To What Extent is FAIMS Beneficial in the Analysis of Proteins?
- 2.1k Downloads
High field asymmetric waveform ion mobility spectrometry (FAIMS), also known as differential ion mobility spectrometry, is emerging as a tool for biomolecular analysis. In this article, the benefits and limitations of FAIMS for protein analysis are discussed. The principles and mechanisms of FAIMS separation of ions are described, and the differences between FAIMS and conventional ion mobility spectrometry are detailed. Protein analysis is considered from both the top-down (intact proteins) and the bottom-up (proteolytic peptides) perspective. The roles of FAIMS in the analysis of complex mixtures of multiple intact proteins and in the analysis of multiple conformers of a single protein are assessed. Similarly, the application of FAIMS in proteomics and targeted analysis of peptides are considered.
KeywordsFAIMS DMS Proteins Proteomics Peptides
What’s in a Name?
This technique, whereby ions at atmospheric pressure are separated on the basis of differences in their mobilities in high and low electric fields, goes by many sobriquets. It is variously known as high field asymmetric waveform ion mobility spectrometry [3, 4] (or simply field asymmetric waveform ion mobility spectrometry ), differential ion mobility spectrometry, differential mobility spectrometry , and the acronyms FAIMS, DMS, and DIMS. Early publications also used the terms transverse field compensation ion mobility spectrometry  and field ion spectrometry , although these have since fallen out of use. Some researchers have attempted to delineate the terms by assigning “FAIMS” to devices comprising cylindrical electrodes and “DMS” to those with planar electrodes. Others have applied c-FAIMS and p-FAIMS to distinguish between cylindrical and planar geometries . The use of multiple names is a barrier to acceptance and adoption by the broader scientific community and this author urges researchers in the field to reach a consensus in the matter. In the current absence of consensus, this article will use the term ‘FAIMS’.
Principles of FAIMS
What Sets FAIMS Apart from Other Ion Mobility Spectrometry Techniques?
In FAIMS, ions are transported through the device by a flow of gas. In conventional IMS, ions are transported by an electric field.
In FAIMS, the electric field is applied perpendicular to the (net) direction of travel (i.e. direction of gas flow). In conventional IMS, the direction of the field and of ion movement are the same.
In FAIMS, the electric field oscillates. In conventional IMS, it is constant. (It should be noted that in TWIMS, the electric field oscillates along the direction of travel, although the overall field along the ion mobility device is constant).
In FAIMS, ions are separated according to electric field (compensation field) whereas in conventional IMS, ions are separated temporally. Consequently, FAIMS provides a continuous flow of ions and conventional IMS provides discrete packets of ions.
To date, FAIMS cannot be used to directly determine the collision cross section (Ω) of an ion. In conventional ion mobility spectrometry (i.e., in the low field limit) several models exist for calculation of Ω including projection approximation (PA) , trajectory model (TM) , exact hard-sphere scattering (EHSS) , and projection superposition approximation (PSA) . No such model exists currently for high field ion mobilities, and differential mobilities cannot be ascribed to ion structures. It is possible to hyphenate FAIMS with other techniques to measure collision cross sections. For example, Smith and co-workers have coupled FAIMS with conventional IMS [19, 20], and Purves et al. coupled an energy loss method in a triple quadrupole mass spectrometer with FAIMS .
FAIMS also differs from differential mobility analysis (DMA) . In DMA, ions travel between two electrodes, in a direction perpendicular to a flow of gas. The gas flow thus displaces the ion. Only ions with specific mobility, determined by the voltage between the electrodes and the flow rate of the gas, are able to traverse the gap. The potential for confusion here is confounded by the fact that FAIMS is often referred to as DMS!
Mechanisms of Separation in FAIMS
At low electric fields, ion mobility is governed by the temperature of the bath gas. At high electric fields, the mobility depends on both the temperature of the gas and the velocity of the ion through the bath gas (and therefore the strength of the field) . A crucial factor is long-range ion-induced dipole attractive forces between the ion and the bath gas. The strength of these interactions depends on the polarizability of the bath gas and the size and charge of the ion and, hence, will vary with gas composition and nature of the ion. If the thermal energy of the bath gas is small relative to the potential well of the attractive force, collisions between the ion and bath gas will result in energy gains by the ion. Hence ion mobility will increase with increasing electric field strength (A-type). Conversely, if the thermal energy is similar or greater than the potential well of the attractive force, collisions will result in energy loss by the ions. That is, ion mobility will decrease with increasing electric field (C-type).
where K X and K Y are the mobilities in gas X and gas Y individually. Blanc’s law only holds when the electric field strength approaches zero. Deviations from Blanc’s law at high fields are relatively small but because FAIMS relies on differences in mobilities, the effects on FAIMS performance are considerable. The effects are most significant when differences in the masses of the gas components and the ion mobilities within the components are large. Shvartsburg and colleagues concluded that a binary mixture comprising He and SF6 would constitute the ultimate FAIMS carrier gas .
Eiceman and co-workers showed that the presence of water in the carrier gas affected the FAIMS spectra of organophosphorus compounds . They proposed a mechanism of separation in which ion-neutral clusters form during the low field segment of the FAIMS waveform and dissociate during the high field segment. The larger collision cross section of the cluster serves to decrease ion mobility during the low-field segment while the ‘naked’ (or near naked) ion has increased ion mobility during the high-field segment. Thus, clustering/declustering amplifies the differential mobility of the ion. An obvious corollary is that addition of modifiers to the carrier gas should improve FAIMS separation capacity. First illustrated for nitro-organic explosives , this approach has since been more widely applied [6, 34, 35, 36, 37]. Blagojevic et al. showed that polar modifiers can suppress proton transfer to other modifiers thereby improving sensitivity . Purves et al. showed that addition of acetonitrile improved peak capacity . Work by Ruotolo et al. suggested that conformational selectivity might be improved more generally in ion mobility spectrometry through judicious selection of drift gas . Further support for the clustering model comes from the work of Russell and co-workers .
A detailed investigation by Schneider et al.  of a range of low molecular weight species (including small peptides) showed that FAIMS separation is a delicate balance of the clustering model and the hard-sphere scattering model. For polar carrier gases and modifiers, the clustering model dominates, and A-type behavior is observed. In the absence of polar modifiers, and particularly in helium, hard-sphere scattering dominates, and C-type behavior is observed. As mentioned above, at sufficiently high fields, mobility will decrease for all ions because of the hard repulsive core of the ion neutral interaction.
A third model has been proposed to explain the behavior of some protein ions in FAIMS . Shvartsburg et al. investigated the FAIMS transmission of 10 proteins within a molecular weight range of 8 to 66 kDa . They found that proteins with mass < ~30 kDa showed a decrease in mobility with increasing electric field (C-type), whereas those with mass >30 kDa showed an increase in mobility with increasing electric field (A- or B-type behavior). It was concluded that the A- (or B-) type behavior was the result of reversible dipole alignment of the high mass ions during the high electric field segment of the FAIMS waveform. That is, all large protein ions have a dipole, which will align in a sufficiently strong electric field. Once aligned, the mobility of the ion will be dictated by the collision cross section of the ion in the plane orthogonal to the dipole, rather than the rotationally-averaged collision cross section. Note that the dipole of the protein ion is not simply that of the native conformation of the protein but may be enhanced or reduced by unfolding in the electrospray process (denaturation in organic solvent and Coulomb repulsion from multiple charges). There are two points to consider relating to dipole alignment in FAIMS:
First, what is the relative orientation of the dipole to the molecular axis with the greatest magnitude (i.e., the longest dimension of the ion)? If it is parallel, the collision cross section in the orthogonal plane will be smaller than the rotationally-averaged collision cross section and the mobility will increase (A- or B-type). If the dipole is orthogonal to the long molecular axis, the reverse is true and C-type behavior will be enhanced. The latter is difficult to prove: How can ‘normal’ C-type behavior (which cannot be predicted) be distinguished from ‘enhanced’ C-type behavior?
Second, the benefits of dipole alignment in FAIMS separation (i.e., by augmenting the differential ion mobility) are only realized when alignment occurs only during the high field segment of the waveform. If the dipole is sufficiently large (or the electric field in the low-field segment sufficiently high), dipole alignment will occur throughout the entire waveform.
Shvartsburg et al. subsequently developed a model for pendular dipole alignment of ions in gases, which considered the effect of rotational heating . The model reveals that for a particular electric field strength, a minimum dipole is required for alignment. Similarly, for a specific dipole, a maximum electric field for alignment exists. These limits arise because at some field strength value the rotational heating resulting from the electric field will counteract the dipole alignment. It was concluded that the minimum dipole required for alignment was ~450 D and that as dipole moment correlates approximately with protein mass, this corresponds to a protein mass of ~30 kDa. That conclusion was further corroborated in experiments performed on ubiquitin (8.6 kDa) and bovine serum albumin (66 kDa) by use of a miniaturized FAIMS device, which enables use of much higher electric fields than standard FAIMS devices . FAIMS analyses of myoglobin (17 kDa) and cytochrome c (12 kDa) have revealed evidence for dipole alignment for particular charge states, presumably due to some of the ensemble of protein conformations having sufficiently high dipole moment .
Analysis of Intact Proteins by FAIMS
FAIMS was first applied to protein ions by Purves and Guevremont . They showed that the charge state distribution observed for cytochrome c was dependent on FAIMS conditions. In further work on ubiquitin, they showed that for some charge states, multiple conformers could be resolved . They speculated “…that the position of the conformer in a CV spectrum is a function of its ability to change structure in response to the applied electric field”; however, if an ion changes structure during transit through the FAIMS device, the new structure may not be transmitted at the same compensation field (i.e., the ion will collide with electrodes as a result of its (new) differential mobility ). This characteristic has been termed the “self-cleaning” mechanism  and is an important consideration when discussing FAIMS of protein ions.
Protein ions exist as an ensemble of low-energy conformers separated by small barriers. Heating the protein ions results in unfolding and population of other conformational states. The high electric fields applied in FAIMS means that ions will be collisionally heated and, therefore, protein ions may be unfolded. This effect was first considered for ubiquitin and cytochrome c ions by Purves and co-workers. They showed that the increase in average temperature during one cycle of the FAIMS waveform was ~7 K for ubiquitin  and ~ 10 K for cytochrome c , concluding that changes in conformational structure should be minimal. Shvartsburg , however, argued that it is the change in maximum temperature rather than average temperature that is relevant, using the neat analogy of egg white protein: the egg white protein will become denatured by intermittent boiling separated by longer periods in cold water, regardless of the fact that the average temperature across the time period is <100 °C. Shvartsburg showed that the effect of FAIMS on ubiquitin ions was equivalent to heating them by ~50–55 K above room temperature (i.e., in agreement with the maximum temperature experienced in the FAIMS cycle).
What is the effect of field-heating of protein ions on FAIMS spectra? The conformational changes occur in short time periods near the waveform peaks. If the unfolding results in small changes in differential ion mobility, small shifts in compensation field may be observed; however, large changes during transit would result in the ion being lost because of the self-cleaning mechanism. If unfolding occurs instantaneously (within the first cycle of the waveform), the ion will not be filtered out, i.e., there will be only one relevant compensation field (that of the unfolded conformer) rather than two (pre- and post-unfolding). This rapid unfolding has been likened to annealing of ions by energetic injection into conventional ion mobility spectrometry . Field heating effects may be advantageous (e.g., improving resolution by reducing spectral complexity [49, 50]), or detrimental (e.g., by distorting conformers intended for subsequent analysis by conventional ion mobility spectrometry), to FAIMS. In the latter case, the effect can be minimized by cooling the carrier gas . Field heating is further augmented by the addition of lighter gases, such as helium, to the carrier gas [27, 28]: The increase in temperature due to field heating is inversely proportional to Ω2 , and collision cross sections are smaller in lighter gases.
One might reasonably ask what the relevance of FAIMS separated protein conformers have to nature, where the structure of a protein is inextricably linked to its function? In particular, what is the significance of the conformers revealed by enhanced field-heating? In the absence of enhanced field heating, it is possible to correlate FAIMS separated conformers with those observed in conventional ion mobility spectrometry . It is also possible to transmit noncovalent protein complexes (dimer of bovine serum albumin) through a FAIMS device . It may be that FAIMS can provide information on native conformations. Indeed, FAIMS has been applied to the analysis of conformers of the amyloidogenic protein β2-microglobulin and an acid destabilized variant, yielding results consistent with circular dichroism spectroscopy and NMR . In terms of the FAIMS field-heating annealing of protein conformers, the likely benefits are in the separation of isomers or isobars (e.g., the separation of post-translationally modified proteins or substitution variants, thus aiding top-down proteomics).
Bottom-Up Protein Analysis
Protein analysis can also be considered from the bottom-up perspective, that is, the analysis of proteolytic peptides. Peptides were first analyzed by FAIMS by Guevremont and Purves [45, 68]. They showed that small peptides exhibit A-type behavior, whereas larger peptides are C-type. Building on that work, they analyzed a tryptic digest of porcine hemoglobin . Their results showed that FAIMS reduced the chemical background, thus increasing signal-to-noise ratios. This observation is attributed to the differing optimum CFs for transmission of the background ions and the peptide ions and/or the background ions dissociating during the FAIMS separation with the fragments having a different optimum CF to the parent ion (the self-cleaning mechanism). Additionally, they showed that isobars (a singly charged ion and a doubly charged ion, both of m/z 532) could be separated . These researchers subsequently coupled FAIMS with MS/MS for the analysis of tryptic peptides . The separation of chemical background from peptide ions resulted in simplified MS/MS spectra, which were more straightforward to interpret. A more extensive study followed, in which tryptic peptides from a range of proteins and mixtures of proteins were analyzed . That study led the way for FAIMS proteomics because it considered theoretically how FAIMS might be coupled with liquid chromatography (LC), specifically the challenge of integrating the discrete packets of analytes produced by LC and the time spent scanning CF in the FAIMS device. They concluded that the most suitable approach for an LC FAIMS MS/MS experiment would be a “stepping” approach in which rather than scanning the full range of CFs, a number of discrete CF values are sampled.
An area in which FAIMS has potential to play a considerable role in the bottom-up analysis of proteins is that of isomer separation. Isomer separation by FAIMS was first demonstrated for amino acids leucine and isoleucine by Barnett et al.  More recently, Mie et al. have shown the separation of enantiomers of amino acids by addition of a chiral reference compound and FAIMS analysis of the resulting diastereomeric complexes . As well as enantiomers and diastereoisomers, in proteomics studies, isomers can arise through sequence inversions, that is, those in which the order of amino acid residues are altered, and localization variants of post-translationally modified peptides (i.e., peptides with identical sequence but differing sites of modification). Separation of the latter by FAIMS was first demonstrated by Xuan et al. for phosphopeptides with sequence APLSFRGSLPKSYVK in which the serine residues are variously modified . In that work, the traverse cylindrical FAIMS device was employed. Improved separation of those phosphopeptides was achieved by use of planar electrodes and helium-rich carrier gas . FAIMS separation of phosphopeptide isomers with adjacent modification sites and both serine and threonine phosphorylation was subsequently demonstrated . Localization variants involving other modifications can also be separated by FAIMS, including glycosylation , acetylation , and methylation . The field is now building on these initial proof-of-principle investigations and starting to exploit this benefit in addressing real biological challenges. Bridon et al. identified several isomeric phosphopeptides in their FAIMS phosphoproteomics study of insulin signaling in Drosophila melanogaster . Ulasi et al. employed LC FAIMS MS/MS to obtain a comprehensive map of glycosylation of the flagellin protein from Campylobacter jejuni . Zhao et al. have applied LC FAIMS MS/MS to study phosphorylation in human breast carcinoma cells .
FAIMS separation of peptide sequence inversions has also been demonstrated. Initially this was shown for nitrated peptides with sequence AxnYAyK (where x = 0–6; y = 6–0; nY is nitrotyrosine) by use of a planar FAIMS device and helium-rich carrier gas . A comprehensive study of this phenomenon considered a phosphopeptide library comprising the sequences GPSGXVpSXAQLX(K/R) and SXPFKXpSPLXFG(K/R), where X = ADEFGLSTVY (i.e., total number of library members = 4000, but total number of unique peptide masses = 556 because of the presence of isomers) . Analysis of the peptide library by reversed-phase (RP) LC MS/MS resulted in identification of 8% of the library; by SCX RPLC MS/MS gave 17% and by RPLC FAIMS MS/MS gave 35%. The maximum number of isomers identified for a single peptide mass by RPLC FAIMS MS/MS was 12 and for SCX RPLC MS/MS were 7 out of a possible 18. That work demonstrates an advantage of FAIMS more generally for proteomics. The separation of co-eluting species by FAIMS affords cleaner mass spectra and consequently higher scoring peptide identifications.
FAIMS has also found applications in targeted analysis of peptides. Although these studies are not strictly protein analysis, they are described in brief here. Ells et al.  showed that FAIMS mass spectrometry could be successfully applied to the analysis of microcystins—peptide hepatotoxins derived from cyanobacteria—with lower limits of detection than ESI mass spectrometry. Jemal and co-workers  and Roemer and co-workers  both developed an LC-FAIMS-selected reaction monitoring (SRM) assay for quantifying proprietary peptide drug candidates in plasma and serum. Bailly-Chouribery et al.  applied the FAIMS-SRM approach to the analysis of tryptic peptides of recombinant human erythropoietin, a glycoprotein used in horse doping. Most recently FAIMS has been applied to the detection to peptide antigens from a leukaemia cell line .
To answer the question posed by this article’s title – “To what extent is FAIMS beneficial in the analysis of proteins?” – FAIMS has the potential to be highly beneficial and this potential is starting to be realized. At present, there are three FAIMS devices commercially available: the Thermo Fisher cylindrical device, the Sciex planar device, and the Owlstone miniaturized planar device. FAIMS cannot currently be used for the direct calculation of collision-cross section but that role is fulfilled by conventional ion mobility spectrometry. The true value of FAIMS comes from its role as a separation device. For example, FAIMS is capable of separating protein conformers, which has potential benefits for top-down proteomics and structural biology. The latter is offered with the caveat that there remains a knowledge gap concerning field-heating and observed conformers. An exciting application of FAIMS is the separation of complex mixtures of proteins extracted from biological substrates, thus enabling direct surface sampling and potentially imaging of previously undetected proteins from those substrates. From a bottom-up perspective, again the advantage of FAIMS lies in its separation capabilities. FAIMS has been coupled with LC MS/MS for proteomics analyses. Several groups have shown that LC FAIMS MS/MS is complementary to the LC MS/MS approach and thus improves proteome coverage. Further benefits may derive from optimization of carrier gas and gas modifiers. In the drive towards total proteome coverage, FAIMS will provide an invaluable tool. The ability of FAIMS to separate post-translationally modified peptide isomers is particularly useful and the field is starting to move towards applying FAIMS to real world challenges.
H.J.C. is an EPSRC Established Career Fellow (EP/L023490/1).
- 13.Pringle, S.D., Giles, K., Wildgoose, J.L., Williams, J.P., Slade, S.E., Thalassinos, K., Bateman, R.H., Bowers, M.T., Scrivens, J.H.: An investigation of the mobility separation of some peptide and protein ions using a new hybrid quadrupole/traveling wave IMS/oa-ToF instrument. Int. J. Mass Spectrom. 261, 1–12 (2007)CrossRefGoogle Scholar
- 21.Purves, R.W., Barnett, D.A., Ells, B., Guevremont, R.: Investigation of bovine ubquitin conformers separated by high-field asymmetric waveform ion mobility spectrometry: cross section measurements using energy-loss experiments with a triple quadrupole mass spectrometer. J. Am. Soc. Mass Spectrom. 11, 738–745 (2000)CrossRefGoogle Scholar
- 57.Li, J.W., Traszka, J.A., Counterman, A.E., Clemmer, D.E.: Influence of solvent composition and capillary temperature on the conformations of electrosprayed ions: unfolding of compact ubiquitin conformers from pseudonative and denatured solutions. Int. J. Mass Spectrom. 185, 37–47 (1999)CrossRefGoogle Scholar
- 65.Borysik, A.J.H., Read, P., Little, D.R., Bateman, R.H., Radford, S.E., Ashcroft, A.E.: Separation of b2-microglobulin conformers by high-field asymmetric waveform ion mobility spectrometry (FAIMS) coupled to electrospray ionisation mass spectrometry. Rapid Commun. Mass Spectrom. 18, 2229–2234 (2004)CrossRefGoogle Scholar
- 66.Sarsby, J., Griffiths, R.L., Race, A.M., Bunch, J., Randall, E.C., Creese, A.J., Cooper, H.J.: Liquid extraction surface analysis mass spectrometry coupled with field asymmetric waveform ion mobility spectrometry for analysis of intact proteins from biological substrates. Anal. Chem. 87, 6794–6800 (2015)CrossRefGoogle Scholar
- 70.Barnett, D.A., Ding, L., Ells, B., Purves, R.W., Guevremont, R.: Tandem mass spectra of tryptic peptides at signal-to-background ratios approaching unity using electrospray ionization high-field asymmetric waveform ion mobility spectrometry/hybrid quadrupole time-of-flight mass spectrometry. Rapid Commun. Mass Spectrom. 16, 676–680 (2002)CrossRefGoogle Scholar
- 77.Bridon, G., Bonneil, E., Muratore-Schroeder, T., Caron-Lizotte, O., Thibault, P.: Improvement of phosphoproteome analyses using FAIMS and decision tree fragmentation. Application to the insulin signaling pathway in Drosophila melanogaster S2 cells. J. Proteome Res. 11, 927–940 (2012)CrossRefGoogle Scholar
- 78.Swearingen, K.E., Hoopmann, M.R., Johnson, R.S., Saleem, R.A., Aitchison, J.D., and Moritz, R.L.: Nanospray FAIMS fractionation provides significant increases in proteome coverage of unfractionated complex protein digests. Mol. Cell Proteom. 11, doi: 10.1074/mcp.M111.014985 (2011)
- 79.Zhao, H., Cunningham, D.L., Creese, A.J., Heath, J.K., Cooper, H.J.: FAIMS and phosphoproteomics of fibroblast growth factor signaling: enhanced identification of multiply phopshorylated peptides. J. Proteome Res. 14(12), 5077–5087 (2015)Google Scholar
- 81.Mie, A., Jornten-Karlsson, M., Axelsson, B.-O., Ray, A., Reimann, C.T.: Enantiomer separation of amino acids by complexation with chiral reference compounds and high-field asymmetric waveform ion mobilty spectrometry: preliminary results and possible limitations. Anal. Chem. 79, 2850–2858 (2007)CrossRefGoogle Scholar
- 82.Xuan, Y., Creese, A.J., Horner, J.A., Cooper, H.J.: Separation of isobaric phosphopeptides by high field asymmetric waveform ion mobility spectrometry confirmed by high resolution electron transfer dissociation mass spectrometry. Rapid Commun. Mass Spectrom. 23, 1963–1969 (2009)CrossRefGoogle Scholar
- 94.Bailly-Chouribery, L., Cormant, F., Garcia, P., Lonnberg, M., Szwandt, S., Bondesson, U., Popot, M.-A., Bonnaire, Y.: A new analytical method based on anti-EPO monolith column and LC-FAIMS-MS/MS for the detection of rHuEPOs in horse plasma and urine samples. Analyst 137, 2445–2453 (2012)CrossRefGoogle Scholar
Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.