Turbodigestion for Top-down Proteomics

By Gordon W. Slysz and David C. Schriemer, PhD

Proteomics, the global characterization of proteins in a complex mixture, has advanced to the point where hundreds of proteins can be identified from a single experiment. This may seem like a large number, but it represents only a small fraction of the proteins in biologically relevant samples. Proteomics focuses primarily on identification, but this is just one facet of protein characterization. For example, proteins undergo post-translational modifications, many of which are reversible and change within seconds or minutes. Determining the location and nature of the structural modifications is critical to understanding the function of many proteins.

Current methods in proteomics have great difficulty in the analysis of low-abundance proteins, and in resolving multiple forms, especially in the context of clinically relevant samples, which are often loaded with contaminants.

New systems are needed that can improve sensitivity and provide solutions to the in-depth characterization of proteins, all within the murky environment of biological samples.

Top-down or Bottom-up? A False Dilemma!

Protein mixture analysis can be accomplished from one of two directions: top-down or bottom-up. Top-down methods focus on the intact protein whereas bottom-up methods concentrate on the peptides that are usually produced from enzymatic digestion. Top-down methods, which include gel electrophoresis and liquid-based protein chromatography, separate proteins according to their molecular weight, isoelectric point, hydrophobicity, and/or binding affinity. These methods can support the resolution of protein states, such as the determination of phosphorylated and unphosphorylated forms of a protein.

In bottom-up methods, the entire batch of proteins is digested to produce a very complex mixture of peptides. This is analogous to jigsaw puzzles, with a given puzzle representing one protein and its pieces representing the peptide components. In bottom-up methods, all the pieces of all the puzzles are mixed together. The protein (or puzzle) is identified by selecting and sequencing the individual peptides (or pieces). More peptides sequenced can give greater confidence that the correct protein is identified.

Identification is the strength of bottom-up methods, but the weakness is full protein characterization. Two similar proteins (i.e., phosphorylated and unphosphorylated) would have many of the same peptides. In bottom-up methods, high sequence coverage is rare, and characterizing individual peptide modifications would be hit-and-miss depending on its selection from the massive pool.

Top-down analysis is capable of differentiating between two similar proteins as they are first separated and then analysed. Unfortunately, the proteins cannot be identified without fragmentation either in solution (enzymatic digestion) or the gas phase. Newer techniques that allow gas-phase protein dissociation in the mass spectrometer support sequencing, but full sequence coverage is not yet possible for most proteins. The best route to fully characterize a protein involves protein separation and a multi-pronged digestion/sequencing scheme — a blend of top-down and bottom-up.

The problem is, a sensitive and robust means of integrating bottom- up with top-down has yet to emerge. Protein mixture separation requires a high-resolution method, which means that proteins are present in very small volumes. Converting the fractionated proteins to peptides involves several sample handling and processing steps. In most protocols, proteins require denaturation, reduction, alkylation, digestion and cleanup. Protein loss at any sample handling step can have a dramatic negative effect on efficiency of protein characterization.

Our lab is working on methods to incorporate protein tryptic digestion within microfluidic systems, to allow “turn-key” integration of top-down and bottom-up methodologies for protein characterization. We believe the key to getting there is to improve the efficiency of tryptic digestion.

Need for Speed

Improving digestion performance can be achieved in two ways. One approach is to increase substrate turnover by using a large excess of enzyme. However, increasing the in-solution concentration of the digestive enzyme is impractical since enzyme autolysis produces large amounts of contaminant peptides. Immobilizing the enzyme on a beaded support can overcome this limitation, although non-specific binding to the support has limited its usefulness. Another approach for increasing substrate turnover is to unfold the protein sample, supporting greater access to all possible cleavage sites. Unfortunately, this can add significant contamination to the sample (e.g., introduces surfactant, chaotropes like urea, cysteine alkylation reagents, etc.).

Our lab has developed a unique digestion approach that we refer to as “turbodigestion.” Turbodigestion uses immobilized trypsin in a microreactor format in the presence of aqueous-organic solvents. This system delivers high trypsin concentration, protein denaturing conditions, and increased enzyme efficiency in a package compatible with both protein separation technologies and mass spectrometry. Protein solutions are mixed with organic solvents and simply infused through the digestion module directly into the mass spectrometer. Turbodigestion has proven to be very fast and efficient. For example, excellent sequence coverage is possible for the digestion of transferrin, a protein normally resistant to tryptic digestion, without necessitating reduction and alkylation steps (Fig. 1). Proteins at low nM concentrations are also well-digested, illustrating the benefits of improved trypsin kinetics in mixed solvents. Protein solutions can be fed through the microreactor at relatively high rates (up to 7,600 cm/hr), corresponding to digestion times of less than a second.

The use of aqueous-organic solvents for the improvement of digestion may seem counter-intuitive since organic solvents are most often detrimental to normal enzyme function, usually causing the protein to unfold. However, trypsin — especially when immobilized — performs well in a range of organic solvents in which the target proteins preferentially unfold, allowing easier access to lysine and arginine cleavage sites. Because organic solvents such as acetonitrile, methanol and isopropanol induce different degrees of protein unfolding, digestion can be tuned to produce a range of peptide maps, which can in turn be altered by flow rate.

Bridging the Gap

The convenience of turbodigestion supports its insertion into liquid-based protein separation systems, effectively bridging the gap between top-down and bottom-up analysis. Instead of overnight incubation, protein can elute from a separation module directly through the digestion module and into the mass spectrometer. The nanoscale size of the digestion column, combined with high-efficiency digestion, prevents the loss of chromatographic integrity. Proteins are separated using a top-down approach but examined in the mass spectrometer using a bottom-up readout.

There are two caveats particular to turbodigestion that needed to be addressed. Solutions entering the turbodigestion module must have a neutral pH, and the organic solvent concentration should remain constant. This is problematic with many forms of upstream protein separations. Reversed-phase protein separation, a common liquid-based protein separation method, is performed under acidic conditions and features a gradient of organic solvent. Microfluidic solution adjustments must therefore be made to bring the pH to neutral levels and maintain the organic solvent at a steady level. At low flow rates, this is no easy task. Fortunately, advances in microfluidic technology in the past few years have made it possible to accurately maintain very steady flow rates as low as 20 nl/min, enabling the fine solution adjustments necessary for optimum digestion.

The protein mixture analysis system is diagrammed in Figure 2A. Proteins are loaded onto a reversed-phase column and eluted using a gradient of acetonitrile under acidic conditions. Effluent from the reversed-phase column is mixed with a mirror gradient, resulting in a steady composition of neutralized acetonitrile/water. Separated proteins are eluted through the turbodigestion module, and analysed by electrospray mass spectrometry at a minimum dilution.

Protein Mixture Analysis in 30 Minutes or Less

The system is capable of separation, digestion and characterization of proteins in a mixture in less than 30 minutes. It should be emphasized that proteins are separated while they are intact, but the separation is monitored in the mass spectrometer through the peptides. Integrating over each protein peak produces a peptide mass map. Figure 2B shows the benefit of this method in observing a ubiquitinated protein. Two peaks/maps are observed for superoxide dismutase (SOD). One peak co-elutes with ubiquitin, while a second peak elutes later, suggesting that superoxide exists in both ubiquitinated and unmodified forms, and free ubiquitin is present in the sample. This is one simple example of the characterization power of a blended approach.

The turbodigestion-based protein mixture analysis is fast, but is it sensitive? Coomassie-blue staining procedures for gel electrophoresis protein detection are limited to 50 ng of protein, while silver or SYPRO Ruby staining can detect as little as 1 ng. Detection is one thing, but identification is another, and the limits of identification usually lie higher than that of detection. Turbodigestion studies with test proteins show excellent sequence coverages, with identifications of the proteins made at amounts as low as 25 fmol (< 1 ng).

Micro Total Analysis Systems

In the chromatographic world of protein separations, smaller is better. The turbodigestion-based protein analysis system marries well the trend towards micro total analysis systems (µTAS). All stages of the protein analysis are interconnected and can be easily automated, from sample injection to data analysis and protein identification. New microchip-based systems for flow path integration will further reduce analysis times and provide additional gains in sensitivity and characterization power.

David Schriemer, PhD is an assistant professor in the department of biochemistry and molecular biology, University of Calgary (Calgary, AB) and director of the Southern Alberta Mass Spectrometry Centre (SAMS) for Proteomics. Gordon Slysz is a PhD candidate in the Schriemer lab.