Insights and prospects for ion mobility-mass spectrometry in clinical chemistry

Abstract

Introduction:

Ion mobility-mass spectrometry is an emerging technology in the clinical setting for high throughput and high confidence molecular characterization from complex biological samples. Ion mobility spectrometry can provide isomer separations on the basis of molecular structure, the ability of which is increasing through technological developments that afford enhanced resolving power. Integrating multiple separation dimensions, such as liquid chromatography-ion mobility-mass spectrometry (LC-IM-MS) provide dramatic enhancements in the mitigation of molecular interferences for high accuracy clinical measurements.

Areas covered:

Multidimensional separations with LC-IM-MS provide better selectivity and sensitivity in molecular analysis. Mass spectrometry imaging of tissues to inform spatial molecular distribution is improved by complementary ion mobility analyses. Biomarker identification in surgical environments is enhanced by intraoperative biochemical analysis with mass spectrometry and holds promise for integration with ion mobility spectrometry. New prospects in high resolving power ion mobility are enhancing analysis capabilities, such as distinguishing isomeric compounds.

Expert Opinion:

Ion mobility-mass spectrometry holds many prospects for the field of isomer identification, molecular imaging, and intraoperative tumor margin delineation in clinical settings. These advantages are afforded while maintaining fast analysis times and subsequently high throughput. High resolving power ion mobility will enhance these advantages further, in particular for analyses requiring high confidence isobaric selectivity and detection.

1. Introduction

Since at least the 1970s, small molecule identification in clinical analyses has been largely dominated by spectroscopic techniques such as Fourier transform infrared spectroscopy, UV-visible absorption spectroscopy, fluorescence spectroscopy, and Raman spectroscopy [1–4]. More recently, mass spectrometry (MS) is increasingly implemented in clinical settings for detection and quantitation of small molecules in complex biological sample types and clinical esoteric analyses [5]. One of the primary challenges associated with optical bioanalytical methods is that these are typically less specific and can be easily confounded by interferences. Consequently, these spectroscopic techniques typically exhibit relatively poorer precision and accuracy as compared with mass spectrometry approaches. With improved precision and accuracy, mass spectrometry is able to detect a larger range of analytes, is capable of analyzing multiple analytes in the same analysis, can operate with increased speed and specificity, and can capture comprehensive omic datasets [6]. For these reasons, MS is expected to be increasingly utilized in routine clinical analyses [7]. However, similar to spectroscopic measurements, interferences such as matrix-derived chemical noise, and the presence of isomeric compounds challenge the use of MS in the clinical setting [8]. The inclusion of additional specificity through online sample separation stages such as gas and liquid chromatography (i.e., GC-MS and LC-MS, respectively) address many of the challenges associated with complex sample analyses and isobaric/isomeric signals, and concurrently provides a means of overcoming issues related to ion source overloading/ion suppression to improve the analytical sensitivity and dynamic range [9–12]. Regardless, chromatographic separations are relatively slow and cannot resolve many isomers. Furthermore, conventional chromatography approaches are not amendable to most MS imaging modes that may be promising for clinical research [13]. To help address these limitations, gas-phase ion mobility (IM) separations can be integrated into MS-based analyses to help mitigate challenges in the analysis of complex biological samples and can be used either as an alternative to chromatography (IM-MS), or for enhanced specificity in combination with LC-MS measurements (e.g., LC-IM-MS) [14–17]. Here, we describe multidimensional analysis with IM-MS as a means to overcome some of the challenges associated with analyzing clinical samples along with current and future prospects for deploying these technologies into the clinical laboratory.

Various sample types (e.g., fluids, cells, tissue, and breath) can be prepared relatively quickly for comprehensive and rapid multidimensional analyses (Figure 1A). Selectivity can then be achieved prior to MS detection through the combination of chromatographic separations using LC, GC, or supercritical fluid (SFC) to disperse the analytes and interferences on a timescale of minutes [18]. These species can then be further separated and detected on the basis of molecular size using IM and mass using MS on timescales of milliseconds and microseconds, respectively. The complex datasets derived from these multistage approaches are subsequently mined for information using a validated bioinformatic pipeline to identify both targeted and untargeted compounds for determining clinical relevance such as disease state, drug concentration levels, or physiological metabolic concentrations of molecular panels (e.g., vitamins, hormones, amino acids, lipids, etc.). Of particular clinical relevance are metabolomics and lipidomics, which focus on relatively low mass range species of ca. 100–1,000 Daltons (Da). Isomeric compounds (species with the same chemical formula, but different molecular arrangements) are prevalent even at these lower masses and pose a challenge for confident small molecule identification efforts which cannot be addressed by MS analyses alone (Figure 1B) [19]. Whereas accurate mass measurements and other mass-derived analytical information such as isotope similarity, mass defect, and ion fragmentation information can assign a unique molecular formula to detected features, translating this chemical composition information into a chemical structure can be daunting. For example, even a relatively small molecule with the chemical formula C₂₀H₂₂N₂O₄ can be easily discerned in an MS measurement at 354.15 Da, but it can be very difficult to ascribe a specific identity with over 10,000 known isomers at this exact mass. Thus, additional orthogonal separation techniques and information are needed for accurate structural determination. Structural mass analysis with IM-MS operates with more peak capacity than MS alone, offers additional capabilities for resolving isomers, and can provide additional analytical information such as the size-to-charge of the molecule as well as empirical size-mass scaling trends. Importantly, several vendors are currently developing next-generation high-resolution ion mobility (HRIM) techniques aimed at improving IM structural selectivity and resolving power. These advances will have direct benefits for increasing the coverage of small molecules in clinical analyses (Figure 1C) [20].

Figure 1.

(A) Sample types used for systems-wide separation techniques with corresponding timescales for various analytical separation dimensions based on the physical characteristic used for separation. (B) A density plot of the number of verified compounds as a function of mass in the PubChem database. The dotted line indicates theoretical compounds with unique molecular formulas, thus a large portion of the entries represent isomeric species. With a 1 ppm mass window, an example is provided where one chemical formula (C₂₀H₂₂N₂O₄) corresponds to over 10,000 unique chemical species. (C) Molecular coverage of the empirical measurements from a large metabolite chemical library based on various mass resolving power instrumentation. The data suggests only about two-thirds of the library could be resolved with MS alone (FT-ICR, R_p>40 million). The addition of current generation IM (R_p ~50) increases the library coverage an additional 8% and with next-generation HRIM techniques (R_p ~250), coverage could increase by 25%, suggesting a total metabolite coverage of over 90% is feasible with IM-MS. Figure 1(A) was adapted from S.D. Sherrod and J.A. McLean, Systems-Wide High-Dimensional Data Acquisition and Informatics Using Structural Mass Spectrometry Strategies, Clin. Chem. (2016). 62(1), 77–83 by permission of Oxford University Press. Figure 1(B) adapted with permission from A.C. Schrimpe-Rutledge, S.G. Codreanu, S.D. Sherrod, J.A. McLean, Untargeted Metabolomics Strategies—Challenges and Emerging Directions. Journal of The American Society for Mass Spectrometry, (2016), 27 (12), 1897–1905. Copyright 2016 American Chemical Society. Figure 1(C) adapted with permission from C.M. Nichols, J.N. Dodds, B.S. Rose, J.A. Picache, C.B. Morris, S.G. Codreanu, J.C. May, S.D. Sherrod, J.A. McLean, Untargeted Molecular Discovery in Primary Metabolism: Collision Cross Section as a Molecular Descriptor in Ion Mobility-Mass Spectrometry. Anal. Chem. (2018), 90(24), 14484–14492. Copyright 2018 American Chemical Society.

Owing to these analytical advantages, the use of IM-MS in clinical settings is increasing rapidly. This review aims to identify current and future directions of IM-MS in clinical analyses by outlining several experimental approaches that may prove advantageous in clinical settings. First, multidimensional analysis of exogenous compounds in complex biological matrices will be described to highlight the analytical advantages of IM-MS for routine testing as well as the bioinformatics required for collision cross section (CCS) library identification. Second, mass spectrometry imaging of tissues for diagnostic purposes is becoming more prevalent in some clinical areas. The integration of IM with matrix-assisted laser desorption ionization (MALDI), or desorption electrospray ionization (DESI) imaging mass spectrometry provides additional filtering and structural information for more confident compound identification for tissue analysis. Finally, deployment of IM and MS in clinical settings is described by two example applications: intraoperative tumor margin delineation with MS and its impact on surgical environments, and breath analysis of volatile organic compounds with IM for identification of biomarkers or disease-state of patients. Both provide rapid analysis to acquire real-time data to improve diagnoses. Prospects and new advances in IM-MS for improved clinical analyses are also described.

2. Ion Mobility Spectrometry Techniques

Ion mobility spectrometry (IMS) was first released commercially in 1970 as a gas-phase separation technique compatible with GC and MS [21]. This technology was initially implemented as a uniform field drift tube operated at ambient pressure and allowed for both positive and negative ions to be dispersed as a separation interface before entering the mass spectrometer. This opened the field of ion mobility to multidimensional separations for small molecule analyses. Separation of ions by IMS depends on several controllable factors, including temperature, pressure of the drift gas (conventionally nitrogen, air, or helium), and strength of the electric field [22]. The drift time of the ion through the electric field is relative to the molecular size, shape, and charge. The time for a particular ion to traverse the drift tube is measured as the drift time, or arrival time at the detector. The measured drift time, in conjunction with the experimental parameters, can then be converted to a CCS value, conventionally reported in square angstroms, either through first-principles determination via the Mason-Schamp equation (for instruments operating with a uniform electric field), or through calibration against chemical standards with known CCS values (refer to Gabelica et al. [31] for derivations of the Mason-Schamp equation and Chouinard et al. [32] for a detailed description of the variables) [23–28]. Physically, the CCS value represents the surface area of the ion presented by a particular structure, which in turn is dictated by the prevailing folding forces the molecule experiences based on its composition and ion type. The CCS values obtained under the same drift gas composition are reproducible across different laboratories and IM techniques and can thus serve as a molecular identifier for the specific compound of interest [29, 30].

Technological improvements over the past decade have refined the resolution of both ion mobility and mass spectrometry approaches, leading to enhanced identification in small molecule analyses [31–33]. Similar to the different types of mass analyzers in MS, there are numerous types of IM analyzers, four of the more commonly encountered IM techniques are: drift tube ion mobility spectrometry (DTIMS), traveling wave ion mobility spectrometry (TWIMS), trapped ion mobility spectrometry (TIMS), and differential mobility spectrometry/high field asymmetric waveform ion mobility spectrometry (DMS/FAIMS) [34–36]. The choice of IMS depends on the type of analysis being performed. DTIMS (commonly found in many research-grade instruments and commercialized by Agilent Technologies in 2015) operates with a drift gas (ca. 2–10 Torr) under a uniform electric field and provides the most accurate means for measuring CCS based on first principles, and is typically used when reference CCS values suitable for calibration are desired [37–40]. TWIMS (Waters Corporation) uses a dynamic and migrating electrical potential which creates a “traveling wave” that moves ions through the buffer gas in a mobility-selective manner [30, 41–44]. In both DTIMS and TWIMS, smaller ions experience fewer collisions with the drift gas and reach the detector faster than larger ions which experience more collisions with the drift gas and reach the detector slower, thus both techniques generate very similar arrival time spectra. TIMS (Bruker Daltonics) conducts IM separations in an “analyzer” cell which holds the ions stationary between a flowing gas and a stopping potential, the latter of which is slowly decreased to release ions based on their gas-phase mobilities, and consequently molecules with the largest CCS values arrive at the detector sooner, followed by smaller molecules later, which is inverted from separations in DTIMS and TWIMS [23, 45–47]. DMS/FAIMS alternates between a low and high electric field orthogonal to a flowing gas in which the ions are entrained, which spatially partitions ions based on their differential response to the low/high field conditions. DMS/FAIMS operates at atmospheric pressure and acts as an IM filtering device which is beneficial for rejecting chemical noise and targeting compounds such as volatile organics in breath [48]. While each of these different IMS techniques are fundamentally distinct in their operational principles, all IMS technologies are selective to physical characteristics of the ions, which provides an additional structural separation strategy complimentary to MS (refer to Delvaux et al. [34] for diagrams of IM electric field configurations described here).

Each of these IM techniques is currently experiencing rapid developments in improving the resolving powers (greater than ~200) that can be achieved in the IM-dimension. For example, TIMS is capable of achieving high IM resolving powers using slow and precise voltage scanning [46]. TWIMS has recently been developed in a cyclic geometry which allows ions to undergo multiple transits to achieve high resolution [49]. Also of note is a recently developed ion optical architecture called structures for lossless ion manipulation (SLIM) which implements TWIMS and has the potential for improved resolution of compounds based on the analyte’s structural properties. The capabilities of SLIM-based IM for conducting broadband and high-resolution mobility separations makes it a prospective technology for deploying into clinical settings and other bioanalytical applications where the enhanced peak capacities are advantageous. SLIM-based IM instrumentation is commercially available as of 2021 (MOBILion Systems) [50–52]. Current prospects for these enhanced resolving power IMS strategies are described further in Section 5.

3. Multidimensional Analysis for Clinical Applications

Additional separation dimensions combined with IM-MS (e.g., LC-IM-MS) provide added capabilities for resolving and characterizing small molecules, because each orthogonal separation dimension provides an additional physical differentiating characteristic that may be unique to the molecules of interest. Utilizing multiple separation stages inline requires a careful matching of timescales, such that the overall analysis time is no longer than the longest separation stage to preserve throughput. For time-of-flight mass analyzers, the mass spectrometry sampling occurs on the order of 100s of microseconds, while IM provides separations on timescales of milliseconds, and commonly coupled pre-separation techniques such as liquid chromatography (LC) or gas chromatography (GC) separates on the timescales of minutes to hours. Thus, it is possible to combine these techniques for multidimensional analysis without sacrificing throughput, because there are sufficient differences in separation timescales of each dimension [18, 32]. Resultingly, the separation capabilities of the analytical platform can be greatly improved while maintaining the breadth of comprehensive analysis in a high throughput manner. Although IM can separate many isomers, column separation with LC or GC is necessary to fully resolve isomers on the basis of hydropathy or volatility [53–55]. Typical liquid chromatography-tandem mass spectrometry (LC-MS/MS) approaches where compounds are fragmented and matched to databases often are insufficient for isomer identification, therefore LC-IM-MS can be utilized to further identify these compounds, and this multidimensional approach is also compatible with tandem MS (e.g., LC-IM-MS/MS) [18, 20, 56]. While LC-IM-MS identifications still require databases in order to match the IM-derived CCS measurements, these analyses do not require lengthy repetitive LC methods and can be used for interlaboratory studies, making them well suited for reproducible clinical workflows [29]. For example, using databases such as the Unified CCS Compendium, mass-mobility correlation (i.e., mass-CCS trends) allows for putative coverage of unknown compounds in complex biological matrices to be identified [57–59]. Furthermore, energy-minimized theoretical structures with CCS predictive algorithms as well as machine learning approaches using molecular descriptors can be used to predict CCS values of unknown compounds based upon empirically known CCS values for classes of compounds [60–63]. There have been recent developments in creating predictive CCS algorithms that are trained against empirical data that currently incorporates larger than desirable error originating from both measurement uncertainty and the limited availability of isomerically pure standards. These algorithms will continue to improve with advancements in ion mobility measurement capabilities and availability of chemical standards.

Clinical samples are typically obtained in a variety of matrices including fluids (e.g., blood, plasma, and urine), tissue, or breath [64]. These complex matrices present a unique challenge for identifying small molecules for each analytical method previously described here (LC, IM, and MS), making it challenging to confidently determine an analyte of interest from the complex background noise of the matrix [64, 65]. By combining these multiple separation dimensions (e.g., LC-IM-MS), analytes of interest can be detected and filtered from the background noise in clinical sample matrices [66–68]. To illustrate the selectivity of multidimensional separations with LC-IM-MS in clinical samples, anabolic androgenic steroid (AAS) sulfate and glucuronide conjugates provide a challenging example because of the large number of isomer species that can confound analyses. AASs are synthesized derivatives of testosterone that promote skeletal muscle growth [69–71]. Originally designed to be used therapeutically, AASs are monitored in athletes to ensure fair and equitable competition in competitive sports [72, 73]. Athlete urine samples are commonly used for testing of steroids due to non-invasive sample collection and easy storage [73]. AASs can be detected in natural abundance and quantified to determine whether an athlete has endogenous levels or exogenous steroids are suspected to be present [74]. As a basis for understanding the levels required to deem a positive test, the minimum required performance level (MRPL) has been determined and set at a specific level/concentration for routine testing by the World Anti-Doping Agency (WADA) for each individual compound tested [75].

To confidently identify and determine whether an athlete has been taking performance enhancing steroids, each drug, metabolite of that drug, and potential long-term small molecule metabolized by the body must first be separated efficiently by analytical methods [73, 76]. This is particularly important for identification of long-term steroid metabolites that can remain in the body for up to months, which are termed phase II steroids [77]. Phase II AAS metabolites have been identified as performance enhancing drugs by WADA and quantification of them above the MRPL allows for routine drug testing laboratories to confidently determine a positive test [71, 78]. As shown in Figure 2, the analysis of three distinct AAS conjugates illustrate the complexity of interferences for determining the analyte of interest [66]. A theoretical isotopic envelope for each AAS represents how the compound would be observed without any measurement interferences. The MS-only analysis demonstrates the inherent noise of the urine matrix and the associated difficulty of compound identification in the presence of isobaric interferences. LC-MS analysis, which is presently used for routine testing, results in reduced background noise and demonstrates improved selectivity of each compound. IM-MS analysis, in some cases, also demonstrates improved selectivity over MS alone, but interferences remain. Only through combining all three separation strategies in a single LC-IM-MS analysis does the MS spectra accurately reproduce the theoretically-expected isotopic envelopes, which improves identification confidence and quantitation efforts.

Figure 2.

Three anabolic androgenic steroids, epi-THMT S3 (3α-sulfoxy-17β-methyl-5β-androstan-17α-ol), drostanolone M1 G (2α-methyl-5α-androstan-3α-ol-17-one-3-β-D-glucuronic acid), and stanozolol 1’N – G (5α-androstan-[3,2-c] pyrazole-3’,17β-diol-17α- methyl-1 ´N-glucuronic acid), with their structures, common names, and minimum required performance level (MRPL) concentrations. The corresponding mass spectra depict the theoretical isotopic distributions (blue traces) and experimental results (black traces) for different combinations of analytical dimensions. Mass measurements that accurately fall within ±5% of theoretically-expected abundances are indicated (green squares) along with abundance mismatches due to isobaric interferents (red squares). The y-axes represent relative abundance within the mass range shown. Reprinted with permission from D.E. Davis, K.L. Leaptrot, D.C. Koomen, J.C. May, G.A. Cavalcanti, M.C. Padilha, H.M.G. Pereira, J.A. McLean, Multidimensional Separations of Intact Phase II Steroid Metabolites Utilizing LC-Ion Mobility-HRMS. Anal. Chem. (2021) (93)31, 10990–10998. Copyright 2021 American Chemical Society.

Although LC-IM-MS analyses can improve selectivity and ultimately provide new methods for identifying molecules in clinical samples, isomers remain a difficult problem, especially with chemically-similar AAS conjugates [79]. Through enzymatic catalysis in the body, phase II steroids form both sulfate and glucuronide conjugates which bind at various locations with each parent steroid, creating additional biologically-derived isomers that challenge MS analysis alone [80, 81]. These species are also chemically-similar and thus are difficult to resolve with LC, and for routine clinical testing laboratories can require costly method development and time. Indeed, method development for an LC baseline separation is typically the limiting factor for routine testing of AASs [82]. Furthermore, conjugated steroid isomers yield similar fragmentation patterns with LC-MS/MS, thus combining LC separation and reference CCS values of analytes matched to standards can provide secondary and tertiary identification strategies that can confirm the analyte of interest [80, 81]. With support from validated CCS values, a relatively short LC run could be utilized, so that the multidimensional separation of compounds by LC-IM-MS could greatly improve turnover time for drug testing results. Experimental frameworks such as these could prove advantageous in many clinical settings [83]. “While this illustrative example is for exogenous species, many clinical analyses are also focused on endogenous levels of key metabolites. IM-MS analytical strategies are equally effective for such species, and we anticipate their use will grow in the future for these purposes. Some examples include vitamin D epimers [84, 85], ganglioside isomers [86], and several isomers from glucose metabolism [87].

4. Mass Spectrometry Imaging

Mass spectrometry imaging (MSI) has been at the forefront of clinical mass spectrometry for the past twenty years and was first pioneered using matrix-assisted laser desorption ionization (MALDI) as an ionization source [88]. Typically, MSI techniques rely on a raster method whereby the ionization probe is scanned over a sample, such as tissue, collecting biochemical data represented by a mass spectrum produced at each sampling location (pixel) of the image [89]. This technique is used to analyze the spatial distribution of biochemical information based upon intensity of the analyte in localized areas of a sample which can be used to determine disease state biomarkers associated with known tissue pathologies [90]. More recently, other ionization techniques such as desorption electrospray ionization (DESI) have been demonstrated for MSI to increase molecular coverage and to extend the application of MSI to other sample types [91, 92]. Importantly, MALDI is conventionally performed at reduced pressure, while DESI is performed at ambient pressure. MSI information is often integrated with complimentary microscopy images to reduce variation between samples and provide additional morphological information that is usually lost due to the limited spatial resolution of MSI images, typically 10–100 micrometers [93–95]. The spatial resolution limitations of MSI have led to the development of secondary ablation techniques, such as MALDI-2 which utilizes a second laser to increase the spatial resolution [96]. Another active area of development for MSI is to incorporate additional orthogonal separation techniques such as IM to increase peak capacity and confidence in chemical identifications.

4.1. Ion Mobility-Mass Spectrometry Imaging

As MSI is becoming increasing prevalent in clinical settings, IM has only recently been included with these experiments [97–99]. The additional peak capacity provided by IM-MS can help address incomplete resolution encountered from the analysis of complex samples, such as biologically-derived clinical samples, which in turn provides benefits for improving the quantitative aspects of MSI experiments. It is important to recognize, however, that the quantitative aspects of MSI are equally challenged by the complexities of the ionization source which both MS-only and IM-MS configurations share. Specifically, quantitative aspects in both configurations are strongly influenced by the ionization efficiency of the specific species being ionized and their chemical properties (i.e., acidity, basicity, functional groups, etc.). These aspects are discussed in detail in several recent manuscripts [100, 101]. Similar to the background noise observed in fluid matrices, biological tissues exhibit numerous and abundant interferences which confound small molecule detection. For conventional MSI using either MALDI or DESI, chromatography cannot be used to separate compounds prior to imaging, therefore combined multimodal 2-dimensional (MSI and microscopy) and 3-dimensional (MSI, microscopy, and fluorescence) imaging techniques are often used to obtain additional analytical information from these studies [102]. However, without chromatography, differentiation between isobaric interferences is difficult to achieve and many identifications are putatively based upon mass measurement alone. IM occurs post-ionization and thus provides an additional separation strategy to MSI that allows for filtering of background noise, and subsequently, improves selectivity of the features identified and enhances the spatial resolution of the image produced [103, 104]. Additionally, isomeric species can potentially be resolved with IM-MSI, improving identification of small molecules for clinical analyses, while also providing CCS measurements that can be used to match detected features to reference CCS values [98, 99, 105, 106].

4.2. IM-MSI: Filtering and Improved Resolution

With additional selectivity, IM-MSI analyses can improve imaging resolution by filtering background noise without loss of concomitant sampling time [106]. Additional IM-MS projections allow background chemical noise to be discerned from signals of interest. This background filtering capability is illustrated in Figure 3 where gangliosides from a murine brain sample are analyzed separately by both MALDI- and DESI-IM-MSI and detected features obtained from both analyses are plotted as an overlay [104]. DESI-IM-MSI was found to yield more total features than MALDI-IM-MSI which likely results from the DESI solvent background (Figure 3(A)). Isobaric species for m/z 862 and 888 are found in both the background noise region as well as the region occupied by known lipid species (Figure 3(B)), which allows the signals of interest to be resolved from noise in the IM-MS analysis, resulting in higher quality MS derived images.

Figure 3.

2-dimensional IM-MS plots of MS mass-to-charge (y-axis) and IM drift time (x-axis) overlaying the features found from MALDI- (blue) and DESI- (red) sample analysis. (A) The full range results demonstrate the dual ionization approaches generates complimentary measurements. Yellow areas overlaid on top of red and blue points represent background MALDI peaks from the matrix. (B) Background noise (purple box) resides in a distinct region from known lipid signal (green box), allowing higher quality IM-filtered MSI images to be obtained (insets). Reprinted from Methods (San Diego, Calif.), 104, Škrášková, K.; Claude, E.; Jones, E. A.; Towers, M.; Ellis, S. R.; Heeren, R. M. A. Enhanced Capabilities for Imaging Gangliosides in Murine Brain with Matrix-Assisted Laser Desorption/Ionization and Desorption Electrospray Ionization Mass Spectrometry Coupled to Ion Mobility Separation, 69–78, Copyright (2016) with permission from Elsevier.

5. Prospects for Ion Mobility Deployment in the Clinic

IM-MS is a field of active development and provides several analytical advantages that are beneficial to clinical applications. Some prospects for IM-MS in the clinical setting are described in the following sections.

5.1. Mass Spectrometry Deployment in the Clinic: Combining Biochemical Tumor Delineation in Surgical Environments

To assess disease state by detecting tumor margins in cancerous tissues, three MS tools with distinct delivery systems have recently been developed and deployed in the clinic: desorption electrospray ionization-MS (DESI-MS), the MasSpec Pen and the intelligent knife (iKnife). DESI-MS has been demonstrated in typically ex vivo contexts to map cancerous tumor margins in pathological tissue samples [92, 107, 108]. First described in 2004, DESI directs charged droplets derived from an electrospray ion source onto a tissue. Through analyte extraction and secondary droplet formation, the analyte species brought to the surface of the tissue are directed into the mass spectrometer. The MasSpec pen utilizes a similar principle to a liquid microjunction surface sampling probe (LMJ-SSP) described by Van Berkel and colleagues for planar surfaces. In the LMJ-SSP a continuously flowing coaxial solvent is directed to a surface via a microjunction for micro liquid extraction of analytes. The extracted analytes are then coaxially directed into the ion source (most commonly ESI) of the MS [109–111]. Rather than a continuous solvent flow, the MasSpec pen uses gas displacement of user defined single droplets for analyte extraction (Figure 4(A)) [112]. This hand-held, liquid-delivery auxiliary tool coupled to MS can be used for real-time biochemical analysis during surgical procedures [100]. The MasSpec Pen allows for rapid, nondestructive molecular analysis to diagnose abnormal tissues. While intraoperative spectroscopic analyses such as Raman have previously been demonstrated for real-time biochemical analysis during surgery, the MasSpec Pen can collect MS-derived molecular information for disease biomarkers in tissue [112, 114, 115]. For example, intraoperative detection of abnormal tissue types can be assessed in ambient conditions by measuring certain onco-metabolites such as 2-hydroxyglutarate in glioma tissue [116]. Additionally, tumor margins can be assessed with the MasSpec pen in thyroid carcinoma (Figure 4(B)) [112]. Elevated abundances of lipid species in the m/z 700–900 range indicate the disease state of the tissue. Among the changes in the metabolic profile of the thyroid carcinoma there is a higher abundance of phosphatidyl inositol (m/z 885), which allows the clinician to determine tumor margins in real-time. While the surgical operation itself is invasive, the MasSpec Pen analysis utilizes discrete water droplets that are nondestructive, which allows for normal tissue to remain intact, improving decision making by the surgeon, and subsequently, the efficiency of the surgical procedure.

Figure 4.

(A) A schematic diagram of the utilization and operational components of the MasSpec pen. The lower right inset in depicts the flow diagram for liquid extraction of analytes from the liquid microjunction of the pen in contact with tissue. When extraction is desired, the foot pedal is depressed and a droplet of water is directed to the top of the pen (t = 0s). After extraction has occurred for a desired time (t = 2s), a vacuum line and gas flow line are opened to transport the droplet and extracted analytes to the MS (t = 3s). (B) MasSpec Pen analysis of normal thyroid tissue (top) and thyroid carcinoma (bottom) showing increased abundance of various lipid species in the m/z 700–900 range in the tumor tissue, specifically phosphatidyl inositol (m/z 885). (Figure 4(A,B) adapted with permission from Zhang, J.; Rector, J.; Lin, J. Q.; Young, J. H.; Sans, M.; Katta, N.; Giese, N.; Yu, W.; Nagi, C.; Suliburk, J.; et al. Nondestructive Tissue Analysis for Ex Vivo and in Vivo Cancer Diagnosis Using a Handheld Mass Spectrometry System. Science Translational Medicine, (2017), 9 (406). Copyright 2017 American Chemical Society.

Another surgery-deployable MS probe termed the iKnife captures the partial combustion products generated during electrocauterization of tissue to collect MS measurements that can be used to determine disease states [117]. The iKnife uses a rapid evaporative ionization probe and a vacuum tube inlet to the MS to collect gas-phase ionic species that are created by the electrosurgical component [118]. This allows for highly specific molecular information unique to the metabolic profile of the tissue type to be collected and analyzed. The rapid evolution of analyte sampling devices for high performance molecular characterization by MS or IM-MS provides new avenues in a variety of clinical settings.

To mitigate damage to tissues during these types of clinical analyses, minimally invasive procedures have also been proposed to collect molecular information with the MasSpec Pen and iKnife for laparoscopic and endoscopic surgeries, respectively [119–121]. Although these techniques provide incredibly precise biochemical analysis, spectral databases need to be improved for further implementation into the clinic since many rare tumors lack validated biomarker information [117]. To improve database coverage, including IMS information in these efforts could help improve the detection and selectivity of molecular features, as well as provide CCS measurement capabilities which can support molecular identifications [29]. Additionally, the IM-MS information could be leveraged for mass-mobility correlations that are associated with known intraoperative molecular targets to prioritize the discovery of unknown compounds associated with specific tissues and disease states.

5.2. Predictive Biomarker Diagnoses with IM Breath Analysis

Volatile organic compounds (VOCs) from saliva offer a fast, non-invasive way to measure biomarkers in several diseases [122]. VOC analyses have been dominated by GC over the past several decades, however these workflows are time consuming [123]. IMS offers a sensitive, fast, and compact design as an alternative approach for VOC analysis [123]. Breath analysis of VOCs using multi-capillary column (MCC) IMS provides rapid detection of compounds in a non-invasive manner to identify disease biomarkers at the bedside in several disease types including pancreatic cancer [124], lung cancer [125], and infections related to pathogenic bacteria [126]. Validated biomarkers and biomarker candidates can diagnose an individual’s current disease state or likelihood of disease development. IM breath analysis can shorten the time needed to diagnosis a patient’s condition which can potentially improve patient outcomes. Additionally, MCC-IMS testing of VOCs is ideal for intensive care neonatal testing where non-invasive techniques are extremely important for immunocompromised newborn babies [122]. These IMS analyses in the clinic are important for quick biomarker detection, however the coverage of VOC molecular analysis is currently limited by the lack of orthogonal analytical dimensions such as LC and MS.

5.3. Next Generation of Ion Mobility

As ion mobility systems continue to improve, the capability to achieve high-resolution ion mobility (HRIM) with resolving power greater than ~200 becomes possible. Several approaches have been developed over the past decade to obtain HRIM including structures for lossless ion manipulation (SLIM), cyclic geometry TWIMS (cIM), and TIMS [49, 127, 128]. Each HRIM technology is implemented with high-resolution time-of-flight MS analyses and boasts resolution capabilities greater than ~200 with the potential to perform tandem IM experiments [129], however each have their own distinct strategies for achieving this level of resolving power. The included HRMS stage is important for resolving molecular features within complex samples such as those encountered in clinical applications, but adding HRIM analysis provides structurally-selective resolution of molecules which HRMS cannot address, namely isobars and isomers. Although the HRIM technologies described below have not yet been directly implemented in clinical applications, HRIM has the potential to significantly impact the field in the coming decade by providing increased peak capacity and resolution sufficient for improved isomer characterization needed for more accurate metabolic profiling of human diseases.

SLIM is an ion optical circuit board architecture supporting extended path lengths of ion trajectories to facilitate higher resolution in ion mobility analyses [130]. SLIM is well suited for interfacing with both MS and LC-MS, as well as providing a method for determining IM-derived CCS values [51, 52, 131]. While conventional mobility technologies such as TWIMS and DTIMS have a maximum resolving power between 40 and 60, SLIM has the capability of operating at resolving powers of ~200 or greater [132–134]. This allows improved resolution of clinically-relevant isomers of different chemical classes at a much higher efficiency than previously described IM techniques. In addition, with SLIM geometries incorporating a cyclic return path for ions, multi-pass SLIM experiments are possible, which can further enhance the resolution of select analytes of interest after each successive pass [132, 133]. Presently, the highest recorded IM resolving power in SLIM was ~340 with a separation power of 1860, demonstrated using 40 passes across a distance of about half a kilometer (~540 m) [133]. This has allowed isotopomers (i.e., amino acids incorporating heavy vs. light isotopes) to be resolved with up to a ~0.4% difference in mobility [135]. Heavy-labeling experiments are currently utilized in many clinical mass spectrometry applications and direct differentiation of isotopically-labeled analytes by HRIM will directly benefit clinical applications. Following extensive development at Pacific Northwest National Laboratories, SLIM is now commercially-available from MOBILion Systems.

Another approach to improve the reproducibility and precision of HRIM is the cyclic geometry TWIMS (cIM) platform, commercially-available from Waters [49]. The cIM geometry incorporates an optional ion optical loop which can operate as a multipass system to improve resolution with each successive pass [136, 137]. This type of technology yields IMSⁿ-type workflows where the number of passes (n) increases the resolving power of the analysis, and the optional cyclic path facilitates structurally-informative tandem IMS experiments where mobility-selected ion populations can be activated and measured with subsequent IMS measurements cycles. The cIM technology has demonstrated resolution for previously indistinguishable stereoisomers such pentasaccharide anomers and fluoroquinolone protomers [138, 139]. Tandem IM capabilities also have the potential to provide more informative stereoselective analysis that could elucidate a specific compound’s previously unknown contribution to human disease.

A third HRIM technique that is commercially-available (Bruker) is TIMS [46, 127, 140–142]. TIMS uses a combination of gas flow and electric fields to trap ions and selectively release them based on their gas phase mobilities. TIMS is capable of operating with high resolving powers by lowering the scanning speed of the trapping field [46]. TIMS has been combined with MALDI MSI to improve the spatial resolution of the imaging analysis, for example, by resolving phosphatidylcholine isobaric species in whole-body mouse tissue [143]. For imaging applications relevant in clinical analyses, the high resolution of TIMS (200 to 400) allows for improved molecular selectivity, which produces more accurate biochemical MS images[144].

5.4. Improved Isomer Separation and CCS Characterization with HRIM

High-resolution IM capabilities have the potential for making a significant impact in clinical analysis, especially for isobaric and isomeric separations where MS alone cannot address. For example, isomers such as reversed sequence peptides (SDGRG and GRGDS) are commonly used to benchmark IM systems, but the singly-charged ion form is unresolvable using current generation IM techniques such as TWIMS and DTIMS (Figure 5A, middle panel) [52, 145]. These reverse sequence peptides show improved isomer separation with both SLIM-IMS and cIM (Figure 5A and 5B), the latter achieved using increasing number of passes through the system [49, 52]. Similarly, TIMS is able to resolve singly charged isobaric peptides with ~200 resolution (Figure 5C) [140]. Each HRIM technology demonstrates improved resolution of isomeric/isobaric peptide small molecules that are challenging to separate with conventional resolving power devices.

Figure 5.

Example isomeric/isobaric peptide separations for three HRIM techniques. (A) Improved resolution of reversed sequence peptide isomers, SDGRG and GRGDS (top panel), by SLIM-IMS. Conventional resolving power DTIMS (middle) yields a single, unresolved peak for the mixture, whereas HRIM SLIM-IMS (bottom) demonstrates baseline resolution. “RA” stands for relative abundance. (B) Cyclic geometry TWIMS showing improved resolution of the same reverse peptides with each iterative pass (successively longer separations) in the system. Here, the resolving power reaches 350 for 16 passes. (C) TIMS analysis of two isobaric peptides which are separated at ~200 resolving power. Figure 5(A) adapted with permission from May, J. C.; Leaptrot, K. L.; Rose, B. S.; Moser, K. L. W.; Deng, L.; Maxon, L.; DeBord, D.; McLean, J. A. Resolving Power and Collision Cross Section Measurement Accuracy of a Prototype High-Resolution Ion Mobility Platform Incorporating Structures for Lossless Ion Manipulation. Journal of the American Society for Mass Spectrometry, 2021, 32 (4), 1126–1137. Copyright 2021 American Chemical Society. Figure 5(B) adapted with permission from Giles, K.; Ujma, J.; Wildgoose, J.; Pringle, S.; Richardson, K.; Langridge, D.; Green, M. A Cyclic Ion Mobility-Mass Spectrometry System. Analytical Chemistry, 2019, 91 (13), 8564–8573. Copyright 2019 American Chemical Society. Figure 5(C) adapted with permission from Garabedian, A.; Benigni, P.; Ramirez, C. E.; Baker, E. S.; Liu, T.; Smith, R. D.; Fernandez-Lima, F. Towards Discovery and Targeted Peptide Biomarker Detection Using NanoESI-TIMS-TOF MS. Journal of The American Society for Mass Spectrometry 2017, 29 (5), 817–826. Copyright 2017 American Chemical Society.

When integrated with additional separation dimensions such as LC, a combined multidimensional LC-HRIM-MS platform should provide sufficient resolution to differentiate the majority of signals originating from complex biological matrices of importance in clinical settings and could potentially lead to the detection of new biomarkers and biological insights [146]. Collectively, analytical approaches incorporating HRIM would be especially useful for untargeted omics applications that seek to discover molecules related to human disease.

6. Conclusion

IM-MS platforms are currently utilized in several clinical areas including the detection of endogenous and exogenous species from complex clinical sample types, biomolecular imaging, intraoperative tumor margin delineation with MS and VOC biomarker detection in IM breath analysis. While clinical applications of MS have been predominantly performed using LC-MS and LC-MS/MS [147], additional multidimensional separation techniques such as LC-IM-MS have shown considerable utility in the omics sciences and are expected to directly impact clinical chemistry into the foreseeable future. Relatively new IM technologies such as SLIM, cIM and TIMS represent the current state-of-the-art in IM-MS technological developments and will contribute to clinical initiatives aimed at identifying and quantifying features in complex biological matrices. Moreover, ongoing efforts to improve HRIM technologies and develop CCS databases in support of compound identification could significantly impact patient outcomes by reducing the time from analysis to physician decision making. The capability of IM-MS to collect a broad range of molecular information with high precision and accuracy will translate to positive contributions to the field of personalized medicine [148–150]. As these technologies continue to improve, the potential for IM-MS to contribute to comprehensive molecular analysis and discovery of biomolecular patterns to address human disease will continue to be bright in clinical chemistry.

7. Expert opinion

Current IM-MS systems provide enhanced selectivity and separation that can increase efficiency of compound detection and identification in complex biological clinical analyses. There is significant promise for using IM-MS for biomarker discovery and validation for assays providing more rapid patient diagnoses. Facilitating rapid and more accurate diagnostics should lead to faster diagnosis times and in turn improve patient outcomes. Development of ion mobility and mass spectrometry technologies for use in the clinic have become more prevalent, however deployment can be hindered by cost of systems, time-consuming training needed to produce skilled technicians, and the need for computational workflows to fully integrate omic datasets. Additionally, development of extensive compound databases with empirical IM-MS measurements derived from chemical standards is costly and compound validation and prediction methods for further computational processing efforts need to be improved. Database coverage of calibrated CCS values are expanding for untargeted omic analyses and hold great promise for the future. The ability of IM-MS to mitigate interferences often encountered in complex biological samples drives research to move these technologies into clinical settings. The breadth of selectivity includes the ability to resolve many isomer species that are common in such analyses. Advances in IMS resolving power should promote implementation of IM-MS systems into hospitals and clinical settings presently and into the future.

A key challenge in molecular discovery and diagnostics are the prevalence of interfering (isobaric) and isomeric species. Contextually, many isomers have yet to be characterized in relation to disease and are in many cases intractable to many analytical methodologies commonly encountered in clinical settings. Ion mobility spectrometry, by virtue of rapid separations on the basis of molecular size and shape, has demonstrated great utility for separating both isobars and isomers in a number of systems. Furthermore, the timescale of IM separations in milliseconds versus that for gas or liquid chromatography in minutes represents a nearly five order of magnitude enhancement in analysis time. High-resolution ion mobility will also improve isomer separation capacity, which would provide higher accuracy to guide decision making. To address reproducibility of clinical analyses and overall molecular coverage, standardized interlaboratory protocols based upon CCS libraries for high-selectivity determination could significantly improve detection of biomolecular compounds for clinical analysis.

Ion mobility-mass spectrometry has improved clinically relevant bioanalytical workflows by minimizing analysis time, expanding molecular coverage, and advancing reproducibility across different laboratories. Future efforts aim to implement HRIM instrument geometries and technologies in clinical settings to resolve previously indistinguishable small molecule isomers, and with this information, discover their role in human disease or to serve as potential diagnostic species. Advances in bioinformatic and biostatistical workflows that leverage the massive chemical information derived in such analysis are needed now and into the future – it is anticipated that with such advances in informatics workflows the translation of multidimensional datasets into clinically relevant information will be key to widespread adoption of these technologies. Standard clinical analyses will be significantly enhanced by HRIM platforms and with the ultimate goal of improving patient outcomes.

Article highlights:

• Ion mobility improves selectivity in clinical analyses.

• Integrated multidimensional separations with LC-IM-MS allow for exogenous isomeric compounds to be resolved.

• Mass spectrometry imaging is improved by ion mobility selectivity.

• Intraoperative deployment of mass spectrometry indicates tissue disease state.

• Next generation high-resolution ion mobility allows for improved isomer separation for high confidence molecular annotation and quantitation.

Abbreviations

MS

Mass spectrometry

IM

ion mobility

IMS

ion mobility spectrometry

LC

liquid chromatography

GC

gas chromatography

SFC

supercritical fluid chromatography

LC-MS

liquid chromatography-mass spectrometry

LC-MS/MS

liquid chromatography-tandem mass spectrometry

LC-IM-MS

liquid chromatography-ion mobility-mass spectrometry

CCS

collision cross section

DTIMS

drift tube ion mobility spectrometry

TWIMS

travelling wave ion mobility spectrometry

TIMS

trapped ion mobility spectrometry

FAIMS

field asymmetric ion mobility spectrometry

SLIM

structures for lossless ion manipulations

AAS

anabolic androgenic steroids

MRPL

minimum required performance level

WADA

World Anti-Doping Agency

MSI

mass spectrometry imaging

MALDI

matrix-assisted laser desorption ionization

DESI

desorption electrospray ionization

LMJ-SPP

liquid microjunction surface sampling probe

HRIM

high-resolution ion mobility

UHRIM

ultra-high-resolution ion mobility

VOCs

volatile organic compounds

MCC

multi-capillary column

cIM

cyclic traveling wave ion mobility