As biological assays increase in number and shrink in size, sample preparation takes on greater significance for getting the most from hard-to-get samples. In particular, sample prep for proteomic mass spectrometry (MS) workflows must consider not only the sample—its availability, matrix, homogeneity, and stability—but also compatibility with the separation method (e.g. liquid chromatography or capillary electrophoresis) and with the mass spec instrument itself.
MS has come to dominate proteomic analysis through its identification and characterization of peptides, amino acids, and their variants with great precision and at high sensitivity. Yet proteomic-MS workflows are incredibly diverse in terms of inputs and expected outputs. Samples vary in abundance, physical form, and homogeneity. While this is not unusual in biology, the sample and prep milieu, or matrix, must be compatible with MS ionization and detection, so "dilute and shoot" is usually not an option. These uncertainties moreover raise issues of method robustness, readout consistency, and quantitative measurement accuracy.
In this article we talk with two experts in proteomic MS: Ryan D. Bomgarden, senior R&D manager, and Sergei Snovida, an MS research scientist, both at Thermo Fisher Scientific. Since an exhaustive guide to proteomic-MS sample prep would take up many volumes, we hope the tips our experts provide will raise awareness of the main issues and set users on the path toward answering critical questions on improving MS-proteomic workflows.
Standardize
Standardizing analytical workflows can take on many forms, beginning with the assay protocols themselves. "Sample preparation for proteomic-MS workflows has traditionally been homebrew," Bomgarden says. Analysts borrow a bit from their knowledge of MS, from experiences with similar sample types, and connect the dots from published protocols and their own experience. "Many of these preps work okay; but they have not been optimized for one particular sample type, or for that matter, standardized such that protocols can be shared and relied upon to provide comparable results in different hands. This has led to demand from researchers for commercially available kits."
Standardization, or lack thereof, has always been a problem in science and is a contributor to the reproducibility crisis in biomedical research. Core labs typically require that scientists submitting samples for MS analysis follow standardized sample prep protocols, which increasingly means commercial kits with optimized, preformulated reagents for protein reduction, alkylation, digestion, and clean up.
Automate
While commercial kits help to achieve consistency at the assay level, they do not address the all-too-common situations of samples that are unstable, heterogeneous, or rare. Kits ensure that every sample experiences the same preparative and assay conditions, but they cannot increase the amount of sample available or the need to stretch it over dozens or hundreds of assays per day. Not surprisingly, as the number of samples increases the physical volumes of individual assays shrink, requiring ever more-precise delivery of vanishingly small volumes of sample and reagent.
"For very small sample quantities—down to individual cells—or for large numbers of samples, automation is the only path to consistency," Bomgarden says.
Emphasizing consistency rather than throughput aligns with the U.S. Food and Drug Administration’s initiatives on quality and risk, but the time savings are nothing to sneeze at either. According to Bomgarden, a typical proteomic sample preparation can be reduced from two or three days to a morning or less, depending on the sample type and level of automation.
Watch your interferences
Omics may be viewed as different ways to study a biological event though global expression of genes, proteins, metabolites, lipids, etc. Tandem analyses, for example proteomics-genomics or proteomics-metabolomics on the same sample, are increasingly popular. A single prep protocol suited to both analysis modes is preferred in these situations, but its realization can be elusive. A unified protocol must account for relative abundances, concentration dynamic range, and interferences—reagents that facilitate one mode but inhibit the other.
"DNA preparations for genomics studies typically include reagents for cell lysis or protein solubilization, which interfere with proteomic MS sample preparation or analysis," says Snovida. "Home-brewed preps are particularly troublesome because they may contain detergents or non-volatile salts. Nonionic detergents used for lysis, for example Triton X-100 and NP-40, are difficult to remove. Phosphate or Tris buffers can be removed but they are not volatile, so they are incompatible with MS. That is why we recommend volatile salts and acids like ammonium bicarbonate or formic acid for MS sample preparation instead of sodium- or phosphate-based buffers."
Watch sample homogeneity
Many proteomic samples, particularly those acquired through biopsy or surgery, are limited in terms of accessibility and quantity. Tumors are notoriously heterogeneous as well, and most resections include a margin of normal cells. Simply cutting a tissue sample in half does not guarantee that each portion will contain the same proteomic analytes.
"Researchers are concerned about sample bias, yet it's common practice to split a sample in two and use one part for proteomics and another for genomics," Bomgarden tells Biocompare.
At the other end of the size spectrum are tiny samples. "Each manipulation step involves losses, which become more critical as samples approach limits of detection and the total available protein becomes vanishingly small, down to the contents of single cells. Remember that unlike DNA or RNA we can't amplify target proteins to compensate for small sample sizes or sample loss." Measurement of trace analytes is also difficult as interferences from high-abundance proteins, for example albumin in blood, may impede detection by MS.
Optimizing proteomic MS sample preparation can pay big dividends in terms of throughput, consistency, and result quality; but these efforts come at the cost of flexibility. Analytical scientists are curious by nature and enjoy experimentation. Adopting standard reagent kits, protocols, and methods developed by product development scientists, and generally following a recipe may appear to diminish their agency or creativity. In fact, optimizing proteomic sample prep does exactly the opposite by freeing them for higher-order problem-solving.