ACS Publications. Most Trusted. Most Cited. Most Read
timsTOF HT Improves Protein Identification and Quantitative Reproducibility for Deep Unbiased Plasma Protein Biomarker Discovery
My Activity
  • Open Access
Article

timsTOF HT Improves Protein Identification and Quantitative Reproducibility for Deep Unbiased Plasma Protein Biomarker Discovery
Click to copy article linkArticle link copied!

  • Dijana Vitko
    Dijana Vitko
    PrognomiQ Inc., 1900 Alameda de las Pulgas, San Mateo, California 94403, United States
    More by Dijana Vitko
  • Wan-Fang Chou
    Wan-Fang Chou
    PrognomiQ Inc., 1900 Alameda de las Pulgas, San Mateo, California 94403, United States
  • Sara Nouri Golmaei
    Sara Nouri Golmaei
    PrognomiQ Inc., 1900 Alameda de las Pulgas, San Mateo, California 94403, United States
  • Joon-Yong Lee
    Joon-Yong Lee
    PrognomiQ Inc., 1900 Alameda de las Pulgas, San Mateo, California 94403, United States
  • Chinmay Belthangady
    Chinmay Belthangady
    PrognomiQ Inc., 1900 Alameda de las Pulgas, San Mateo, California 94403, United States
  • John Blume
    John Blume
    PrognomiQ Inc., 1900 Alameda de las Pulgas, San Mateo, California 94403, United States
    More by John Blume
  • Jessica K. Chan
    Jessica K. Chan
    PrognomiQ Inc., 1900 Alameda de las Pulgas, San Mateo, California 94403, United States
  • Guillermo Flores-Campuzano
    Guillermo Flores-Campuzano
    PrognomiQ Inc., 1900 Alameda de las Pulgas, San Mateo, California 94403, United States
  • Yuntao Hu
    Yuntao Hu
    PrognomiQ Inc., 1900 Alameda de las Pulgas, San Mateo, California 94403, United States
    More by Yuntao Hu
  • Manway Liu
    Manway Liu
    PrognomiQ Inc., 1900 Alameda de las Pulgas, San Mateo, California 94403, United States
    More by Manway Liu
  • Mark A. Marispini
    Mark A. Marispini
    PrognomiQ Inc., 1900 Alameda de las Pulgas, San Mateo, California 94403, United States
  • Megan G. Mora
    Megan G. Mora
    PrognomiQ Inc., 1900 Alameda de las Pulgas, San Mateo, California 94403, United States
  • Saividya Ramaswamy
    Saividya Ramaswamy
    PrognomiQ Inc., 1900 Alameda de las Pulgas, San Mateo, California 94403, United States
  • Purva Ranjan
    Purva Ranjan
    PrognomiQ Inc., 1900 Alameda de las Pulgas, San Mateo, California 94403, United States
    More by Purva Ranjan
  • Preston B. Williams
    Preston B. Williams
    PrognomiQ Inc., 1900 Alameda de las Pulgas, San Mateo, California 94403, United States
  • Robert J. X. Zawada
    Robert J. X. Zawada
    PrognomiQ Inc., 1900 Alameda de las Pulgas, San Mateo, California 94403, United States
  • Philip Ma
    Philip Ma
    PrognomiQ Inc., 1900 Alameda de las Pulgas, San Mateo, California 94403, United States
    More by Philip Ma
  • Bruce E. Wilcox*
    Bruce E. Wilcox
    PrognomiQ Inc., 1900 Alameda de las Pulgas, San Mateo, California 94403, United States
    *Email: [email protected]
Open PDFSupporting Information (1)

Journal of Proteome Research

Cite this: J. Proteome Res. 2024, 23, 3, 929–938
Click to copy citationCitation copied!
https://doi.org/10.1021/acs.jproteome.3c00646
Published January 15, 2024

Copyright © 2024 The Authors. Published by American Chemical Society. This publication is licensed under

CC-BY-NC-ND 4.0 .

Abstract

Click to copy section linkSection link copied!

Mass spectrometry (MS) is a valuable tool for plasma proteome profiling and disease biomarker discovery. However, wide-ranging plasma protein concentrations, along with technical and biological variabilities, present significant challenges for deep and reproducible protein quantitation. Here, we evaluated the qualitative and quantitative performance of timsTOF HT and timsTOF Pro 2 mass spectrometers for analysis of neat plasma samples (unfractionated) and plasma samples processed using the Proteograph Product Suite (Proteograph) that enables robust deep proteomics sampling prior to mass spectrometry. Samples were evaluated across a wide range of peptide loading masses and liquid chromatography (LC) gradients. We observed up to a 76% increase in total plasma peptide precursors identified and a >2-fold boost in quantifiable plasma peptide precursors (CV < 20%) with timsTOF HT compared to Pro 2. Additionally, approximately 4.5 fold more plasma peptide precursors were detected by both timsTOF HT and timsTOF Pro 2 in the Proteograph analyzed plasma vs neat plasma. In an exploratory analysis of 20 late-stage lung cancer and 20 control plasma samples with the Proteograph, which were expected to exhibit distinct proteomes, an approximate 50% increase in total and statistically significant plasma peptide precursors (q < 0.05) was observed with timsTOF HT compared to Pro 2. Our data demonstrate the superior performance of timsTOF HT for identifying and quantifying differences between biologically diverse samples, allowing for improved disease biomarker discovery in large cohort studies. Moreover, researchers can leverage data sets from this study to optimize their liquid chromatography–mass spectrometry (LC–MS) workflows for plasma protein profiling and biomarker discovery. (ProteomeXchange identifier: PXD047854 and PXD047839).

This publication is licensed under

CC-BY-NC-ND 4.0 .
  • cc licence
  • by licence
  • nc licence
  • nd licence
Copyright © 2024 The Authors. Published by American Chemical Society

Note Added after ASAP Publication

This paper was published ASAP on January 15, 2024, with corrections throughout the paper. The corrected version was reposted on January 30, 2024.

Introduction

Click to copy section linkSection link copied!

The discovery of novel and reliable biomarkers for disease diagnosis and classification has consistently been of intense interest for oncological applications, where effective screening and early detection can both guide clinical management and significantly improve overall survival. (1−3) Plasma proteome analysis is a promising avenue for identifying potential disease biomarkers due to the wide range of tissue-specific proteins accessible through the circulation and the noninvasive nature of blood collection. (4,5) Mass spectrometry (MS) has been the method of choice for unbiased profiling of proteins in complex biological samples like blood. (6) However, the dynamic range of protein concentrations in plasma spans across several orders of magnitude, making it difficult to detect low-abundance proteins. (7) The discovery of reliable and specific protein biomarkers requires comprehensive and large-scale studies involving well-characterized patient cohorts. (8) These factors represent distinct challenges to obtaining reproducible and meaningful results in large biomarker discovery studies.
Advanced timsTOF MS technologies such as trapped ion mobility spectrometry (TIMS) (9,10) and parallel accumulation serial fragmentation (PASEF) (11) allow for rapid and efficient data acquisition with higher sensitivity and increased coverage of the proteome. The recent development of the Bruker timsTOF HT further advances this technology by utilizing a newly designed fourth-generation TIMS-XR analyzer coupled with an acquisition workstation equipped with a faster PCIe bus and a faster digitizer with higher fidelity. The new TIMS-XR analyzer doubles the ion storage volume and replaces the funnel exit pumping region with a quadrupolar exit, allowing for better ion focusing. The gas velocity profile has been altered, and ion charge capacity has been increased 5× compared to the prior version of the TIMS analyzer. Storing ions at low gas velocity permits increased ion capacity, while ion mobility separation at high gas velocity increases ion mobility resolution. The low noise and higher amplification of the new digitizer allow for the detection of single ion events with a lower amplitude output from the detector. As a result, the detector gain decreases from 100 to 65%. The new digitizer improves vertical resolution from 10 to 14 bits and the sampling rate from 5 to 8 GS/s. Noise in the least significant bits is higher due to faster sampling rates and a higher number of bits; therefore, the effective number of bits is 2, providing up to a 4× vertical dynamic range improvement.
Recent reports have demonstrated the improved performance of the EvosepOne-timsTOF HT mass spectrometer system over the previously released timsTOF Pro 2 system in cell lines and tissue profiling, (11) but no study has comprehensively evaluated timsTOF HT performance for plasma proteomics profiling generally or plasma biomarker discovery specifically. Compared to cell lines and tissue samples, MS analysis of plasma samples poses a greater technical challenge because of the wide dynamic range of plasma protein concentrations spanning over 10 orders of magnitude; (6) therefore, an independent evaluation of the timsTOF MS systems’ performance is needed. Our data offer a comprehensive comparison of the timsTOF HT and Pro 2 systems for neat and Proteograph plasma proteomic profiling across a wide range of plasma peptide loading masses (100–1200 ng) and liquid chromatography (LC) gradients [100 samples per day (SPD), 60 SPD, 30 SPD]. The Proteograph workflow with the Proteograph Assay Kit (v1.2) leverages 5 distinct nanoparticles (NP1–5), in an automated workflow reproducibly compressing the dynamic range of the plasma proteome and thus enabling deeper proteome profiling. (12−17) In this paper, we propose an optimal range of liquid chromatography–mass spectrometry (LC–MS) conditions for deep and reproducible plasma proteomics analysis. Our data showcase the potential of timsTOF HT together with the Proteograph to identify putative cancer biomarkers in plasma using cohorts designed to exhibit distinct proteomic profiles [i.e., samples from 20 patients with stage IV nonsmall cell lung cancer (NSCLC) and 20 noncancer controls]. The qualitative and quantitative performance of timsTOF HT and Pro 2 was directly compared through Proteograph plasma proteomics analysis of lung cancer and control samples to demonstrate their relative potential to advance plasma biomarker research and its clinical applications.

Materials and Methods

Click to copy section linkSection link copied!

Neat and Proteograph Plasma Sample Preparation

Pooled human control plasma samples (pooled control; BioIVT) were denatured, alkylated, and digested using the PreOmics iST HT 192x kit according to the manufacturer’s instructions (neat [unfractionated] plasma, PreOmics Inc.). A separate set of pooled control samples was processed via the Proteograph workflow with Proteograph Assay (Seer, Inc.). Neat and Proteograph plasma samples were diluted in 0.5 mL Protein LoBind Tubes (Eppendorf AG) to a final concentration yielding 100–1200 ng per 20 μL (Table S1). Each sample was spiked with a constant amount of PQ500 peptides (Biognosys AG), 0.75 μL of stock solution prepared according to manufacturer’s instructions in 20 μL of final plasma peptide sample across all different peptide loading masses. For both timsTOF HT and Pro 2 systems, samples for comparison were loaded onto the Evotips (Evosep Biosystems) from the same peptide dilution. An additional set of plasma samples from 20 patients with stage IV NSCLC cancer and 20 noncancer controls (Table S2), which were expected to differ in proteomic profiles, underwent processing via the Proteograph Product Suite. Samples were loaded onto the Evotips prior to LC–MS acquisition. Additional information on sample preparation is available in the Supporting Information.

LC–MS Data Acquisition

The pooled control samples were analyzed in technical triplicate for each sample type (neat and Proteograph plasma NP1–5), peptide loading mass (100–1200 ng, except NP4 100–900 ng), and LC gradient (100 SPD, 60 SPD, and 30 SPD) on an EvosepOne (Evosep Biosystems) system coupled with either the timsTOF HT or Pro 2 mass spectrometer (Bruker), equipped with a TIMS-XR or TIMS-Pro 2 cartridge, respectively. Individual lung cancer and control plasma samples were analyzed separately on the same LC–MS systems at a 60 SPD gradient. A single peptide mass was loaded onto the column for each sample: 600 ng of peptide for NP1–3 and NP5 and 300 ng for NP4 due to a limited peptide yield. A PepSep column of 8 cm x 150 μm (C18, 1.5 μm; Bruker) was used for plasma peptide separation at 100 and 60 SPD at 40 °C (Sonation column oven, Lab Sweeden AB), and a PepSep column of 15 cm x 150 μm (C18, 1.9 μm; Bruker) was used for 30 SPD at room temperature. All data were acquired under the dia-PASEF mode (18) with a MS1 scan range of 100–1700 m/z. Twelve isocratic m/z and IM windows were selected for serial MS2 fragmentation ranging from 390 to 1250 m/z and 0.84 to 1.31 1/K0 for mass and IM range, respectively (Table S3).

Data Search and Postanalysis

All raw files were analyzed with DIA-NN (version 1.8.1) (19,20) using an in-house spectral library. PQ500 peptide precursors were searched with Spectronaut (version 18.1) against a PQ500 spectral library provided by Biognosys AG. To independently identify stable isotope-labeled (SIL) and endogenous PQ500 peptide precursors, default settings with multichannel workflow definition set from library annotation with spike-in fallback option and reference-based identification were enabled. Trypsin/P protease cleavage was specified, with a maximum of two missed cleavages allowed. Carbamidomethylation of cysteine residues was set as a fixed modification, while methionine oxidation and N-terminal protein acetylation were set as variable modifications. The 1% peptide precursor false discovery rate (FDR) cutoff was applied, and MS1 and MS2 mass tolerances were automatically determined by DIA-NN. For other settings, default DIA-NN settings were used. DIA-NN outputs were analyzed and visualized with the Python Jupyter Notebook and Python packages pandas (1.5.1), scipy (1.10.1), numpy (1.23.5), seaborn (0.12.2), and matplotlib (3.5.1). For visualization of the timsTOF HT vs. Pro 2 peak area ratio distribution (MS1 and MS2), the kdeplot function available in Seaborn (version 0.12.2) was employed. For the kernel density estimation, the parameters bw_method = Scott and bw_adjust = 1 were utilized. Only peptide precursors that were identified in all three replicates of a pooled control sample under the same sample, loading mass, and LC conditions were considered. To test for statistically significant differences in the number and median MS2 intensity values of SIL PQ500 peptide precursors across peptide loading masses, Welch’s t-test was applied. For the differential expression analysis between 20 lung cancer and 20 control plasma samples, we performed the Wilcoxon signed-rank test with median-normalized peptide precursors detected in at least 25% of samples in each cohort per NP. The Benjamini–Hochberg correction was applied for FDR control. Additional data acquisition methodology details are available in the Supporting Information.

Results

Click to copy section linkSection link copied!

Qualitative Performance of timsTOF HT Exceeds timsTOF Pro 2 across a Wide Range of Plasma Peptide Loading Masses and LC Gradients

To evaluate the qualitative performance of timsTOF HT compared to Pro 2, we measured the total number of peptide precursors identified from neat and pooled control Proteograph plasma samples across a range of peptide loading masses and LC gradients (Figure 1). A higher number of peptide precursors was consistently identified with timsTOF HT compared to Pro 2 with loading masses ≥ 600 ng for neat and Proteograph plasma across all 3 LC gradients (Figure 1B). This pattern of increased qualitative performance with timsTOF HT was maintained for each NP individually (Figure S1). For neat plasma samples, timsTOF HT identified a higher number of peptide precursors compared to timsTOF Pro 2 across all loading masses at 100 and 30 SPD, but for 60 SPD, a qualitative gain in performance was observed only for peptide masses ≥ 600 ng (Figure 1). The greatest increase in peptide precursor identification with timsTOF HT relative to Pro 2 was 76%, which was observed with 1200 ng of neat peptide mass analyzed at 100 SPD (Table S4). Similarly, in Proteograph plasma samples, timsTOF HT identified substantially more precursors at peptide loading masses ≥ 600 ng across all 3 LC gradients, with a maximum gain of 46% at 1200 ng of peptide load at 100 SPD relative to timsTOF Pro 2. Notably, at low Proteograph plasma peptide loads of 100–300 ng, timsTOF HT and Pro 2 showed comparable plasma proteomics coverage depth.

Figure 1

Figure 1. Qualitative performance of timsTOF HT exceeded that of timsTOF Pro 2 across a wide range of plasma peptide loading masses and LC gradients. Compared to timsTOF Pro 2, HT identified up to 76 and 46% more peptide precursors in (A) neat and (B) Proteograph plasma, respectively, with up to 4.5-fold more precursors detected in Proteograph plasma than neat plasma. Red (timsTOF HT) and blue (timsTOF Pro 2) lines represent the number of peptide precursors uniquely detected in all replicate injections (n = 3) of neat and Proteograph plasma samples (NP1–3,5). Proteograph plasma NP4 was not considered for plotting due to insufficient yield for 1200 ng peptide mass load. Abbreviations: LC, liquid chromatography; NP, nanoparticle, SPD, samples per day.

Consistent with the enhanced performance associated with the increased ion capacity of the TIMS-XR cartridge, higher peptide loading masses resulted in increased peptide precursor identification rates with timsTOF HT. Conversely, we observed a reduction in precursors identified at higher loading masses with timsTOF Pro 2, suggesting TIMS cartridge saturation in timsTOF Pro 2 (Figures 1 and S1). To investigate this further, we analyzed pooled control plasma samples spiked with PQ500 peptides. The number of PQ500 peptide precursors identified with timsTOF HT was comparable across different peptide loading masses; however, with timsTOF Pro 2, the number of PQ500 peptide precursors identified decreased as the plasma peptide load increased (Figure S2A,B). This effect was more prominent for the short LC gradient (100 SPD) compared to the longer gradients (60 and 30 SPD), which was expected given that more ions are analyzed within the TIMS cartridge at a given time, further suggesting TIMS cartridge saturation. For neat plasma, a statistically significant decrease of identified PQ500 peptide precursors in ≥ 300 ng peptide loading masses compared to 100 ng peptide load at 100 SPD was observed in timsTOF Pro 2 but not in HT (Figure S2A). In representative Proteograph plasma NP2 data, a statistically significant decrease of identified PQ500 peptide precursors was observed across all 3 LC gradients with timsTOF Pro 2, while only high loading masses, i.e., 900 and 1200 ng, compared to 100 ng at 100 SPD, have a significantly lower number of PQ500 peptide precursors identified with timsTOF HT (Figure S2B). We also compared PQ500 peptide quantitation in both MS systems. The median MS2 quantity of PQ500 peptide precursors identified decreased with higher peptide loading masses in both timsTOF HT and Pro 2, but the decrease was greater in the timsTOF Pro 2 data set (Figure S2C,D). It is noteworthy that at neat and Proteograph plasma peptide loading masses of ≥300 ng, quantitation of the PQ500 peptide precursors was more reproducible in timsTOF HT compared to Pro 2, visualized by the tighter distribution and lower median coefficient of variation (CV) values of triplicate PQ500 MS2 quantities (Figure S2E,F).

timsTOF HT Allows for a Superior Quantitative Linear Range Compared to timsTOF Pro 2

We further investigated quantitative differences between timsTOF HT and Pro 2 by a comprehensive comparison of MS2 responses across a wide range of peptide loading masses and LC gradients. Figure 2 shows representative data for Proteograph plasma NP2 with loading masses ranging between 100 and 1200 ng and peptide separation at 60 SPD gradient. The density distribution of peptide precursor MS2 quantitative ratios for Proteograph plasma NP2 shows that a majority of the MS2-based peak areas were higher with timsTOF HT compared to Pro 2 (Figure 2A). A broadening of the ratio distribution at higher peptide loading masses suggests better quantitative accuracy of timsTOF HT for capturing peak area increases compared with timsTOF Pro 2. (Figure 2A). For visualization purposes, the MS2 peak area ratio cutoff was set to 5. Similar density plots for MS1 (Figure S3) and MS2 (Figure S4) peak area ratios between timsTOF HT and Pro 2 were seen for each NP. We randomly selected one precursor from the third quartile of the total precursor intensity range (FLVGPDGIPIMR, 2+) to illustrate the MS2 intensity difference. With timsTOF Pro 2, overall MS2 peak intensities of the y9 fragment ion are lower compared to the same ion intensities measured by timsTOF HT (Figure 2B). Moreover, MS2 intensity with timsTOF Pro 2 was already saturated at 600 ng, whereas with timsTOF HT, the MS2 intensity increased with loading mass up to 1200 ng. Comparison of peak areas from all MS2 precursor fragment ions shows that timsTOF HT had superior sensitivity and quantitative response between 100 and 1200 ng of peptide loading mass compared to timsTOF Pro 2, regardless of the sample type and LC gradient tested (Figure S5). The total ion chromatograms showed a similar trend in ion intensity profiles (Figure S6), suggesting that, overall, the increased capacity of the TIMS-XR cartridge allows better linearity across a wide range of peptide loading masses.

Figure 2

Figure 2. timsTOF HT had a greater quantitative linear range compared to timsTOF Pro 2. (A) Distribution of peptide precursor MS2 peak area (triplicate average) ratios quantified with timsTOF HT vs Pro 2. For visualization purposes, the MS2 peak area ratio cutoff was set to 5. (B) Extracted ion chromatogram of MS2 fragment ion (y9) for a randomly selected precursor, FLVGPDGIPIMR (2+), from within the 3rd quartile of the total precursor intensity range. (C) R-squared distribution for the precursors quantified in all three replicate measurements of each peptide loading mass for Proteograph plasma NP2 at 60 SPD gradient with timsTOF HT (n = 4256) and timsTOF Pro 2 (n = 3331). Abbreviations: NP, nanoparticle; SPD, samples per day.

To verify the quantitative linearity performance of the entire data set acquired on timsTOF HT and Pro 2, the R-squared (R2) distribution was plotted to examine the extent to which a linear regression model fit the experimentally measured quantitative response (Figure S7). As a representative example, the median R2 values for 3331 peptide precursors quantified by timsTOF Pro 2 from Proteograph plasma NP2 analyzed at 60 SPD with loading masses of 100–600, 100–900, and 100–1200 ng were 0.79, 0.62, and 0.53, respectively. Conversely, when the same peptide loading mass ranges were extracted from the timsTOF HT data, the median R2 values were 0.93, 0.91, and 0.86, for a total of 4256 precursors considered, demonstrating an improved linearity response across a wide range of peptide loading masses (Figure 2C). This result demonstrates that the improved ion capacity and extended dynamic range of the TIMS cartridge in the timsTOF HT results in a superior linear dynamic range compared to timsTOF Pro 2. The pattern observed in the representative data from Proteograph plasma NP2 (Figure 2C) was also observed across neat and Proteograph plasma samples (Figure S7).

timsTOF HT Has Enhanced Reproducibility Compared to timsTOF Pro 2

Next, we investigated whether the increased peptide precursor identification and enhanced sensitivity of timsTOF HT relative to Pro 2 translated to greater reproducibility of quantified peptide precursors. To remove precursors with unreliable quantitation, only precursors with an MS2-based quantity CV below 20% across triplicate measurements of the same sample, peptide loading mass, and LC gradient were considered. For these peptide precursors, timsTOF HT exhibited greater reproducibility compared to timsTOF Pro 2 across 100–1200 ng peptide loading masses and 3 LC gradients (Figure 3). This improvement is particularly evident when examining neat plasma samples with loading masses of ≥ 300 ng and Proteograph plasma samples with peptide loading masses of ≥600 ng. As an example, Proteograph plasma NP2 (900 ng at 100 SPD) had up to an 81% increase in reproducibly quantified peptide precursors with timsTOF HT relative to Pro 2 (Figure 3B). The number of reproducibly quantified peptide precursors gradually increases with higher peptide loading mass in timsTOF HT but stays comparable across peptide loading masses in timsTOF Pro 2 (Figure S8). A similar pattern was observed across individual NPs when considering precursors with CV < 20 and <10% (Figures S8, S9 and Table S4). For samples analyzed with timsTOF HT, compared to samples analyzed by timsTOF Pro 2, the greatest increase in the number of reproducibly quantified precursors was 127% in neat plasma (600 ng at 100 SPD) and 169% in Proteograph plasma (NP4, 1200 ng load at 30 SPD, Table S5). However, at low plasma peptide loading mass, i.e., 100 ng, there was no meaningful difference in quantitative performance between timsTOF HT and Pro 2 or timsTOF Pro 2 exceeded the performance of timsTOF HT (Figures 3, and S8, and S9).

Figure 3

Figure 3. timsTOF HT had greater reproducibility compared to timsTOF Pro 2. The number of reproducibly quantified peptide precursors (CV < 20% of triplicate measurements) was increased by up to 127% in (A) neat and 81% in (B) Proteograph plasma (NP2) with timsTOF HT relative to Pro 2 across different LC gradients. As the loading mass increased, the quantitative reproducibility of timsTOF HT improved significantly compared to timsTOF Pro 2. Abbreviations: CV, coefficient of variation; LC, liquid chromatography; NP, nanoparticle; SPD, samples per day.

Qualitative and Quantitative Improvements of timsTOF HT Allow for Higher Sensitivity and Reproducibility of Putative Lung Cancer Biomarkers Detected in Plasma

We tested both MS systems’ potential for plasma protein biomarker discovery by analyzing plasma from an exploratory case-control study consisting of Proteograph plasma samples from 20 stage IV NSCLC and 20 noncancer control patients (Supporting Information, Figure S10 and Table S2). We compared the differences in both the total number and number of statistically significant peptide precursors obtained from the same set of samples analyzed with either the timsTOF HT or Pro 2 MS systems. The same peptide loading masses (600 ng for NP1–3,5; 300 ng for NP4) and 60 SPD gradient were used for both systems. We observed a gain in total and statistically significant peptide precursors (Wilcoxon test corrected p-value, q < 0.05) between lung cancer and control cohorts when samples were analyzed by timsTOF HT compared to Pro 2. For randomly selected representative Proteograph plasma NP2, the total precursors increased from 15,075 with timsTOF Pro 2 to 22,039 with HT. The number of statistically significant precursors increased from 1010 with timsTOF Pro 2 to 1751 with HT (Figure 4A). Similar trends were observed across all 5 NPs (Figure S11). When examining differences in peptide precursor fold-change between lung cancer and control cohorts, we observed higher fold-change ratios with timsTOF HT compared with Pro 2 (Figure 4A). The combined analysis across all NPs revealed that the timsTOF HT had a 48% increase in the total number of peptide precursors identified and a 52% increase in the number of statistically significant precursors compared to timsTOF Pro 2. In all experiments, the use of the Proteograph improved the performance of the mass spectrometry platform consistent with the significant improvement in depth of coverage described in prior publications that have ranged from 5 - 8X(Figure 4B).

Figure 4

Figure 4. Qualitative and quantitative improvements of timsTOF HT allowed for higher sensitivity and reproducibility of lung cancer biomarkers detected in plasma. (A) Volcano plot comparison of precursors quantified in the control (n = 20) vs. lung cancer (n = 20) samples for representative Proteograph plasma NP2 when data were analyzed by timsTOF HT or Pro 2. (B) Improved sensitivity and reproducibility of timsTOF HT resulted in 48% more identified peptide precursors compared to timsTOF Pro 2, which translated to 52% more statistically significant precursors (q < 0.05) across 5 NPs. Abbreviations: NP, nanoparticle.

Discussion

Click to copy section linkSection link copied!

The ability to deeply profile the plasma proteome and reproducibly quantify protein biomarkers at low concentrations is crucial, particularly when dealing with low-abundance biomarkers, which have important implications for disease diagnosis and monitoring. This is especially relevant when studying large patient cohorts, where technical and biological variability poses a significant challenge. The timsTOF HT advanced technology offers several advantages for plasma proteomic profiling, including higher sensitivity and speed of data acquisition, resulting in increased proteome coverage and reproducible quantitation. (11) State-of-the-art TIMS cartridges and digitizer technology have been shown to enhance the qualitative and quantitative capabilities of the system, providing a high dynamic range and analytical depth for quantitative cell and tissue proteomics. (11) Our data complement the current knowledge and demonstrate the potential for timsTOF HT to advance the field of plasma proteomics and biomarker discovery.
We performed extensive comparisons of timsTOF HT and Pro 2 for neat and Proteograph plasma ranging from 100 to 1200 ng peptide mass load. When comparing the same sample type between the 2 systems, we demonstrated improved precursor identification of up to 76 and 46% in neat and Proteograph plasma, respectively, when samples were analyzed with timsTOF HT as opposed to Pro 2. Proteograph utilizes nanoparticle technology to compress the dynamic range of protein concentrations in plasma, enabling deeper protein profiling. We were able to identify approximately 4.5 times more peptide precursors in Proteograph plasma than in neat plasma across both MS systems. It is intriguing that higher sensitivity of the timsTOF HT does not lead to a reduction in fold increase of peptide precursors suggesting that many more precursors remain to be identified below the current limit of detection. The plasma peptide mass range in this study was limited by the Evotip peptide binding capacity of approximately 1 μg. (21) An alternative LC system will be needed to demonstrate further qualitative and quantitative performance gains of timsTOF HT compared to Pro 2 beyond the 1200 ng peptide mass load.
A comparison of spike-in PQ500 peptide precursor quantification between the 2 MS systems suggested that the higher peptide loading mass reduced precursor ion intensities, resulting in fewer precursors identified with timsTOF Pro 2 but not with HT. Given that the LC system, analytical column, and ionization source were kept consistent across MS systems, our data support the hypothesis that the decrease in precursors identified at higher peptide loading masses observed with timsTOF Pro 2 was due to TIMS-Pro 2 cartridge saturation. Moreover, the increased ion capacity and dynamic range of timsTOF HT resulted in more reproducibly quantified precursors compared to timsTOF Pro 2. The timsTOF HT yielded over 2-fold more reproducible precursors (triplicate LC–MS injections, CV < 20%) compared to timsTOF Pro 2. Additionally, timsTOF HT generally showed improved quantitative capabilities when ≥300 or ≥600 ng peptide mass of neat or Proteograph plasma, respectively, was analyzed.
Our comprehensive assessment of the performance of timsTOF HT and Pro 2 should enable researchers to choose the optimal LC conditions and MS systems for their unique experimental requirements. For example, if a limited amount of sample is available (e.g., <300 ng for neat and <600 ng for Proteograph plasma), comparable quality data can be achieved by both timsTOF MS systems. On the other hand, if peptide loading mass is not limited, a higher peptide load at 100 SPD with timsTOF HT yields comparable results to lower peptide loads at 60 SPD or 30 SPD. With this information, it is possible to increase the throughput without sacrificing data quality.
Using a combination of Proteograph sample processing and timsTOF HT for our exploratory biomarker discovery study, we showed a nearly 50% increase in total plasma precursors identified across the patient cohorts, indicating a substantial enhancement in proteomic depth compared to timsTOF Pro 2. In addition, approximately 50% more statistically significant precursors (q < 0.05) between lung cancer and control cohorts were identified with timsTOF HT compared to timsTOF Pro 2. When considering only commonly identified precursors across both MS systems, we still observed an approximately 50% gain in statistically significant precursors in timsTOF HT compared to that in Pro 2 (Figure S11B). This confirmed that the gain in statistically significant precursors observed by timsTOF HT was due to improvements on both the qualitative and quantitative levels. Notably, our findings on the timsTOF HT qualitative and quantitative precursor gains translate to changes in the peptide and protein level, demonstrating that the increase in dynamic range of timsTOF HT allowed for more accurate measurement of quantitative differences in biologically diverse samples (Figures S11 and S12).
The incorporation of a TIMS-XR cartridge with a higher ion capacity and a higher-speed digitizer improved the ion statistics and measurement fidelity, which resulted in improved sensitivity, precursor identification, reproducibility, and quantitative capability. We demonstrated that the increased sensitivity and precision of timsTOF HT allowed for higher throughput compared to timsTOF Pro 2 without compromising data quality. Our data underscore the potential of the timsTOF HT, in conjunction with Proteograph, to advance plasma biomarker discovery and contribute to the development of novel diagnostic tests for disease detection and monitoring. The Proteograph technology enables deep plasma profiling, while the improved sensitivity and qualitative capabilities of timsTOF HT expand the pool of potential biomarker candidates for further investigation. Critically, more reproducible quantitation of peptide precursors at low concentrations in large-scale discovery studies could provide a more reliable assessment of their potential as diagnostic or prognostic indicators, eventually increasing the success rate of follow-up validation experiments and their translation into a functional clinical test.

Conclusions

Click to copy section linkSection link copied!

In this study, we profiled neat and Proteograph plasma proteomes on the EvosepOne LC system coupled to timsTOF HT and Pro 2 mass spectrometers. We demonstrated that timsTOF HT identified up to 76% more plasma peptide precursors and resulted in a >2-fold increase in reproducible precursors (CV < 20%) compared to timsTOF Pro 2. In all experiments, the use of the Proteograph improved the performance of the mass spectrometry platform consistent with the significant improvement in depth of coverage described in prior publications that have ranged from 5 − 8X. (12−17) Proteograph plasma analyzed at 600–1200 ng peptide loading mass on an EvosepOne-timsTOF HT enabled deep plasma proteome profiling with exceptional quantitative reproducibility and linearity. Qualitative and quantitative improvements of timsTOF HT translated into an approximate 50% increase in total and statistically significant (q < 0.05) plasma precursors between lung cancer and control cohorts when plasma processed by the Proteograph were analyzed by timsTOF HT compared to Pro 2. Together, our results provide new insights into the advances in reliable plasma protein profiling at unprecedented depths, demonstrating the potential for improved plasma biomarker discovery studies and higher translational potential in the clinics.

Data Availability

Click to copy section linkSection link copied!

MS raw files (.d folders) and processed DIA-NN v18.1 outputs are deposited in a public repository PRIDE/ProteomeXchange under the identifiers PXD047854 and PXD047839.

Supporting Information

Click to copy section linkSection link copied!

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.jproteome.3c00646.

  • Plasma sample processing and data acquisition; plasma peptide precursors identified in neat and Proteograph plasma (NP1–5) by timsTOF HT and Pro 2 across a wide range of plasma peptide loading masses and LC gradients (Figure S1); qualitative and quantitative assessment of spike-in PQ500 QC peptides (Figure S2); density distribution of MS1 fold-change quantities for neat and Proteograph plasma (NP1–5) precursors measured in timsTOF HT vs. Pro 2 (Figure S3); density distribution of MS2 fold-change quantities for neat and Proteograph plasma (NP1–5) precursors measured in timsTOF HT vs. Pro 2 (Figure S4); density distribution of MS2 peak area quantities for neat and Proteograph plasma (NP1–5) peptide precursors measured in timsTOF HT vs. Pro 2 (Figure S5); timsTOF HT showed increased sensitivity compared to timsTOF Pro 2 (Figure S6); timsTOF HT demonstrated a superior linear response compared to timsTOF Pro 2 for neat and Proteograph plasma (NP1-NP5) across a wide range of peptide loading masses and LC gradients (Figure S7) timsTOF HT resulted in a higher number of reproducibly quantified precursors (CV < 10% and CV < 20%) in neat and Proteograph plasma (NP1-NP5) compared to timsTOF Pro 2 across a wide range of peptide loading masses and LC gradients (Figure S8); comparison of reproducibly quantified precursors (CV < 20%) in neat and Proteograph plasma (NP1–5) analyzed with timsTOF HT and Pro 2 (Figure S9); demographic plots for lung cancer and control cohorts (Figure S10); timsTOF HT increased total and statistically significant peptide precursors identified in lung cancer and control cohorts (Figure S11); overlap of statistically significant precursors, peptides, and proteins in a cohort of 20 lung cancer and 20 control samples analyzed with timsTOF HT and Pro 2 (Figure S12); peptide dilution schema for neat and Proteograph plasma (Table S1); patient demographic information for the cohorts of 20 lung cancer and 20 control patient samples (Table S2); DIA m/z and IM window ranges within the MS2 fragmentation cycle (Table S3); peptide precursors identified in timsTOF HT and Pro 2 with CV < 10% (Table S4A); peptide precursors identified in timsTOF HT and Pro 2 with CV < 20% (Table S4B); total peptide precursor identified in timsTOF HT vs. Pro 2 in neat and Proteograph plasma; total peptide precursors identified in timsTOF HT vs. Pro 2 in Proteograph plasma NP1–5 (Table S5) (PDF)

Terms & Conditions

Most electronic Supporting Information files are available without a subscription to ACS Web Editions. Such files may be downloaded by article for research use (if there is a public use license linked to the relevant article, that license may permit other uses). Permission may be obtained from ACS for other uses through requests via the RightsLink permission system: http://pubs.acs.org/page/copyright/permissions.html.

Author Information

Click to copy section linkSection link copied!

  • Corresponding Author
  • Authors
    • Dijana Vitko - PrognomiQ Inc., 1900 Alameda de las Pulgas, San Mateo, California 94403, United States
    • Wan-Fang Chou - PrognomiQ Inc., 1900 Alameda de las Pulgas, San Mateo, California 94403, United States
    • Sara Nouri Golmaei - PrognomiQ Inc., 1900 Alameda de las Pulgas, San Mateo, California 94403, United States
    • Joon-Yong Lee - PrognomiQ Inc., 1900 Alameda de las Pulgas, San Mateo, California 94403, United States
    • Chinmay Belthangady - PrognomiQ Inc., 1900 Alameda de las Pulgas, San Mateo, California 94403, United States
    • John Blume - PrognomiQ Inc., 1900 Alameda de las Pulgas, San Mateo, California 94403, United States
    • Jessica K. Chan - PrognomiQ Inc., 1900 Alameda de las Pulgas, San Mateo, California 94403, United States
    • Guillermo Flores-Campuzano - PrognomiQ Inc., 1900 Alameda de las Pulgas, San Mateo, California 94403, United States
    • Yuntao Hu - PrognomiQ Inc., 1900 Alameda de las Pulgas, San Mateo, California 94403, United States
    • Manway Liu - PrognomiQ Inc., 1900 Alameda de las Pulgas, San Mateo, California 94403, United States
    • Mark A. Marispini - PrognomiQ Inc., 1900 Alameda de las Pulgas, San Mateo, California 94403, United States
    • Megan G. Mora - PrognomiQ Inc., 1900 Alameda de las Pulgas, San Mateo, California 94403, United States
    • Saividya Ramaswamy - PrognomiQ Inc., 1900 Alameda de las Pulgas, San Mateo, California 94403, United States
    • Purva Ranjan - PrognomiQ Inc., 1900 Alameda de las Pulgas, San Mateo, California 94403, United States
    • Preston B. Williams - PrognomiQ Inc., 1900 Alameda de las Pulgas, San Mateo, California 94403, United States
    • Robert J. X. Zawada - PrognomiQ Inc., 1900 Alameda de las Pulgas, San Mateo, California 94403, United States
    • Philip Ma - PrognomiQ Inc., 1900 Alameda de las Pulgas, San Mateo, California 94403, United States
  • Author Contributions

    D.V. reviewed and adjusted the experimental plan, contributed to LC–MS sample preparation, supervised LC–MS data acquisition, proposed the data analysis plan, and wrote the manuscript with inputs (methodology) from W.C., S.N.G., J.L., and M.A.M. W.C. prepared LC–MS samples, performed LC–MS data acquisition and coordinated data transfer between R&D and data science departments. S.N.G. and J.L. analyzed the data, visualized results, and provided critical bioinformatics support. S.R. reviewed and adjusted the experimental plan and contributed to LC–MS sample preparation and LC–MS data acquisition. M.A.M. upgraded timsTOF Pro 2 to timsTOF HT and supported LC–MS data acquisition. P.B.W. prepared neat (unfractionated) plasma digest and quality control samples. J.C. and M.G.M. processed pooled control plasma samples via Proteograph. M.L. designed lung cancer and control cohort. R.J.X.Z. coordinated lung cancer and control sample processing via Proteograph. P.R. coordinated the manuscript review process and communication with the Prescott Medical Communication Group. J.B. contributed to the data analysis of lung cancer vs. control samples. Y.H. helped with the quality control assessment. B.W. designed the study, provided essential technical guidance throughout the study, and critically reviewed data. C.B., R.J.X.Z, B.E.W., and P.M. provided a critical review of the manuscript.

  • Funding

    This study was funded in its entirety by PrognomiQ, Inc.

  • Notes
    The authors declare the following competing financial interest(s): Authors with PrognomiQ, Inc. affiliation are (or were) employees of PrognomiQ, Inc. at the time of study completion and receive salary and equity compensation as such. Philip Ma, Bruce E. Wilcox and Robert J.X. Zawada are Seer, Inc. shareholders. Philip Ma is a Bruker Corporation shareholder. Philip Ma is appointed to the Board of Directors of Bruker Corporation; Bruce Wilcox is appointed to the Scientific Advisory Board of Seer, Inc.

Acknowledgments

Click to copy section linkSection link copied!

We want to thank the clinical operation, research & development, and data science departments at PrognomiQ Inc. for collaborative work and fruitful data discussions, particularly Ghristine Boundalian for clinical quality assurance, Brian Koh for clinical documentation review, and Rabab Karimjee for providing technical details on filter-based plasma protein digestion. We extend thanks to Jimmy Zeng, Hao Qian, and Rabab Karimjee for assistance with generating proprietary spectral libraries created independently of this project and Ehdieh Khaledian and Jin Choi for help in performing univariate analysis. We acknowledge Seer, Inc. (Redwood City, CA) for Proteograph sample processing of cancer and control plasma samples. We would also like to thank Bruker Daltonics Inc. (Fremont, CA) for early access to the timsTOF HT technology and directions regarding timsTOF Pro 2 to timsTOF HT upgrade, particularly Oliver Raether, Eduardo Carrascosa, and Narayanaganesh Balasubramanian. We appreciate guidance from Monika Puchalska (Biognosys AG) regarding the PQ500 data search. Editorial and graphical assistance was provided by the Prescott Medical Communication Group (Chicago, IL).

References

Click to copy section linkSection link copied!

This article references 21 other publications.

  1. 1
    WHO. A Short Guide to Cancer Screening: Increase Effectiveness, Maximize Benefits and Minimize Harm; WHO: Copenhagen, 2022.
  2. 2
    World Health Organization. Guide to Cancer Early Diagnosis; World Health Organization: Geneva, 2017.
  3. 3
    Brantley, K.; Sikoa, K. Earlier Cancer Detection Improves Quality of Life and Patient Outcomes, Avalere, 2023. https://avalere.com/insights/earlier-cancer-detection-improves-quality-of-life-and-patient-outcomes.
  4. 4
    Enroth, S. Plasma Proteins and Cancer. Cancers 2021, 13 (5), 1062  DOI: 10.3390/cancers13051062
  5. 5
    Landegren, U.; Hammond, M. Cancer diagnostics based on plasma protein biomarkers: hard times but great expectations. Mol. Oncol. 2021, 15 (6), 17151726,  DOI: 10.1002/1878-0261.12809
  6. 6
    Ignjatovic, V.; Geyer, P. E.; Palaniappan, K. K. Mass Spectrometry-Based Plasma Proteomics: Considerations from Sample Collection to Achieving Translational Data. J. Proteome Res. 2019, 18 (12), 40854097,  DOI: 10.1021/acs.jproteome.9b00503
  7. 7
    Geyer, P. E.; Holdt, L. M.; Teupser, D.; Mann, M. Revisiting biomarker discovery by plasma proteomics. Mol. Syst. Biol. 2017, 13 (9), 942,  DOI: 10.15252/msb.20156297
  8. 8
    Nakayasu, E. S.; Gritsenko, M.; Piehowski, P. D. Tutorial: best practices and considerations for mass-spectrometry-based protein biomarker discovery and validation. Nat. Protoc. 2021, 16 (8), 37373760,  DOI: 10.1038/s41596-021-00566-6
  9. 9
    Naylor, C. N.; Reinecke, T.; Ridgeway, M. E.; Park, M. A.; Clowers, B. H. Validation of Calibration Parameters for Trapped Ion Mobility Spectrometry. J. Am. Soc. Mass Spectrom. 2019, 30 (10), 21522162,  DOI: 10.1007/s13361-019-02289-1
  10. 10
    Michelmann, K.; Silveira, J. A.; Ridgeway, M. E.; Park, M. A. Fundamentals of trapped ion mobility spectrometry. J. Am. Soc. Mass Spectrom. 2015, 26 (1), 1424,  DOI: 10.1007/s13361-014-0999-4
  11. 11
    timsTOF-HT─Expanding the Capabilities of High-Throughput, 4D-proteomics, Bruker Scientific LLC, 2023. https://www.bruker.com/en/products-and-solutions/mass-spectrometry/timstof/timstof-ht.html.
  12. 12
    Blume, J. E.; Manning, W. C.; Troiano, G. Rapid, deep and precise profiling of the plasma proteome with multi-nanoparticle protein corona. Nat. Commun. 2020, 11 (1), 3662  DOI: 10.1038/s41467-020-17033-7
  13. 13
    Huang, T.; Wang, J.; Stukalov, A. Protein Coronas on Functionalized Nanoparticles Enable Quantitative and Precise Large-Scale Deep Plasma Proteomics. bioRxiv 2023,  DOI: 10.1101/2023.08.555225
  14. 14
    Ferdosi, S.; Tangeysh, B.; Brown, T. R.; Everley, P. A.; Figa, M.; McLean, M.; Elgierari, E. M.; Zhao, X.; Garcia, V. J.; Wang, T.; Chang, M. E. K.; Riedesel, K.; Chu, J.; Mahoney, M.; Xia, H.; O’Brien, E. S.; Stolarczyk, C.; Harris, D.; Platt, T. L.; Ma, P.; Goldberg, M.; Langer, R.; Flory, M. R.; Benz, R.; Tao, W.; Cuevas, J. C.; Batzoglou, S.; Blume, J. E.; Siddiqui, A.; Hornburg, D.; Farokhzad, O. C. Engineered nanoparticles enable deep proteomics studies at scale by leveraging tunable nano-bio interactions. Proc Natl Acad Sci U S A 2022, 119, e2106053119  DOI: 10.1073/pnas.2106053119
  15. 15
    Donovan, M. K. R.; Huang, Y.; Blume, J. E.; Wang, J.; Hornburg, D.; Ferdosi, S.; Mohtashemi, I.; Kim, S.; Ko, M.; Benz, R. W.; Platt, T. L.; Batzoglou, S.; Diaz, L. A.; Farokhzad, O. C.; Siddiqui, A. Functionally distinct BMP1 isoforms show an opposite pattern of abundance in plasma from non-small cell lung cancer subjects and controls. PLos One 2023, 18, e0282821  DOI: 10.1371/journal.pone.0282821
  16. 16
    Ferdosi, S.; Stukalov, A.; Hasan, M.; Tangeysh, B.; Brown, T. R.; Wang, T.; Elgierari, E. M.; Zhao, X.; Huang, Y.; Alavi, A.; Lee-McMullen, B.; Chu, J.; Figa, M.; Tao, W.; Wang, J.; Goldberg, M.; O'Brien, E. S.; Xia, H.; Stolarczyk, C.; Weissleder, R.; Farias, V.; Batzoglou, S.; Siddiqui, A.; Farokhzad, O. C.; Hornburg, D. Enhanced Competition at the Nano-Bio Interface Enables Comprehensive Characterization of Protein Corona Dynamics and Deep Coverage of Proteomes. Adv Mater. 2022, 34, e2206008  DOI: 10.1002/adma.202270307
  17. 17
    Koh, B.; Liu, M.; Almonte, R. Multi-omics profiling with untargeted proteomics for blood-based early detection of lung cancer. medRxiv. 2024,  DOI: 10.1101/2024.01.03.24300798
  18. 18
    Meier, F.; Brunner, A. D.; Frank, M. diaPASEF: parallel accumulation-serial fragmentation combined with data-independent acquisition. Nat. Methods 2020, 17 (12), 12291236,  DOI: 10.1038/s41592-020-00998-0
  19. 19
    Demichev, V.; Messner, C. B.; Vernardis, S. I.; Lilley, K. S.; Ralser, M. DIA-NN: neural networks and interference correction enable deep proteome coverage in high throughput. Nat. Methods 2020, 17 (1), 4144,  DOI: 10.1038/s41592-019-0638-x
  20. 20
    Guergues, J.; Wohlfahrt, J.; Stevens, S. M., Jr. Enhancement of Proteome Coverage by Ion Mobility Fractionation Coupled to PASEF on a TIMS-QTOF Instrument. J. Proteome Res. 2022, 21 (8), 20362044,  DOI: 10.1021/acs.jproteome.2c00336
  21. 21
    Sample Loading Protocol for Evotips, Evosep 2020, 2023. https://www.evosep.com/wp-content/uploads/2020/03/Sample-loading-protocol.pdf.

Cited By

Click to copy section linkSection link copied!
Citation Statements
Explore this article's citation statements on scite.ai

This article is cited by 9 publications.

  1. Hui Jing, Paul L. Richardson, Gregory K. Potts, Sameera Senaweera, Violeta L. Marin, Ryan A. McClure, Adam Banlasan, Hua Tang, James E. Kath, Shitalben Patel, Maricel Torrent, Renze Ma, Jon D. Williams. Automated High-Throughput Affinity Capture-Mass Spectrometry Platform with Data-Independent Acquisition. Journal of Proteome Research 2025, Article ASAP.
  2. Che-Fan Huang, Michael A. Hollas, Aniel Sanchez, Mrittika Bhattacharya, Giang Ho, Ambika Sundaresan, Michael A. Caldwell, Xiaoyan Zhao, Ryan Benz, Asim Siddiqui, Neil L. Kelleher. Deep Profiling of Plasma Proteoforms with Engineered Nanoparticles for Top-Down Proteomics. Journal of Proteome Research 2024, 23 (10) , 4694-4703. https://doi.org/10.1021/acs.jproteome.4c00621
  3. Kai Li, Guo Ci Teo, Kevin L. Yang, Fengchao Yu, Alexey I. Nesvizhskii. diaTracer enables spectrum-centric analysis of diaPASEF proteomics data. Nature Communications 2025, 16 (1) https://doi.org/10.1038/s41467-024-55448-8
  4. William F. Beimers, Katherine A. Overmyer, Pavel Sinitcyn, Noah M. Lancaster, Scott T. Quarmby, Joshua J. Coon. A Technical Evaluation of Plasma Proteomics Technologies. 2025https://doi.org/10.1101/2025.01.08.632035
  5. Heather C. Murray, Jonathan Sillar, Maddison Chambers, Nicole M. Verrills. Proteogenomic profiling of acute myeloid leukemia to identify therapeutic targets. Expert Review of Proteomics 2024, 21 (12) , 515-528. https://doi.org/10.1080/14789450.2024.2431272
  6. Rajesh Kumar Soni. Frontiers in plasma proteome profiling platforms: innovations and applications. Clinical Proteomics 2024, 21 (1) https://doi.org/10.1186/s12014-024-09497-2
  7. Jian Song, Hebin Liu, Chengpin Shen, Xiaohui Wu. Beta-DIA: Integrating learning-based and function-based feature scores to optimize the proteome profiling of single-shot diaPASEF mass spectrometry data. 2024https://doi.org/10.1101/2024.11.19.624419
  8. Eva Borràs, Federica Anastasi, Olga Pastor, Marc Suárez-Calvet, Eduard Sabidó. Enhanced proteome profiling of human cerebrospinal fluid using a commercial plasma enrichment strategy. 2024https://doi.org/10.1101/2024.10.07.616086
  9. Vanessa M. Beutgen, Veronika Shinkevich, Johanna Pörschke, Celina Meena, Anna M. Steitz, Elke Pogge von Strandmann, Johannes Graumann, María Gómez-Serrano. Secretome Analysis Using Affinity Proteomics and Immunoassays: A Focus on Tumor Biology. Molecular & Cellular Proteomics 2024, 23 (9) , 100830. https://doi.org/10.1016/j.mcpro.2024.100830
  10. Eleonora Camilleri, Mirjam Kruijt, Paul L. den Exter, Suzanne C. Cannegieter, Nienke van Rein, Christa M. Cobbaert, Bart J.M. van Vlijmen, L. Renee Ruhaak. Quantitative protein mass spectrometry for multiplex measurement of coagulation and fibrinolytic proteins towards clinical application: What, why and how?. Thrombosis Research 2024, 241 , 109090. https://doi.org/10.1016/j.thromres.2024.109090
  11. Hui Jing, Paul L. Richardson, Gregory K. Potts, Sameera Senaweera, Violeta L Marin, Ryan McClure, Adam Banlasan, Hua Tang, James E. Kath, Renze Ma, Jon D. Williams. An Automated High-throughput Affinity Capture-Mass Spectrometry Platform with Data- Independent Acquisition. 2024https://doi.org/10.1101/2024.08.13.607785
  12. Matthew E. K. Chang, Jane Lange, Jessie May Cartier, Travis W. Moore, Sophia M. Soriano, Brenna Albracht, Michael Krawitzky, Harendra Guturu, Amir Alavi, Alexey Stukalov, Xiaoyuan Zhou, Eltaher M. Elgierari, Jessica Chu, Ryan Benz, Juan C. Cuevas, Shadi Ferdosi, Daniel Hornburg, Omid Farokhzad, Asim Siddiqui, Serafim Batzoglou, Robin J. Leach, Michael A. Liss, Ryan P. Kopp, Mark R. Flory. A Scaled Proteomic Discovery Study for Prostate Cancer Diagnostic Markers Using ProteographTM and Trapped Ion Mobility Mass Spectrometry. International Journal of Molecular Sciences 2024, 25 (15) , 8010. https://doi.org/10.3390/ijms25158010
  13. Che-Fan Huang, Michael A. Hollas, Aniel Sanchez, Mrittika Bhattacharya, Giang Ho, Ambika Sundaresan, Michael A. Caldwell, Xiaoyan Zhao, Ryan Benz, Asim Siddiqui, Neil L. Kelleher. Deep Profiling of Plasma Proteoforms with Engineered Nanoparticles for Top-down Proteomics. 2024https://doi.org/10.1101/2024.07.20.604425
  14. Steven M. Yannone, Vikas Tuteja, Olena Goleva, Donald Y.M. Leung, Aleksandr Stotland, Angel J. Keoseyan, Nathan G. Hendricks, Jennifer E. Van Eyk, Simion Kreimer. Blood to Biomarker Quantitation in Under One Hour with Rapid Proteomics using a Hyperthermoacidic Protease. 2024https://doi.org/10.1101/2024.06.01.596979
  15. Jens R. Coorssen, Matthew P. Padula. Proteomics—The State of the Field: The Definition and Analysis of Proteomes Should Be Based in Reality, Not Convenience. Proteomes 2024, 12 (2) , 14. https://doi.org/10.3390/proteomes12020014
  16. Kai Li, Guo Ci Teo, Kevin L. Yang, Fengchao Yu, Alexey I. Nesvizhskii. diaTracer enables spectrum-centric analysis of diaPASEF proteomics data. 2024https://doi.org/10.1101/2024.05.25.595875
  17. Ryo Konno, Masaki Ishikawa, Daisuke Nakajima, Sho Hagiwara, Kaori Inukai, Osamu Ohara, Yusuke Kawashima. Ultra-deep proteomics by Thin-diaPASEF with a 60-cm long column system. 2024https://doi.org/10.1101/2024.04.26.591246
  18. Rajesh Kumar Soni. Frontiers in Plasma Proteome Profiling Platforms: Innovations and Applications.. 2024https://doi.org/10.21203/rs.3.rs-4193960/v1

Journal of Proteome Research

Cite this: J. Proteome Res. 2024, 23, 3, 929–938
Click to copy citationCitation copied!
https://doi.org/10.1021/acs.jproteome.3c00646
Published January 15, 2024

Copyright © 2024 The Authors. Published by American Chemical Society. This publication is licensed under

CC-BY-NC-ND 4.0 .

Article Views

6343

Altmetric

-

Citations

Learn about these metrics

Article Views are the COUNTER-compliant sum of full text article downloads since November 2008 (both PDF and HTML) across all institutions and individuals. These metrics are regularly updated to reflect usage leading up to the last few days.

Citations are the number of other articles citing this article, calculated by Crossref and updated daily. Find more information about Crossref citation counts.

The Altmetric Attention Score is a quantitative measure of the attention that a research article has received online. Clicking on the donut icon will load a page at altmetric.com with additional details about the score and the social media presence for the given article. Find more information on the Altmetric Attention Score and how the score is calculated.

  • Abstract

    Figure 1

    Figure 1. Qualitative performance of timsTOF HT exceeded that of timsTOF Pro 2 across a wide range of plasma peptide loading masses and LC gradients. Compared to timsTOF Pro 2, HT identified up to 76 and 46% more peptide precursors in (A) neat and (B) Proteograph plasma, respectively, with up to 4.5-fold more precursors detected in Proteograph plasma than neat plasma. Red (timsTOF HT) and blue (timsTOF Pro 2) lines represent the number of peptide precursors uniquely detected in all replicate injections (n = 3) of neat and Proteograph plasma samples (NP1–3,5). Proteograph plasma NP4 was not considered for plotting due to insufficient yield for 1200 ng peptide mass load. Abbreviations: LC, liquid chromatography; NP, nanoparticle, SPD, samples per day.

    Figure 2

    Figure 2. timsTOF HT had a greater quantitative linear range compared to timsTOF Pro 2. (A) Distribution of peptide precursor MS2 peak area (triplicate average) ratios quantified with timsTOF HT vs Pro 2. For visualization purposes, the MS2 peak area ratio cutoff was set to 5. (B) Extracted ion chromatogram of MS2 fragment ion (y9) for a randomly selected precursor, FLVGPDGIPIMR (2+), from within the 3rd quartile of the total precursor intensity range. (C) R-squared distribution for the precursors quantified in all three replicate measurements of each peptide loading mass for Proteograph plasma NP2 at 60 SPD gradient with timsTOF HT (n = 4256) and timsTOF Pro 2 (n = 3331). Abbreviations: NP, nanoparticle; SPD, samples per day.

    Figure 3

    Figure 3. timsTOF HT had greater reproducibility compared to timsTOF Pro 2. The number of reproducibly quantified peptide precursors (CV < 20% of triplicate measurements) was increased by up to 127% in (A) neat and 81% in (B) Proteograph plasma (NP2) with timsTOF HT relative to Pro 2 across different LC gradients. As the loading mass increased, the quantitative reproducibility of timsTOF HT improved significantly compared to timsTOF Pro 2. Abbreviations: CV, coefficient of variation; LC, liquid chromatography; NP, nanoparticle; SPD, samples per day.

    Figure 4

    Figure 4. Qualitative and quantitative improvements of timsTOF HT allowed for higher sensitivity and reproducibility of lung cancer biomarkers detected in plasma. (A) Volcano plot comparison of precursors quantified in the control (n = 20) vs. lung cancer (n = 20) samples for representative Proteograph plasma NP2 when data were analyzed by timsTOF HT or Pro 2. (B) Improved sensitivity and reproducibility of timsTOF HT resulted in 48% more identified peptide precursors compared to timsTOF Pro 2, which translated to 52% more statistically significant precursors (q < 0.05) across 5 NPs. Abbreviations: NP, nanoparticle.

  • References


    This article references 21 other publications.

    1. 1
      WHO. A Short Guide to Cancer Screening: Increase Effectiveness, Maximize Benefits and Minimize Harm; WHO: Copenhagen, 2022.
    2. 2
      World Health Organization. Guide to Cancer Early Diagnosis; World Health Organization: Geneva, 2017.
    3. 3
      Brantley, K.; Sikoa, K. Earlier Cancer Detection Improves Quality of Life and Patient Outcomes, Avalere, 2023. https://avalere.com/insights/earlier-cancer-detection-improves-quality-of-life-and-patient-outcomes.
    4. 4
      Enroth, S. Plasma Proteins and Cancer. Cancers 2021, 13 (5), 1062  DOI: 10.3390/cancers13051062
    5. 5
      Landegren, U.; Hammond, M. Cancer diagnostics based on plasma protein biomarkers: hard times but great expectations. Mol. Oncol. 2021, 15 (6), 17151726,  DOI: 10.1002/1878-0261.12809
    6. 6
      Ignjatovic, V.; Geyer, P. E.; Palaniappan, K. K. Mass Spectrometry-Based Plasma Proteomics: Considerations from Sample Collection to Achieving Translational Data. J. Proteome Res. 2019, 18 (12), 40854097,  DOI: 10.1021/acs.jproteome.9b00503
    7. 7
      Geyer, P. E.; Holdt, L. M.; Teupser, D.; Mann, M. Revisiting biomarker discovery by plasma proteomics. Mol. Syst. Biol. 2017, 13 (9), 942,  DOI: 10.15252/msb.20156297
    8. 8
      Nakayasu, E. S.; Gritsenko, M.; Piehowski, P. D. Tutorial: best practices and considerations for mass-spectrometry-based protein biomarker discovery and validation. Nat. Protoc. 2021, 16 (8), 37373760,  DOI: 10.1038/s41596-021-00566-6
    9. 9
      Naylor, C. N.; Reinecke, T.; Ridgeway, M. E.; Park, M. A.; Clowers, B. H. Validation of Calibration Parameters for Trapped Ion Mobility Spectrometry. J. Am. Soc. Mass Spectrom. 2019, 30 (10), 21522162,  DOI: 10.1007/s13361-019-02289-1
    10. 10
      Michelmann, K.; Silveira, J. A.; Ridgeway, M. E.; Park, M. A. Fundamentals of trapped ion mobility spectrometry. J. Am. Soc. Mass Spectrom. 2015, 26 (1), 1424,  DOI: 10.1007/s13361-014-0999-4
    11. 11
      timsTOF-HT─Expanding the Capabilities of High-Throughput, 4D-proteomics, Bruker Scientific LLC, 2023. https://www.bruker.com/en/products-and-solutions/mass-spectrometry/timstof/timstof-ht.html.
    12. 12
      Blume, J. E.; Manning, W. C.; Troiano, G. Rapid, deep and precise profiling of the plasma proteome with multi-nanoparticle protein corona. Nat. Commun. 2020, 11 (1), 3662  DOI: 10.1038/s41467-020-17033-7
    13. 13
      Huang, T.; Wang, J.; Stukalov, A. Protein Coronas on Functionalized Nanoparticles Enable Quantitative and Precise Large-Scale Deep Plasma Proteomics. bioRxiv 2023,  DOI: 10.1101/2023.08.555225
    14. 14
      Ferdosi, S.; Tangeysh, B.; Brown, T. R.; Everley, P. A.; Figa, M.; McLean, M.; Elgierari, E. M.; Zhao, X.; Garcia, V. J.; Wang, T.; Chang, M. E. K.; Riedesel, K.; Chu, J.; Mahoney, M.; Xia, H.; O’Brien, E. S.; Stolarczyk, C.; Harris, D.; Platt, T. L.; Ma, P.; Goldberg, M.; Langer, R.; Flory, M. R.; Benz, R.; Tao, W.; Cuevas, J. C.; Batzoglou, S.; Blume, J. E.; Siddiqui, A.; Hornburg, D.; Farokhzad, O. C. Engineered nanoparticles enable deep proteomics studies at scale by leveraging tunable nano-bio interactions. Proc Natl Acad Sci U S A 2022, 119, e2106053119  DOI: 10.1073/pnas.2106053119
    15. 15
      Donovan, M. K. R.; Huang, Y.; Blume, J. E.; Wang, J.; Hornburg, D.; Ferdosi, S.; Mohtashemi, I.; Kim, S.; Ko, M.; Benz, R. W.; Platt, T. L.; Batzoglou, S.; Diaz, L. A.; Farokhzad, O. C.; Siddiqui, A. Functionally distinct BMP1 isoforms show an opposite pattern of abundance in plasma from non-small cell lung cancer subjects and controls. PLos One 2023, 18, e0282821  DOI: 10.1371/journal.pone.0282821
    16. 16
      Ferdosi, S.; Stukalov, A.; Hasan, M.; Tangeysh, B.; Brown, T. R.; Wang, T.; Elgierari, E. M.; Zhao, X.; Huang, Y.; Alavi, A.; Lee-McMullen, B.; Chu, J.; Figa, M.; Tao, W.; Wang, J.; Goldberg, M.; O'Brien, E. S.; Xia, H.; Stolarczyk, C.; Weissleder, R.; Farias, V.; Batzoglou, S.; Siddiqui, A.; Farokhzad, O. C.; Hornburg, D. Enhanced Competition at the Nano-Bio Interface Enables Comprehensive Characterization of Protein Corona Dynamics and Deep Coverage of Proteomes. Adv Mater. 2022, 34, e2206008  DOI: 10.1002/adma.202270307
    17. 17
      Koh, B.; Liu, M.; Almonte, R. Multi-omics profiling with untargeted proteomics for blood-based early detection of lung cancer. medRxiv. 2024,  DOI: 10.1101/2024.01.03.24300798
    18. 18
      Meier, F.; Brunner, A. D.; Frank, M. diaPASEF: parallel accumulation-serial fragmentation combined with data-independent acquisition. Nat. Methods 2020, 17 (12), 12291236,  DOI: 10.1038/s41592-020-00998-0
    19. 19
      Demichev, V.; Messner, C. B.; Vernardis, S. I.; Lilley, K. S.; Ralser, M. DIA-NN: neural networks and interference correction enable deep proteome coverage in high throughput. Nat. Methods 2020, 17 (1), 4144,  DOI: 10.1038/s41592-019-0638-x
    20. 20
      Guergues, J.; Wohlfahrt, J.; Stevens, S. M., Jr. Enhancement of Proteome Coverage by Ion Mobility Fractionation Coupled to PASEF on a TIMS-QTOF Instrument. J. Proteome Res. 2022, 21 (8), 20362044,  DOI: 10.1021/acs.jproteome.2c00336
    21. 21
      Sample Loading Protocol for Evotips, Evosep 2020, 2023. https://www.evosep.com/wp-content/uploads/2020/03/Sample-loading-protocol.pdf.
  • Supporting Information

    Supporting Information


    The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.jproteome.3c00646.

    • Plasma sample processing and data acquisition; plasma peptide precursors identified in neat and Proteograph plasma (NP1–5) by timsTOF HT and Pro 2 across a wide range of plasma peptide loading masses and LC gradients (Figure S1); qualitative and quantitative assessment of spike-in PQ500 QC peptides (Figure S2); density distribution of MS1 fold-change quantities for neat and Proteograph plasma (NP1–5) precursors measured in timsTOF HT vs. Pro 2 (Figure S3); density distribution of MS2 fold-change quantities for neat and Proteograph plasma (NP1–5) precursors measured in timsTOF HT vs. Pro 2 (Figure S4); density distribution of MS2 peak area quantities for neat and Proteograph plasma (NP1–5) peptide precursors measured in timsTOF HT vs. Pro 2 (Figure S5); timsTOF HT showed increased sensitivity compared to timsTOF Pro 2 (Figure S6); timsTOF HT demonstrated a superior linear response compared to timsTOF Pro 2 for neat and Proteograph plasma (NP1-NP5) across a wide range of peptide loading masses and LC gradients (Figure S7) timsTOF HT resulted in a higher number of reproducibly quantified precursors (CV < 10% and CV < 20%) in neat and Proteograph plasma (NP1-NP5) compared to timsTOF Pro 2 across a wide range of peptide loading masses and LC gradients (Figure S8); comparison of reproducibly quantified precursors (CV < 20%) in neat and Proteograph plasma (NP1–5) analyzed with timsTOF HT and Pro 2 (Figure S9); demographic plots for lung cancer and control cohorts (Figure S10); timsTOF HT increased total and statistically significant peptide precursors identified in lung cancer and control cohorts (Figure S11); overlap of statistically significant precursors, peptides, and proteins in a cohort of 20 lung cancer and 20 control samples analyzed with timsTOF HT and Pro 2 (Figure S12); peptide dilution schema for neat and Proteograph plasma (Table S1); patient demographic information for the cohorts of 20 lung cancer and 20 control patient samples (Table S2); DIA m/z and IM window ranges within the MS2 fragmentation cycle (Table S3); peptide precursors identified in timsTOF HT and Pro 2 with CV < 10% (Table S4A); peptide precursors identified in timsTOF HT and Pro 2 with CV < 20% (Table S4B); total peptide precursor identified in timsTOF HT vs. Pro 2 in neat and Proteograph plasma; total peptide precursors identified in timsTOF HT vs. Pro 2 in Proteograph plasma NP1–5 (Table S5) (PDF)


    Terms & Conditions

    Most electronic Supporting Information files are available without a subscription to ACS Web Editions. Such files may be downloaded by article for research use (if there is a public use license linked to the relevant article, that license may permit other uses). Permission may be obtained from ACS for other uses through requests via the RightsLink permission system: http://pubs.acs.org/page/copyright/permissions.html.