nDIA HT
nDIA HT
Abstract
In this study we further optimized a dia-PASEF® acquisition mode by using narrow isolation Keywords:
windows of 5 Da combined with focused m/z and ion mobility window placement, resulting in Proteomics, high throughput,
dia-PASEF, timsTOF HT,
the identification of 8850 protein groups and 134,728 peptides for a human cell line sample,
Spectronaut
14,366 and 170,917 from a complex triple proteome sample and 707 protein groups and 4818
peptides from neat plasma with a 15-minute gradient.
Introduction
In proteomics studies, accurate and precise relative protein quantitation, together with high protein
sequence coverage, is the key to unravel the complex secrets of biological processes. Scientists
in the pharmaceutical and biopharmaceutical industry, as well as in the life sciences are committed
to advancing knowledge, often relying on using cutting-edge technologies to achieve this goal.
Data-independent acquisition (DIA) offers the advantage of comprehensive proteome coverage
and reliable quantitation of proteins, while also reducing missing information, leading to more
robust results. dia-PASEF is an advanced variant of DIA, capitalizing on the additional dimension
of separation unlocked on the timsTOF platform by trapped ion mobility separation (TIMS). Due to
the two-dimensional mass and mobility space in combination with varying TIMS scan/accumulation
times, dia-PASEF enables method creation with tailored window schemes and duty cycles. When
introduced in 2020 [1], the standard dia-PASEF method consisted of constant 25 Da m/z window
widths covering the mass range of interest. Already with the introduction of py_diAID [2] variable
Stephanie Kaspar-Schoenefeld, Andreas Schmidt, Torsten Mueller, Jonathan Krieger; Bruker Daltonics GmbH & Co. KG.
window schemes became frequently used, showing a benefit of smaller isolation windows in
the very dense ion cloud region. Recently a new flavor of dia-PASEF has been introduced, called
thin-PASEF [3], which applies very small isolation windows of 10 Da focusing exclusively on regions
of high ion density. The application of thin-PASEF resulted in the identification of nearly 11,000 protein
groups from human cell line digests within a 100-minute active gradient. While identification of
11,000 proteins is impressive, the field is tending towards shorter gradients as high throughput is
crucial in proteomics. Thus, we further optimized the approach for shorter gradients of 15-minutes,
resulting in a narrow-window dia-PASEF method with 5 Da windows.
For plasma proteomics measurements, 54 individual plasma samples (18 from citrate, 36 from EDTA
plasma from 20 different healthy donors) were digested with the PreOmics iST-BCT kit for biofluids.
A pooled sample of all digestions was generated and diluted to 500 ng/µL for LC-MS/MS
analysis.
LC conditions
nanoLC method parameters
LC system nanoElute
Mobile Phase A 0.1% formic acid (FA) in water
Mobile Phase B 0.1% FA in acetonitrile (ACN)
Flow rate 300 nL/min
Aurora Ultimate 25 x 75 C18 UHPLC column
Column
(IonOpticks)
Column temperature 50°C
0 min: 2% B
2 min: 5% B
13 min: 29% B
Gradient 15 min: 35% B
15.5 min: 95% B
18 min: 95% B
MS conditions
MS method parameters
Source Captive Spray ionization source
MS system timsTOF HT
Acquisition mode dia-PASEF
5 Da window size
IM Range: 0.8 to 1.1 1/K0
m/z range: 350 to 900
112 windows a 44 frames plus 10 MS1 scans
Accu/Ramp time: 30 ms
Collision Energy: 20 (0.60 1/K0) to 59eV (1.6 1/K0)
Data processing
Processing parameters
Software Spectronaut ® v19 (Biognosys)
1. File conversion via HTRMS converter
Deselect „Sum multiple MS1 scans per
Cycle“
Workflow
2. directDIA+ workflow
Default settings
Use MS1 level quant
Results
We optimized a dia-PASEF method to cover the high ion density region of interest combined
with high specificity by using narrow windows of 5 Da, resulting in reduced complexity of MS2
spectra. In detail, we developed a method covering a significantly smaller m/z range from 350
to 900 m/z combined with a focused mobility range from 0.8 to 1.1 1/K0 (Figure 1). 110 MS/MS
windows with 5 Da window width were covered in 44 TIMS frames with 10 MS1 frames in
between to ensure good peak coverage on MS1 level for quantitation (on average 8 data points
per peak). The multiplexing capabilities of dia-PASEF allowing measurements of more than one
window per frame combined with the ability to use very short accumulation and ramp times of
30 ms in the TIMS cartridge resulted in an overall cycle time of 1.76 s on MS/MS level and 0.45 s
on MS level.
When applying the optimized narrow-window dia-PASEF method to a human cell line digest,
we were able to identify 8850 protein groups and 134,728 peptides from triplicate injections
with a 15-minute active gradient (Figure 2) using library-free directDIA+ in Spectronaut (v.19,
Biognosys). These results relate to an increase in number of identifications of 12% on protein
group level and 30% on peptide level, compared to a previously published study using the same
setup [4]. This significant improvement can be accounted to the combination of a focused
dia-PASEF window placement combined with narrow 5 Da windows, very short accumulation
and ramp times in the TIMS dimension of 30 ms as well as an improved data processing by using
the latest Spectronaut version (v19). Impressively, 88% of the protein groups were quantified
1.4
1.2
1.2
Mobility 1/K0
1.05
1.0
1.0
Mobility 1/K0
1.00
1.00
0.95
0.95
0.90
0.90
0.85
0.85
0.80
0.80
0.8
0.8 400
400 500
500 600
600 700
700 800
800 900m/z
m/z
200
200 400
400 600
600 800
800 1000
1000 1200
1200 1400
1400 1600
1600
m/z
Figure 1
Overview narrow-window dia-PASEF method.
To optimize deep proteome coverage a method has been generated with focus on most intense ion region regarding m/z and ion mobility pane.
with a coefficient of variation (CV) below 10%. Even though mass and mobility ranges were
dramatically reduced compared to typical dia-PASEF default methods (standard range: 400
to 1200 m/z and 0.6 to 1.3 1/K0), we were still able to identify more than 134,720 peptides
from triplicate injections. Applying the narrow-window dia-PASEF method to a less complex
proteome, namely yeast, resulted in the identification of 4677 whole proteome. This is an
increase of 6% on protein group level and more than 60% on peptide level compared to our
previous study running a 15-minute active gradient [4].
For both samples, median CV values were 3% on protein group level and 8% on peptide level
(Figure 2).
10,000 150,000
100
90
8000
80
100,000 70
6000
60
HeLa
50
8850 134,728
8484
4000 7759 40
105,513
50,000
30
74,048
2000 20
10
0 0 0
Total CV <20% CV <10% Total CV <20% CV <10% Protein groups Peptides
5000 75,000
100
90
4000
80
50,000 70
3000
60
Yeast
4677 50
4554 68,744
4305
2000 40
54,297
25,000
30
39,206
1000 20
10
0 0 0
Total CV <20% CV <10% Total CV <20% CV <10% Protein groups Peptides
Figure 2
Reproducible and in-depth peptide and protein identification from human A and yeast B protein digest with total protein group and peptide numbers
shown together with number of quantified with CV values below 20% and 10% and box plot visualisation of CV distribution on protein groups and
peptide level.
In the next step we analyzed a complex hybrid proteome sample consisting of HeLa, yeast, and
E. coli digests combined in defined ratios (HeLa: 1:1, Yeast: 3:2, E. coli: 2:3). The narrow-window
dia-PASEF method resulted in an on average identification of 14,366 (7978 HeLa, 3885 yeast,
2473 E. coli) protein groups and 170,917 peptides from six injections per sample within a
15-minute gradient (Figure 3). The median coefficient of variation for six runs per sample was
around 6% on protein group level and 12% on peptide level.
A B
Number of protein groups
15,000
2468 2462
2.0
2485 2486 2486 2484 2484 2480 2464 2461 2457 2463
1.5
10,000 3819 3840 3839 3856 3848 3855 3923 3921 3932 3925 3942 3921 1,0
0.5
Log2 ratios
5000 -0.5
8004 7990 8000 7996 7972 7962 7971 7965 7965 7958 7998 7954
-1.0
-1.5
0 -2.0
A_1 A_2 A_3 A_4 A_5 A_6 B_1 B_2 B_3 B_4 B_5 B_6
HeLa Yeast E. coli HeLa Yeast E. coli
Number of peptides
200,000
150,000
100,000
171,864 170,705 170,220 169,897 169,767 169,497 171,801 171,571 171,826 171,373 171,123 171,357
50,000
0
A_1 A_2 A_3 A_4 A_5 A_6 B_1 B_2 B_3 B_4 B_5 B_6
Figure 3
In-depth protein identification and quantitation from a hybrid proteome sample using library-free data processing (directDIA+, Spectronaut).
More than 14,000 protein groups have been reproducibly identified and quantified in each of the 12 samples. A Total numbers of identified protein
groups and peptides from the mixed proteome samples (sample “A”: H50Y20E30, Sample “B”: H50Y30E20) for each of the three species (HeLa,
yeast, E.coli). B Box and Whisker plot of obtained log10 ratios for all three species (HeLa, yeast and E.coli).
Neat plasma is one of the most challenging samples in proteomics due to its high dynamic
range. When applying the generated narrow-window dia-PASEF method to a plasma sample,
we were able to identify on average 707 protein groups from 4818 peptides with a 15-minute
gradient. 661 protein groups were identified in all three replicates displaying the high recovery
rate of the method (Figure 4 A). Nearly 6 orders of magnitude can be covered by the used setup
(Figure 4 B).
800
744
715
700 661
Protein group identifications
600
500
400
300
200
100
0
1.0 1.5 2.0 2.5 3.0
Minimum number of samples a protein group was identified in
6
Log10 median protein group quantity
0
0 100 200 300 400 500 600 700 800
Rank
Figure 4
In-depth protein identification from a neat plasma proteome sample using library-free data processing (directDIA+, Spectronaut).
More than 700 protein groups have been reproducibly identified and quantified in triplicate runs. A Protein group identifications from plasma proteome
sample. Shown are the number of protein groups identified in all three runs (661) and the cumulative number of protein groups identified in at least one
run (744). B Ranked protein group plot visualizing log10 median protein group quantity across all quantified protein groups.
Conclusions
The two-dimensional mass and mobility space of dia-PASEF enables generation of various methods tailored to
sample complexities and throughput demands.
The presented method using 5 Da isolation windows over a condensed ion mobility and m/z region resulted in
high proteome coverage and accurate quantitation in short gradients of 15 minutes.
Nearly full yeast proteome coverage can be achieved with on average 4677 protein groups identified.
Analysis of complex mixed proteomes resulted in identification of 14,366 protein groups using library-free data
processing.
References
[1] Meier et al. (2020). Nat Methods. 17(12):1229-1236
[2] Skowronek et al. (2022). Molecular & Cellular Proteomics. 21(9):100279
[3] Konno et al. (2024). bioXriv. https://doi.org/10.1101/2024.04.26.591246
[4] Bruker application note LCMS 218. https://www.bruker.com/en/products-and-solutions/mass-spectrometry/timstof/timstof-ht/_jcr_content/root/
sections/more_information/sectionpar/search.download-asset.pdf/4922715e-6af9-4abd-a16c-070b6cb70ed4/1905946-lcms-218-timstof-ht-and-hela-
ebook.pdf
Further reading
Setting the standard for Protein Immediately dive into results of your
Analysis with PreOmics. experiment.
www.preomics.com/ www.bruker.com/en/products-and-
solutions/mass-spectrometry/ms-
software/proteoscape.html
Learn more about the timsTOF HT. The gold standard for DIA proteomics
www.bruker.com/en/products-and- analysis: Spectronaut®.
solutions/mass-spectrometry/timstof/ biognosys.com/software/spectronaut/
timstof-ht.html
Benefit Feature
Bruker Daltonics is continually improving its products and reserves the right
Greater Complete proteome search for 10,000+
discoveries protein groups in tissue or cell lines with
higher throughput with Biognosys
TrueDiscovery™.
For Research Use Only. Not for use in clinical diagnostic procedures.