Ultrafast enzymatic digestion of proteins by microdroplet mass spectrometry

The study addresses the long-standing bottleneck in bottom-up proteomics: the lengthy and often incomplete enzymatic digestion step required to generate peptides for mass spectrometric sequencing. Prior observations show that reactions in micron-sized droplets can proceed much faster than in bulk solutions, likely due to unique interfacial environments. The authors hypothesize that forming aqueous microdroplets containing proteins and trypsin will dramatically accelerate proteolysis without pre-treatment, improving sequence coverage and reducing digestion times from hours to sub-millisecond. The goal is to demonstrate ultrafast, efficient tryptic digestion in microdroplets directly coupled to mass spectrometry and evaluate its utility on challenging proteins and a therapeutic antibody.

Previous work has documented reaction-rate acceleration in microdroplets generated by DESI, ESI, microfluidics, theta tips, paper spray, and related ionization methods, with many organic and materials syntheses proceeding far faster than in bulk. Proposed contributing factors include droplet size, surface charge, reagent confinement, solvent composition, evaporation, and distinct interfacial pH or redox environments. In proteomics, conventional strategies to accelerate tryptic digestion include elevated temperature, microwave or ultrasound energy, high pressure, immobilized enzyme reactors (on-column, on-chip, magnetic particles), organic co-solvents, and infrared heating, yet typical digestion still requires minutes to hours and often yields incomplete coverage. Microdroplet chemistry has been suggested as compatible with biochemical systems, but its application to protein digestion had been little explored. The authors build on these observations to test microdroplet-enabled proteolysis and compare with established acceleration techniques.

The authors implemented electrosonic spray ionization (ESSI) to generate aqueous microdroplets from a homemade sprayer comprising a fused silica capillary (50 µm i.d., 148 µm o.d.) surrounded by rapidly flowing dry N₂ at 120 psi. A high voltage of ±3 kV was applied. Typical solutions contained protein or peptide at 10 µM and trypsin at 5 µg mL⁻¹ in 5 mM ammonium bicarbonate (NH₄HCO₃, pH 8) or 5 mM ammonium acetate (NH₄OAc, pH 8). The infusion rate was 10 µL min⁻¹. Droplet sizes were characterized by laser particle analysis: ESSI produced ~6–9 µm droplets; a commercial heated ESI probe (500 µm needle, 10 psi sheath gas) produced ~60 µm droplets. Microdroplet velocities were ~84 ± 18 m s⁻¹ (from high-speed imaging). By varying the spray-to-MS inlet distance (2–50 mm; typical 20 mm), the in-flight reaction time was tuned (up to ~0.6 ms at 50 mm), while digestion ceased upon droplet evaporation in the heated inlet (275 °C), confirmed by varying inlet temperature with no change in digestion extent. MS analyses were performed on an LTQ Orbitrap Elite (Thermo Scientific). For peptide sequencing, CID MS/MS was used (isolation width 1 m/z; collision energy 25 a.u.; resolution 12,000). Data handling used Xcalibur and IGOR Pro. Peptide identification used Protein Prospector MS-digest (UniprotKB database, trypsin, up to three missed cleavages; variable modifications as appropriate; 5 ppm mass tolerance). Controls included: (1) commercial ESI-MS analysis of the same solutions; (2) bulk-phase digestions at 37 °C (typically 3–14 h) after denaturation (95 °C, 5 min), with aliquots frozen at set times and analyzed by commercial ESI-MS; and (3) buffer and pH variation studies, including NH₄HCO₃ versus NH₄OAc, and pH dependence. Model systems: ACTH(1–24), myoglobin, cytochrome c, αS1-casein, a synthetic peptide with a slow tryptic site (LYAA-DTR-LYAVR), and trastuzumab antibody light and heavy chains. Post-PAGE workflows involved SDS–PAGE separation, Coomassie staining, gel extraction using a commercial kit, precipitation/desalting, and microdroplet digestion. For antibody sequencing, iodoacetamide was used to block thiols; elastase was also tested for unspecific digestion to improve coverage. Both positive and negative ion modes were evaluated; negative mode with NH₄HCO₃ was found advantageous for detecting singly deprotonated peptides and improving coverage.

- Microdroplet-enabled tryptic digestion proceeds to near-completion within sub-millisecond flight times. Digestion extent scales with droplet travel distance/time; reactions effectively stop upon entry into the heated inlet. - ACTH(1–24): Increasing sprayer–inlet distance from 2 to 20–50 mm markedly increased digestion; 26 fragment peaks covering the whole sequence were identified in microdroplets. At 50 mm (~0.6 ms), intact ACTH signals became minimal. Bulk digestion at 37 °C for 3 h yielded only 16 peptide peaks and comparable but less comprehensive digestion than microdroplets. - Myoglobin (153 aa, protease-resistant): Commercial ESI-MS showed minimal digestion; bulk 14 h at 37 °C gave ~60% sequence coverage (13 peptides). Microdroplet ESSI in positive mode yielded ~86% coverage (19 peptides). In negative mode (−3 kV), 55 peaks corresponding to 38 peptides provided 100% sequence coverage in <1 ms; all trypsin-cleavable sites (after K/R, except when followed by P) were cleaved, and starting material was nearly fully digested. - Cytochrome c: ~83% coverage in positive mode; 100% in negative mode. - Post-PAGE digestion: αS1-casein ~90.3% coverage and cytochrome c ~99% coverage achieved via microdroplets after gel extraction. - Synthetic peptide with slow kinetics (k = 0.24×10⁻³ s⁻¹; LYAA-DTR-LYAVR): Microdroplets achieved 100% digestion; commercial ESI showed ~5% digestion; bulk 1 h at 37 °C reached ~80%. A related peptide LYAA-DK-LYAVR could not be digested in microdroplets, indicating sequence-context limitations. - Antibody trastuzumab: After PAGE separation and microdroplet digestion (−3 kV), light chain achieved 100% coverage; heavy chain ~86% coverage. Using elastase increased heavy-chain coverage to ~98%. Conventional overnight bulk digestion at 37 °C gave ~74% heavy-chain coverage, demonstrating a clear advantage for microdroplets. - Operational insights: Negative ion mode and NH₄HCO₃ buffer improved detection and coverage (singly deprotonated peptides, less spectral complexity). Inlet temperature and spray voltage had little effect on the digestion extent but aided detection. Initial droplet size critically impacted acceleration (ESSI ~9 µm vs commercial ESI ~60 µm), consistent with limited flight time and minimal fission/evaporation before inlet under standard ESI conditions. - Mechanistic clues: Potential contributors include higher interfacial OH⁻ availability and redox activity (hydroxyl radicals, H₂O₂ formation), ammonium bicarbonate-induced CO₂ outgassing and bubble-mediated unfolding, and chain ejection model-driven enrichment of unfolded proteins at droplet surfaces.

The findings demonstrate that aqueous microdroplets generated by ESSI can drive tryptic proteolysis from an hours-long bulk process to sub-millisecond reaction times while improving sequence coverage, even for challenging proteins like myoglobin and large biologics such as trastuzumab. The enhanced digestion likely arises from distinct microdroplet interfacial phenomena: altered pH distribution, strong interfacial electric fields, and facile redox activity leading to hydroxyl radical and H₂O₂ formation; increased surface concentration; and physical effects such as ammonium bicarbonate-induced bubble formation promoting unfolding, combined with chain ejection that enriches unfolded proteins at the droplet surface. Negative ion mode and NH₄HCO₃ buffer further streamline detection, yielding predominantly singly deprotonated peptides with reduced spectral congestion. Comparisons with bulk digestion and commercial ESI underscore the importance of small initial droplet size and limited flight time/fission in conventional sources. Collectively, these results directly address the proteolysis bottleneck in bottom-up proteomics and suggest broad applicability to protease-resistant targets and post-gel workflows, with immediate implications for rapid peptide mapping and therapeutic antibody sequence confirmation.

This work introduces a simple, room-temperature microdroplet-MS approach that achieves near-instantaneous, high-coverage tryptic digestion, reducing reaction time from hours to less than a millisecond and improving sequence coverage across diverse targets, including complete coverage of myoglobin and trastuzumab light chains. The method integrates seamlessly as an ionization/digestion interface and is compatible with gel-extracted samples and alternative enzymes (e.g., elastase) to further boost coverage. Future research should focus on deeper mechanistic elucidation of interfacial effects, optimizing droplet generation and buffer composition for different proteases and protein classes, expanding to complex biological matrices, and integrating with upstream sample preparation and downstream LC-MS/MS workflows for comprehensive proteomics.

- The acceleration and coverage gains depend on initial droplet size and microdroplet flight conditions; commercial ESI sources with larger droplets show minimal acceleration. - Sequence context can still limit cleavage (e.g., D immediately preceding K resisted digestion), and trypsin’s site specificity constrains maximum coverage for some proteins (e.g., trastuzumab heavy chain without alternative enzymes). - Very low pH (<4) inhibits acceleration; optimal performance depends on buffer and ion mode (NH₄HCO₃ and negative mode often superior). - While digestion is ultrafast, overall analytical throughput and utility depend on integration with other sample preparation steps; practical performance may be case dependent. - The precise mechanistic basis of acceleration remains incompletely established.