Methods of Clinical Sample Preparation
For clinical tissue studies, proper preservation is critical to ensure the amount of protein from surgical excision to proteolytic digestion. Several methods are available: fresh frozen (FF), formalin-fixed paraffin-embedded (FFPE), and OCT-embedded.FF allows for the detection of more proteins compared to FFPE, but available FFPE has been stored for several years or even a decade, and is an important source of samples for retrospective studies, such as clinical follow-up. While tissue proteomics studies can explore information on biological mechanisms, clinical proteomics studies have the primary goal of discovering new biomarkers, and therefore "body fluid" samples such as blood (serum, plasma), urine, saliva, tear fluid, and cerebrospinal fluid are ideal sample formats, and are also used for the detection of cancers and the development of therapeutic response in longitudinal studies. longitudinal studies. When clinical samples are not of sufficient quality to support the study, model systems such as transgenic animal models, cancer cell lines, xenograft models (CDX, PDX), organoids, etc., may be considered.
There is no uniform protocol for proteomic sample preparation, and it is important to select the appropriate method and optimize it based on sample complexity, sample volume, and research objectives. The main methods for preparation include FASP, MStern, S-trap, SP3 and iST. FASP is introduced here as an example.FASP, the filter-assisted sample preparation method, starts with solubilization of proteins using the anionic surfactant sodium dodecyl sulfate (SDS), followed by molecular weight (MW) filtration to bind proteins onto nitrocellulose filters while lower MW material is filtered out.Successive urea washes help to better remove SDS, and lastly proteolytic digestion on the filter and elution to obtain the peptide product.
MS Detection Principle and Procedure
In order to increase the proteome coverage in the assay, the peptide samples are first separated into different fractions by, for example, reversed-phase liquid chromatography, and then enter the MS analysis. The peptides are ionized using soft ionization techniques (ESI or MALDI), and the atomized peptides can be further separated by ion mobility, thus reducing the complexity of primary mass spectrometry (MS1) and contamination of secondary mass spectrometry (MS2) and ultimately achieving greater proteome coverage. Such techniques include ion mobility (TIMS) and high intensity field ion mobility spectroscopy and (FAIMS). In terms of the choice of mass spectrometry scanning modes, the traditional Data Dependent Acquisition (DDA) mode is well established in proteomics research and is compatible with label-based quantification techniques. In DDA, only the top n parent ions with the strongest signals in the MS1 scan results are selected and subjected to sequential fragmentation and MS2 detection. However, this mode has poor assay reproducibility and suffers from the problem of high abundance peptides in MS1 affecting the detection of low abundance peptides. Due to the shortcomings of DDA, proteomics studies are beginning to favor the use of the data-independent acquisition (DIA) mode. In this mode, sequential fragmentation of all parent ions in multiple small-scale mass-to-charge ratio windows produces more complex MS2 results. These results are then matched against a pre-defined library of spectra to achieve maximum proteome depth through extensive peptide hierarchies.
Protein Quantification Methods
Protein quantification techniques are diverse and can be categorized into targeted and non-targeted techniques according to the scope of the assay, and into relative or absolute quantification techniques according to the method of quantification. Among them, relative quantitative techniques can be divided into labeling techniques (TMT and iTRAQ) and non-labeling techniques (label-free, DIA). TMT labeling in labeled relative quantification techniques can increase the sample throughput to 16. However, TMT methods require multiple levels of peptide grading to obtain in-depth proteomic profiles, and 1-2 TMT channels are commonly used to assay a mix of all samples to minimize batch-to-batch variation, which reduces the ability to make valid comparisons between individual programs and increases assay costs. The label-free technique, on the other hand, thanks to the development of data analysis software, allows the relative abundance of proteins to be calculated from the peptide ion peak fraction of MS1. Compared with labeling techniques, non-labeling techniques have a wider dynamic range but can be slightly less accurate. Therefore, for clinical samples with large inter- and intra-patient protein expression differences, label-free quantification techniques are more suitable for identifying more differentially expressed proteins.
Target proteins screened by non-targeted relative quantitative detection techniques require expression validation, such as antibody-based ELISA and MS-based targeted analysis techniques. Among them, two MS-based targeted quantitative techniques are multiple reaction detection (MRM) and parallel reaction detection (PRM). MRM uses a triple quadrupole mass spectrometer for analysis, which needs to determine the mass-to-charge ratios of the target parent ions and fragment ions, and the combinations of the parent ions and 3-5 relevant fragment ions are selected by the quadrupole and quantitatively analyzed. PRM, on the other hand, utilizes high-resolution mass spectrometry to improve specificity. all fragment ions in PRM are generated and recorded during the analysis, so only the mass-to-charge ratio of the target parent ion needs to be determined and the best fragment ions are selected directly from the secondary mass spectrum for quantitative analysis. If peptide standards labeled with stable isotopes are added as a control, both targeting techniques can reach an absolute level of quantification. PRM reliably monitors more targets than either technique.
Directions for clinical proteomics
In oncology research, tissue analysis can most accurately reflect the physiological state of the tumor, discover biomarkers, biological pathways, and integrate with existing genomics and transcriptomics results for multi-omics analysis. These studies typically use cancer tissue samples from the same patient and "healthy" control samples next to the cancer to compare potential diagnostic biomarkers, as well as comparing patients with different cancer stages to obtain prognostic information. Once a small number of candidate proteins have been identified, pathway analyses can be used to gain insights into how these proteins are associated with tumorigenesis, proliferation, metastasis, and other cancer-driven processes, which can be supplemented with validation experiments on differentially expressed proteins in large independent cohorts. Summarizing the current state of scientific research, cancer proteomics is mainly directed towards finding markers for risk prediction, cancer grading and prognosis, identifying effective therapeutic targets and post-translational modifications such as phosphorylation, acetylation and glycosylation. In addition, the problem of tumor heterogeneity demands protein research at the single-cell level. Mass spectrometry-based mass spectrometry flow-through technology allows monitoring of dozens of protein markers in a single cell, where antibody probes and unique heavy metal isotopes are ligated together and incubated with the cells, which are then nebulized by inductively coupled plasma (ICP), and the metal ions provide the mass spectrometer with a quantitative readout of the target protein in the sample.
In the article "Integrated Proteogenomic Characterization of HBV-Related Hepatocellular Carcinoma" published in Cell in October 2019
Research Prospects for Clinical Proteomics
As standardized, high-throughput proteomics technologies continue to evolve, clinical studies will progress toward larger cohorts, which will make proteomics findings more statistically significant and improve the efficiency of clinical translation of protein markers and drug targets. On the other hand, proteomics will become an important component of cancer systems biology by integrating multi-omics data from genomics, epigenomics, transcriptomics and post-translational modification genomics.
References
Macklin, Andrew et al. "Recent advances in mass spectrometry based clinical proteomics: applications to cancer research." Clinical proteomics vol. 17 17. 24 May. 2020.
Zhang, Yaoyang et al. "Protein analysis by shotgun/bottom- up proteomics." Chemical reviews vol. 113,4 (2013): 2343-94.
Gao, Qiang et al. "Integrated Proteogenomic Characterization of HBV-Related Hepatocellular Carcinoma." Cell vol. 179,2 (2019): 561-577.e22.