evaluation of srm-, prm-and dia-based targeted quantification · 2017. 8. 2. · evaluation of...
TRANSCRIPT
1
Evaluation of SRM-, PRM- and DIA-based Targeted Quantification
Ehwang Song1, Tujin Shi1, Hui Wang1, Anil K. Shukla1, Thomas L. Fillmore1, Yuqian Gao1, Song Nie1, Ronald J. Moore1, Matthew J. Gaffrey1, Athena A. Schepmoes1, Karin D. Rodland1,
Richard D. Smith1, Wei-Jun Qian1, and Tao Liu1
1Pacific Northwest National Laboratory, Richland, WA, USA
Biological Sciences Facility (BSF) & Computational Sciences Facility (CSF)
Environmental Molecular Sciences Laboratory (EMSL)
Outline
Introduction
Initial Evaluation of Targeted Proteomics Performance
Large Scale Demonstration- Quantification of EGFR Pathway Proteins With/Without Fractionation
Summary
2
Targeted Proteomics Workflow
3
Peptides
…
Transition Selection &
Optimization
SRM
PRM & DIA
K^R^
Isotopically labeled peptides
Sample
46.0 46.2 46.4Retention Time
0
200
Inte
nsi
t y(1
0^3)
46.1
46.0 46.2 46.4
Retention Time
0
1
Inte
nsi
t y(1
0^6)
46.1
Light (or endogenous)
Heavy
47.4 47.6 47.8 48.00
200
400
Inte
nsity
(10^
3 )
47.7+0.3 ppm
47.4 47.6 47.8 48.0
Retention Time
0
0.5
1
Inte
nsity
(10^
6 )
47.7+0.8 ppm
47.4 47.6 47.8 48.00
200
400
Inte
ns i
ty(1
0^3 )
47.7+0.3 ppm
47.4 47.6 47.8 48.0
Retention Time
0
0.5
1
Inte
ns i
ty(1
0^6 )
47.7+0.8 ppm
Method including Precursor (Isolation
Window) m/z & Retention Time
Post-data Refinement
Method including Transition m/z,
Collision Energy, and Retention Time
SRM: Selected Reaction MonitoringPRM: Parallel Reaction MonitoringDIA: Data Independent Acquisition
Light (or endogenous)
Heavy
Aim: Evaluation of Targeted Quantification Performance
4
1. SRM and PRM: A.C. Peterson et al., Mol Cell Proteomics, 2012, 11, 1475.2. SRM and PRM: S. Gallien et al., J Proteomics, 2013, 81, 148.3. SRM and DIA: T. Schmidlin et al., Proteomics, 2016, 16, 2193.4. SRM, PRM and DIA: T. Kockmann et al., Proteomics, 2016, 16, 2183.5. IS-PRM: S. Gallien et al., Proteomics, 2015, 14, 1630.6. TOMAHAQ: B.K. Erickson et al., Mol Cell, 2017, 65, 361.
PRISM Fractionation
Same LC System
Sensitivity
Selectivity
AccuracyReproducibility
Throughput
Experimental Design
5
1. Reverse Response Curve Matrix: Shewanella oneidensis (bacteria)
2. Large Scale DemonstrationTargets: 133 EGFR Peptides
23 Heavy Stock Concentration
(fmol/uL)
0
0.0025
0.0125
0.025
0.125
0.25
2.5 Throughput Sensitivity
Sensitivity Selectivity Accuracy
Reproducibility
SRM: TSQ Vantage; PRM&DIA: Q Exactive-HFTechnical triplicate
PRISM Fractionation
No Fractionation
vs.
PRISM: Shi et al., Proc Natl Acad Sci, 2012, 109, 15395.
Linear Dynamic Range of Response: SRM ≥ PRM > DIA
6
y = 1.0184x - 0.0043R² = 0.9979
y = 1.0125x + 0.0128R² = 0.9902
y = 0.9709x - 0.0136R² = 0.9841
9.77E-04
3.91E-03
1.56E-02
6.25E-02
2.50E-01
1.00E+00
4.00E+00
9.77E-04 3.91E-03 1.56E-02 6.25E-02 2.50E-01 1.00E+00 4.00E+00
Log2
[H/L
]
Averaged Log2[H/L]
SRM
PRM
DIA
SRM PRM DIA
1.00x2-9
1.00x2-7
1.00x2-5
1.00x2-3
1.00x2-1
1.00x21
1.00x23
1.00x25
Log2
[H/L
]
3 Orders
LOQ Values: SRM ≤ PRM < DIA
7
1.00E+00
4.00E+00
1.60E+01
6.40E+01
2.56E+02
1.02E+03
LOQ
(log
2[pg
/mL]
)
SRMPRMDIA
SRM, PRM <(=) DIA
LOQ Values: SRM ≤ PRM < DIA
8
1.00E+00
4.00E+00
1.60E+01
6.40E+01
2.56E+02
1.02E+03
LOQ
(log
2[pg
/mL]
)
SRMPRMDIA
SRM=PRM :10 SRM < PRM: 7 PRM < SRM : 5
PRM: Selectivity Sensitivity
9
y = 1292.0x - 169.33R² = 0.9935
y = 1305.4x - 135.91R² = 0.9926
y = 1326.2x - 138.32R² = 0.9926
3.91E-03
3.13E-02
2.50E-01
2.00E+00
1.60E+01
1.28E+02
1.02E+03
8.19E+03
1.95E-03 1.17E-02 7.03E-02 4.22E-01 2.53E+00
Mea
sure
d H
eavy
pep
tide
(fg/m
L)
Spiked Heavy Peptide (fmol/uL)
SRM
PRM
DIA
ERRFI1_NSPSLFPCAPLCERSRM PRM DIA
Retention Time
30 310
0.51
1.530.5
+1.8 ppm
30 31Retention Time
0500
10001500 30.5
-1.1 ppm ND
y9+y8+y5+
Inte
nsity
(10^
6)In
tens
ity
30.5
0
530.5
+0.8 ppm
30.5
0
1030.5
+0.5 ppm
y9+y8+y5+
Inte
nsity
(10^
6)In
tens
ity (1
0^3)
29.0 29.50
100
200
30029.1
29.0 29.5Retention Time
0
200
400 29.1
Inte
nsity
(10^
3)In
tens
ity
Y9+y8+y12++
S/N:6
30.0 30.50
1
2 30.3+1.7 ppm
30.0 30.5Retention Time
0
4000
800030.3
+0.2 ppm
Inte
nsity
(10^
6)In
tens
ity
30.0 30.50
530.3
+0.8 ppm
30.0 30.5
Retention Time
0
5030.3
+0.7 ppm
Inte
nsity
(10^
6)In
tens
ity (1
0^3)
29.0 29.50
100200300400
29.1
29Retention Time
0100020003000
29.1
Inte
nsity
(10^
3)In
tens
ity
PRM: Selectivity Quantification Accuracy
Potential matrix interference in SRMLow mass resolution and accuracy afforded by QqQ mass spectrometer
High accuracy for PRMHigh resolution and tight mass tolerance applied for data acquisition and extraction 10
in HMEC cell line
1.56E-02
6.25E-02
2.50E-01
1.00E+00
4.00E+00
1.60E+01
Log2
[H/L
]
Heavy Peptide Conc. (ng/mL)
SRMPRMDIA
HRAS_LVVVGAGGVGK
3.91E-03
3.13E-02
2.50E-01
2.00E+00
1.60E+01
Log2
[H/L
]
Heavy Peptide Conc. (ng/mL)
KRAS_SYGIPFIETSAK
Reproducibility among Three Techniques
11
Spiked in Shewanella oneidensis Spiked in HMEC
SRM PRM DIA
40
30
20
10
0
SRM PRM DIA
CV
(%)
40
30
20
10
0
Large Scale Demonstration: Quantification of EGFR Pathway Proteins
12
EGFR pathway One of the most important pathwaysOne of the most well-studied signaling networksUnderlying diseases (dysregulation of signaling networks)Key players (low abundance regulatory proteins) and their PTMs
133 peptides/ 32 proteins (including 6 phosphopeptides)Quantification with and without PRISM fractionationSample
PRISM: 25 ug of peptidesNo PRISM: 0.5 ug of peptides
www.qiagen.com/productsOda et al., Mol Syst Biol 2005, 1:2005.0010.
Shi et al., Sci Signal 2016, 9(436):rs6.
PRISM
13
PRISM: high-pressure high-resolution separations with intelligent selection and multiplexing
200 nL/min
2 µL/min
2.2 µL/min
Heavy stock of 133 EGFR pathway peptides spiked in HMEC cell line digests
Time (min)
QQQ MS
Q1 Q2 Q33 µm, C18 column(200 µm i.d. × 50 cm)
1.7 µm, C18 column(75 µm i.d. × 25 cm)
iSelection
96 well plate
On-line monitoring heavy peptides
Capillary LC (high pH)
Nano LC (low pH)
#2 #6 #10
Multiplexing: 12 concatenated fractions
SRM PRM & DIA
Shi et al., Proc Natl Acad Sci, 2012, 109, 15395.
Quantification of EGFR Pathway Proteinsby LC vs. PRISM-S/PRM and -DIA
14
* grey: non-detected
53 10546 9916 64
-14.90
-11.58
-8.260
-4.940
-1.620
1.700
Log2[L/H]
PRIS
M-S
RM
PRIS
M-P
RM
PRIS
M-D
IA
LC-S
RM
LC-P
RM
LC-D
IA20
40
60
80
100
120N
umbe
r of P
eptid
es
Dynamic Range of Quantification: PRM ≥ SRM >DIA
15
LC-S
RM
LC-P
RM
LC-D
IA-18
-16
-14
-12
-10
-8
-6
-4
-2
0
2
4
Log2
[L/H
]
3 Orders 5 Orders
PRIS
M-S
RM
PRIS
M-P
RM
PRIS
M-D
IA
-18
-16
-14
-12
-10
-8
-6
-4
-2
0
2
4
0505 0 ( 02 468 ( 00005 012 ( 3 0000 2 4 0 00 (8
XICs of Very Low Abundant Peptide Quantified by PRISM
16
L/H: 0.0045L/H: 0.0028
PRISM-SRM PRISM-PRM
SPRY4_LQHPLTILPIDQVK (copies/cell: ~900)
N/D
PRISM-DIAy6+y4+y6++
y5+y4+y6++
y6+y5+b7+
Summary
Sensitivity: SRM ≥ PRM > DIASelectivity: PRM > DIA > SRMAccuracy: PRM > DIA > SRMReproducibility: SRM ≥ PRM > DIAThroughput: SRM ≥ PRM > DIA
In combination with the front-end fractionation, overall quantification performance was improved for all three techniques. 17
SRM PRM DIAPost-Data
Refinement
Front-end Assay
Development
MultiplexingThroughput
Reproducibility
Accuracy
SelectivitySensitivity Post-Data
Refinement
Front-end Assay
Development
MultiplexingThroughput
Reproducibility
Accuracy
SelectivitySensitivity Post-Data
Refinement
Front-end Assay
Development
MultiplexingThroughput
Reproducibility
Accuracy
SelectivitySensitivity
Acknowledgements
18
The Integrative Omics Group at PNNL