Advances in Bottom-Up Proteomics

Senior Research Scientist in the Protein Analysis group at Promega

Chris Hosfield is a Senior Research Scientist in the Protein Analysis group at Promega. He has led the development of tools for proteomics and antibody characterization and has broad expertise in working with proteolytic enzymes and preparing protein samples for analysis by mass spectrometry. Prior to joining Promega, Chris held senior scientist positions at several biotechnology companies where he applied proteomic methods to the discovery and development of both small molecules and biotherapeutics. He received his Ph.D. in biochemistry from Queen’s University. Promega Corporation is a global leader in providing innovative solutions and technical support to the life sciences industry, offering more than 4,000 products used in research, diagnostics, forensics, and applied testing worldwide.

December 30, 2025

Add On Google

In bottom-up proteomics, the ultimate goal is to precisely identify and quantify every protein and post-translational modification (PTM) in a single cell. This goal has yet to be fully realized, but recent innovations in sample preparation, instrument technologies, and data analysis strategies are converging to dramatically expand the depth of coverage of the proteome and generate more robust insights into important biological questions.

Sample preparation is critical to data quality

Generating high-quality proteomic data through liquid-chromatography-mass spectrometry (LC-MS) begins with a high-quality sample. An expanding sample prep toolbox is enabling researchers to improve efficiency and build new experimental designs, from PTM-mapping to analysis of single cells.

Trypsin and beyond: New and improved proteases for more complete proteomes

The process begins by using a protease to digest a protein sample into smaller peptides that are subsequently sequenced and identified by LC-MS/MS. Trypsin has long been the protease of choice for bottom-up proteomics for its balance of specificity and efficiency, as well as the compatibility of tryptic-peptides with mass spectrometry. Still, numerous innovations have been made in recent years to improve the quality of tryptic peptides. Recombinant trypsin has shown improved specificity and better resistance to autoproteolysis. Furthermore, digestion efficiency is significantly improved by supplementing trypsin with complementary proteases such as Lys-C and/or Arg-C. In addition, fast digestion kits and heat stable formulations can reduce sample prep time without sacrificing performance.

Search Mass Spectrometry
Search Now Search our directory to find mass spec tools for your research needs.

More researchers are also utilizing trypsin alternatives like Lys-C, Glu-C, Asp-N, Arg-C, ProAlanase, chymotrypsin, and others, which offer distinct cleavage specificities that help uncover parts of proteins missed by trypsin alone.¹ This strategy increases depth and confidence in protein identification and can enhance analysis of protein isoforms and PTMs, such as phosphorylation, acetylation, glycosylation, ubiquitination, and many others. These proteases can be combined with PTM-specific enrichment technologies to analyze and quantify PTMs even at very low levels.

Beyond proteases, breakthroughs in sample preparation have largely been achieved using suspension-trapping of proteins on quartz filters, or solvent-induced protein capture on magnetic beads as in the SP3/PAC-based methods. These techniques are very useful since they allow protein samples to be solubilized in harsh detergents or denaturants, which can then be washed away prior to protease-mediated digestion and downstream MS analysis. Magnetic bead-based technologies are especially promising due to their relatively low cost and compatibility with numerous automated sample preparation systems, which have also become more ubiquitous in recent years.^2,3

Low-input and single-cell proteomics

Cells of the same type can differ dramatically in protein expression, modification states, and biological function. Traditional proteomics approaches utilize populations of cells, which average out these differences, masking rare or transient cellular behaviors that could be biologically or clinically significant. Single-cell proteomics enables researchers to resolve this variability, but given that a typical mammalian cell contains only ~200 picograms of protein, analysis is particularly challenging. Workflows such as nPOP, nanoPOTS, and Chip-Tip minimize sample loss and maximize sensitivity via automated workflows that enable digestion and isobaric labeling in nanoliter volumes, allowing identification of thousands of proteins from individual cells, particularly when paired with ultra-sensitive mass spectrometry technologies.^4-6

Automation enables reproducibility and throughput

Automation is increasingly integral to modern proteomics workflows. Robotic sample handling minimizes human error and enhances throughput, allowing large-scale, reproducible analyses across hundreds of samples. When paired with AI-driven data processing, automated pipelines accelerate discovery and free researchers to focus on biological interpretation rather than manual optimization.

Instruments that see more, faster

Hardware advances have matched or even surpassed the gains in sample prep. Improvements in nanoflow LC separation technologies (throughput, robustness, column particles, pressure limits) have been paired with next-generation mass spectrometers to dramatically increase speed, sensitivity, and ion-utilization efficiency.⁷ These innovations have allowed researchers to achieve deeper proteome coverage even from very low sample amounts and are now routinely reporting >10,000 proteins from human cell lines or >5,000 proteins from single cells.^6-10 These improvements are not just incremental. They’re enabling entirely new experimental designs that were impractical or impossible just a few years ago.

Smarter data: Acquisition and analysis breakthroughs

DIA becomes the norm

Data-independent acquisition (DIA) has evolved from an emerging technique to a mainstream standard. Unlike data-dependent acquisition (DDA), which targets selected (usually the most abundant) precursors, DIA fragments all ions in predefined mass windows, thereby capturing a comprehensive and unbiased snapshot of the proteome. This broad sampling increases detection of low-abundance peptides, ensures consistent quantification across replicates, and delivers greater completeness for large cohorts or PTM-focused studies where reproducibility is essential.

AI and deep learning in data analysis

Artificial intelligence and deep learning are transforming how proteomic data are interpreted. These models can now predict peptide fragmentation, retention time, and even ion mobility with near-experimental accuracy, allowing spectra to be matched more confidently and modified peptides to be identified with greater precision.¹¹ They enable library-free DIA searches, automate PTM detection, and extract meaningful signals from complex datasets that would previously have gone unnoticed. By learning from experimental variability, AI-driven tools also reduce missing values and improve quantitation—especially valuable in single-cell and PTM studies.

Conclusion

Advances in sample preparation, LC-MS instrumentation, and bioinformatics are converging to deliver deeper, more reproducible proteomic data. Whether mapping the phosphoproteome of a tumor biopsy, profiling ubiquitin chains in response to drug treatment, or quantifying thousands of proteins from a single cell, today’s workflows are enabling what seemed out of reach just a few years earlier. And while instrumentation continues to deliver groundbreaking platforms, improvements in sample prep methodologies have likewise been critical in achieving new levels of proteomic depth.