C D 5. Byonic: advanced peptide and protein identification software. Curr Protoc Bioinformatics.2012 Dec; Chapter 13:Unit13.20. Bern M, Kil YJ, Becker C80

90

e

869.3573z=1 80

90

e 1399.0565

zoom into Full MSzoom into Full MSup and top down analysis of a monoclonal antibody. The bottom up workflow (Figure1 A) consists of the combination of the major modules: the database search using C D

http://www.ncbi.nlm.nih.gov/pubmed/2325515360

70

ndan

ce

1039.7874 60

70

ndan

ce

1399.0565z=2

987.0570z=3

Sequest HT and Byonic as well as the precursor ion quantification for the assessmentof relative abundances of peptides

6 Acknowledgements5060

e A

bun z=3

904.5111z=1

50

60

e A

bun z=3of relative abundances of peptides.

The top down workflow (Figure 1 B) includes the ProSight High/High Crawler that 6.Acknowledgements30

40

Rel

ativ

e z=1

30

40

Rel

ativ

e

1480.08242

1135.4626z=2

The top-down workflow (Figure 1 B) includes the ProSight High/High Crawler thatperforms the spectra deconvolution based on high resolution mass spectra combined

We would like to thank Martin Samonig for providing the LC-MS data file of the antibody digest

20

R

1147.8228z=3972.09413723 0647

20

R z=21297.5157z=2

1865 40741678 8691

performs the spectra deconvolution based on high resolution mass spectra combinedwith the ProSight absolute mass search.

digest.© 2015 Thermo Fisher Scientific Inc All rights reser ed Seq est is a trademark of the Uni ersit of Washington

0

10 z=3723.0647z=40

10 1865.4074z=3

1678.8691z=5

© 2015 Thermo Fisher Scientific Inc. All rights reserved. Sequest is a trademark of the University of Washington. ProSight is a trademark of Proteinaceous Inc. Byonic is a trademark of Protein Metrics Inc. All other trademarks

800 1000 1200m/z

1000 1500m/z

are the property of Thermo Fisher Scientific and its subsidiaries. This information is not intended to encourage use of these products in any manner that might infringe the intellectual property rights of others.

C bi ti f B tt U d T D Ch t i ti f Bi l i U i Hi h Th h t C bl W kfl iCombination of Bottom-Up and Top-Down Characterization of Biologics Using a High Throughput Capable Workflow in p p g g g g p pP t Di S ftProteome Discoverer Software

S ff 1 2 C 2 2Kai Scheffler1, Torsten Ueckert2, Carmen Paschke2, Bernard Delanghe2, , , gTh Fi h S i tifi 1D i i h 2B GThermo Fisher Scientific, 1Dreieich, 2Bremen, Germany, , , y

O i Results continued FIGURE 6: Sequence coverage of the antibodies' light chain based on the topFIGURE 3. Top-down data of the reduced monoclonal antibody sample showing R lt ti dOverview Results continued FIGURE 6: Sequence coverage of the antibodies' light chain based on the topdown spectrum. Blue lines with kinks towards the left/right indicate theFIGURE 3. Top down data of the reduced monoclonal antibody sample showingthe Full MS spectrum (A), the MS/MS HCD spectrum of a single charge state of the Results continued

Purpose: Demonstrate the benefits of Thermo ScientificTM Proteome DiscovererTM 2.0 For the complete characterization of biologics it is essential to confirm the expectedi i t 100% d i dditi th it i f l ti

do spect u ue es t s to a ds t e e t/ g t d cate t eassignment of corresponding b/y-type fragment ions.antibodies' light chain (B). Trace C shows a zoom into the top-down spectrum

highlighting a series of fragments of charge state z=6 and a number of low Figure 8 lists the identified modifications found in the bottom up approach: oxidation,software for combining a top-down and bottom up protein analysis workflow for in-depth protein characteri ation

sequence aiming at 100% sequence coverage and in addition the monitoring of relativeabundances of glycoforms as well as the levels of modifications such as oxidation

highlighting a series of fragments of charge state z=6 and a number of lowabundant higher charge states z=10,12. Trace D highlights the deconvoluted

gdeamidation and the conversion of the N-termini of both the light and heavy chain from Glnto p ro Gl Fig re 9 highlights the relati e q antitation of the C terminal peptide of thedepth protein characterization.

M th d A S t HT d t b h ith l l ti

abundances of glycoforms as well as the levels of modifications such as oxidation,deamidation and truncation of the heavy chains’ C-terminal lysine residue. Figure 2A

g g g gportion of the top-down spectrum shown in trace C highlighting the partialseq ence aa81 83 DAA of the light chain

to pyro-Glu. Figure 9 highlights the relative quantitation of the C-terminal peptide of theheavy chain with and without the terminal lysine. For the full length peptide both chargeMethods: A Sequest HT database search with precursor area calculation was

combined with a ProSightPD and Byonic searchshows the base peak chromatogram and the separation achieved from the antibodydigest sample with a separation time of 20min Figure 2B is showing the XIC of the

sequence aa81-83 DAA of the light chain. heavy chain with and without the terminal lysine. For the full length peptide both chargestates +1 and +2 were considered, the truncated peptide was detected only in charge state 1combined with a ProSightPD and Byonic search.

Results: A 100% sequence coverage was obtained by combining bottom up and topdigest sample with a separation time of 20min. Figure 2B is showing the XIC of theoxonium ion indicative for the presence of glycopeptides that are shown in Figures 2C

due to the lack of the proton-capturing lysine. The areas obtained for the two versions of thepeptide allow the determination of the ratio truncated:untruncated peptide to be ~23100 1536.74AResults: A 100% sequence coverage was obtained by combining bottom-up and top-down experiments.

p g y p p gand 2D on the Full MS level highlighting the presence of the glycopeptides in its variousi f Th b i d f S HT d B i h i

peptide allow the determination of the ratio truncated:untruncated peptide to be ~23,corresponding to ~4.3% abundance of the untruncated form.Full MS Spectrum80

100

ance

1646.591280 881152 83 1440 83

Ap

I d iisoforms. The obtained sequence coverage from Sequest HT and Byonic are shown inFigures 4 and 5 for the light and heavy chain respectively Figure 7 lists the identified

p g

Figure 3 shows the Full MS spectrum of the reduced antibody with the overlapping charge60Abun

da 1280.881152.83 1440.831048.13 1773.181920 821690 20Introduction Figures 4 and 5 for the light and heavy chain respectively. Figure 7 lists the identifiedglycoforms of the heavy chains’ glycopeptide in two forms, identified with and without a

g p y pp g genvelopes for the light and heavy chains. For top-down analysis and fragmentation a singleh i l d i h 10D i l i i d i d f i20

40

ativ

e A 1920.821690.20 2095.35 2304.77

1950.112327 70

2112.48

In the development as well as the production phase of biologics such as monoclonaltib di th ifi ti f th t i id ll th FIGURE 1 P t Di 2 0 ft kfl f th d t l i f

missed cleavage. charge state was isolated with a 10Da isolation window in order to get a fragment ionspectrum of either light chain (Figure 3B) or the heavy chain (not shown) was obtained0

20

Rel

a 2327.702204.21

antibodies the verification of the correct amino acid sequence as well as theassessment of the presence of the correct type and relative amount of glycosylation

FIGURE 1. Proteome Discoverer 2.0 software workflows for the data analysis ofbottom-up (A) and top-down (B) spectra. The bottom-up workflow consist of

spectrum of either light chain (Figure 3B) or the heavy chain (not shown) was obtained.When top-down analyses are performed involving chromatographic separation the isolation

1000 1200 1400 1600 1800 2000 2200 2400m/zassessment of the presence of the correct type and relative amount of glycosylation

are essential steps and only a few out of many analytical methods applied. Mostbottom up (A) and top down (B) spectra. The bottom up workflow consist ofnodes for protein ID using the Sequest HT search algorithm, the precursor ion window can be widened up to several hundreds of Da in order to co-isolate and fragment

several charge states of either species in parallel Figure 3C highlights the number and100

e

1322.901176 02

commonly, enzymatic digests of the proteins are analyzed in a peptide mapping typeexperiment In addition top down experiments provide additional information regarding

area detector for relative quantitation and the new Byonic node. The top-downworkflow contains the new ProSight TopDown High/High Crawler for spectra

several charge states of either species in parallel. Figure 3C highlights the number anddensity of fragment ions observed, and in particular fragment ions in higher charges statesTop Down MS/MS Spectrum80nda

nce 1176.02 1310.06

1358.981161.60Bexperiment. In addition, top-down experiments provide additional information regarding

mass, amino acid and further modifications. Whereas previously data analysis had toworkflow contains the new ProSight TopDown High/High Crawler for spectradeconvolution combined with the Prosight absolute mass search node.

density of fragment ions observed, and in particular fragment ions in higher charges statesdue to the use of the protein mode available on the Q Exactive mass spectrometer. Figure

p p40

60

e A

bun

1058.52 1511.601235.53 1378 82

1161.60

1011 40, p y y

be run through different software packages, Proteome Discoverer software nowid d th t bl th l i f b th b tt d t d

g3D shows the deconvoluted spectrum of the zoomed MS/MS spectrum highlighting thefragment ion series b -b supporting the amino acid sequence “DAA” The assignment of all20

40

Rel

ativ

e 1378.82 1744.021526.141073.23974.97 1723.01 1781.33

1643 84

1011.40

provides new nodes that enable the analysis of both bottom-up and top-downexperiments in one workflow on one platform A B

fragment ion series b80-b83 supporting the amino acid sequence DAA . The assignment of alldetected fragment ions to the sequence of the light chain is highlighted in Figure 6. The1000 1100 1200 1300 1400 1500 1600 1700 1800 1900

0

R 1643.84 1807.64 1911.63

experiments in one workflow on one platform. A B g q g g g gisoform with Gln->pyro-Glu (D mass -17.0264 Da) was selected so this modification wasdd d t th O ll 48% f ll id l fi d ith th t

000 00 00 300 00 500 600 00 800 900m/z

1403 71 Zoom into MS/MS SpectrumMethods added to the sequence. Overall 48% of all residue cleavages were confirmed with the top-down approach This resulted in a combined sequence coverage of 100 %.100ce

1403.71z=6 1446.40

z=61434.56z=61422 72

Zoom into MS/MS Spectrum

ConclusionSample Preparation FIGURE 7: List of the identified glycoforms of the antibodies’ glycopeptidedown approach This resulted in a combined sequence coverage of 100 %.

60

80

unda

nc z=61422.72z=61411.20

z 10

1447.23z=6C ConclusionSample PreparationFor bottom-up experiments, the monoclonal antibody Rituximab was denatured for 30

g y g y p pEEQYNSTYR and TKPREEQYNSTYR using the Byonic search node.40

60

ve A

bu z=101437.71

z=121417.12

z=12 1427.71101444.72

61406.71 1431.721400 19

Here we demonstrate the advantages of the possibilities to implement nodes for Byoincy

min in 7 M Urea and 50 mM Tris HCL at pH 8.00. The sample was reduced with 5 mMDTT for 30 min at 37°C and alkylated by addition of 10 mM IAA for 30 min at room 0

20

Rel

ati z=10 z=6z=6

1431.72z=6

1400.19z=?

and ProSightPC into Proteome Discoverer software for a high-throughput capableanalysis of biologics In the example detailed on this poster results were obtained for:

DTT for 30 min at 37°C, and alkylated by addition of 10 mM IAA for 30 min at roomtemperature. The reaction was quenched by addition of 10 mM DTT. Thermo

1400 1405 1410 1415 1420 1425 1430 1435 1440 1445 1450m/z

0

analysis of biologics. In the example detailed on this poster results were obtained for:

§ The sequence coverage for the light (100%) and heavy chain (100%)

temperature. The reaction was quenched by addition of 10 mM DTT. ThermoScientificTM Pierce™ Trypsin Protease (MS Grade) was added and digestion allowed Second Level Head Deconvolution

m/z

8667 31b80 bb82 § The sequence coverage for the light (100%) and heavy chain (100%).

The identification of various glycoforms of the heavy chains’ glycopeptideto proceed over night at 37°C. Digests were stopped by addition of TFA toapproximately pH 3 00 Body text. AAD80

100

ance

8667.318596.328410.188525.25

b80 b81 b8382

§ The identification of various glycoforms of the heavy chains’ glycopeptide.

V i difi ti f li ht d h h i i ti l id ti d id ti

approximately pH 3.00.The sample for the top-down characterization of the reduced antibody was produced

A83A82D816080

Abu

nda 8525.25

D§ Various modifications of light and heavy chain, in particular oxidation, deamidation

and Gln->pyroGluc conversion

p p y pby offline buffer exchange against 50mM ammonium acetate followed by incubation in10 M TCEP f 1 h t 60°C di tl i t l i 20

40

lativ

e A

8477 158392 08 8428 20 and Gln >pyroGluc conversion.

§ The assessment of relative ratio of the truncation of the terminal lysine residue on the10 mM TCEP for 1 h at 60°C directly prior to analysis.

8400 8450 8500 8550 8600 86500

20

Rel 8477.158392.08 8428.20 8580.26 8688.31

§ The assessment of relative ratio of the truncation of the terminal lysine residue on the heavy chain.Liquid Chromatography FIGURE 8: This table shows the list of identified modifications based on the

FIGURE 2. Base Peak Chromatogram of the monoclonal antibody digest (toptrace) The peptide mixture was separated over a 20min gradient The XIC of m/z

8400 8450 8500 8550 8600 8650m/z y

§ The light chain top-down data confirmed the sequence and supported the bottom-up Thermo Scientific™ Vanquish™ UHPLC system with a 2.1 x 250 mm i.d. ThermoFIGURE 8: This table shows the list of identified modifications based on theincluded search criteria for the bottom-up database search:Gln->pyroGlu,

trace). The peptide mixture was separated over a 20min gradient. The XIC of m/z204.0685 shows where the glycopeptides elute (middle trace). A zoom into the g p q pp p

analysis with 48% obtained residue cleavages.Scientific™ Acclaim™ 120 C18 column and gradients of water and ACN with 0.1% formicacid each were used to separate the peptide mixture For the gradient 2 45% ACN

p pydeamidation and oxidation.

204.0685 shows where the glycopeptides elute (middle trace). A zoom into theaveraged Full MS spectra shows the variety and relative abundances of the various FIGURE 4: The sequence coverage obtained from the bottom-up approach for the

tib di li ht h i i 98 6% O l th t i tid EAK i d d t th

Referencesacid each were used to separate the peptide mixture. For the gradient 2-45% ACNwere ramped for 30min at a flow rate of 0.4 mL/min.

glycoforms and different charge states of the glycopeptide EEQYNSTYR. antibodies light chain is 98.6%. Only the tripeptide EAK was missed due to thesearch algorithms’ capabilities to identify peptides with at least 4 amino acids Referencesp

Mass Spectrometry Base Peak Chromatogramsearch algorithms capabilities to identify peptides with at least 4 amino acids.

A 1. LeDuc, R. D.; Taylor, G. K.; Kim, Y. B.; Januszyk, T. E.; Bynum, L. H.; Sola, J. V.;Mass Spectrometry

Data acquisition was performed on the Thermo ScientificTM Q ExactiveTM Plus and 100 8.42 NL: 2.11E9Base Peak ChromatogramA

Garavelli, J. S.; Kelleher, N. L. ProSight PTM: an integrated environment for proteinidentification and characterization by top down mass spectrometry Nucleic Acids

Data acquisition was performed on the Thermo Scientific Q Exactive Plus andThermo ScientificTM Q ExactiveTM HF mass spectrometers, both of which were 80

100

danc

e

10.739.93

10 77 14 51

Base Peak F: ms MS QE HF Vanquish identification and characterization by top-down mass spectrometry. Nucleic Acids

Res. 2004, 32: W340–W345.equipped with the protein mode feature. Bottom-up experiments were performedusing a HESI source; top down experiments were performed by direct infusion with a 4060

Abu

nd 10.77 14.512.707.99

4 50 5 60

_ _ q_Rituximab_20min

,

2. Durbin KR, Tran JC, Zamdborg L, Sweet SM, Catherman AD, Lee JE, Li M, Kellie JF,using a HESI source; top-down experiments were performed by direct infusion with aThermo Scientific™ Nanospray Flex™ ion source with the static nanospray needle 20

40

elat

ive 4.50 5.602.22 15.10

4.00 6.58 14.029.371.581.48 12 26 16 417 35 18 72 2. Durbin KR, Tran JC, Zamdborg L, Sweet SM, Catherman AD, Lee JE, Li M, Kellie JF,Kelleher NL. Intact mass detection, interpretation, and visualization to automate Top-

p y p ysetup.

FIGURE 9 The truncation of the C terminal lysine residue of the heavy chain is1000R

e 12.26 16.417.35 18.723.93 NL: 1.11E7

m/z= Down proteomics on a large scale. Proteomics. 2010 Oct;10(20):3589-97.Data AnalysisFIGURE 9 The truncation of the C-terminal lysine residue of the heavy chain isquite commonly observed. Using the precursor ion quantification in the bottom-80

100

3.904 58

m/z= 204.08568-204.08772 MS QE HF V i h

XIC m/z 204.0685B3. Leduc RD, Kelleher NL. Using ProSight PTM and related tools for targeted protein

identification and characterization with high mass accuracy tandem MS data CurrData analysis was performed using Proteome Discoverer 2.0 software with the quite commonly observed. Using the precursor ion quantification in the bottomup workflow an assessment of relative quantitation of the peptide with and

40

60 4.58 QE_HF_Vanquish_Rituximab_20min

identification and characterization with high mass accuracy tandem MS data. CurrProtoc Bioinformatics. 2007 Sep;Chapter 13:Unit 13.6.

y p gSequest HT search algorithm and ProsightPD [1-4] and Byonic [5] nodes enabled. without the C-terminal lysine can be performed including different charge states.

The ratio of the truncated to untruncated peptide based on the precursor are is2040

4.595 553 85 12 362 11 12 22 14 486 14 7 78 8 701 39 18 4317 10 p; p

4. Zamdborg L, LeDuc RD, Glowacz KJ, Kim YB, Viswanathan V, Spaulding IT, EarlyBP,The ratio of the truncated to untruncated peptide based on the precursor are is6.2e8:(2.6e6+2.5e7)=23 corresponding to ~4.3% of the untruncated form.0 2 4 6 8 10 12 14 16 18 20

05.553.85 12.362.11 12.22 14.486.14 7.78 8.701.39 18.4317.10 FIGURE 5. The sequence coverage obtained from the bottom-up approach for the

antibodies heavy chain is 100%Results 4. Zamdborg L, LeDuc RD, Glowacz KJ, Kim YB, Viswanathan V, Spaulding IT, EarlyBP,Bluhm EJ, Babai S, Kelleher NL. ProSight PTM 2.0: improved protein identification( ) p g

Time (min)QE_HF_Vanquish_Rituximab_20min #1838-2017 RT: 3.68-4.03 QE_HF_Vanquish_Rituximab_20min #2212-2260 RT: 4.47-4.55

antibodies heavy chain is 100%.Resultsand characterization for top down mass spectrometry. Nucleic Acids Res. 2007 Jul;35:W701 6

_ _ q _ _T: FTMS + p ESI Full ms [200.00-2000.00]

820.60664

_ _ q _ _T: FTMS + p ESI Full ms [200.00-2000.00]

933.03943

In this study we have evaluated the new options in the Proteome Discoverer 2.0ft t i l d kfl d f B i d P Si htPC i bi d b tt :W701-6.

5 Byonic: advanced peptide and protein identification software Curr Protoc Bioinformatics90100 z=4

90

100 z=3

zoom into Full MSzoom into Full MSsoftware to include workflow nodes for Byonic and ProSightPC in a combined bottom-up and top-down analysis of a monoclonal antibody. The bottom-up workflow (Figure C D 5. Byonic: advanced peptide and protein identification software. Curr Protoc Bioinformatics.2012 Dec; Chapter 13:Unit13.20. Bern M, Kil YJ, Becker C80

90

e

869.3573z=1 80

90

e 1399.0565

zoom into Full MSzoom into Full MSup and top down analysis of a monoclonal antibody. The bottom up workflow (Figure1 A) consists of the combination of the major modules: the database search using C D

http://www.ncbi.nlm.nih.gov/pubmed/2325515360

70

ndan

ce

1039.7874 60

70

ndan

ce

1399.0565z=2

987.0570z=3

Sequest HT and Byonic as well as the precursor ion quantification for the assessmentof relative abundances of peptides

6 Acknowledgements5060

e A

bun z=3

904.5111z=1

50

60

e A

bun z=3of relative abundances of peptides.

The top down workflow (Figure 1 B) includes the ProSight High/High Crawler that 6.Acknowledgements30

40

Rel

ativ

e z=1

30

40

Rel

ativ

e

1480.08242

1135.4626z=2

The top-down workflow (Figure 1 B) includes the ProSight High/High Crawler thatperforms the spectra deconvolution based on high resolution mass spectra combined

We would like to thank Martin Samonig for providing the LC-MS data file of the antibody digest

20

R

1147.8228z=3972.09413723 0647

20

R z=21297.5157z=2

1865 40741678 8691

performs the spectra deconvolution based on high resolution mass spectra combinedwith the ProSight absolute mass search.

digest.© 2015 Thermo Fisher Scientific Inc All rights reser ed Seq est is a trademark of the Uni ersit of Washington

0

10 z=3723.0647z=40

10 1865.4074z=3

1678.8691z=5

© 2015 Thermo Fisher Scientific Inc. All rights reserved. Sequest is a trademark of the University of Washington. ProSight is a trademark of Proteinaceous Inc. Byonic is a trademark of Protein Metrics Inc. All other trademarks

800 1000 1200m/z

1000 1500m/z

are the property of Thermo Fisher Scientific and its subsidiaries. This information is not intended to encourage use of these products in any manner that might infringe the intellectual property rights of others.

C bi ti f B tt U d T D Ch t i ti f Bi l i U i Hi h Th h t C bl W kfl iCombination of Bottom-Up and Top-Down Characterization of Biologics Using a High Throughput Capable Workflow in p p g g g g p pP t Di S ftProteome Discoverer Software

S ff 1 2 C 2 2Kai Scheffler1, Torsten Ueckert2, Carmen Paschke2, Bernard Delanghe2, , , gTh Fi h S i tifi 1D i i h 2B GThermo Fisher Scientific, 1Dreieich, 2Bremen, Germany, , , y

O i Results continued FIGURE 6: Sequence coverage of the antibodies' light chain based on the topFIGURE 3. Top-down data of the reduced monoclonal antibody sample showing R lt ti dOverview Results continued FIGURE 6: Sequence coverage of the antibodies' light chain based on the topdown spectrum. Blue lines with kinks towards the left/right indicate theFIGURE 3. Top down data of the reduced monoclonal antibody sample showingthe Full MS spectrum (A), the MS/MS HCD spectrum of a single charge state of the Results continued

Purpose: Demonstrate the benefits of Thermo ScientificTM Proteome DiscovererTM 2.0 For the complete characterization of biologics it is essential to confirm the expectedi i t 100% d i dditi th it i f l ti

do spect u ue es t s to a ds t e e t/ g t d cate t eassignment of corresponding b/y-type fragment ions.antibodies' light chain (B). Trace C shows a zoom into the top-down spectrum

highlighting a series of fragments of charge state z=6 and a number of low Figure 8 lists the identified modifications found in the bottom up approach: oxidation,software for combining a top-down and bottom up protein analysis workflow for in-depth protein characteri ation

sequence aiming at 100% sequence coverage and in addition the monitoring of relativeabundances of glycoforms as well as the levels of modifications such as oxidation

highlighting a series of fragments of charge state z=6 and a number of lowabundant higher charge states z=10,12. Trace D highlights the deconvoluted

gdeamidation and the conversion of the N-termini of both the light and heavy chain from Glnto p ro Gl Fig re 9 highlights the relati e q antitation of the C terminal peptide of thedepth protein characterization.

M th d A S t HT d t b h ith l l ti

abundances of glycoforms as well as the levels of modifications such as oxidation,deamidation and truncation of the heavy chains’ C-terminal lysine residue. Figure 2A

g g g gportion of the top-down spectrum shown in trace C highlighting the partialseq ence aa81 83 DAA of the light chain

to pyro-Glu. Figure 9 highlights the relative quantitation of the C-terminal peptide of theheavy chain with and without the terminal lysine. For the full length peptide both chargeMethods: A Sequest HT database search with precursor area calculation was

combined with a ProSightPD and Byonic searchshows the base peak chromatogram and the separation achieved from the antibodydigest sample with a separation time of 20min Figure 2B is showing the XIC of the

sequence aa81-83 DAA of the light chain. heavy chain with and without the terminal lysine. For the full length peptide both chargestates +1 and +2 were considered, the truncated peptide was detected only in charge state 1combined with a ProSightPD and Byonic search.

Results: A 100% sequence coverage was obtained by combining bottom up and topdigest sample with a separation time of 20min. Figure 2B is showing the XIC of theoxonium ion indicative for the presence of glycopeptides that are shown in Figures 2C

due to the lack of the proton-capturing lysine. The areas obtained for the two versions of thepeptide allow the determination of the ratio truncated:untruncated peptide to be ~23100 1536.74AResults: A 100% sequence coverage was obtained by combining bottom-up and top-down experiments.

p g y p p gand 2D on the Full MS level highlighting the presence of the glycopeptides in its variousi f Th b i d f S HT d B i h i

peptide allow the determination of the ratio truncated:untruncated peptide to be ~23,corresponding to ~4.3% abundance of the untruncated form.Full MS Spectrum80

100

ance

1646.591280 881152 83 1440 83

Ap

I d iisoforms. The obtained sequence coverage from Sequest HT and Byonic are shown inFigures 4 and 5 for the light and heavy chain respectively Figure 7 lists the identified

p g

Figure 3 shows the Full MS spectrum of the reduced antibody with the overlapping charge60Abun

da 1280.881152.83 1440.831048.13 1773.181920 821690 20Introduction Figures 4 and 5 for the light and heavy chain respectively. Figure 7 lists the identifiedglycoforms of the heavy chains’ glycopeptide in two forms, identified with and without a

g p y pp g genvelopes for the light and heavy chains. For top-down analysis and fragmentation a singleh i l d i h 10D i l i i d i d f i20

40

ativ

e A 1920.821690.20 2095.35 2304.77

1950.112327 70

2112.48

In the development as well as the production phase of biologics such as monoclonaltib di th ifi ti f th t i id ll th FIGURE 1 P t Di 2 0 ft kfl f th d t l i f

missed cleavage. charge state was isolated with a 10Da isolation window in order to get a fragment ionspectrum of either light chain (Figure 3B) or the heavy chain (not shown) was obtained0

20

Rel

a 2327.702204.21

antibodies the verification of the correct amino acid sequence as well as theassessment of the presence of the correct type and relative amount of glycosylation

FIGURE 1. Proteome Discoverer 2.0 software workflows for the data analysis ofbottom-up (A) and top-down (B) spectra. The bottom-up workflow consist of

spectrum of either light chain (Figure 3B) or the heavy chain (not shown) was obtained.When top-down analyses are performed involving chromatographic separation the isolation

1000 1200 1400 1600 1800 2000 2200 2400m/zassessment of the presence of the correct type and relative amount of glycosylation

are essential steps and only a few out of many analytical methods applied. Mostbottom up (A) and top down (B) spectra. The bottom up workflow consist ofnodes for protein ID using the Sequest HT search algorithm, the precursor ion window can be widened up to several hundreds of Da in order to co-isolate and fragment

several charge states of either species in parallel Figure 3C highlights the number and100

e

1322.901176 02

commonly, enzymatic digests of the proteins are analyzed in a peptide mapping typeexperiment In addition top down experiments provide additional information regarding

area detector for relative quantitation and the new Byonic node. The top-downworkflow contains the new ProSight TopDown High/High Crawler for spectra

several charge states of either species in parallel. Figure 3C highlights the number anddensity of fragment ions observed, and in particular fragment ions in higher charges statesTop Down MS/MS Spectrum80nda

nce 1176.02 1310.06

1358.981161.60Bexperiment. In addition, top-down experiments provide additional information regarding

mass, amino acid and further modifications. Whereas previously data analysis had toworkflow contains the new ProSight TopDown High/High Crawler for spectradeconvolution combined with the Prosight absolute mass search node.

density of fragment ions observed, and in particular fragment ions in higher charges statesdue to the use of the protein mode available on the Q Exactive mass spectrometer. Figure

p p40

60

e A

bun

1058.52 1511.601235.53 1378 82

1161.60

1011 40, p y y

be run through different software packages, Proteome Discoverer software nowid d th t bl th l i f b th b tt d t d

g3D shows the deconvoluted spectrum of the zoomed MS/MS spectrum highlighting thefragment ion series b -b supporting the amino acid sequence “DAA” The assignment of all20

40

Rel

ativ

e 1378.82 1744.021526.141073.23974.97 1723.01 1781.33

1643 84

1011.40

provides new nodes that enable the analysis of both bottom-up and top-downexperiments in one workflow on one platform A B

fragment ion series b80-b83 supporting the amino acid sequence DAA . The assignment of alldetected fragment ions to the sequence of the light chain is highlighted in Figure 6. The1000 1100 1200 1300 1400 1500 1600 1700 1800 1900

0

R 1643.84 1807.64 1911.63

experiments in one workflow on one platform. A B g q g g g gisoform with Gln->pyro-Glu (D mass -17.0264 Da) was selected so this modification wasdd d t th O ll 48% f ll id l fi d ith th t

000 00 00 300 00 500 600 00 800 900m/z

1403 71 Zoom into MS/MS SpectrumMethods added to the sequence. Overall 48% of all residue cleavages were confirmed with the top-down approach This resulted in a combined sequence coverage of 100 %.100ce

1403.71z=6 1446.40

z=61434.56z=61422 72

Zoom into MS/MS Spectrum

ConclusionSample Preparation FIGURE 7: List of the identified glycoforms of the antibodies’ glycopeptidedown approach This resulted in a combined sequence coverage of 100 %.

60

80

unda

nc z=61422.72z=61411.20

z 10

1447.23z=6C ConclusionSample PreparationFor bottom-up experiments, the monoclonal antibody Rituximab was denatured for 30

g y g y p pEEQYNSTYR and TKPREEQYNSTYR using the Byonic search node.40

60

ve A

bu z=101437.71

z=121417.12

z=12 1427.71101444.72

61406.71 1431.721400 19

Here we demonstrate the advantages of the possibilities to implement nodes for Byoincy

min in 7 M Urea and 50 mM Tris HCL at pH 8.00. The sample was reduced with 5 mMDTT for 30 min at 37°C and alkylated by addition of 10 mM IAA for 30 min at room 0

20

Rel

ati z=10 z=6z=6

1431.72z=6

1400.19z=?

and ProSightPC into Proteome Discoverer software for a high-throughput capableanalysis of biologics In the example detailed on this poster results were obtained for:

DTT for 30 min at 37°C, and alkylated by addition of 10 mM IAA for 30 min at roomtemperature. The reaction was quenched by addition of 10 mM DTT. Thermo

1400 1405 1410 1415 1420 1425 1430 1435 1440 1445 1450m/z

0

analysis of biologics. In the example detailed on this poster results were obtained for:

§ The sequence coverage for the light (100%) and heavy chain (100%)

temperature. The reaction was quenched by addition of 10 mM DTT. ThermoScientificTM Pierce™ Trypsin Protease (MS Grade) was added and digestion allowed Second Level Head Deconvolution

m/z

8667 31b80 bb82 § The sequence coverage for the light (100%) and heavy chain (100%).

The identification of various glycoforms of the heavy chains’ glycopeptideto proceed over night at 37°C. Digests were stopped by addition of TFA toapproximately pH 3 00 Body text. AAD80

100

ance

8667.318596.328410.188525.25

b80 b81 b8382

§ The identification of various glycoforms of the heavy chains’ glycopeptide.

V i difi ti f li ht d h h i i ti l id ti d id ti

approximately pH 3.00.The sample for the top-down characterization of the reduced antibody was produced

A83A82D816080A

bund

a 8525.25

D§ Various modifications of light and heavy chain, in particular oxidation, deamidation

and Gln->pyroGluc conversion

p p y pby offline buffer exchange against 50mM ammonium acetate followed by incubation in10 M TCEP f 1 h t 60°C di tl i t l i 20

40

lativ

e A

8477 158392 08 8428 20 and Gln >pyroGluc conversion.

§ The assessment of relative ratio of the truncation of the terminal lysine residue on the10 mM TCEP for 1 h at 60°C directly prior to analysis.

8400 8450 8500 8550 8600 86500

20

Rel 8477.158392.08 8428.20 8580.26 8688.31

§ The assessment of relative ratio of the truncation of the terminal lysine residue on the heavy chain.Liquid Chromatography FIGURE 8: This table shows the list of identified modifications based on the

FIGURE 2. Base Peak Chromatogram of the monoclonal antibody digest (toptrace) The peptide mixture was separated over a 20min gradient The XIC of m/z

8400 8450 8500 8550 8600 8650m/z y

§ The light chain top-down data confirmed the sequence and supported the bottom-up Thermo Scientific™ Vanquish™ UHPLC system with a 2.1 x 250 mm i.d. ThermoFIGURE 8: This table shows the list of identified modifications based on theincluded search criteria for the bottom-up database search:Gln->pyroGlu,

trace). The peptide mixture was separated over a 20min gradient. The XIC of m/z204.0685 shows where the glycopeptides elute (middle trace). A zoom into the g p q pp p

analysis with 48% obtained residue cleavages.Scientific™ Acclaim™ 120 C18 column and gradients of water and ACN with 0.1% formicacid each were used to separate the peptide mixture For the gradient 2 45% ACN

p pydeamidation and oxidation.

204.0685 shows where the glycopeptides elute (middle trace). A zoom into theaveraged Full MS spectra shows the variety and relative abundances of the various FIGURE 4: The sequence coverage obtained from the bottom-up approach for the

tib di li ht h i i 98 6% O l th t i tid EAK i d d t th

Referencesacid each were used to separate the peptide mixture. For the gradient 2-45% ACNwere ramped for 30min at a flow rate of 0.4 mL/min.

glycoforms and different charge states of the glycopeptide EEQYNSTYR. antibodies light chain is 98.6%. Only the tripeptide EAK was missed due to thesearch algorithms’ capabilities to identify peptides with at least 4 amino acids Referencesp

Mass Spectrometry Base Peak Chromatogramsearch algorithms capabilities to identify peptides with at least 4 amino acids.

A 1. LeDuc, R. D.; Taylor, G. K.; Kim, Y. B.; Januszyk, T. E.; Bynum, L. H.; Sola, J. V.;Mass Spectrometry

Data acquisition was performed on the Thermo ScientificTM Q ExactiveTM Plus and 100 8.42 NL: 2.11E9Base Peak ChromatogramA

Garavelli, J. S.; Kelleher, N. L. ProSight PTM: an integrated environment for proteinidentification and characterization by top down mass spectrometry Nucleic Acids

Data acquisition was performed on the Thermo Scientific Q Exactive Plus andThermo ScientificTM Q ExactiveTM HF mass spectrometers, both of which were 80

100

danc

e

10.739.93

10 77 14 51

Base Peak F: ms MS QE HF Vanquish identification and characterization by top-down mass spectrometry. Nucleic Acids

Res. 2004, 32: W340–W345.equipped with the protein mode feature. Bottom-up experiments were performedusing a HESI source; top down experiments were performed by direct infusion with a 4060

Abu

nd 10.77 14.512.707.99

4 50 5 60

_ _ q_Rituximab_20min

,

2. Durbin KR, Tran JC, Zamdborg L, Sweet SM, Catherman AD, Lee JE, Li M, Kellie JF,using a HESI source; top-down experiments were performed by direct infusion with aThermo Scientific™ Nanospray Flex™ ion source with the static nanospray needle 20

40

elat

ive 4.50 5.602.22 15.10

4.00 6.58 14.029.371.581.48 12 26 16 417 35 18 72 2. Durbin KR, Tran JC, Zamdborg L, Sweet SM, Catherman AD, Lee JE, Li M, Kellie JF,Kelleher NL. Intact mass detection, interpretation, and visualization to automate Top-

p y p ysetup.

FIGURE 9 The truncation of the C terminal lysine residue of the heavy chain is1000R

e 12.26 16.417.35 18.723.93 NL: 1.11E7

m/z= Down proteomics on a large scale. Proteomics. 2010 Oct;10(20):3589-97.Data AnalysisFIGURE 9 The truncation of the C-terminal lysine residue of the heavy chain isquite commonly observed. Using the precursor ion quantification in the bottom-80

100

3.904 58

m/z= 204.08568-204.08772 MS QE HF V i h

XIC m/z 204.0685B3. Leduc RD, Kelleher NL. Using ProSight PTM and related tools for targeted protein

identification and characterization with high mass accuracy tandem MS data CurrData analysis was performed using Proteome Discoverer 2.0 software with the quite commonly observed. Using the precursor ion quantification in the bottomup workflow an assessment of relative quantitation of the peptide with and

40

60 4.58 QE_HF_Vanquish_Rituximab_20min

identification and characterization with high mass accuracy tandem MS data. CurrProtoc Bioinformatics. 2007 Sep;Chapter 13:Unit 13.6.

y p gSequest HT search algorithm and ProsightPD [1-4] and Byonic [5] nodes enabled. without the C-terminal lysine can be performed including different charge states.

The ratio of the truncated to untruncated peptide based on the precursor are is2040

4.595 553 85 12 362 11 12 22 14 486 14 7 78 8 701 39 18 4317 10 p; p

4. Zamdborg L, LeDuc RD, Glowacz KJ, Kim YB, Viswanathan V, Spaulding IT, EarlyBP,The ratio of the truncated to untruncated peptide based on the precursor are is6.2e8:(2.6e6+2.5e7)=23 corresponding to ~4.3% of the untruncated form.0 2 4 6 8 10 12 14 16 18 20

05.553.85 12.362.11 12.22 14.486.14 7.78 8.701.39 18.4317.10 FIGURE 5. The sequence coverage obtained from the bottom-up approach for the

antibodies heavy chain is 100%Results 4. Zamdborg L, LeDuc RD, Glowacz KJ, Kim YB, Viswanathan V, Spaulding IT, EarlyBP,Bluhm EJ, Babai S, Kelleher NL. ProSight PTM 2.0: improved protein identification( ) p g

Time (min)QE_HF_Vanquish_Rituximab_20min #1838-2017 RT: 3.68-4.03 QE_HF_Vanquish_Rituximab_20min #2212-2260 RT: 4.47-4.55

antibodies heavy chain is 100%.Resultsand characterization for top down mass spectrometry. Nucleic Acids Res. 2007 Jul;35:W701 6

_ _ q _ _T: FTMS + p ESI Full ms [200.00-2000.00]

820.60664

_ _ q _ _T: FTMS + p ESI Full ms [200.00-2000.00]

933.03943

In this study we have evaluated the new options in the Proteome Discoverer 2.0ft t i l d kfl d f B i d P Si htPC i bi d b tt :W701-6.

5 Byonic: advanced peptide and protein identification software Curr Protoc Bioinformatics90100 z=4

90

100 z=3

zoom into Full MSzoom into Full MSsoftware to include workflow nodes for Byonic and ProSightPC in a combined bottom-up and top-down analysis of a monoclonal antibody. The bottom-up workflow (Figure C D 5. Byonic: advanced peptide and protein identification software. Curr Protoc Bioinformatics.2012 Dec; Chapter 13:Unit13.20. Bern M, Kil YJ, Becker C80

90

e

869.3573z=1 80

90

e 1399.0565

zoom into Full MSzoom into Full MSup and top down analysis of a monoclonal antibody. The bottom up workflow (Figure1 A) consists of the combination of the major modules: the database search using C D

http://www.ncbi.nlm.nih.gov/pubmed/2325515360

70

ndan

ce

1039.7874 60

70

ndan

ce

1399.0565z=2

987.0570z=3

Sequest HT and Byonic as well as the precursor ion quantification for the assessmentof relative abundances of peptides

6 Acknowledgements5060

e A

bun z=3

904.5111z=1

50

60

e A

bun z=3of relative abundances of peptides.

The top down workflow (Figure 1 B) includes the ProSight High/High Crawler that 6.Acknowledgements30

40

Rel

ativ

e z=1

30

40

Rel

ativ

e

1480.08242

1135.4626z=2

The top-down workflow (Figure 1 B) includes the ProSight High/High Crawler thatperforms the spectra deconvolution based on high resolution mass spectra combined

We would like to thank Martin Samonig for providing the LC-MS data file of the antibody digest

20

R

1147.8228z=3972.09413723 0647

20

R z=21297.5157z=2

1865 40741678 8691

performs the spectra deconvolution based on high resolution mass spectra combinedwith the ProSight absolute mass search.

digest.© 2015 Thermo Fisher Scientific Inc All rights reser ed Seq est is a trademark of the Uni ersit of Washington

0

10 z=3723.0647z=40

10 1865.4074z=3

1678.8691z=5

© 2015 Thermo Fisher Scientific Inc. All rights reserved. Sequest is a trademark of the University of Washington. ProSight is a trademark of Proteinaceous Inc. Byonic is a trademark of Protein Metrics Inc. All other trademarks

800 1000 1200m/z

1000 1500m/z

are the property of Thermo Fisher Scientific and its subsidiaries. This information is not intended to encourage use of these products in any manner that might infringe the intellectual property rights of others.

2 Combination of Bottom-Up and Top-Down Characterization of Biologics Using a High Throughput Capable Workflow in Proteome Discoverer Software

C bi ti f B tt U d T D Ch t i ti f Bi l i U i Hi h Th h t C bl W kfl iCombination of Bottom-Up and Top-Down Characterization of Biologics Using a High Throughput Capable Workflow in p p g g g g p pP t Di S ftProteome Discoverer Software

S ff 1 2 C 2 2Kai Scheffler1, Torsten Ueckert2, Carmen Paschke2, Bernard Delanghe2, , , gTh Fi h S i tifi 1D i i h 2B GThermo Fisher Scientific, 1Dreieich, 2Bremen, Germany, , , y

O i Results continued FIGURE 6: Sequence coverage of the antibodies' light chain based on the topFIGURE 3. Top-down data of the reduced monoclonal antibody sample showing R lt ti dOverview Results continued FIGURE 6: Sequence coverage of the antibodies' light chain based on the topdown spectrum. Blue lines with kinks towards the left/right indicate theFIGURE 3. Top down data of the reduced monoclonal antibody sample showingthe Full MS spectrum (A), the MS/MS HCD spectrum of a single charge state of the Results continued

Purpose: Demonstrate the benefits of Thermo ScientificTM Proteome DiscovererTM 2.0 For the complete characterization of biologics it is essential to confirm the expectedi i t 100% d i dditi th it i f l ti

do spect u ue es t s to a ds t e e t/ g t d cate t eassignment of corresponding b/y-type fragment ions.antibodies' light chain (B). Trace C shows a zoom into the top-down spectrum

highlighting a series of fragments of charge state z=6 and a number of low Figure 8 lists the identified modifications found in the bottom up approach: oxidation,software for combining a top-down and bottom up protein analysis workflow for in-depth protein characteri ation

sequence aiming at 100% sequence coverage and in addition the monitoring of relativeabundances of glycoforms as well as the levels of modifications such as oxidation

highlighting a series of fragments of charge state z=6 and a number of lowabundant higher charge states z=10,12. Trace D highlights the deconvoluted

gdeamidation and the conversion of the N-termini of both the light and heavy chain from Glnto p ro Gl Fig re 9 highlights the relati e q antitation of the C terminal peptide of thedepth protein characterization.

M th d A S t HT d t b h ith l l ti

abundances of glycoforms as well as the levels of modifications such as oxidation,deamidation and truncation of the heavy chains’ C-terminal lysine residue. Figure 2A

g g g gportion of the top-down spectrum shown in trace C highlighting the partialseq ence aa81 83 DAA of the light chain

to pyro-Glu. Figure 9 highlights the relative quantitation of the C-terminal peptide of theheavy chain with and without the terminal lysine. For the full length peptide both chargeMethods: A Sequest HT database search with precursor area calculation was

combined with a ProSightPD and Byonic searchshows the base peak chromatogram and the separation achieved from the antibodydigest sample with a separation time of 20min Figure 2B is showing the XIC of the

sequence aa81-83 DAA of the light chain. heavy chain with and without the terminal lysine. For the full length peptide both chargestates +1 and +2 were considered, the truncated peptide was detected only in charge state 1combined with a ProSightPD and Byonic search.

Results: A 100% sequence coverage was obtained by combining bottom up and topdigest sample with a separation time of 20min. Figure 2B is showing the XIC of theoxonium ion indicative for the presence of glycopeptides that are shown in Figures 2C

due to the lack of the proton-capturing lysine. The areas obtained for the two versions of thepeptide allow the determination of the ratio truncated:untruncated peptide to be ~23100 1536.74AResults: A 100% sequence coverage was obtained by combining bottom-up and top-down experiments.

p g y p p gand 2D on the Full MS level highlighting the presence of the glycopeptides in its variousi f Th b i d f S HT d B i h i

peptide allow the determination of the ratio truncated:untruncated peptide to be ~23,corresponding to ~4.3% abundance of the untruncated form.Full MS Spectrum80

100

ance

1646.591280 881152 83 1440 83

Ap

I d iisoforms. The obtained sequence coverage from Sequest HT and Byonic are shown inFigures 4 and 5 for the light and heavy chain respectively Figure 7 lists the identified

p g

Figure 3 shows the Full MS spectrum of the reduced antibody with the overlapping charge60Abun

da 1280.881152.83 1440.831048.13 1773.181920 821690 20Introduction Figures 4 and 5 for the light and heavy chain respectively. Figure 7 lists the identifiedglycoforms of the heavy chains’ glycopeptide in two forms, identified with and without a

g p y pp g genvelopes for the light and heavy chains. For top-down analysis and fragmentation a singleh i l d i h 10D i l i i d i d f i20

40

ativ

e A 1920.821690.20 2095.35 2304.77

1950.112327 70

2112.48

In the development as well as the production phase of biologics such as monoclonaltib di th ifi ti f th t i id ll th FIGURE 1 P t Di 2 0 ft kfl f th d t l i f

missed cleavage. charge state was isolated with a 10Da isolation window in order to get a fragment ionspectrum of either light chain (Figure 3B) or the heavy chain (not shown) was obtained0

20

Rel

a 2327.702204.21

antibodies the verification of the correct amino acid sequence as well as theassessment of the presence of the correct type and relative amount of glycosylation

FIGURE 1. Proteome Discoverer 2.0 software workflows for the data analysis ofbottom-up (A) and top-down (B) spectra. The bottom-up workflow consist of

spectrum of either light chain (Figure 3B) or the heavy chain (not shown) was obtained.When top-down analyses are performed involving chromatographic separation the isolation

1000 1200 1400 1600 1800 2000 2200 2400m/zassessment of the presence of the correct type and relative amount of glycosylation

are essential steps and only a few out of many analytical methods applied. Mostbottom up (A) and top down (B) spectra. The bottom up workflow consist ofnodes for protein ID using the Sequest HT search algorithm, the precursor ion window can be widened up to several hundreds of Da in order to co-isolate and fragment

several charge states of either species in parallel Figure 3C highlights the number and100

e

1322.901176 02

commonly, enzymatic digests of the proteins are analyzed in a peptide mapping typeexperiment In addition top down experiments provide additional information regarding

area detector for relative quantitation and the new Byonic node. The top-downworkflow contains the new ProSight TopDown High/High Crawler for spectra

several charge states of either species in parallel. Figure 3C highlights the number anddensity of fragment ions observed, and in particular fragment ions in higher charges statesTop Down MS/MS Spectrum80nda

nce 1176.02 1310.06

1358.981161.60Bexperiment. In addition, top-down experiments provide additional information regarding

mass, amino acid and further modifications. Whereas previously data analysis had toworkflow contains the new ProSight TopDown High/High Crawler for spectradeconvolution combined with the Prosight absolute mass search node.

density of fragment ions observed, and in particular fragment ions in higher charges statesdue to the use of the protein mode available on the Q Exactive mass spectrometer. Figure

p p40

60

e A

bun

1058.52 1511.601235.53 1378 82

1161.60

1011 40, p y y

be run through different software packages, Proteome Discoverer software nowid d th t bl th l i f b th b tt d t d

g3D shows the deconvoluted spectrum of the zoomed MS/MS spectrum highlighting thefragment ion series b -b supporting the amino acid sequence “DAA” The assignment of all20

40

Rel

ativ

e 1378.82 1744.021526.141073.23974.97 1723.01 1781.33

1643 84

1011.40

provides new nodes that enable the analysis of both bottom-up and top-downexperiments in one workflow on one platform A B

fragment ion series b80-b83 supporting the amino acid sequence DAA . The assignment of alldetected fragment ions to the sequence of the light chain is highlighted in Figure 6. The1000 1100 1200 1300 1400 1500 1600 1700 1800 1900

0

R 1643.84 1807.64 1911.63

experiments in one workflow on one platform. A B g q g g g gisoform with Gln->pyro-Glu (D mass -17.0264 Da) was selected so this modification wasdd d t th O ll 48% f ll id l fi d ith th t

000 00 00 300 00 500 600 00 800 900m/z

1403 71 Zoom into MS/MS SpectrumMethods added to the sequence. Overall 48% of all residue cleavages were confirmed with the top-down approach This resulted in a combined sequence coverage of 100 %.100ce

1403.71z=6 1446.40

z=61434.56z=61422 72

Zoom into MS/MS Spectrum

ConclusionSample Preparation FIGURE 7: List of the identified glycoforms of the antibodies’ glycopeptidedown approach This resulted in a combined sequence coverage of 100 %.

60

80

unda

nc z=61422.72z=61411.20

z 10

1447.23z=6C ConclusionSample PreparationFor bottom-up experiments, the monoclonal antibody Rituximab was denatured for 30

g y g y p pEEQYNSTYR and TKPREEQYNSTYR using the Byonic search node.40

60

ve A

bu z=101437.71

z=121417.12

z=12 1427.71101444.72

61406.71 1431.721400 19

Here we demonstrate the advantages of the possibilities to implement nodes for Byoincy

min in 7 M Urea and 50 mM Tris HCL at pH 8.00. The sample was reduced with 5 mMDTT for 30 min at 37°C and alkylated by addition of 10 mM IAA for 30 min at room 0

20

Rel

ati z=10 z=6z=6

1431.72z=6

1400.19z=?

and ProSightPC into Proteome Discoverer software for a high-throughput capableanalysis of biologics In the example detailed on this poster results were obtained for:

DTT for 30 min at 37°C, and alkylated by addition of 10 mM IAA for 30 min at roomtemperature. The reaction was quenched by addition of 10 mM DTT. Thermo

1400 1405 1410 1415 1420 1425 1430 1435 1440 1445 1450m/z

0

analysis of biologics. In the example detailed on this poster results were obtained for:

§ The sequence coverage for the light (100%) and heavy chain (100%)

temperature. The reaction was quenched by addition of 10 mM DTT. ThermoScientificTM Pierce™ Trypsin Protease (MS Grade) was added and digestion allowed Second Level Head Deconvolution

m/z

8667 31b80 bb82 § The sequence coverage for the light (100%) and heavy chain (100%).

The identification of various glycoforms of the heavy chains’ glycopeptideto proceed over night at 37°C. Digests were stopped by addition of TFA toapproximately pH 3 00 Body text. AAD80

100

ance

8667.318596.328410.188525.25

b80 b81 b8382

§ The identification of various glycoforms of the heavy chains’ glycopeptide.

V i difi ti f li ht d h h i i ti l id ti d id ti

approximately pH 3.00.The sample for the top-down characterization of the reduced antibody was produced

A83A82D816080

Abu

nda 8525.25

D§ Various modifications of light and heavy chain, in particular oxidation, deamidation

and Gln->pyroGluc conversion

p p y pby offline buffer exchange against 50mM ammonium acetate followed by incubation in10 M TCEP f 1 h t 60°C di tl i t l i 20

40

lativ

e A

8477 158392 08 8428 20 and Gln >pyroGluc conversion.

§ The assessment of relative ratio of the truncation of the terminal lysine residue on the10 mM TCEP for 1 h at 60°C directly prior to analysis.

8400 8450 8500 8550 8600 86500

20

Rel 8477.158392.08 8428.20 8580.26 8688.31

§ The assessment of relative ratio of the truncation of the terminal lysine residue on the heavy chain.Liquid Chromatography FIGURE 8: This table shows the list of identified modifications based on the

FIGURE 2. Base Peak Chromatogram of the monoclonal antibody digest (toptrace) The peptide mixture was separated over a 20min gradient The XIC of m/z

8400 8450 8500 8550 8600 8650m/z y

§ The light chain top-down data confirmed the sequence and supported the bottom-up Thermo Scientific™ Vanquish™ UHPLC system with a 2.1 x 250 mm i.d. ThermoFIGURE 8: This table shows the list of identified modifications based on theincluded search criteria for the bottom-up database search:Gln->pyroGlu,

trace). The peptide mixture was separated over a 20min gradient. The XIC of m/z204.0685 shows where the glycopeptides elute (middle trace). A zoom into the g p q pp p

analysis with 48% obtained residue cleavages.Scientific™ Acclaim™ 120 C18 column and gradients of water and ACN with 0.1% formicacid each were used to separate the peptide mixture For the gradient 2 45% ACN

p pydeamidation and oxidation.

204.0685 shows where the glycopeptides elute (middle trace). A zoom into theaveraged Full MS spectra shows the variety and relative abundances of the various FIGURE 4: The sequence coverage obtained from the bottom-up approach for the

tib di li ht h i i 98 6% O l th t i tid EAK i d d t th

Referencesacid each were used to separate the peptide mixture. For the gradient 2-45% ACNwere ramped for 30min at a flow rate of 0.4 mL/min.

glycoforms and different charge states of the glycopeptide EEQYNSTYR. antibodies light chain is 98.6%. Only the tripeptide EAK was missed due to thesearch algorithms’ capabilities to identify peptides with at least 4 amino acids Referencesp

Mass Spectrometry Base Peak Chromatogramsearch algorithms capabilities to identify peptides with at least 4 amino acids.

A 1. LeDuc, R. D.; Taylor, G. K.; Kim, Y. B.; Januszyk, T. E.; Bynum, L. H.; Sola, J. V.;Mass Spectrometry

Data acquisition was performed on the Thermo ScientificTM Q ExactiveTM Plus and 100 8.42 NL: 2.11E9Base Peak ChromatogramA

Garavelli, J. S.; Kelleher, N. L. ProSight PTM: an integrated environment for proteinidentification and characterization by top down mass spectrometry Nucleic Acids

Data acquisition was performed on the Thermo Scientific Q Exactive Plus andThermo ScientificTM Q ExactiveTM HF mass spectrometers, both of which were 80

100

danc

e

10.739.93

10 77 14 51

Base Peak F: ms MS QE HF Vanquish identification and characterization by top-down mass spectrometry. Nucleic Acids

Res. 2004, 32: W340–W345.equipped with the protein mode feature. Bottom-up experiments were performedusing a HESI source; top down experiments were performed by direct infusion with a 4060

Abu

nd 10.77 14.512.707.99

4 50 5 60

_ _ q_Rituximab_20min

,

2. Durbin KR, Tran JC, Zamdborg L, Sweet SM, Catherman AD, Lee JE, Li M, Kellie JF,using a HESI source; top-down experiments were performed by direct infusion with aThermo Scientific™ Nanospray Flex™ ion source with the static nanospray needle 20

40

elat

ive 4.50 5.602.22 15.10

4.00 6.58 14.029.371.581.48 12 26 16 417 35 18 72 2. Durbin KR, Tran JC, Zamdborg L, Sweet SM, Catherman AD, Lee JE, Li M, Kellie JF,Kelleher NL. Intact mass detection, interpretation, and visualization to automate Top-

p y p ysetup.

FIGURE 9 The truncation of the C terminal lysine residue of the heavy chain is1000R

e 12.26 16.417.35 18.723.93 NL: 1.11E7

m/z= Down proteomics on a large scale. Proteomics. 2010 Oct;10(20):3589-97.Data AnalysisFIGURE 9 The truncation of the C-terminal lysine residue of the heavy chain isquite commonly observed. Using the precursor ion quantification in the bottom-80

100

3.904 58

m/z= 204.08568-204.08772 MS QE HF V i h

XIC m/z 204.0685B3. Leduc RD, Kelleher NL. Using ProSight PTM and related tools for targeted protein

identification and characterization with high mass accuracy tandem MS data CurrData analysis was performed using Proteome Discoverer 2.0 software with the quite commonly observed. Using the precursor ion quantification in the bottomup workflow an assessment of relative quantitation of the peptide with and

40

60 4.58 QE_HF_Vanquish_Rituximab_20min

identification and characterization with high mass accuracy tandem MS data. CurrProtoc Bioinformatics. 2007 Sep;Chapter 13:Unit 13.6.

y p gSequest HT search algorithm and ProsightPD [1-4] and Byonic [5] nodes enabled. without the C-terminal lysine can be performed including different charge states.

The ratio of the truncated to untruncated peptide based on the precursor are is2040

4.595 553 85 12 362 11 12 22 14 486 14 7 78 8 701 39 18 4317 10 p; p

4. Zamdborg L, LeDuc RD, Glowacz KJ, Kim YB, Viswanathan V, Spaulding IT, EarlyBP,The ratio of the truncated to untruncated peptide based on the precursor are is6.2e8:(2.6e6+2.5e7)=23 corresponding to ~4.3% of the untruncated form.0 2 4 6 8 10 12 14 16 18 20

05.553.85 12.362.11 12.22 14.486.14 7.78 8.701.39 18.4317.10 FIGURE 5. The sequence coverage obtained from the bottom-up approach for the

antibodies heavy chain is 100%Results 4. Zamdborg L, LeDuc RD, Glowacz KJ, Kim YB, Viswanathan V, Spaulding IT, EarlyBP,Bluhm EJ, Babai S, Kelleher NL. ProSight PTM 2.0: improved protein identification( ) p g

Time (min)QE_HF_Vanquish_Rituximab_20min #1838-2017 RT: 3.68-4.03 QE_HF_Vanquish_Rituximab_20min #2212-2260 RT: 4.47-4.55

antibodies heavy chain is 100%.Resultsand characterization for top down mass spectrometry. Nucleic Acids Res. 2007 Jul;35:W701 6

_ _ q _ _T: FTMS + p ESI Full ms [200.00-2000.00]

820.60664

_ _ q _ _T: FTMS + p ESI Full ms [200.00-2000.00]

933.03943

In this study we have evaluated the new options in the Proteome Discoverer 2.0ft t i l d kfl d f B i d P Si htPC i bi d b tt :W701-6.

5 Byonic: advanced peptide and protein identification software Curr Protoc Bioinformatics90100 z=4

90

100 z=3

zoom into Full MSzoom into Full MSsoftware to include workflow nodes for Byonic and ProSightPC in a combined bottom-up and top-down analysis of a monoclonal antibody. The bottom-up workflow (Figure C D 5. Byonic: advanced peptide and protein identification software. Curr Protoc Bioinformatics.2012 Dec; Chapter 13:Unit13.20. Bern M, Kil YJ, Becker C80

90

e

869.3573z=1 80

90

e 1399.0565

zoom into Full MSzoom into Full MSup and top down analysis of a monoclonal antibody. The bottom up workflow (Figure1 A) consists of the combination of the major modules: the database search using C D

http://www.ncbi.nlm.nih.gov/pubmed/2325515360

70

ndan

ce

1039.7874 60

70

ndan

ce

1399.0565z=2

987.0570z=3

Sequest HT and Byonic as well as the precursor ion quantification for the assessmentof relative abundances of peptides

6 Acknowledgements5060

e A

bun z=3

904.5111z=1

50

60

e A

bun z=3of relative abundances of peptides.

The top down workflow (Figure 1 B) includes the ProSight High/High Crawler that 6.Acknowledgements30

40

Rel

ativ

e z=1

30

40

Rel

ativ

e

1480.08242

1135.4626z=2

The top-down workflow (Figure 1 B) includes the ProSight High/High Crawler thatperforms the spectra deconvolution based on high resolution mass spectra combined

We would like to thank Martin Samonig for providing the LC-MS data file of the antibody digest

20

R

1147.8228z=3972.09413723 0647

20

R z=21297.5157z=2

1865 40741678 8691

performs the spectra deconvolution based on high resolution mass spectra combinedwith the ProSight absolute mass search.

digest.© 2015 Thermo Fisher Scientific Inc All rights reser ed Seq est is a trademark of the Uni ersit of Washington

0

10 z=3723.0647z=40

10 1865.4074z=3

1678.8691z=5

© 2015 Thermo Fisher Scientific Inc. All rights reserved. Sequest is a trademark of the University of Washington. ProSight is a trademark of Proteinaceous Inc. Byonic is a trademark of Protein Metrics Inc. All other trademarks

800 1000 1200m/z

1000 1500m/z

are the property of Thermo Fisher Scientific and its subsidiaries. This information is not intended to encourage use of these products in any manner that might infringe the intellectual property rights of others.

C bi ti f B tt U d T D Ch t i ti f Bi l i U i Hi h Th h t C bl W kfl iCombination of Bottom-Up and Top-Down Characterization of Biologics Using a High Throughput Capable Workflow in p p g g g g p pP t Di S ftProteome Discoverer Software

S ff 1 2 C 2 2Kai Scheffler1, Torsten Ueckert2, Carmen Paschke2, Bernard Delanghe2, , , gTh Fi h S i tifi 1D i i h 2B GThermo Fisher Scientific, 1Dreieich, 2Bremen, Germany, , , y

O i Results continued FIGURE 6: Sequence coverage of the antibodies' light chain based on the topFIGURE 3. Top-down data of the reduced monoclonal antibody sample showing R lt ti dOverview Results continued FIGURE 6: Sequence coverage of the antibodies' light chain based on the topdown spectrum. Blue lines with kinks towards the left/right indicate theFIGURE 3. Top down data of the reduced monoclonal antibody sample showingthe Full MS spectrum (A), the MS/MS HCD spectrum of a single charge state of the Results continued

Purpose: Demonstrate the benefits of Thermo ScientificTM Proteome DiscovererTM 2.0 For the complete characterization of biologics it is essential to confirm the expectedi i t 100% d i dditi th it i f l ti

do spect u ue es t s to a ds t e e t/ g t d cate t eassignment of corresponding b/y-type fragment ions.antibodies' light chain (B). Trace C shows a zoom into the top-down spectrum

highlighting a series of fragments of charge state z=6 and a number of low Figure 8 lists the identified modifications found in the bottom up approach: oxidation,software for combining a top-down and bottom up protein analysis workflow for in-depth protein characteri ation

sequence aiming at 100% sequence coverage and in addition the monitoring of relativeabundances of glycoforms as well as the levels of modifications such as oxidation

highlighting a series of fragments of charge state z=6 and a number of lowabundant higher charge states z=10,12. Trace D highlights the deconvoluted

gdeamidation and the conversion of the N-termini of both the light and heavy chain from Glnto p ro Gl Fig re 9 highlights the relati e q antitation of the C terminal peptide of thedepth protein characterization.

M th d A S t HT d t b h ith l l ti

abundances of glycoforms as well as the levels of modifications such as oxidation,deamidation and truncation of the heavy chains’ C-terminal lysine residue. Figure 2A

g g g gportion of the top-down spectrum shown in trace C highlighting the partialseq ence aa81 83 DAA of the light chain

to pyro-Glu. Figure 9 highlights the relative quantitation of the C-terminal peptide of theheavy chain with and without the terminal lysine. For the full length peptide both chargeMethods: A Sequest HT database search with precursor area calculation was

combined with a ProSightPD and Byonic searchshows the base peak chromatogram and the separation achieved from the antibodydigest sample with a separation time of 20min Figure 2B is showing the XIC of the

sequence aa81-83 DAA of the light chain. heavy chain with and without the terminal lysine. For the full length peptide both chargestates +1 and +2 were considered, the truncated peptide was detected only in charge state 1combined with a ProSightPD and Byonic search.

Results: A 100% sequence coverage was obtained by combining bottom up and topdigest sample with a separation time of 20min. Figure 2B is showing the XIC of theoxonium ion indicative for the presence of glycopeptides that are shown in Figures 2C

due to the lack of the proton-capturing lysine. The areas obtained for the two versions of thepeptide allow the determination of the ratio truncated:untruncated peptide to be ~23100 1536.74AResults: A 100% sequence coverage was obtained by combining bottom-up and top-down experiments.

p g y p p gand 2D on the Full MS level highlighting the presence of the glycopeptides in its variousi f Th b i d f S HT d B i h i

peptide allow the determination of the ratio truncated:untruncated peptide to be ~23,corresponding to ~4.3% abundance of the untruncated form.Full MS Spectrum80

100

ance

1646.591280 881152 83 1440 83

Ap

I d iisoforms. The obtained sequence coverage from Sequest HT and Byonic are shown inFigures 4 and 5 for the light and heavy chain respectively Figure 7 lists the identified

p g

Figure 3 shows the Full MS spectrum of the reduced antibody with the overlapping charge60Abun

da 1280.881152.83 1440.831048.13 1773.181920 821690 20Introduction Figures 4 and 5 for the light and heavy chain respectively. Figure 7 lists the identifiedglycoforms of the heavy chains’ glycopeptide in two forms, identified with and without a

g p y pp g genvelopes for the light and heavy chains. For top-down analysis and fragmentation a singleh i l d i h 10D i l i i d i d f i20

40

ativ

e A 1920.821690.20 2095.35 2304.77

1950.112327 70

2112.48

In the development as well as the production phase of biologics such as monoclonaltib di th ifi ti f th t i id ll th FIGURE 1 P t Di 2 0 ft kfl f th d t l i f

missed cleavage. charge state was isolated with a 10Da isolation window in order to get a fragment ionspectrum of either light chain (Figure 3B) or the heavy chain (not shown) was obtained0

20

Rel

a 2327.702204.21

antibodies the verification of the correct amino acid sequence as well as theassessment of the presence of the correct type and relative amount of glycosylation

FIGURE 1. Proteome Discoverer 2.0 software workflows for the data analysis ofbottom-up (A) and top-down (B) spectra. The bottom-up workflow consist of

spectrum of either light chain (Figure 3B) or the heavy chain (not shown) was obtained.When top-down analyses are performed involving chromatographic separation the isolation

1000 1200 1400 1600 1800 2000 2200 2400m/zassessment of the presence of the correct type and relative amount of glycosylation

are essential steps and only a few out of many analytical methods applied. Mostbottom up (A) and top down (B) spectra. The bottom up workflow consist ofnodes for protein ID using the Sequest HT search algorithm, the precursor ion window can be widened up to several hundreds of Da in order to co-isolate and fragment

several charge states of either species in parallel Figure 3C highlights the num