Isomeric and isobaric species are prolific within the lipidome, originating from fatty acyl (FA) position, functional group orientation (e.g. bis(monoacylglycerol)phos- phate [BMP] versus phosphatidylglycerol [PG]), sn- position, double- bond position and orientation, regioisomers, and even species in different classes and subclasses with the same atoms but different head group compositions. Since IMS offers mul- tiple analytical benefits including orthogonality to chromatographic and MS analy- ses, the reduction of spectral complexity of biological samples, and the measurement
of CCS to support molecular identification and infer structural information, its use in lipidomic studies keeps on increasing (Table 6.2) [90]. This section will further detail the IMS benefits of (i) chemical separation, (ii) analyte identification and characterization, and (iii) structural analysis for lipidomic assays.
6.1.4.1 Chemical Space Separation with IMS
Biomolecule classes such as oligonucleotides, lipids, metabolites, and proteins all share similar compositions of C, H, N, P, O atoms; however, each differs in its mono- meric atomic structural arrangements that are then polymerized to form complex structures [91]. As such, each of these classes exhibits a unique structural density.
Given the correlation of analyte size and mass, the separation of analytes by IMS- MS results in the formation of distinct trendlines when comparing the observed mass- to- mobility relationships, compensation voltage (CV), and/or CCS values (Figure 6.2). For complex sample types, the IMS size separation is particularly advantageous to filter biomolecule classes prior to MS analyses. This capability has resulted in an increase in the S/N ratio of lipids at lower abundances and enhanced peak capacity through the reduction of chemical noise interferences [92]. The linear trends of CCS and m/z further translate within biomolecular classes where varia- tions in lipid backbone facilitate the separation of lipids by category (Figure 6.6a) [25, 92]. Additionally, within lipid categories, the fluctuations in the head group moiety such as those in phospholipids also showcase unique trends in both FAIMS CV ver- sus m/z plots [88] and CCS and m/z plots (Figure 6.6b) [25, 92]. Further zooming into these mass–mobility trends has also demonstrated a correlation of fatty acyl Table 6.2 Benefits of IMS for lipidomics.
Analytical use of ion
mobility Description Additional
requirements Example application areas
1. Chemical
separation Partition signal from chemical noise and increase the peak capacity of analysis
None Imaging and shotgun
lipidomics, chemical space reduction, and discovery of low- level lipid ions
2. Analyte identification and characterization
Use CCS values to characterize unknowns by correlation
Reference values from databases/libraries Normalized drift times, measured reduced mobilities, and/or calculated CCS values
Lipid identification
3. Structural
analysis Utilize
experimental CCS values to infer structural information
Computational methods to link theoretical structure to CCS values
Fundamental understanding of IMS separations gas- phase chemical arrangement insight
Source: Adapted from May et al. [90].
6 Ion Mobility Spectrometry 166
information where lipids with the same head group and total carbon count exhibit a linear relationship (Figure 6.6c) [25, 92]. Thus, as the number of double bonds increases from 0 to 1 and so on, the observed CCS value decreases in a linear fash- ion [25, 92], and the saturated version of a lipid, i.e. PC(44:0), has the longest observed drift time and largest m/z value. As more unsaturated sites are added, both mass and drift time subsequently decrease. This relationship is incredibly powerful for postulating putative identifications of unknown spectral features.
6.1.4.2 Lipid Identification and Characterization with CCS
The integration of IMS and MS for lipidomic analyses has proven extremely useful for facilitating the separation of chemical space. However, the abundance of lipid isomers and the correlation of mass with size limit the characterization capabilities of the entire lipidome solely through these two dimensions, even with the integra- tion of additional separation dimensions such as LC. Therefore, the amount of lipid speciation possible can be variable depending on the separation capabilities and co- existing isomers (Figure 6.7a) [25]. IMS analyses for numerous lipid isomer pair standards have noted Rp of several hundreds to distinguish variations in functional
250
200 200250300350
300Predicted CCS (Å2)Predicted CCS (Å2) Predicted CCS (Å2)400250270310290
500 750 400 600
PC PE PG PS PI PA
PC PE PG PS PI PA
600 1000 Glycerolipid (GL)
Lipid category Lipid class
(a) (b)
Sphingolipid (SP) Glycerophospholipid (GP)
1000
650 700 750
(c)
Length of acyl chains
800 850
Lipid species Number of
doub le bo
nds PC(42 : 0)PC(44 : 0) PC(44 : 1)
900 950
m/z m/z
m/z 1250
Figure 6.6 Mass versus mobility relationships for lipids. The separation of lipids by category (a), class (b), and summed fatty acyl composition (c) is facilitated by the relationship of CCS and/or mobility with observed m/z. Source: Figure adapted from Tu et al. [93].
group position, acyl chain position, chain length, and double- bond geometry and orientation (Figure 6.7b) [25]. This has however been accomplished. For instance, FAIMS platforms have demonstrated the successful separation of isomer types aris- ing from fatty acyl differences [89]. Similarly, multiple lipid isomer types including PG and BMP isomers, which differ in their head group placement in the sn- 3 (PG) or sn- 2 (BMP) position, have been effectively resolved with TIMS [94]. Advancements in temporal IMS systems such as SLIM have shown effective separation of cis/trans and double- bond- positional isomers through increased separation path length [95].
However, biological sample complexity can preclude the separation efficiency that has previously been showcased with lipid isomer standards, demonstrating that size- based separation alone may not be enough to resolve species and make identi- fications in real- world samples. However, as CCS values are an intrinsic property of each molecule, they can be determined using low- field IMS systems and provide an additional descriptor for making identifications to enhance annotation confidence and facilitate isomer distinction within complex biological samples. Thus, the more CCS values collected for lipids, the better our confidence and annotation capabili- ties will be.
Databases of experimental CCS and mobility values have been released to support the identification of biomolecules. Notably, the CCS Compendium out of the McLean laboratory at Vanderbilt University houses 3800 experimental CCS values across 80 molecular classes determined from a DTIMS system operating under nitrogen gas [20]. Additionally, AllCCS [96] and LipidCCS [25] by the Zhou lab and DeepCCS [97] offer a means of computationally predicting lipid CCS values under nitrogen gas for a number of positive- and negative- mode adducts. CCSbase is another database containing over 14 000 computationally predicted and experimen- tally derived CCS values from a variety of IMS platforms and sources [98]. Other computational algorithms such as ISiCLE [99] also offer CCS prediction capabilities
Lipid category
Lipid class
GP
PC 300
277
284 289 292 296 300
284 289 292 296 300 284 289 292 296 300 284 289 292 296 300 281 285 289 293 277 281 285 289 293 304 308
ΔCCS = 0.73%
Rp= 50
ΔCCS = 0.20%
Rp= 50
ΔCCS = 0.74%
Rp= 50
ΔCCS = 0.43%
Rp = 50 313 317 300 304
CCS (Å2) CCS (Å2)
CCS (Å2) CCS (Å2)
CCS (Å2)
CCS (Å2) CCS (Å2)
CCS (Å2)
308 313 317 Rp = 500
Rp= 500
Rp= 500
Rp= 500 Acyl-chain isomers
Sn-positional isomers
Double-bond positional isomers
Cis/trans isomers PC(18 : 2/18 : 2) PC(16 : 0/20 : 4)
PC(14 : 0/18 : 0)
PC(18 : 1(9Z)/18 : 1(9Z))
PC(18 : 1(9Z)/18 : 1(9Z)) PC(18 : 1(9E)/18 : 1(9E)) PC(18 : 1(6Z)/18 : 1(6Z))
PC(18 : 0/14 : 0)
PC(34 : 1)
PC(16 : 0_18 : 1) versus PC(16 : 1_18 : 0)
PC(16 : 0/18 : 1) versus PC(18 : 1/16 : 0)
PC(16 : 0/18 : 1(9z)) versus PC(16 : 0/18 : 1(9E)) Lipid species
(a) (b)
Acyl-chain isomer
Positional isomer
Classification hierarchy
Stereo- isomer
Figure 6.7 IMS separation capabilities for lipid isomers and the resulting effects on lipid speciation. Source: Figure reproduced with permission from Tu et al. [93].
6 Ion Mobility Spectrometry 168
under helium gas. Taken together, the extensive culmination of CCS values in data- bases offers researchers a wealth of information that is readily accessible for analyte identification. Experimentally, CCS values are also of great importance for separat- ing isomers and comparing data to reference values. CCS values have also been used as a filter to set tolerance windows to deconvolute spectra by observed drift time that has previously been demonstrated to reduce both false positive and negative identi- fications [100]. Vendor- neutral software platforms such as Skyline are also begin- ning to incorporate CCS and ion mobility spectrum filtering for small and large molecules [101]. More details on the benefits of CCS are presented in Section 6.1.5.
6.1.4.3 CCS for Lipid Structural Analysis
Since CCS is a descriptor for a gas- phase structure of a molecule, it has also been used in conjunction with cryo- EM, NMR, and CD to compare to the solution phase and elucidate structural information [102, 103]. The most obvious application of CCS in structural analysis lies with large biomolecules such as proteins [103, 104].
However, the measurement of CCS in tandem with in silico modeling can also pro- vide a fundamental understanding of lipid adduct formation with observed differ- ences in CCS elucidating the structural variations driving mass versus mobility trends and helping infer molecular structure when standards are unavailable. Kim et al. performed molecular dynamics simulations to elucidate the structural occur- rences of saturated and unsaturated PC cations that detailed the preferential gas- phase structure of PC lipids as globular or extended conformations at varying tail lengths and degrees of unsaturation [105]. Likewise, modeling has been employed to understand the structural data of fatty acid isomer separations [7]. Taken together, CCS measurements provide researchers a valuable reference that can be used to detail the benefits of IMS for lipidomic analysis that also translates to optimizing IMS separation efficiency of an isomeric pair by leveraging in silico predictions.