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.
