A Simplified High Throughput Cell-Based Assay for Increased Proteomic Coverage of Cardiomyocytes
At ASMS 2023, Saeed Seyedmohammad, Ph.D., Staff Scientist at Cedars-Sinai Medical Institute, presented the results of an ultrasonication-assisted tryptic digestion method to increase protein coverage using Covaris’ Adaptive Focused Acoustics® (AFA®) technology. Read on to learn how incorporating AFA technology into a simplified detergent-free protein extraction workflow enabled deeper proteome coverage of biologically perturbed cardiomyocytes, opening the door to biomarker and drug discovery.
Introduction
Clinical proteomics—or the quantification of an individual’s proteome—can inform the prevention, diagnosis, and treatment of complex conditions like heart disease. This research requires high throughput, reproducibility, streamlined workflows, and deep coverage to extract meaningful high-quality data that can advance precision medicine approaches.
To this end, Dr. Seyedmohammad and his group at the Jenny Van Eyk Lab investigated better methods to identify and quantify proteins within biologically perturbed cells such as ischemic cardiomyocytes. To improve their results, they turned to Covaris, known for their sample prep solutions for NGS and other -omics, including protein disruption, lysis, extraction, purification and digestion. In this talk, Dr. Seyedmohammad reviews how they significantly improved protein coverage and discovered key insights about lower abundant proteins in these cells by incorporating an AFA-assisted tryptic digestion step using Covaris’ AFA technology into their workflow.
An Automated Protein Extraction Workflow Identifies Protein Expression in Cardiac Injury
Dr. Seyedmohammad’s research builds on his previous work in which he developed and validated an automated, detergent-free high-throughput (HTP) cell-based screening workflow for investigating how protein expression changed when cardiomyocytes became hypoxic and then re-perfused. This approach harnessed bottom-up proteomics, which requires the digestion of intact proteins into peptides before analysis.
Analysis of differentially expressed proteins (DEPs) validated the stability of the workflow; downregulation was observed in the hypoxic state. As expected, the top-ranked downregulated DEPs corresponded to proteins that are depleted in hypoxia. Additional results showed that four of the top DEPs were upregulated and associated with adaptation to oxygen depletion and reduction of cholesterol and fatty acid accumulation.
To deepen their understanding of cardiac modulators in ischemic cells, the team tested whether adding AFA-assisted tryptic digestion to their workflow would help identify and quantify more proteins, especially lower abundant ones.
AFA-Assisted Tryptic Digestion Increased Detection of Lower Abundant Proteins
To test this hypothesis, the team analyzed AC16 cells with and without sonication at two hours of digestion. In the AFA-assisted condition, they alternated between sonication and digestion in a continuous cycle, optimizing high throughput sample processing in a 96-row format (Figure 1).
Figure 1. Full workflow with the tryptic digestion step highlighted.
Initial statistical analyses showed that the AFA-assisted digested samples demonstrated a tighter association and more positive correlation compared to the non-sonicated classically digested samples. Principal component analysis confirmed that biological replicates from the AFA-assisted digestion samples were more correlated with each other than classically digested biological replicates.
To examine differences in detection of lower abundant proteins in AFA-assisted and classically digested samples, Dr. Seyedmohammad constructed a protein order for both conditions and ranked them against their mass spectrometry intensities, then built a polynomial best fit to match the multilateral distribution of the dataset (Figure 2). Lower abundant proteins were better detected with a higher intensity signal in the AFA-assisted tryptic digestion condition.
Figure 2. Lower abundant proteins were better detected in sonicated samples (green curves) than non-sonicated samples (blue curves).
Findings were similar at the peptide level when intensities were plotted on a dynamic range graph. 18,500 peptides were detected in both conditions over the corresponding biological replicates. In Figure 3, the dotted line intersecting the polynomial trend line demonstrates the ability to detect specific high or low abundant proteins and is higher for the AFA-assisted digestion condition.
Figure 3. Lower abundant peptides were better detected in sonicated samples (blue curves) than non-sonicated samples (orange curves).
AFA-Assisted Tryptic Digestion Detected Three Times More Upregulated Proteins
Differential expression analysis revealed that 47% of the total detected protein was significant with AFA-assisted tryptic digestion, an indicator of strong workflow efficiency. A volcano plot showed three times more upregulated proteins detected when samples were sonicated after two hours compared to controls (Figure 4).
Figure 4. Volcano plot showing differences in DEP detection across both conditions.
Violin plots of top DEPs showed that only two of the top 10 DEPs were downregulated, while the remaining upregulated proteins showed a tight spread of data. This low variability among biological replicates in upregulated proteins was confirmed by heat map analysis (Figure 5).
Figure 5. Violin plots and heatmap indicating data variability for DEPs.
To better probe the molecular function of DEPs in their samples, the team used Gene Ontology annotations. One hundred significantly impacted proteins were involved in cell substrate junction and focal adhesion and annotated as cell surface markers that could be targeted for drug delivery. Another 90 impacted proteins were transmembrane proteins, allowing cells to adhere to each other. These findings could inform future biomarker discovery work.
Increasing Peptide Load Enhanced Protein Coverage
To investigate the upper limit of detection, or how much more they could increase protein coverage, the team increased peptide load threefold, up to 390 ng. Results showed that this threefold increase in protein load increased the detection of DEPs by about 40%. Further DEP analysis revealed a similar twofold increase in protein detection with a 390 ng injection. Notably, this data was produced over a 45-minute acquisition gradient, suggesting that a longer acquisition gradient would yield an even greater protein coverage.
AFA-Assisted Tryptic Digestion Improved Digestion Efficiency
Missed cleavages, or undigested peptide bonds, are an important consideration when evaluating digestion performance. In AFA-assisted tryptic digestion, 60% of peptides displayed no missed cleavages involving lysine and arginine and a 90% combined total ratio of peptides showed just one or none (Figure 6).
Figure 6. Missed cleavage results with a smaller and higher peptide load.
AFA-Assisted Digestion Reveals Upregulated Mitochondrial Proteins
Most of the identified upregulated proteins between the AFA-assisted and classically digested conditions played pivotal roles in mitochondrial organization and transport. There were more upregulated proteins with greater full change values based on mass spectrometry intensities than downregulated ones.
Further Refinement of AFA-Assisted Digestion Conditions Expands Protein Coverage
Given that AFA-assisted digestion improved protein coverage, the team wanted to explore if modifying AFA conditions could deepen coverage even further. Six conditions involving peak intensity power (PIP), duty factor (DF) and assay incubation type were tested. While details will be discussed in a forthcoming publication, Dr. Seyedmohammad shared that one representative condition detected an additional 300 proteins compared to others.
A differential expression analysis of the top two AFA-assisted conditions showed that the representative condition increased protein coverage by about 15%. Gene Ontology annotation showed that membrane-associated proteins were detected in this representative condition, and the dominant cluster of proteins in the two conditions involved focal adhesion and cell substrate junctions.
Violin plots and heat maps reveal that in the representative condition, the top 10 DEPs were associated with important processes like the mitochondrial respiratory chain and other high-level processes (Figure 7). These findings were enabled by AFA-assisted digestion and could inform future efforts to target the role of these proteins.
Figure 7. Heat map and violin plots for DEP analysis in the AFA-assisted digestion representative condition.
Future Directions Applies AFA Technology to Streamline and Improve Protein Detection
Dr. Seyedmohammad and his group plan to harness the benefits of AFA-assisted digestion to make strides towards their long-term goal of developing new validated assays for clinical proteomics. One plan is to integrate this high throughput workflow with azidohomoaline (AHA) labeling, a method of characterizing newly synthesized proteins, enabling the study of multiple protein folding and post-translational modifications.
Conclusion
Dr. Seyedmohammad’s team have successfully added AFA-assisted digestion powered by Covaris’ AFA technology to an automated protein extraction workflow to increase proteomic coverage and enhance detection of lower detection proteins that may be future targets for drug discovery.
Are you interested in learning more about Covaris’ AFA technology for protein sample preparation? Click here to reach out to one of our specialists.