Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

4.1 MSE and Other Data-Independent Acquisition Strategies

Shotgun proteomics is a strategy with broad applicability, and is based on digestion of proteins with proteolytic enzymes and analysis of the resulting peptides by liquid chromatography–mass spectrometry (LC/MS). The method of choice for data collection has long been a data-dependent acquisition (DDA) method in which acquisition parameters are modified in real time by selecting a narrow mass window in a quadrupole analyzer to allow precursor ions to pass through for fragmentation after an initial survey scan [1]. This was an important enhancement to the initial approach of using stored mass and retention time information from previous liquid chromatography runs [2]. More recently, a strategy has been proposed [3, 4] and finally implemented [5, 6] using data-independent acquisition (DIA) in a method called MSE to perform data collection in a massively parallel, time-based schedule rather than attempting to sample masses sequentially. MSE has been extensively developed as a method in which alternating low and high energy scans are recorded across the full mass range and was implemented on quadrupole time-of-flight mass spectrometers from Waters Corporation [5, 6]. The original approach has been very successful and has stimulated the development of variants modeled after the MSE protocol. These include sequential proteomic method precursor acquisition independent from ion count (PAcIFIC) [7], windowed data-independent acquisition of total high-resolution (SWATH) [8], all-ion fragmentation (AIF) [9], and multiplexed data-independent acquisition (MSX) [10]. These recent adaptations can be utilized on other instrument platforms, although only MSE is unique in that it covers the entire mass range in each scan. Our group has used MSE effectively on a routine basis for cells, tissues, and affinity preparations in a label-free approach to large-scale protein profiling [11–15]. The validity, power, and effectiveness of MSE are now firmly established and extensively validated independently by many groups [16–18]. One example is our work on human somatic stem cells, where we were able to use this technique to demonstrate clear differential expression of proteins such as aldose reductase and many other proteins (Fig. 4.1) which responded to the combination of growth factor priming and increased osmotic pressure [13]. The roles of some of these proteins are characteristic of the physiological states studied [13]. In that experiment, 5′-nucleotidase and transgelin were detected as differentially expressed, and have been previously linked to cell differentiation state. Data-independent label-free profiling was demonstrated in that work to be a useful tool in characterizing cellular responses to treatment regimes, and as an aid to optimization of cell priming protocols for cartilage tissue engineering.

Fig. 4.1
figure 1

Example of expression data for selected proteins in agglomerative hierarchical cluster of Z-score transformed intensity data as processed by the Rosetta Elucidator program. These proteins have at least 1.5-fold response to the treatment, and are represented by at least two peptides for both identification and quantification at ratio P-values as calculated by Elucidator. Mean relative abundance of each protein in control and treated as well as abundance ratio is listed. Cluster coloration indicates protein abundance in the sample (red indicates higher abundance; green indicates lower abundance and black equal abundance). Reprinted with permission from Oswald et al. [13]. Copyright (2011) American Chemical Society

Full size image

4.2 DIA Strategies Enhanced by Accurate Mass and Retention Time Matching Across Multiple Chromatographic Runs

Our work on adipose-derived stem cells [13] uses the Elucidator Protein Expression Data Analysis System from Rosetta (and Ceiba Solutions), a commercial program to match accurate mass and retention time intensities of precursor ions across multiple chromatographic runs. This enabled measurement of the intensities of precursor peptides even if fragmentation is not particularly successful for a particular peptide in a particular chromatographic run. This approach to a large extent counters the limitation imposed by the stochastic nature of mass spectrometry that otherwise results in a large amounts of missing data in large-scale experiments. A similar strategy is used by the TransOmics Informatics for Progenesis QI (TOIP) from Waters Corporation [19], a program also specifically enhanced to take advantage of ion mobility separations (see below). Coupling mass and retention time matching programs such as these with DIA technologies including MSE provides a particularly powerful label-free platform that enables large and complex experiments otherwise difficult to conduct by other methods.

These software approaches are enabled by new faster instruments capable of data-independent scanning, but owe their conceptual inspiration to the concept of an accurate mass tags database originally conceived by Smith’s group [20].

4.3 Enhancement of MSE: Coupling to Traveling Wave ion Mobility Spectrometry (TWIMS)

A variation of MSE coupled with traveling wave ion mobility spectrometry (TWIMS) provides the capability to perform separations of peptides based on cross-sectional area and shape in addition to their mass/charge (m/z) ratio. The use of TWIMS for shotgun proteomics is now well established and has been validated independently by a number of groups [21–23]. A two-dimensional separation is achieved in the gas phase, providing considerably increased protein coverage, typically a doubling of the number of proteins detected, a significant advantage [12, 24]. The complex TWIMS data requires a powerful computing platform, typically with a graphics processing unit (GPU)-equipped computer with 448 cores or more [12, 24]. The TWIMS principle is illustrated in Fig. 4.2 where integrating ion mobility drift time enhances identification of peptides [25]. In a study of a bacteriophage virion proteome [26], MSE with TWIMS was consistently more effective than more conventionally employed DDA method. Current versions of the ProteinLynx Global Server commercial program [5, 6] (Waters Corp.) have been enabled to process TWIMS data.

Fig. 4.2
figure 2

Diagram illustrating the role of ion mobility spectrometry in addressing the challenge of peptides with both the same elution time in liquid chromatography (co-eluting) and having the same mass (isobaric). Without ion mobility, fragmentation of isobaric peptides in the collision cell results in a chimeric spectrum which is difficult to interpret. When ion mobility is activated, peptides are separated in the mobility cell on the basis of their drift time. The end result is cleaner spectra with reduced chimeracy [26]. Reprinted from Journal of Virological Methods, Vol 195, Moran, Deborah; Cross, Trevor; Brown, Lewis M.; Colligan, Ryan M; Dunbar, David, Data-independent acquisition (MSE) with ion mobility provides a systematic method for analysis of a bacteriophage structural proteome, pp. 9–17, 2014, with permission from Elsevier

Full size image

We have applied these techniques in many other systems [15, 26] including patient samples [27]. Such analyses include fold-change and p-value determinations, providing an unbiased view of phenotype or biological responses to experimental treatments. An example is shown in Fig. 4.3 (mouse hippocampus) [12]. This figure represents fragmentation spectra collected with and without TWIMS activated. It can be clearly seen from this example that the interfering peaks from other peptides (derived from other proteins, magenta) and unassigned peaks (grey) are largely separated out by TWIMS. These sorts of comparisons can be made in data where spectra with or without TWIMS can be readily compared for the same peptide.

Fig. 4.3
figure 3

Fragment ion spectra of peptide ADQLADESLESTR with TWIMS off and TWIMS on. Purple and grey peaks represent interferences and noise, respectively [12]

Full size image

With this technique it is possible to quantify large numbers of proteins and generate abundances (examples in Table 4.1) for each LC/MS/MS run and for each biological replicate, and calculate means and standard deviations with coefficient of variation as low as 10–20 % for replicate analyses [12]. Large-scale experiments can be done on a routine basis with recording of many thousands of protein abundance values for multiple biological and technical replicates.

Table 4.1 Example protein quantitation data from WT mouse hippocampus recorded using TWIMS and elucidator post-processing [12]
Full size table

MSE combined with TWIMS is being applied to a remarkable variety of practical biological problems including the interactome of the RNA-binding protein RALY [28], analysis of chaperonins and biosynthetic enzymes in the unculturable bacterial endosymbiont Blochmannia, [29] and quantitative analysis of human embryonic kidney cells proteome following sialic acid overproduction [30].

In another practical example of this technique, Drosophila melanogaster was evaluated for the role of dMyc in the larval fat body (Table 4.2). The fat body is a tissue that functions as a sensor of circulating nutrients to control the release of Drosophila insulin-like peptides (Dilps) influencing growth and development. Using MSE and TWIMS, it was demonstrated that dMyc affected expression of hexokinase C and pyruvate kinase, key regulators of glycolysis, as well as of stearoyl-CoA desaturase (Desat1). Desat1 is an enzyme that is necessary for monosaturation and production of fatty acids, and its reduction affects dMyc and the ability to induce fat storage and resistance to animal survival.

Table 4.2 Quantitation of some proteins in fat bodies from cg-dMyc and cg-control D. melanogaster larvae
Full size table

These techniques can also be applied very effectively to chemoproteomic affinity experiments. For example, we have identified a protein target of a small molecule (RSL3) that is an inducer of a novel form of cell death (ferroptosis) and a potential novel anticancer agent [24]. In that work we also used DIA (MSE) with TWIMS. A total of 1,353 proteins were detected in 27 HDMSE chromatograms. Analysis focused on 979 proteins with identification and quantitation supported by two or more unique peptides. Most detected proteins were not differentially bound to fluorescein-RSL3 beads compared to controls. Only one protein, glutathione peroxidase 4 (GPX4_HUMAN), was significantly enriched (P < 0.01) and had the highest fold-change (26-fold in the affinity preparation compared to the inactive analogue and 13-fold compared to the control of preincubation with free RSL3). These studies identified GPX4 as an essential regulator of ferroptotic cancer cell death using data-independent profiling with TWIMS and these results were confirmed by an extensive series of corroborative experiments [24]. Isotopic clusters of an example peptide from GPX4 are given in Fig. 4.4

Fig. 4.4
figure 4

3D visualization of an isotopic cluster of an example peptide ILAFPCNQFGK from GPX4 from an affinity preparation and two controls. Cell lysates were prepared from cells treated with active probe, inactive probe, or active probe in the presence of competitor. Reprinted from Cell Vol. 156, Yang, Wan S.; SriRamaratnam, R.; Welsch, Matthew E.; Shimada, K.; Skouta, R.; Viswanathan, Vasanthi S.; Cheah, Jaime H.; Clemons, Paul A.; Shamji, Alykhan F.; Clish, Clary B.; Brown, Lewis M.; Girotti, Albert W.; Cornish, Virginia W.; Schreiber, Stuart L.; Stockwell, Brent R., Regulation of Ferroptotic Cancer Cell Death by GPX4, pp., 317–331, 2014, with permission from Elsevier

Full size image

4.4 Conclusions

In mass spectrometry-based proteomics, there has been an evolution from initial pioneering data-dependent approaches where slower instruments could not collect enough precursor intensity data to allow robust label-free quantitation. The massively parallel MSE and other DIA methodologies have enabled label-free quantitation and have been applied to a wide variety of sample types from cells to tissues, and have been applied to viral, microbial, plant, animal, and patient samples. This evolution has continued with the introduction of software to match accurate mass and retention time data in large experiments, increasing sophistication of ultrapressure liquid chromatography separations and finally through orthogonal separation with technologies such as TWIMS that significantly increase the number of proteins detected in an experiment. This pipeline provides exceptional flexibility for large and complex experiments and complements isotopic labeling approaches described in other chapters of this volume.