Session MOG. There are 4 abstracts in this session.


Microscaled methods for proteogenomic analysis of patient-derived tumors

Steve Carr

Current gene-expression-based approaches for cancers are largely for prognostication, not prediction of individual drug responses. A major obstacle to progress in NGS-based diagnostics is a fundamental one: we poorly understand how complex cancer somatic genomes drive clinical phenotypes and drug vulnerabilities. Key issues such as therapeutic resistance, the contribution of the tumor microenvironment and the metastatic process belie single gene/mutation explanations. The new field of proteogenomics provides an opportunity to generate new insights by melding the complexity of cancer genomics with cancer proteomics to more completely understand how somatic genomes activate aberrant signal transduction events that drive cancer pathogenesis. Our group tackles this with both broad discovery efforts and targeted assays in the context of preclinical and clinical core biopsies, well-annotated cohorts and clinical trial samples in an iterative design to identify predictive tests that can be feasibly developed into clinical tests. We postulate that this microscaled and integrated approach will produce sounder therapeutic hypotheses and a new generation of accurate predictive biomarkers. Here we will present on novel protocols for simultaneous extraction of DNA, RNA and protein from single core biopsies and microscaled proteomics approaches that enable multiplexed, quantitative and deep-scale profiling of the extracted protein from the core biopsy tissues at proteome and phosphoproteome levels. We will also describe our recent advances in multiplexed, deep-scale analysis of the acetylome and ubiquitylome of tissue samples and the utility of FAIMS for improving quantitative accuracy.
Tips and Tricks:


Sample Processing and Analysis Platform for In-Depth Single-Cell Proteomics

Ryan Kelly

Biological tissues are highly heterogeneous, consisting of a variety of cell types, states and subpopulations, and understanding heterogeneity at the single cell level is of great interest for biomedical research. We have developed a microfluidic platform to minimize sample losses normally incurred during sample processing which, in combination ultra-low-flow nanoLC separations and latest-generation MS, achieves in-depth proteome coverage for samples comprising few or single cells. The sample processing platform, termed nanoPOTS (Nanodroplet Processing in One pot for Trace Samples), utilizes a robotic platform to dispense nanoliter volumes of reagents into photolithographically patterned nanowell reaction vessels, greatly reducing adsorptive losses. NanoPOTS-prepared samples are then analyzed using low-flow nanoLC-MS. To maximize sensitivity, in-house-packed nanoLC columns having internal diameters of 20–30 µm are employed. To date, we have identified >1000 proteins from single mammalian cells. Single cells and other small samples are readily isolated into nanowells using widely available fluorescence-activated cell sorting (FACS) and laser capture dissection (LCM). We have applied the platform to profiling proteins in tissue subsections with high spatial resolution (≤100 µm) and high depth of coverage (up to ~2600 protein groups) including profiling pancreatic islets from type 1 diabetic and nondiabetic donors, brain, liver, uterine and plant tissues. The ability to profile single cells and map the proteome with high spatial resolution across tissue regions provides a fundamental way to understand the tissue microenvironment, substructure, and cellular organization from a global proteome perspective. We will discuss recent advances in sample processing, separations and mass spectrometry, as well as prospects for further improving the platform for single-cell and other low-input proteomics studies.
Tips and Tricks:


Single Cell Proteomics and the Carrier Proteome Effect

Atticus McCoy; Christopher M. Rose
Genentech, South San Francisco, CA

Single cell proteomics (SCP) by mass spectrometry (SCoPE-MS) is a recently introduced method that utilizes isobaric labels to quantify multiplexed single cell proteomes.  While this technique has generated great excitement, the enabling technologies underlying SCoPE-MS - isobaric labels and mass spectrometry - comprise technical limitations with the potential to unfavorably impact data quality and biological interpretation if not considered properly. Here, we provide a detailed characterization of the relationship between the level of carrier proteome and quantitative accuracy of SCoPE-MS.

Through the analysis of a sample containing four aliquots each of HeLa and K562 cells and various levels of carrier proteome we find that the relationship between the number of ions sampled and the quantitative accuracy of the measurement is dependent on the level of the carrier proteome.  Low levels of the carrier proteome (e.g., 5x) display accurate quantification (≤20% CV) even when a small number of ions are sampled, while high carrier proteome levels (e.g., 400x) require many more ions to be sampled for accurate quantification.  By diluting our samples to levels at or below the protein amount found in single cells, we also demonstrate that common mass spectrometer settings tend to under sample ions for SCP and lead to inaccurate data when utilizing high levels of carrier proteome.

Taken together, these data demonstrate that increasing the level of carrier proteome requires a concomitant increase in the number of ions sampled in order to maintain quantitative accuracy within SCoPE-MS experiments – we term this the “carrier proteome effect”. This observation has implications for SCP experiments where high levels of carrier proteome could lead to insufficient sampling of ions and inaccurate quantification.  This is particularly troublesome for single cell proteomics analyses where variability in single cell proteomes may falsely be attributed to cell heterogeneity, rather than poor measurements.

Tips and Tricks:


Multidimensional cross-linking with a tetra-reactive cross-linker

Jared Mohr; Juan Chavez; James E. Bruce
University of Washington, Seattle, Washington

Cross-linking mass spectrometry (XL-MS) is a powerful and increasingly popular tool for the discovery and characterization of protein-protein interactions. Cross-linker molecular design is not inherently limited to two reactive groups and additional groups can yield multidimensional distance constraints and enable the unambiguous identification of protein complexes with more than two proteins.

We will present “Bisby”, a new CID-labile cross-linker with four amine-reactive n-hydroxyphthalamide (NHP) functionalities, as well as a method for enriching and identifying tetralinked species. Bisby can link up to four proximal lysine residues, providing enhanced distance constraints and greater insight into protein complexes compared to a bifunctional linker. A novel instrument control method was written in ion trap control language to efficiently identify a set of four peaks in MS2 spectra that fulfill an expected mass relationship.  Each released peptide was then targeted with MS3 analysis to produce fragment data. Tetralinking experiments with purified histones yield numerous fully identified 4 peptide cross-links. These four peptide cross-links offer enhanced structural information compared to their six constitutive binary cross-links as they are generated from a single molecule at one point in time, guaranteeing six distance constraints compatible with at least a single conformer or complex. Tetralinks between H1 and up to 3 core nucleosome proteins allow for tightly constrained docking from only a single tetralink. Beyond these fully identified multidimensional links, there are more than 100 additional cross-links with 2 or 3 peptides identified, showing that a tetralinker provides structural data despite hydrolysis or unidentified spectra. Preliminary tests of whole proteome tetralinking of bacterial samples yields identified multi-dimensional links in abundant proteins, indicating application in live cells is feasible for at least some set of proteins. Live cell applications are particularly appealing due to the ability to unambiguously characterize novel complex interfaces of three or more proteins without additional information.

Tips and Tricks: