High Field Asymmetric Waveform Ion Mobility Spectrometry (FAIMS)expands the comprehensiveness and precision of multiplex proteomics

1Ins�tute for Research in Immunology and Cancer (IRIC); 2Department of Chemistry, Université de Montréal, Montréal, QC; 3ThermoFisher Scien�fic, San Jose, CA

Sibylle Pfamma�er1,2; Eric Bonneil1; Francis P. McManus1; Satendra Prasad3; Derek J. Bailey3; Michael Belford3; Jean-Jacques Dunyach3; Pierre Thibault1,2

OVERVIEWAIM:Improved coverage and quantification in proteomic analysis using differential ion mobility.METHODS: A novel high field asymmetric waveform ion mobility (FAIMS) interface was coupled to an Orbitrap Fusion Tribrid (Thermo Fisher Scientific). Isobaric labeling of peptides was used to profile dynamic changes in protein abundance upon heat shock.RESULTS: LC-FAIMS-MS2 provided 2.5-fold higher number of quantifiable peptides compared to the SPS-MS3 strategy for TMT labeling with comparable fold changes. measurements.INTRODUCTION

AKNOWLEDGMENTS

1. Pfammatter et al., J. Proteome Res., 2016, 15 (12), pp 4653–46652. Keshishian et al., Nat. Protoc., 2017, 12, 1683-17013. Bonneil et al., J. Mass Spectrometry., 2015, 50 (11), pp 1181-1195

METHOD

RESULTS RESULTS

SF3A1

SRRM2

CDC5L

SNRPG

PRPF6CRNKL1

SF3B2

PRPF31

LSM2

EIF3J

EIF3L

TPM1

THOC1

DIS3

EIF3GACTN1

THOC6

EIF3B

DDX27

PTCD3

GNL3

RPS17L

WDR33

ZC3H18

PKN2

HNRNPC

HNRNPA0

CPSF1

YTHDC2

DDX6

PNNRBMX

USP10

C14orf166

DDX1

CHD1SRP14

PAF1

C22orf28

TIMM10

CTR9

ALDH18A1 SRP9

SDE2

LUC7L3

CYP51A1THRAP3

SUPT16H

PFAS

PRPF3

PRPF19

NHP2L1

DHX38

SNRPB2DDX23

NAA38

SLIRP

DDX46DDX3X

TIMM13

TIMM8BU2AF2PPP1R8

GMPS

MFAP1

SMN1

EIF1B

COPG1

CLINT1

DNAJA1

LMAN1

CCT2

DNAJB4

HSPE1

HUWE1

SSR1

RANBP2PHB2

NUCKS1

DNAJB1

NDUFA4

COX4I1

COX5A VDAC3

NDUFA2

MYBBP1A

UTP6

DHX37NOP14NAP1L1

WDR75

WDR43

H1FXUTP20

NOL6

CDK1

HSPA6STUB1

SUPT6H

CDK4SUGT1

LMAN2SDHB

RCC1

SCP2UQCRC1

NDUFA7

GTF2ITHOC2

MATR3

NDUFA5

NDUFS3

UNC45A

HSPA1A

COPA

HSPA8

AHSA1

HSP90AA1

RUVBL1SYAP1

PRKDCXRCC6

NTPCR

XRCC4 SDHA

TERF2IP

XRCC5

ARPC1BACTR3

CPSF6

PDS5A

WAPAL

ASUN

BCL7A

ARHGAP17

ARPC5

SMARCA5

SMARCA1

SMARCC2

BAZ1B

SMARCC1

POLE3

SMARCB1

PDCD10

HEATR2

PPP2R1A

INTS3

SLC25A3

ANKHD1

PPP4R2

AFG3L2

TMCO1

PPP4C

PPP2R5D

CAPZB

PLS3

CFL1

WDR1TXN

TYMS

DCTN4

CAPZA2

PSMB3UCHL5

UBFD1

PSMB1

PSMA4

G3BP1

PSMA7

ADRM1

PSMB4

NVL

NCL RRP12

POLA2

UBA1

TSR1PDCD11

UBE2N

BMS1

RPL19

RPS18

RPL23ARPL6

RPL38

MRPS31

RPS16

MRPL46RSL1D1 TXLNA

MRPL37MRPL38

MRPL39

RPL7

MRPS7MRPL12RPL31

RPL4

RPL13A

RPL24

RPS15ARPL35

RPL13RPL18ADAP3

RPS19

RPL5

EEF2

SF1

RPL7A

BOP1

RRP1B

RPS28

RPL35ARPL34

RPS13RPS15RPL17

RPL14

RPL3

RPS26

SRSF11

TRA2A

DBR1AARS

IK

MSH2TSR3

MSH6

YBX1

SRSF7SRSF3

SLC3A2

GTF3C5

TRA2B

GTF3C1

SMU1

MRPL55

RTFDC1LARS

IARS2

EPRS

GARS

MRPS30

GMNNCCNA2

HSPH1

COX6B1 ATP5C1

RPA3

SRRT

MCM3

FASN

RFC4

MCM2

Ribosomep-value 1.9e-26

Proteasome complexp-value 4.98e-5

Response tounfolded proteinp-value 1.01e-4

Mitochondrial respiratory chainp-value 9.27e-4

Spliceosomal complexp-value 7.64e-14

Chromatin remodelingp-value 2.76e-4 Mitochondrial

Ribosomes

CytoplasmicRibosome

Time (h)

9h1h

EARLY Response LATE Response

Down RegulatedUp Regulated

43°C

a

b

ChromatinRemodelling

MitochondrialProteins

Cell-Celladhesion

Cell-Celladhesion

MitochondrialProteins

ProteasomeComplex

HSPs aggregation

missfolding

Responseto unfolded

proteins

RibosomeSpliceosome

Exon 1 Intron Exon 2

Exon 1 Exon 2

Spliceosome

Exon 1 Intron Exon 2

Exon 1 Exon 2

Spliceosome/SpliceosomeSpliceosome/Spliceosome

SPS-MS3 FAIMS-MS2# MS 20437 22450# MS2# MS3

105820105797

124005

# identified PSM(peptide sequences)

27146(13642)

49968(38374)

# identfied protein groups(2 uniques peptides)

1260 2673

# quantified PSM(peptide sequences)

25142(12400)

44490(30848)

# quantified protein groups(2 uniques peptides)

1229 2646

# dynamic proteins 375 902

b

a

# un

ique

pep

tides

iden

tifie

d

# Injections

965022.8%

2872467.8%

39929.4%

withoutFAIMS

withFAIMS

0

50001000015000200002500030000350004000045000

1 2 3

119543.6%

147854.0%

652.4%

withoutFAIMS

withFAIMS

peptides

protein groups

* quantified = present in min 7 TMT channels

# cluster proteins 149 502

−2−1

01

2

n=54

− − −

without FAIMS with FAIMS

Prot

ein

grou

ps (c

lust

er m

embr

eshi

p >0

.9)

Clu

ster

4

− − −

-1.5-1.0

-0.5 0.0 0.5 1.0 1.5

−

log2 FC (HS/CTR)

0 1 2 3 4 5 6 7 8 9Time( h)

0

100

200

300

400

500

Clu

ster

3C

lust

er 2

Clu

ster

1

Clu

ster

4C

lust

er 3

Clu

ster

2C

lust

er 1

0 1 2 3 4 5 6 7 8 9Time( h)

0

20

40

60

120

140

80

100

Prot

ein

grou

ps (c

lust

er m

embr

eshi

p >0

.9)

00.20.40.60.81.01.21.41.61.8

HSPA1A

Clu

ster

1

-0.5

-0.4

-0.3

-0.2

-0.1

0

0.1

LRPPRC

Clu

ster

3

log2

FC

(HS/

CTR

)lo

g2 F

C(H

S/C

TR)

SPS-MS3

FAIMS-MS2

SPS-MS3

FAIMS-MS2

0 1 2 3 4 5 6 7 8 90 1 2 3 4 5 6 7 8 9

0 1 2 3 4 5 6 7 8 90 1 2 3 4 5 6 7 8 9

0 1 2 3 4 5 6 7 8 90 1 2 3 4 5 6 7 8 9

0 1 2 3 4 5 6 7 8 90 1 2 3 4 5 6 7 8 9

without FAIMS

Cluster 1Late Up

Cluster 2Early up

Cluster 3

Late Down

Cluster 4

Early down

Time HS (h)

log 2 F

old

Cha

nge

(ord

er n

=2)

−2−1

01

2−2

−10

12

−2−1

01

2−2

−10

12

−2−1

01

2−2

−10

12

n=116

n=153

n=56

n=177

n=41

−2−1

01

2

n=24

n=30

rela

tive

Fold

Cha

nge

0.0

0.2

0.4

0.6

0.1

0.2

0.3

−0.4

−0.3

−0.2

−0.1

0.0

−0.5

−0.4

−0.3

−0.2

−0.1

0.0

0.1

0.2

0.3

0.4

0.1

0.2

−0.3

−0.2

−0.1

0.0

−0.4

−0.3

−0.2

−0.1

Time HS (h)

1 2 3 4 5 6 7 8 9

1 2 3 4 5 6 7 8 9

with FAIMSc d

log2 FC(SPS-MS3)

log2

FC

(FA

IMS-

MS2

)

−2

−1

0

1

2

3

−2 −1 0 1 2 3

8h HS

protein groups

e

f

Fig.1: Experimental workflow. (a) All experiments were performed using LC-MS/MS on a Orbitrap tribrid Fusion mass spectrometer with a two hours LC gradient, 500ng/injection with 3s cycles. For SPS-MS3, 10 notches were selected2. Without FAIMS we did 3 replicates, with FAIMS we combined CVs together to cover in 3 injections the whole transmission range. (b) Optimised FAIMS method shows little overlap between the different injections and comparable number of identification. (c) Yeast and HEK293 cell extracts were reduced, alkylated and digested with trypsin prior to labeling with TMT 6-plex. (d) HEK293 cells were incubated at 43°C, and collected at 1h intervals up to 9 h prior to digestion and TMT 10-plex labeling.

d

TMT labeling

TMT10 126TMT10 127N

TMT10 128N

TMT10 129N

TMT10 130N

TMT10 131

Protein extractionTryptic digestion

0h

1h

2h

...

8h

9h

Human

Heat Shock43°C TMT10 127C

TMT10 128C

TMT10 129C

TMT10 130C

1:1:1:1:1:1:1:1:1:1

LC-MS/MS/MS (SPS)

LC-FAIMS-MS/MS3 x 2h

MS InletESI

MS Inlet

CVESI

IE 100°C - OE 100°C

aMS-MS/MS MS-MS/MS

MS-MS/MS

MS-MS/MS

CV1

CV Transmission range: -37V to -93V

MS-MS/MSCV2

MS-MS/MSCV3

MS-MS/MSCV4

MS-MS/MSCV5

MS-MS/MSCV6

MS-MS/MSCV7

MS-MS/MSCV8

MS-MS/MSCV9

-37V /-44V /-51V

-58V /-65V

-73V /-80V /-87V /-93 V

Rep1

Rep3

Rep2

Inj #1

Inj #3

Inj #2

without FAIMS with FAIMS

-58V/-65V

-73V/-80V/-87V/-93V

-37V/-44V/-51V

1%

26%

4%

29%

4%

7% 29%

b

Yeast

TMT labeling

Yeas

t

4 : 6 : 4 : 0 : 0 : 0

Hum

an

0 : 2.5: 4 : 10: 4 : 2.5

Human

Tryptic digestion

WIT

HO

UT

Inte

rfere

nces

WIT

H In

terfe

renc

es

Interferences

1 : 1.5 : 1

1 : 1.6 : 4 : 1.6 : 1

TMT6 126

TMT6 131

TMT6 127

TMT6 128

TMT6 129

TMT6 130

c

Isobaric labeling of peptides provides a convenient approach to enhance the throughput of quantitative proteomic measurements, and can be achieved using different reagents including Tandem Mass Tag (TMT) labeling. However, the fragmentation of co-eluting isobaric ions can lead to chimeric MS/MS spectra and distorted reporter ion ratios. Synchronous precursor selection (SPS)-based MS3 method can alleviate this problem, though this approach can result in a reduced number of quantifiable proteins compared to traditional MS2 method. Here, we compared the analytical merits of SPS-MS3 to that of LC-MS/MS that combines a new high field asymmetric waveform ion mobility spectrometry (FAIMS) interface. LC-MS/MS experiments performed using FAIMS enhanced peak capacity and sensitivity while reducing peptide co-fragmentation thus extending the coverage of multiplex proteomic measurements1.

• The New FAIMS interface provides significant advantages in terms of instrument speed compared to the old generation FAIMS, allowing in three injections to cover the CV transmission range with optimized CV distribution (Fig.1a and Fig.1b).

• Using a known two-proteome model of Yeast - Human peptides (Fig.1b), FAIMS MS2 improves peak capacity3 and reduces the occurence of co-fragmentation of isobaric precursors with precision comparable to the SPS MS3 approach (Fig.2).

• The dynamic changes of HEK293 cells exposed to hyperthermia was analyzed with total 1.5ug of peptides. FAIMS increases the number of quantifiable TMT-labeled peptides by 3-fold and proteins by 2-fold compare to the SPS MS3-based strategy for LC-MS/MS analyses (Fig.3b).

• Proteins showing changes in abundance upon heat shock were separated in 4 groups: early up- or down-regulation and late up- or down-regulation and showed similar trends for FAIMS and SPS (Fig.3c-d).

• Hyperthermia affects several key cellular processes that impact protein homeostasis, such as responses to unfolded proteins (Fig.4).

b

* quantified = present in min 2 TMT channels2 instrumental replicates for SPS and FAIMS

Human126 Human127 Human128 Human129 Human130

Yeast127 Yeast128 Yeast129 Yeast130 Yeast131

0 2 4 6 0 2 4 6 0 2 4 6 0 2 4 6 0 2 4 6

0

1000

2000

3000

0

1000

2000

3000

0.22

1.221.62

4.12

1.541

11.57

1.130.53

0.18 0.10

0

2

4

6

0.08

1.121.65

4.32

1.601

11.55

1.06

0.24 0.06 0.03

TMT61260

2

4

6

SPS-MS3FAIMS-MS2

TMT6127 TMT6128 TMT6129 TMT6130 TMT6131

TMT6126 TMT6127 TMT6128 TMT6129 TMT6130 TMT6131

Interferences Human

Yeast Interferences

0 2 4 6 0 2 4 6 0 2 4 6 0 2 4 6 0 2 4 6

Interferences

Interferences

ratio to TMT6 131

ratio to TMT6 126

# un

ique

pep

tides

# un

ique

pep

tides

ratio

to T

MT 6 1

31ra

tio to

TM

T 6 126

c dYeast

Yeas

t

4 : 6 : 4 : 0 : 0 : 0

Hum

an

0 : 2.5: 4 : 10: 4 : 2.5

Human

1 : 1.5 : 1 1 : 1.6 : 4 : 1.6 : 1

500ng/injectiona3 x 2h RP LC

2 instrumental replicates

SPS-MS3 FAIMS-MS2

# identified PSM

• peptide sequences 28232

• 9354 79302

• 40631

# identified protein

(# uniques peptides)

1120 (>2);

301 (=2);

485 (=1)

3987 (>2);

798 (=2);

1226 (=1)

# quantified PSM

• peptide sequences 26158

• 8685 77373

• 39636

# quantified protein groups 1062 3424

# quantified unique

peptide sequences 6074 28088

• Yeast • 2240 • 9444

• Human • 3834 • 16844

Fig.2: Precision of TMT quantification (a) To determine the extent of co-fragmentation using FAIMS, we labeled separate aliquots of yeast and HEK293 tryptic digests with TMT reagents, and mixed those aliquots to obtain final TMT ratios of 1:1.5:1:0:0:0 for yeast and 0:1:1.6:4:1.6:1 for human extracts. Human peptides were not labeled with TMT-126, whereas channels TMT-129 to TMT-131 were not used for yeast peptides. This labeling scheme facilitated the identification of co-fragmentation arising from interfering peptides of each species. 70% of all unique sequences in SPS-MS3 were also quantified in FAIMS-MS2. (b) Summary table comparing MS analysis between SPS-MS3 (middle column) and FAIMS-MS2 (right column) for identified and quantified features. (c) Distortion of TMT ion ratios and extent of ion contamination for the two-proteome model analysed with SPS-MS3 (grey) and FAIMS-MS2 (green). Box plot and (d) frequency distribution of TMT reporter ion ratios normalized using TMT-126 and TMT-131 for for unique yeast and human peptide sequences, respectively.

Fig.3: FAIMS improves TMT quantification of the human proteome. HEK293 cells were exposed to a 43˚C heat stress for up to 9 h in 1h increments. (a) Cumulative number of unique peptides identified as a function of repeat injections for SPS-MS3 (black) or CV stepping program with FAIMS-MS2 (green). Overlap in peptide and protein identifications between the two methods are depicted in Venn diagrams to the right of the curves.(b) Summary table comparing MS analysis parameters between SPS-MS3 (middle column) and FAIMS-MS2 (right column). (c) Dynamic clusters for heat shock regulated proteins without FAIMS (left) and with FAIMS (right). The grey lines show the relative fold changes of the individual proteins with high membership (≥0.9), the blue lines the average fold changes of all the proteins in the corresponding cluster. (d) Corresponding heat map for all proteins in the clusters from (c). (e) Representative dynamic profile of HSPA1A (assigned to a late up regulation) and LRPPRC (assigned to a late up downregulation) for SPS-MS3 (grey) and FAIMS-MS2 analysis (green), highlighting virtually identical profiles/quantifications for both acquisition methods. (f) Scatterplot representations for the common dynamic (c) proteins at time point 8h.

CONCLUSION

Fig.4: Heat stress affects several key cellular processes that impact protein homeostasis. (a) Interaction network for upregulated proteins (green) and down regulated proteins (red) based on the clustering shown in Figure 3c. Proteins that belong to enriched GO-terms are outlined by coloured shapes. (b) Cellular processes affected in early and late response to heat shock.

• A compact FAIMS interface combined with an Orbitrap tribrid mass spectrometer improves accuracy and coverage in multiplex quantification.

• TMT-based MS2 quantitation made possible with FAIMS enabled a 2-3 fold gain in identification with high TMT ratio accuracy compared to SPS-MS3 quantitation.

REFERENCES

4260 238281814

SPS-MS3

withFAIMS-MS2

unique peptides

This work was carried out with financial support from the Natural Sciences and Engineering Research Council and the Genomic Applications Partnership Program (GAPP) of Genome Canada. IRIC receives infrastructure support from IRICoR, the Canadian Foundation for Innovation, and the Fonds de Recherche du Québec - Santé (FRQS). IRIC proteomics facility is a Genomics Technology platform funded in part by the Canadian Government through Genome Canada.