Alkali Metal Cationization of Tumor-associated Antigen Peptides for Improved Dissociation and Measurement by Differential Ion Mobility-Mass Spectrometry

Abstract

Tandem mass spectrometry (MS/MS) is a highly sensitive and selective method for detection of tumor associated peptide antigens. These short, non-tryptic sequences may lack basic residues, resulting in formation of predominantly [peptide+H]⁺ ions in electrospray. These singly charged ions tend to undergo inefficient dissociation, leading to issues in sequence determination. Addition of alkali metal salts to the electrospray solvent can drive formation of [peptide+H+metal]²⁺ ions that have enhanced dissociation characteristics relative to [peptide+H]⁺ ions. Both previously identified tumor-associated antigens and predicted neoantigen sequences were investigated. The previously reported rearrangement mechanism in MS/MS of sodium cationized peptides is applied here to demonstrate complete C-terminal sequencing of tumor associated peptide antigens. Differential ion mobility spectrometry (DIMS) is shown to selectively enrich [peptide+H+metal]²⁺ species by filtering out singly charged interferences at relatively low field strengths, offsetting the decrease in signal intensity associated with the use of alkali metal cations.

Graphical Abstract

INTRODUCTION

Tumor-associated or specific peptide antigens, a class of 8–12 residue immunopeptides that are presented on the cell surface by the major histocompatibility complex (MHC), are promising immunotherapy targets.^1–3 Sequences of tumor peptide antigens can be predicted through genomics and confirmed by mass spectrometry, enabling the development of highly specific T-cell therapies.^4–10 Although improvements in genomics and related bioinformatics techniques have enabled these predictive models, confirming the presence of tumor peptide antigens in individual patients requires the continued improvement of mass spectrometric methods. The principal method for identification of peptides, including immunopeptides, by mass spectrometry is matching peptide sequence information obtained by a tandem mass spectrometry (MS/MS) experiment to databases containing relevant protein sequences (e.g., UniProt).^11–14 Peptide identifications rely on effective dissociation to provide sequence information of the target sequences. Immunopeptides are uniquely challenging in this regard, as they do not necessarily contain the C-terminal basic residue (Lys or Arg) that is present in tryptic peptides. This lack of a characteristic basic residue results in approximately 30% of immunopeptides being detected as [M+H]⁺ ions rather than multiply-charged species in standard LC-ESI-MS/MS approaches.^15–17

In electrospray ionization (ESI), the most widely used method for peptide ionization, the charge state of a given ion is dependent primarily on the gas-phase basicity of possible charge sites and the excess charge conditions near the end of droplet lifetime.^18–20 For small peptides, the number of basic residues generally dictates the charge state of the resulting peptide ion. Singly charged protonated peptides are known to undergo less comprehensive dissociation by the low energy collision-based techniques that dominate proteomic/peptidomic analyses, leading to ambiguity in sequence determination.^21,22 Methods have been developed using alkali metal cations (e.g., Na⁺, Li⁺) in electrospray of peptides to form [M+Na]⁺ or [M+Li]⁺ ions rather than [M+H]⁺ ions. These alkali metal cationized peptides undergo dissociation that favors the formation of y-type ions, which can be complementary to the b-type ions (N-terminus inclusive) detected from dissociation of the [M+H]⁺ ion of the same peptide.²³ However, the presence of these alkali metal cations divides the signal intensity of a given peptide across multiple species. In the analysis of peptides that readily form multiply protonated species (e.g., tryptic peptides) the presence of alkali metal cations is a significant disadvantage in terms of signal intensity and usually not an advantage in terms of sequence coverage. This has led to the development of strategies to limit the abundance of alkali metal cationized peptides, either by solution-phase preparatory techniques or gas-phase chemistries to deplete metal cations.^24–26

However, in the case of immunopeptides, the use of alkali metal cations can provide the sequence coverage necessary to make confident identifications. We explore two methods for the sequencing of tumor peptide antigens that lack basic residues: multiple-stage tandem mass spectrometry (MSⁿ) of [M+metal]⁺ species and MS/MS of [M+H+metal]²⁺ species, with lithium, sodium, and potassium as the tested metals. The former was reported early in the development of proteomics, where MS/MS of [M+metal]⁺ peptides yields a rearrangement product ion that structurally resembles a peptide that has been enzymatically cleaved on the C-terminal side of the penultimate residue, and it retains the metal cation.^24,25 This rearrangement product, [b_n-1+metal+OH]⁺ (n = total number of residues in the peptide), forms with high efficiency and can be collisionally-activated again for formation of the rearrangement product with the next residue removed from the C-terminus. Repeated collisional-activation (MSⁿ) can be used to obtain complete sequence information of a peptide.^29,30 Longer peptides can be challenging to fully sequence due to the high number of MS stages required, though up to MS⁸ has been reported for sequencing of a sodiated decapeptide.³⁰ Proposed mechanisms suggest that the alkali metal cation interacts with the C-terminal COOH and the carbonyl oxygen of the penultimate residue prior to activation.^31,32

The formation and use of [M+H+metal]²⁺ peptide species for enhanced sequence coverage by collision-based dissociation was reported later and has been discussed less extensively. The first report of [M+H+Na]²⁺ investigated applications to non-tryptic peptides and found that the sodium-adduct directed fragmentation towards formation of [y_n+Na]⁺ ions.³³ The [M+H+Na]²⁺ species was also investigated in the context of singly charged tryptic peptides, reporting that collisional activation yields alkali metal cationized b-ions and protonated y-ions based on interaction of the alkali metal cation near the N-terminus.³⁴ Here we use two previously reported tumor-associated peptide antigens, FLLPTGAEA and VLDIEQFSV, and predicted leukemia spliceosome mutant sequences (neoantigens) as model sequences.^35,36 We observe key differences in the dissociation behavior of [M+H+metal]²⁺ species of these peptide antigen sequences relative to previous studies of non-tryptic peptides. Specifically, dissociation of the 8–11 residue sequences yields large alkali metal cationized b and y ions and small protonated b and y ions. Comparatively, dissociation of the [M+H]⁺ species of these same peptide antigen sequences yields large protonated b ions, often only providing information on 3 or 4 residues. Thus, collisional activation of [M+H+metal]²⁺ tumor peptide antigens provides substantially improved sequence coverage relative to activation of the [M+H]⁺ species.

The improvement in sequence coverage comes at the cost of signal intensity, as the signal for a given peptide is split between protonated and alkali metal cationized species. However it should be noted that because of salts in samples, often there is a significant signal for alkali metal cationized species without any further addition of alkali metals. The disadvantage of the signal being split across multiple peaks, whether naturally occurring or due to addition of alkali metal salts can be mitigated by coupling differential ion mobility spectrometry (DIMS) to the mass spectrometer. Typically implemented at atmospheric pressure between the ionization source and mass spectrometer inlet, DIMS separates ions based on the difference between their high and low electric field mobilities.³⁷ Unlike drift tube ion mobility, which separates packets of ions by the time required for each species to transit a given distance against an inert gas in a constant electric field, DIMS utilizes a combination of electric fields to transmit only ions of a given differential mobility until the field is changed. This functionality makes DIMS more analogous to a mobility-based quadrupole rather than a gas-phase chromatographic separation. As the separation is not time-dependent, DIMS is readily coupled to all mass analyzers.

Beyond providing a separation that is orthogonal to mass analysis, DIMS can dramatically improve signal-to-noise by filtering out interfering ions before they are able to enter the mass spectrometer. Based on these advantages, we have successfully applied DIMS-MS/MS to the measurement and discovery of tumor antigens.^36,38 Also referred to as field asymmetric ion mobility spectrometry (FAIMS) in the literature, this ion mobility technology has seen a recent increase in popularity for proteomics applications.^39–41 Here we show that the target [M+H+metal]²⁺ species for a given peptide can be selectively transmitted to the mass spectrometer. Though DIMS does not impact the sequence coverage improvements that are the primary focus of the method demonstrated herein, its ability to enhance the signal-to-noise of MS measurement may enable more widespread adoption of a metal cationization strategy.

MATERIALS AND METHODS

Samples

Peptides were selected from leukemia splice variant sequence mutations based on binding affinity and immunogenicity (as predicted by NetMHC). The peptides used in this work were filtered to include only those that lack basic residues. These predicted peptide antigens and previously identified tumor associated peptide antigens were purchased from Bio-Synthesis (Lewisville, TX). Methanol (LC-MS grade), water (LC-MS grade), and formic acid (LC-MS grade) were purchased from Fischer Scientific (Fairlawn, NJ). Lithium acetate was purchased from Sigma-Aldrich (St. Louis, MO). Peptides were diluted to 1 μM in 60:40 methanol:water with 0.1% formic acid for direct infusion experiments. Lithium acetate, sodium acetate, or potassium acetate were added to the electrospray solution at 100 μM for generation of their respective [M+metal]⁺ and [M+H+metal]²⁺ species. Peptides were diluted in water to 100 ng/mL per peptide for liquid chromatography experiments.

Instrumentation

Experiments were performed on a Bruker HCT ion trap mass spectrometer in the positive ion mode. Direct infusion electrospray ionization was performed with a 2 μL/min flow rate and a capillary voltage of −4.5 kV. Collision-induced dissociation was used for all MS/MS experiments with a 40 ms activation time and low mass cutoff of 27% of the precursor ion m/z. MS/MS spectra were manually interpreted. The nitrogen drying gas was set to 5 L/min and 300°C. A custom-built differential ion mobility spectrometer was placed on the inlet capillary of the mass spectrometer using the nitrogen drying gas as the ion mobility carrier gas.³⁸ A schematic of the DIMS can be found in the Supporting Information (Figure 1). An Agilent 1260 Infinity II HPLC was used for liquid chromatography experiments, with mobile phase A as water (Optima grade) and mobile phase B as methanol (Optima grade), both containing 100 μM lithium acetate and 0.1% formic acid.

RESULTS AND DISCUSSION

Collision-induced dissociation of [M+H+Li]²⁺ and [M+Li]⁺ of confirmed tumor-associated peptides

The sequence coverage obtained by MS/MS of [M+H]⁺ was compared to the sequence coverage obtained using the [M+H+Li]²⁺ and [M+Li]⁺ precursor ions for the previously identified peptide antigens FLLPTGAEA and VLDIEQFSV in Figure 1.

Figure 1:

A) MS/MS spectra of FLLPTGAEA and VLDIEQFSV as [M+H]⁺, [M+Li]⁺, and [M+H+Li]²⁺ precursors. Blue text is used to label rearrangement product ions. Blue arrows represent losses of 18 Daltons (corresponding to neutral loss of H₂O) from a given product ion. Ions are singly charged unless otherwise noted. B) Sequence ions observed from dissociation of the [M+H]⁺ (green triangles), [M+Li]⁺ (red circles), and [M+H+Li]²⁺ (blue squares) species of FLLPTGAEA and VLDIEQFSV.

The [M+H]⁺ ion of FLLPTGAEA preferentially dissociates to the y₆ ion, which is expected based on the location of the proline residue and the established proline effect in CID.^42,43 The rearrangement product [b₈+Li+OH]⁺ was the most abundant product ion in CID of the [M+Li]⁺ species, with [b₈+Li]⁺ detected at 64% of the intensity of the rearrangement product. Further activation (MS³) of the [b₈+Li+OH]⁺ product ion can provide additional sequence coverage relative to MS² of the [M+Li]⁺ species and will be discussed in a later section. The rearrangement mechanism requires the alkali metal cation to interact with the C-terminal COOH while formation of the [b₈+Li]⁺ requires the loss of the C-terminal COOH group. The presence of both ions at high relative abundances in the same MS/MS spectra suggests that there are multiple gas-phase structures of [FLLPTGAEA+Li]⁺. The greatest sequence coverage is observed in dissociation of [FLLPTGAEA+H+Li]²⁺, with detection of larger (4 or more residues) lithium cationized b and y ions and smaller (3 or fewer residues) protonated b and y ions. Notably, the proline in the fourth position of FLLPTGAEA has less of an influence on the product ion distribution from MS/MS of [M+H+Li]²⁺ than in MS/MS of the [M+H]⁺ species, as the base peak in the MS/MS spectrum of [M+H+Li]²⁺ is a lithium cationized y₇ ion. Also related to this proline residue, the improvement in sequence coverage observed in dissociation of the [M+H+Li]²⁺ species relative to the [M+H]+ species is less pronounced in FLLPTGAEA than other sequences used in this work. This proline provides efficient dissociation nearer to the N-terminus of FLLPTGAEA than is typical in MS/MS of immunopeptide [M+H]⁺ species.

The extent of dissociation of the [M+H+Li]²⁺ of VLDIEQFSV exemplifies the improvement in sequence coverage obtainable with the use of alkali metal cations in the analysis of tumor peptides. While the dissociation of the [M+Li]⁺ species is more extensive than that of [M+H]⁺, it is still less complete than what is obtained by collisional activation of the [M+H+Li]²⁺ of VLDIEQFSV. One potential explanation for the improvement in dissociation of the [M+H+Li]²⁺ species is the additional charge repulsion between the protonated and alkali metal cationized sites on the peptide relative to the singly charged species. The additional charge repulsion could enable more comprehensive dissociation by effectively lowering the critical energy for backbone dissociation. Another explanation for the improvement in sequence coverage is the ability to detect b/y ion pairs from [M+H+Li]²⁺ because both sides of the peptide can retain a positive charge.

Figure 1B is a comparison of the three ion types in regards to sequence coverage rather than all detected product ions. This presentation of results is more easily interpreted in terms of sequence coverage. Importantly, the detection of the [b₈+Li+OH]⁺ and [b₇+Li+OH]⁺ rearrangement product ions are not noted separately from the detection of the [b₈+Li]⁺ and [b₇+Li]⁺ product ions in the dissociation of [FLLPTGAEA+Li]⁺ as they provide the same level of information in terms of sequence coverage.

Collision-induced Dissociation of [M+H+Li]²⁺ and [M+Li]⁺ of Predicted Neoantigens

Both acute myeloid and chronic lymphocytic leukemia (AML and CLL) have frequent mutations in spliceosome associated genes, and the effect of those mutations on mRNA splicing were used in combination with prediction of binding affinity to the MHC to propose neoantigen target sequences.^44–46 Many of these predicted epitopes lack a basic residue, thus preferentially forming [M+H]⁺ species in electrospray ionization. The often inadequate dissociation of [M+H]⁺ peptides results in an undesirable bias towards detection of tumor peptide antigens that contain basic residues rather than towards the most likely candidates based on biological criteria like the binding affinity to the MHC or likelihood of a given mutation. The extent of fragmentation for a selection of the predicted neoantigens that lack basic residues is shown in Figure 2. For clarity, Figure 2 does not differentiate between lithium cationized and protonated product ions or between singly and doubly charged product ions, as all of these product ion types are similarly informative regarding peptide sequences. Tabulated results including the specific product ions detected from dissociation of each ion type for the predicted sequences are provided in Supporting Table 1.

Figure 2:

Product ions observed from dissociation of the [M+H]⁺ (green triangles), [M+Li]⁺ (red circles), and [M+H+Li]²⁺ (blue squares) species of predicted neoantigen sequences from leukemia splice variant mutations.

The extent of dissociation observed for the [M+H]⁺, [M+Li]⁺, and [M+H+Li]²⁺ species for the predicted neoantigens follows the same trend as the previously identified sequences. The singly protonated peptides tend to form large (5–8 residue) b-ions based on dissociation near the C-terminus. This is especially true for sequences containing a penultimate proline residue (e.g., ILGITSLPL). The preference for N-terminal cleavage relative to the position of proline results in highly abundant b_n-2 ions and limits detection of other product ions for these sequences. The deleterious effect of a penultimate proline on peptide dissociation is a potential cause for the bias against MS detection of neoantigens containing this motif relative to predictions based on MHC binding affinity.^47,48 Dissociation of the [M+H+Li]²⁺ species provides MS/MS spectra containing the greatest number of product ions, with generation of both b-type and y-type ions. From [M+H+Li]²⁺, the larger product ions (5–8 residues) tend to be lithiated, whereas the smaller product ion (2–4 residues) tend to be protonated. This tendency is consistent regardless of the terminus that the product ions contain. In particular, the large y-ions generated through dissociation of the [M+H+Li]²⁺ species provide sequence information that is unique relative to dissociation of the [M+H]⁺ species, highlighting the usefulness of alkali metal cationization of these immunopeptides that lack basic residues. Dissociation of the [M+Li]⁺ species is the most variable in terms of final sequence coverage, though all of the predicted neoantigens favor the formation of the [b_n-1+Li+OH]⁺ rearrangement product ion. Only backbone cleavage C-terminal to acidic residues (Asp, Glu) is competitive in terms of relative abundance with the rearrangement product ion in the dissociation of [M+Li]⁺, with other product ions at significantly lower abundances. The preference for cleavage C-terminal to acidic residues is expected based on a salt-bridge mechanism previously reported in the dissociation of [M+Na]⁺ species, where the alkali metal cation stabilizes proton transfer from the Asp sidechain to the backbone amide prior to dissociation.⁴⁹

The results from dissociation of these predicted neoantigens suggests a potential reason for the improved coverage in dissociation of [M+H+Li]²⁺ relative to [M+H]⁺ or [M+Li]⁺ is the reduced bias towards proline, aspartic acid, and glutamic acid-directed cleavages or formation of the [b_n-1+Li+OH]⁺ rearrangement product ion. This reduced bias and demonstration of improved coverage makes the [M+H+Li]²⁺ species a promising target for detection of tumor peptides that lack basic residues.

Effect of Alkali Metal Cation on Dissociation of [M+metal]⁺ and [M+H+metal]²⁺

The dissociation behavior of the peptide-alkali metal complex was further explored using lithium, sodium, and potassium adducts. Product ion intensities from dissociation of singly and doubly charged peptides as lithium, sodium, and potassium adducts are shown in Figure 3.

Figure 3:

A) Comparison of product ion intensities from dissociation of the [M+metal]⁺ species using lithium, sodium, and potassium as alkali metal cations. B) Comparison of product ion intensities from dissociation of the [M+H+metal]²⁺ species using lithium, sodium, and potassium as alkali metal cations.

Dissociation of the singly charged [M+metal]⁺ species always results in an abundant rearrangement product ion, with acidic residue directed cleavage as the most competitive pathways. As the size of the metal cation increases, the abundance of the rearrangement product decreases relative to the abundance of acidic directed (both Asp and Glu) cleavages. This is observed in VLDIEQFSV, YVVTDQIPV, and TIVEGILEV, though the relative abundances of the acidic cleavage to the rearrangement product ion within a given sequence is different. For example, [VLDIEQFSV+Li]⁺ dissociates into the y₆ and rearrangement product ion with very similar abundance, whereas the rearrangement product ion from [YVVTDQIPV+Li]⁺ is more than three times the abundance of the b₅ ion. This may be related to the residues that surround a given acidic residue, or the residues near the C-terminus. The rearrangement product ion is the only major peak detected in dissociation of sequences that lack acidic residues, as shown in the [YIIMFWPV+metal]⁺ data for all three alkali metals tested.

Dissociation of the doubly charged [M+H+Li]²⁺ and [M+H+Na]²⁺ species primarily result in backbone dissociation that is not related to the C-terminal rearrangement or acidic directed cleavages. Only the [M+H+K]²⁺ species undergo the rearrangement, with the abundance of the rearrangement product ions dominating other sequence ions. The minimal differences in dissociation of the [M+H+Li]²⁺ and [M+H+Na]²⁺ species supports the idea that the nature of the metal interactions is similar between lithium and sodium, whereas potassium either binds to a different location on the peptide or is able to support a mechanism that the smaller metals cannot. The similarity of dissociation in [M+H+Li]²⁺ and [M+H+Na]²⁺ suggests that either metal is suitable for use in a method targeting [M+H+metal]²⁺ species using MS/MS. However, [M+Li]⁺ is preferable for targeting repeated collisional-activation of the [b_n-1+metal+OH]⁺ ion, as [M+Na]⁺ slightly favors acidic-residue directed cleavage relative the rearrangement product ion.

Multi-stage mass spectrometry (MSⁿ) of tumor-associated peptides

As reported previously, MS/MS of the rearrangement product ions can be repeated to sequence a peptide in a step-wise manner.³⁰ Here we demonstrate this method can be applied to alkali metal cationized tumor associated peptide antigens. The results of multi-stage mass spectrometry for full C-terminal sequencing of the predicted neoantigen sequence YIIMFWPV is shown in Figure 4.

Figure 4:

Multi-stage tandem mass spectrometry (up to MS⁸) of the predicted splice variant neoantigen sequence YIIMFWPV. The rearrangement product ion [b_n-1+Li+OH]⁺ was re-isolated and activated in each step of MS.

The use of MS² to MS⁸ in the dissociation of the 9-residue peptide YIIMFWPV enables detection of each rearrangement product ion, [b_n-1+Li+OH]⁺, starting from the C-terminus until lithiated tyrosine is detected as [b₁+Li+OH]⁺. Although the ability to fully sequence a peptide through MSⁿ (where n is equal to the total number of residues the peptide) is powerful, the reduction in signal intensity on each step of MS/MS requires relatively high starting abundances. The MS² to MS⁸ experiment shown here was performed with 1 μM YIIMFWPV, and the signal intensity at the MS⁸ is low enough to suggest that lower peptide concentrations would be limited to fewer MS stages. For targeted detection of genomically-predicted epitopes, this limitation could be partially mitigated by use of a multiple frequency activation waveform that would activate each of the rearrangement product ions from the same initial packet of isolated ions without requiring additional isolation and activation steps. However, the conversion efficiency to each subsequent rearrangement product will be the ultimate constraint, as this is not limited by the number of MS/MS steps but rather by competition of other dissociation pathways. For example, in the MS⁶ of [b₄+Li+OH]⁺ of YIIMFWPV, the [a₄+Li]⁺, [b₃+Li]⁺, and [a₃+Li]⁺ product ions are detected in high abundance with the [b₄+Li+OH]⁺ rearrangement product ion. The time and signal require to take advantage of this rearrangement strategy is likely not possible with traditional shotgun proteomics approaches, where the goal is to obtain information on the largest number of peptides from a small sample. However, the unique advantage of complete C-terminal sequencing with this strategy may be worthwhile in targeted applications where traditional approaches fail to provide adequate sequence information.

DIMS for Reduced Chemical Background

The primary disadvantage of adding alkali metals to the electrospray solvent is that signal for a given peptide is split between protonated and alkali metal cationized species. This splitting of signal can limit ability to detect the target peptides. To offset this limitation, DIMS was implemented between the electrospray source and the mass spectrometer inlet capillary. Typical operation of DIMS uses a constant E_D and changes E_C to transmit target species, with E_D selected to provide the separation needed for a given application. A relatively low (24 kV/cm) E_D provides separation between the target [M+H+Li]²⁺ species and the [M+H]⁺ and other singly charged interferences without significant separation between [M+H+Li]²⁺ species of different peptides. This allows E_C to be held constant during a LC gradient rather than cycling through a number of values. Increasing E_D improves the separation between [M+H+Li]²⁺ species at the cost of ion transmission, which is characteristic of DIMS. A comparison of chromatograms collected with and without the use of DIMS after injection of 5 pg per peptide on column using [M+H+Li]²⁺ species as targets are shown in Figure 5.

Figure 5:

A) Extracted ion chromatograms with DIMS off (left) and DIMS on (right) of [M+H+Li]²⁺ species of FLLPTGAEA, VTDQIPVFV, and VLDIEQFSV shown in red, black, and blue traces, respectively. B) Mass spectra at the retention times of FLLPTGAEA, VTDQIPVFV, and VLDIEQFSV with and without use of DIMS. DIMS was operated with E_D = 24 kV/cm and E_C = 145 V/cm.

The reduced noise in the extracted ion chromatograms for each peptide are explained by the dramatic decrease in chemical noise observed in the related mass spectra. As the [M+H+Li]²⁺ species are the preferred targets for MS/MS based on their enhanced dissociation characteristics, the improvement in signal-to-noise makes DIMS an important addition to a method that utilizes alkali metal cationization.

CONCLUSIONS

The use of alkali metal cationization provides a significant improvement in sequence coverage during low-energy collisional activation of tumor associated peptide antigens that lack basic residues. This improvement in dissociation enables more confident identifications of peptide sequences that are predicted to occur in known mutations of specific tumor types. This method also serves to lessen the bias in peptide detection based on mass spectrometric performance, as peptides that undergo more effective dissociation are not necessarily the same as those with the highest binding affinity to the MHC. As mentioned previously, the presence of alkali metals is often undesirable in electrospray, as it results in the splitting of analyte signal across multiple peaks. However, as confident peptide identifications rely heavily on achieving effective dissociation, we suggest that the splitting of analyte signal across multiple peaks is worthwhile in many cases. Also, this problem can be mitigated by the use of differential ion mobility spectrometry, which allows for selective transmission and thereby enhanced signal-to-noise in the measurement of [M+H+metal]²⁺ species.

Though the data provided in this manuscript demonstrates the fundamental advantages of a metal cationization strategy, further development is necessary before it can be adopted as an alternative to traditional proteomics approaches. Comparison of this strategy to a traditional antigen discovery strategy with larger samples sizes of peptides isolated from relevant biological sources will require modifications to current MS/MS spectral database. Databases and search algorithms are currently targeted towards protonated ions, and would need to be modified to include alkali metal cationized product ions. As tumor associated peptide antigens remain challenging to detect by traditional MS methods, continued development of methods in this category is necessary for translation of genomics tools to clinical use.

Supplementary Material

Supplementary Information

Supporting Figure 1: Schematic of custom-built differential ion mobility spectrometer.

Supporting Table 1: Product ions from all ion types of predicted neoantigens