B American Society for Mass Spectrometry, 2017 J. Am. Soc. Mass Spectrom. (2017) 28:1203Y1215DOI: 10.1007/s13361-017-1635-x

RESEARCH ARTICLE

Defining Gas-Phase Fragmentation Propensitiesof Intact Proteins During Native Top-Down MassSpectrometry

Nicole A. Haverland, Owen S. Skinner, Ryan T. Fellers, Areeba A. Tariq, Bryan P. Early,Richard D. LeDuc, Luca Fornelli, Philip D. Compton, Neil L. KelleherDepartment of Chemistry and Molecular Biosciences and the Proteomics Center of Excellence, Northwestern University, 2170Campus Drive, Evanston, IL 60208, USA

more directed

D GW A

F Wfragmentation

9+

Denatured vs. NativeTop-down MS

16+

D P Abstract. Fragmentation of intact proteins in the gas phase is influenced by aminoacid composition, the mass and charge of precursor ions, higher order structure, andthe dissociation technique used. The likelihood of fragmentation occurring between apair of residues is referred to as the fragmentation propensity and is calculated bydividing the total number of assigned fragmentation events by the total number ofpossible fragmentation events for each residue pair. Here, we describe generalfragmentation propensities when performing top-down mass spectrometry (TDMS)using denaturing or native electrospray ionization. A total of 5311 matched fragmen-tation sites were collected for 131 proteoforms that were analyzed over 165 exper-iments using native top-downmass spectrometry (nTDMS). These datawere used to

determine the fragmentation propensities for 399 residue pairs. In comparison to denatured top-down massspectrometry (dTDMS), the fragmentation pathways occurring either N-terminal to proline or C-terminal toaspartic acid were even more enhanced in nTDMS compared with other residues. More generally, 257/399(64%) of the fragmentation propensities were significantly altered (P ≤ 0.05) when using nTDMS compared withdTDMS, and of these, 123 were altered by 2-fold or greater. The most notable enhancements of fragmentationpropensities for TDMS in native versus denatured mode occurred (1) C-terminal to aspartic acid, (2) betweenphenylalanine and tryptophan (F|W), and (3) between tryptophan and alanine (W|A). The fragmentation propen-sities presented here will be of high value in the development of tailored scoring systems used in nTDMS of bothintact proteins and protein complexes.Keywords: Native mass spectrometry, Top-down mass spectrometry, Residue fragmentation propensity, Frag-mentation propensity, Native electrospray ionization, Native ESI, Tandem mass spectrometry

Received: 28 November 2016/Revised: 17 February 2017/Accepted: 20 February 2017/Published Online: 3 April 2017

Introduction

The analysis of intact proteins via Btop-down^ proteomics[1, 2] has grown in popularity in recent years and has been

used to show proteomic differences associated with senescence[3–5], a model of bacterial infection [6], and myocardial in-farction [7]. A unique advantage of top-down proteomics is theaccurate measurement of the mass of an intact proteoform. The

term Bproteoform^ describes the biological variability for aprotein by taking into account gene or protein processingevents (e.g., mutations, post-translational modifications) [8].To fully characterize the proteoform and localize the modifi-cations, it is necessary to controllably disassemble protein ionsin the gas phase using tandem mass spectrometry (MS/MS).

Since the first correlation of gaseous protein fragment ions toprimary sequence in 1990 [9], MS/MS has been widely used toenable mass spectrometry-based analysis of peptides [10–12]and proteins [13, 14]. The process requires the measurement ofthe intact mass of the precursor, which is then isolated in the gasphase and activated to break its inter-residue bonds. Manymethods for activation and fragmentation are available(reviewed in [15, 16]), but one of the most common is via

Electronic supplementary material The online version of this article (doi:10.1007/s13361-017-1635-x) contains supplementary material, which is availableto authorized users.

Correspondence to: Neil L. Kelleher; e-mail: n–kelleher@northwestern.edu

energetic collisions with a neutral gas, which is termedcollisionally-induced dissociation (CID). Variations of CID ex-ist, with the two major types being (1) ion trap CID, which usesresonant excitation to excite ions and potentiate radial movementthat slowly dissociates trapped ions after many collisions, or (2)beam-style CID, which is performed using higher-energy axialacceleration to rapidly activate and dissociate the ions. Whileboth of these techniques produce b- and y-type protein backbonecleavages [17], each can produce a different set of cleavageproducts [18, 19]. As opposed to CID, electron-based and ultra-violet photon-based fragmentation techniques rely on muchfaster processes, breaking different backbone bonds and produc-ing potentially more than 10 different fragment types [20–22].Regardless of the method used for dissociation, the fragmentions can be used to determine the residue fragmentation propen-sities, which are described here as the chance that a fragmenta-tion event will occur between any given pair of residues.

The efforts that have been put forth to analytically describethe gas-phase residue fragmentation propensities for peptidesusing a variety of fragmentation techniques have contributed tothe development of the mobile proton model [23] (reviewed in[24, 25]). This model describes fragmentation of peptides in thegas phase with considerations for residue composition [23, 26–28], gas-phase basicity [23, 28–30], secondary structure [30],relative location of the basic residue [31, 32], and the theoret-ical mobility of the proton (e.g., sequestered versus mobile)[23, 32]. For instance, following ionization, a peptide that has agreater number of charges than the number of residues withhigh gas-phase basicity (e.g., arginine [33]) will have protonsreadily mobilize to amide bonds to induce backbone fragmen-tation [23]. The energy requirements for fragmentation ofpeptides with mobile protons are relatively low [23, 28, 29].In contrast, if the peptide has fewer charges than the number ofbasic residues, the proton(s) is sequestered to the basic resi-due(s) and the activation energy needed to mobilize the protonand induce backbone fragmentation is increased substantially[23, 28–30, 34]. The conventional fragmentation pathways thatare best described by the mobile proton model are grouped intofour major clusters, including fragmentation occurring (i) N-terminal to proline (X|P), (ii) C-terminal to isoleucine, leucine,or valine (I/L/V|X), (iii) C-terminal to aspartic acid or glutamicacid (D/E|X), and (iv) Bb&y^ fragmentation [31, 32]. ThisBb&y^ fragmentation pathway generally includes ions witharginine near the N-terminus (e.g., in the case of missed trypticcleavages) or internal fragment ions (ions that form by re-fragmentation of a b- or y-type ion) [31, 32, 35]. Fragmentationthrough these four major channels has also been investigatedfor intact proteins analyzed under denaturing conditions [35]. Itis important to note that development and application of thismodel primarily used peptides and proteins that were analyzedusing acidified, denatured electrospray ionization (dESI) con-ditions. Although it is well established that the ESI solvent canaffect peptide and protein conformation and protonation [36–38], the extension of the mobile proton model as applied to theanalysis of intact proteins and complexes using native ESI(nESI) has had limited exploration.

Beyond the improved understanding of the general trends offragmentation for whole protein cations, characterization ofresidue fragmentation propensities for different protein activa-tion types can also be used as priors to enhance protein iden-tification [39] and characterization of proteoforms [40] in top-down proteomics. However, since the vast majority ofMS/MS-based top-down proteomics has been performed on denaturedproteins, the scoring systems used for identification and char-acterization of proteoforms have naturally been developed andtested using datasets derived from denatured top-down massspectrometry (dTDMS) experiments. Native top-down massspectrometry (nTDMS) approaches are proliferating becausenESI can preserve many native properties of proteins andprotein complexes, including aspects of structure [41, 42],interactions [41, 43–47], and activity [41, 48] (these topicsare also reviewed in [49–51]). Further, nTDMS has an addi-tional benefit of increased signal-to-noise due to fewer overallcharge states [52]. All in all, nTDMS is readily becoming apowerful technique for the analysis of intact proteins. In fact,the Kaltashov and Ge groups have recently made advance-ments in on-line separations coupled with native mass spec-trometry for the analysis of native proteins and complexes [53–55], which is an important step in establishing nTDMS as atechnique capable of supporting discovery mode studies. How-ever, a scoring system will need to be optimized for nTDMS asthe fragmentation patterns for native proteins appear to bealtered compared to those observed for denatured proteins[33, 37, 54, 56–58]. An essential step for optimizing thesescoring systems is to better define the differences in fragmen-tation propensity when analyzing proteins using nTDMS ordTDMS. Here, we provide the first statistically-backed analysisof the differences in residue fragmentation propensities be-tween intact, endogenous proteins analyzed with nTDMS ascompared to dTDMS.

ExperimentalSample Preparation

Four human cell lines were obtained from the German Collec-tion of Microorganisms and Cell Cultures GmbH (www.dsmz.de) or the American Type Culture Collection (www.atcc.org),including Hg-3 (DSMZACC 765), Ramos (ATCCCRL-1596), Jurkat (ATCC TIB-152), and HEK-293 T (ATCC CRL-11268) cell lines. Cells were grown following manufacturerrecommended culture methods using 150 mm2 flasks to obtain>100 million cells per cell line. Cells were collected by centri-fugation (300 × g for 5 min at 4 °C), washed using phosphatebuffered saline (PBS), re-pelleted, and stored at –80 °C untilprocessed.

All chemicals were obtained from Sigma-Aldrich (St. Louis,MO; USA), unless otherwise noted. Cells were resuspended in ahypotonic buffer composed of 15 mMTris-HCl (pH 7.5), 60 mMKCl, 15mMNaCl, 5mMMgCl2, 1mMCaCl2, 250mMsucrose,10mM sodium butyrate, and 1× HALT protease and phosphataseinhibitor (Thermo Fisher Scientific (Waltham, MA; USA)). Cells

1204 N. A. Haverland et al.: Gas-Phase Fragmentation with Native MS

were allowed to swell for 45–60 min. Cells were then lysed usingmild sonication with a probe sonicator set to 30% amplitude and 2× 30 s pulse. Cellular debris, including the nuclei and other intactorganelles, were pelleted using centrifugation (11,000 × g for5 min at 4 °C). The soluble cytosolic proteins were then filteredusing aMillex HV PVDF 0.45 μm syringe filter (EMDMillipore;Billerica, MA; USA). Next, the samples were desalted usingAmicon Ultra-0.5 centrifugal filter units (EMD Millipore) in astepwisemanner: First, the lysate was concentrated using 100 kDanominal molecular weight limit (NMWL) filter. The flow-throughwas retained and concentrated using a 3 kDa NMWL centrifugalfilter, effectively creating a fractionated sample with proteins (andcomplexes) ranging from ~3–100 kDa. This 3–100 kDa samplewas then buffer exchanged using a 10 mM ammonium acetate(Sigma-Aldrich) buffer prepared using Optima LC/MS water(Thermo Fisher Scientific) within the 3 kDa NMWL centrifugalfilter; >5 rounds of desalting was performed to maximize down-stream separation.

Ion-Exchange Chromatography (IEX)

Samples were fractionated by IEX using the Agilent series1100 (degasser, pumps, column compartment, UV module)and 1200 (fraction collector, fraction chiller). The columnwas a mixed-bed IEX column composed of an equal proportionof PolyCAT A and PolyWAX LP (PolyLC; Columbia, MD,USA). Two columns were used over the course of this project:12 μm resin with a 1500 Å pore size and 5 μm resin with a1000 Å pore size. Optimized methods for separation weredeveloped using the Agilent Chemstation control softwareusing a gradient of Buffer A (10 mM ammonium acetate,pH 7) and Buffer B (1 mM ammonium acetate, pH 7). Frac-tions were collected into a 96 deep-well plate using 1.5 mintime slices, which were not dependent on the elution profile.The separation was monitored using UV absorbance at 280 nm.The column compartment was maintained at a constant tem-perature of 21 °C for the duration of the run. Fractions wereconcentrated using a 3 kDa NMWL centrifugal filter anddesalted using a 150 mM ammonium acetate buffer, pH 7.

Mass Spectrometry

Electrospray ionization was performed using a custom nanoelectrospray source [59] applying between 0.8 and 1.6 kVvoltage. All analyses were performed on a modified Q-Exactive HF mass spectrometer [60] using methods as previ-ously described [57, 58]. Briefly, the intact mass spectrum(MS1) was acquired using low (~5 V) source-induced dissoci-ation to relieve salt adducts from the precursor ions. Next, asingle charge state of a precursor was quadrupole-isolated andfragmented (MS2) using higher-energy collisional dissociation(HCD), which is a type of beam-style CID [61]. For mostanalyses, a resolving power of >120,000 (at 200m/z) was usedfor both MS1 and MS2. Deconvolution of isotopically resolvedprecursor and fragment ions was performed using the Xtractsoftware (Thermo). When the precursor was not isotopicallyresolved, the average mass was manually calculated from an

MS1 collected using either a resolving power of 15,000 or30,000 (at 200m/z).

Informatics and Statistics

The ProSightPC 4.0 software suite (Thermo Fisher Scientific)was used for identification and characterization of proteoformsusing the precursor mass (monoisotopic or average) and mon-oisotopic fragment ion masses, which were searched using adatabase consisting of SwissProt entries from UniProt (data-base retrieved on October 27, 2015). The precursor masstolerance was set between 200 and 4000 Da to account forunexpected mass shifts resulting from ligand/co-factor binding,splice variants, post-translational modifications, and/or trunca-tion events. The fragment ion mass tolerance was set to≤20 ppm. Proteoform data were exported from ProSightPC toProSight Lite [62] to generate graphical fragment maps, whichwere saved as .pcml files. These .pcml files were used to countthe total number of assigned fragmentation events for eachresidue pair as well as the total number of possible fragmenta-tion events, the latter of which was multiplied by 2 to accountfor bidirectional cleavages. Ratios were calculated as the totalassigned divided by total possible fragmentation events foreach residue pair; these ratios were used as the fragmentationpropensities.

Sampling of the nTDMS dataset and dTDMS dataset wereperformed using (1) the entire nTDMS dataset, and (2) HCDhits of 13,034 proteoforms that were identified with a 1%global false discovery rate (FDR) from the dTDMS dataset inreference [4]. For each round of sampling per dataset, 50% ofthe hits were randomly selected and the fragmentation propen-sity was calculated. A total of 10 passes were performed foreach dataset and the average fragmentation propensity andstandard error were calculated for each residue pair. Differ-ences between the nTDMS dataset and the dTDMS datasetwere determined using an independent samples Student’s t-testwith Bonferroni correction to account for multiple compari-sons. The percentage point change is the absolute differencebetween two percentages and was calculated by subtracting thedTDMS fragmentation propensities from the nTDMS propen-sities. In contrast, the percent change is the relative change inresidue fragmentation propensities and was calculated by di-viding the percentage point difference of the nTDMS anddTDMS fragmentation propensities by the dTDMS fragmen-tation propensities [({nTDMS – dTDMS}/dTDMS) × 100%].

Results and DiscussionAcquisition of nTDMS Data on Protein Monomers(4–70 kDa)

While conducting previous work in the analysis of proteincomplexes using nTDMS [58], we observed a trend in frag-mentation patterns: matched fragment ions in nTDMS wereseemingly even more prevalent for the C-terminal side ofaspartic and glutamic acid residues and the N-terminal side of

N. A. Haverland et al.: Gas-Phase Fragmentation with Native MS 1205

prolines than previously established for dTDMS [63]. In orderto quantitatively examine this trend, we compiled a datasetcontaining fragmentation events collected from human proteinsthat were analyzed using nTDMS/MS with HCD. This datasetincludes 85 human proteins with 131 unique proteoforms that

were collected during 165 independent acquisitions of MS/MSdata. To be considered an independent acquisition, the datamust have been acquired from a unique charge state, fromseparate sample preparations, and/or from distinct cell lines.The molecular weight distribution of proteoforms ranged from

X|X’X|X’

Fra

gm

enta

tio

n e

ven

ts (

#)

0

400

800

400

800

1200

(a)

A C D E F G H I K L M N P Q R S T V W Y

A C D E F G H I K L M N P Q R S T V W Y

A C D E F G H I K L M N P Q R S T V W YACDEFGHIKLMNPQRSTVWY

(b)

0 13 26 52 104+nTDMS HCD fragmentation events (#)

X|X’

X|X

’

Figure 1. (a) The distribution of assigned residue fragmentation events in nTDMS. The total number of events that were assigned foreach residue N-terminal to fragmentation (top) or C-terminal to fragmentation (bottom). The expected number of assignedfragmentation events assuming a uniform fragmentation propensity across all 20 residues is shown as a dashed black line (n =266 events). (b) Assigned fragmentation events deconstructed by residue pair. The expected number of fragmentation eventsassuming uniform propensities across all 400 possible pairs is shown in white (n = 13 events). For all panels, X|X′ refers tofragmentation occurring C-terminal to the amino acid residue whereas X|X′ refers to fragmentation occurring N-terminal to theamino acid residue

1206 N. A. Haverland et al.: Gas-Phase Fragmentation with Native MS

4 to over 70 kDa. Our approach for the analysis of proteincomplexes [58, 60, 64] uses energy to eject monomers in the

source region of the mass spectrometer. Often, these ejectedmonomers will be observed with a disproportionately highnumber of charges compared with the intact precursor in aprocess known as asymmetric charge partitioning [46, 65],which can be attributed to partial or complete unfolding ofthe monomer [46, 66, 67] or to heterolytic scission of ion pairs[68]. Because the preferred fragmentation pathways may beaffected by this increased protonation, monomers that wereejected from complexes were not included in this study.

Overall, a total of 28,250 fragment ion masses were ac-quired and of these, 5311 (18.8%) were assigned to the precur-sor sequence positions with high confidence. The number offragmentation events was not evenly distributed across allresidue pairs; rather, 60% of the ~5300 assigned fragmentationevents occurred C-terminal to aspartic acid, glutamic acid, orlysine residues, or N-terminal to proline or glycine (Fig. 1a).When deconstructed to characterize the effect of unique residuepairs on fragmentation, large biases in residue fragmentationwere observed (Fig. 1b). Assuming an even distribution offragmentation across all 400 residue pairs, ~13 fragmentationevents would map to each pair on average. However, thenumber of fragmentation events ranged from 0 on the lowend to over 100 events mapped to a single pair (8-fold greaterthan expected for an even distribution, Fig. 1b).

The fragmentation propensity describes the likelihood offragmentation occurring between a specific set of residues whilenormalizing for differences in residue pair frequency occurringfor proteins within the dataset. The residue fragmentation pro-pensities were calculated by dividing the total number of ob-served matching fragment ions for the residue pair by the totalnumber of possible fragmentation events for that same residuepair. The average fragmentation propensity across all pairs was8%, and ranged from 0% to 64% (Fig. 2a, b). The fragmentationpathways that occurred either C-terminal to aspartic acid or N-terminal to proline were well above the average with propensitiesof 38% and 24% (Fig. 2a), respectively. These preferred cleav-age pathways have been previously documented in nTDMSwhen assessing the relationship between the precursor’s chargestate and residue fragmentation propensities, but have beenlimited in scope to one or a few purified protein species [35,37, 56, 69–73]. In addition to fragmentation pathways involvingproline or aspartic acid, several residue pairs displayed notablyhigher or lower than expected fragmentation propensities definedas >20% or <5%, respectively (Table 1). Altogether, these resultsindicate that specific residues can impact—in a strong and localfashion—the residue fragmentation propensity during a nTDMS/MS experiment using HCD.

Fragmentation Propensities are Significantly Dif-ferent for nTDMS Versus dTDMS

Next, the HCD fragmentation propensities for nTDMS werecompared with those for dTDMS. First, the fragmentationpropensities for all residue pairs were calculated using anextensive dTDMS dataset published previously (Fig. 2a) [4].This dataset was filtered to include 13,034 proteoforms, which

ACDEFGHIKLMNPQRSTVWY

(b)

dTDMS HCD fragmentation propensity

X|X’

X|X

’

(a)

0%10%20%30%

0%10%20%30%40%

Fra

gm

enta

tio

n p

rop

ensi

ty

X|X’ (nTDMS) X|X’ (dTDMS)X|X’ (nTDMS) X|X’ (dTDMS)

(c)

X|X’

X|X

’

nTDMS HCD fragmentation propensity

0% 8% 16% 32% >48%

A C D E F G H I K L M N P Q R S T V W Y

A C D E F G H I K L M N P Q R S T V W Y

0% 8% 16% 32% >48%

A C D E F G H I K L M N P Q R S T V W

A C D E F G H I K L M N P Q R S T V W Y

YACDEFGHIKLMNPQRSTVWY

Figure 2. (a) The HCD fragmentation propensities observed forthe nTDMS dataset (purple) and dTDMS (turquoise) fromCatherman et al. [4]. The dashed black line at 8% representsthe average fragmentation propensity across all residues fromthe nTDMS dataset. The top graph includes the fragmentationpropensities for events occurring C-terminal to a given residue,whereas the bottom graph includes the fragment propensitiesfor events occurring N-terminal to a given residue. (b) Residuefragmentation propensities for the nTDMS dataset. The asteriskon the cysteine|tryptophan pair indicates that no possible frag-mentation event existed for that pair within the dataset. (c)Residue fragmentation propensities for the dTDMS dataset.For all panels, X|X′ refers to fragmentation occurring C-terminal to the amino acid residue, whereas X|X′ refers to frag-mentation occurring N-terminal to the amino acid residue

N. A. Haverland et al.: Gas-Phase Fragmentation with Native MS 1207

Table 1. The nTDMS Fragmentation Propensities Ordered from the Most Frequently Cleaved Residue Pairs (Top) to the Least Frequently Cleaved Residues Pairs(Bottom)

Residue pairs with fragmentation propensities >20% and number of assigned fragmentation events >13 are shown above the gray bar. In contrast, residue pairs withlower than expected fragmentation propensities (<5%) and >275 possible fragmentation events are shown below the gray bar

1208 N. A. Haverland et al.: Gas-Phase Fragmentation with Native MS

were identified with a global FDR of <1% and fragmentedusing HCD. Few dramatic differences were observed whencomparing the fragmentation pathways for each residue innTDMS and dTDMS (Fig. 2a); the notable exception wasfragmentation occurring C-terminal to aspartic acid, whichwas increased 2-fold in the nTDMS dataset compared withthe dTDMS dataset. When deconstructed by residue pair, atrend for increased fragmentation propensity was observedwithin the dTDMS dataset among the more hydrophobic

residue pairs (leucine, isoleucine, valine, phenylalanine, andmethionine; Fig. 2c), which is not as evident within thenTDMS dataset (Fig. 2b). Overall, the residue fragmentationpropensities for each pair appears to be more evenly distributedin dTDMS compared with nTDMS with an average residuefragmentation propensity of 9% and a range of 0.02% to 37%.

To further quantify the differences in residue fragmentationpropensities for nTDMS and dTDMS, bootstrapping was per-formed for each dataset to obtain summary statistics (standard

A C D E F G H I K L M N P Q R S T V W YACDEFGHIKLMNPQRSTVWY

* * * * * * * * * * * *

* * * * * * * * * * * * * * *

* * * * * * * * * * * * * * * * *

* * * * * * * * * * * *

* * * * * * * * * * * * * *

* * * * * * * * * * * * * * *

* * * * * * * * * * * *

* * * * * * * * * * *

* * * * * * * * * * *

* * * * * * * * *

* * * * * * * * * * * * * * *

* * * * * * * * * * * * *

* * * * * * * * * * * * * *

* * * * * * * * * * * * *

* * * * * * * * * *

* * * * * * * * * *

* * * * * * * * * *

************

* * * * * * * * * * * * * * *

* * * * * * * * * * * * * * * * * *

−2 0 2 5 10 fragmentation propensity

(fold-change, nTDMS vs. dTDMS)

X|X’X

|X’

Figure 3. nTDMS has a fewer number of highly preferred fragmentation pathways compared with dTDMS. The fold-change in residuefragmentation propensity for nTDMS compared with dTDMS. Blue indicates a decrease in fragmentation propensity in nTDMScompared with dTDMS, whereas red indicates an increase for nTDMS. Significant differences (P ≤ 0.05) are indicated with an asterisk.X|X′denotes fragmentation occurringC-terminal to the amino acid residue, whereas X|X′ refers to fragmentation occurringN-terminal tothe amino acid residue

N. A. Haverland et al.: Gas-Phase Fragmentation with Native MS 1209

error and mean fragmentation propensity). Using the Student’s t-test with the Bonferroni correction for multiple comparisons (n =399), the fragmentation propensities for a total of 257 residuepairs were significantly different (P ≤ 0.05) between the twodatasets, and of these, 123 residue pairs were altered by 2-foldor greater (Fig. 3). The most extreme relative change wasproline|histidine with an increase of >12-fold in nTDMS com-pared with dTDMS. In contrast, the largest relative decrease was–2-fold, which was the case when no matched fragmentationevents were observed for a residue pair. Additionally, severaltrends were observed in nTDMS compared with dTDMS, in-cluding enhanced fragmentation occurring C-terminal to asparticacid and diminished fragmentation between the more bulky andhydrophobic residues (leucine, isoleucine, valine, phenylalanine,methionine, and tryptophan; Fig. 3). The complete dataset con-taining residue fragmentation propensities from both nTDMSand dTDMS is provided in Online Resource 1. These resultsindicate that even while using the same type of dissociation,ionization of the protein in either native or denatured mode willsignificantly influence the fragmentation propensities for back-bone amide bonds flanked by themajority of amino acid pairings.

Correlating Fragmentation with Charge State,Sources of Mobile Protons, and Tertiary Structures

The number of protonation events for a protein can be influencedin part by the intrinsic properties of a protein and the ionizationtechnique that is used [74, 75] (reviewed in [76]). In fact, one ofthe more profound observable differences between dESI andnESI is the effect of ESI solvent on the charge state distributionof a protein [37]. For example, a protein that has been denaturedis able to carry additional protons, which is indicative of anelongated structure with a higher surface area and increasedaccessibility to ionizable residues. In contrast, a globular proteinthat retains a native-like fold during the transfer from solution tothe gas phase will likely have a decreased surface area. In thecase of globular proteins, the theoretical maximum number ofcharges that can be imparted to the protein during electrospray isdescribed by the Rayleigh charge (zR) limit theory, which hasbeen defined and discussed elsewhere [74, 75, 77, 78] and islargely influenced by the surface tension of the solvent and theradius of the ion. Thus, because a denatured protein will have amuch larger surface area than its native (compact) form, thedenatured protein will not necessarily conform to the Rayleighcharge limit theory and may be observed with charge statesgreater than the Rayleigh charge (z > zR). (For a review onprotein charging and supercharging with a focus on ESI, see[79]). Based on the results presented here and elsewhere [35, 51,63, 73, 80], we posit that the residue fragmentation propensitiesfor intact proteins are primarily influenced by the thermalizedstructures, the net charge, dissociation technique, and the residuecomposition of precursor ions.

A given charge state for a protein is typically designated as ahigh, intermediate, or low charge state, which is a historicreference to the McLuckey group’s designation for ubiquitin,

0

5

10

15

20

5 15 25 35 45 55 65 75

Ch

arg

e (z

)

Mass (kDa)

(a)

nTDMS: P23528, COF1; p-score = 3.3 × 10−43; (z/zR = 0.85), 9+

(b)

nTDMS: P23528, COF1; p-score = 2.2 × 10−43; (z/zR = 1.23), 13+

(c)

nTDMS: P23528, COF1; p-score = 1.2 × 10−4; (z/zR = 1.52), 16+

(d)

TDMS: P23528, COF1; p-score = 9.5 × 10−13; (z/zR = 1.52), 16+

(e)

TDMS: P23528, COF1; p-score = 2.2 × 10−9; (z/zR = 1.79), 19+

(f)

intermediate

low

high

Figure 4. (a) A relationship between charge and precursormass. Precursor ions were selected for fragmentation withoutconsideration for charge state designation, which resulted inthe inclusion of fragmentation data from precursors with high(pink) and intermediate (yellow) charge state designations. Theyellow curve (y = 0.0778 m½) represents the Rayleigh chargelimit for each precursor, zR (see text and [75]). The pink curve(y = 0.1081 m½) is the lower bound for the high charge statedesignation [56], which was determined by calculating thetheoretical zHigh (zHigh = zR × 1.43) for each protein in the dataset.The blue curve (y = 0.0667 m½) is the upper bound for the lowcharge state designation [56], which was determined bycalculating the theoretical zLow (zLow = zR × 0.86) for each proteinin the dataset. The graphical fragmentation maps for COF1obtained by nTDMS of a “low” (b), “intermediate” (c), and “high”(d) charge state precursor ion compared with dTDMS [4] of an“intermediate” (e) and “high” (f) charge state precursor ion. Forpanels (b)–(f), blue flags represent matched fragment ions withmass tolerance of ≤15 ppm

1210 N. A. Haverland et al.: Gas-Phase Fragmentation with Native MS

with high charge states defined as 13+ to 10+, intermediate as 9+to 7+, and low as 6+ and below [56]. It is possible to extrapolatethese designations to the proteins analyzed here by nTDMS bydetermining the ratio of observed charge to Rayleigh charge(z/zR); the latter can be calculated as zR = 0.0778 m½, but hasthe assumption that the structure of the ion is a compactedglobular sphere in an aqueous environment and, as such, theion density is ρ = 1 g/cm3 and the surface tension is that of water,γ = 0.072N/m [75]. Therefore, given that themass of ubiquitin is8.5 kDa, the zR is calculated as 7 (rounded to the nearest wholeinteger) and the z/zR for the high charge states is >1.43 and thez/zR for low charge states is <0.86, with the intermediate chargestates falling in between these values. In calculating the z/zR ratiofor each of the precursor ions used in the nTDMS datasetreported here, we found 102 had low charge state precursor ions,57 had intermediate charge state precursor ions, and six had highcharge state precursor ions (Fig. 4a). These results indicate thatthe majority of precursor ions used in this analysis conform tothe Rayleigh charge limit theory predictions. However, a non-trivial number of precursors have z > zR, especially amongprecursors with a mass of <30 kDa (Fig. 4a). We anticipatedthat these high and intermediate charge state precursor ions maybe an outcome of non-sphericity due to the intrinsic fold of theprotein (e.g., inclusion of fibrous and disordered protein precur-sor ions). In fact, among the precursor ions with the high chargestate designation was STMN1 (P16949), which is a foundingmember for the class of intrinsically disordered/unstructuredproteins [81–83]. Additionally, COF1 (P23528), which has anunstructured N-terminus when unmodified [84], had precursorions that spanned all three charge state designations and wereincluded in the nTDMS dataset. The fragmentation patterns forCOF1 are distinct for each charge state designation (Fig. 4b,c, d), which can be attributed to the differences in net charge.However, one should also consider that starting precursor inten-sity can be a confounding factor and influence the ability todetect fragment ions, which may explain the relatively fewfragment ions that were observed in the high charge state desig-nation of COF1 (Fig. 4d). Additionally, the fragmentation pat-terns observed for COF1 during dTDMS are distinct from eventhe high charge state designation in nTDMS (Fig. 4e, f), whichsuggests that aspects of tertiary structure could correlate stronglyto the observed gas-phase fragmentation patterns.

Although one cannot assume that the gas-phase structure ofa protein will mimic its crystal structure, we investigated thepossible role of tertiary structure on fragmentation by mappingthe fragmentation events to the crystal structure for a variety ofproteins. The largest protein that had multiple charge statedesignations was GDIR1 (P52565; 23 kDa), which has aglobular core with an extended, flexible arm and a disorderedregion (Fig. 5a). The two charge states that were included in thenTDMS portion of this study represent a low and intermediatecharge state with z/zR ratios of 0.76 (9+) and 1.18 (14+). Thefragmentation patterns for these charge states are similar withseveral notable differences, including an increased number ofcleavages occurring at aspartic acid in the 9+ precursor com-pared with the 14+ precursor (Fig. 5b, c). The fragmentation

patterns for both of the nTDMS precursors sharply contrast thepatterns observed by dTDMS (Fig. 5d). Interestingly, mappingthe fragmentation events observed for the 9+ precursor ion ofGDIR1 to its crystal structure reveals a trend for fragmentation

nTDMS: P52565, GDIR1; p-score = 1.2 × 10−23; ; (z/z

R = 0.76), 9+

dTDMS: P52565, GDIR1; p-score = 6.5 × 10−10 ; (z/z

R = 1.78), 21+

(d)

nTDMS: P52565, GDIR1; p-score = 7.5 × 10−7 ; (z/z

R = 1.18), 14+

(c)

(b)

(a)

Figure 5. (a) The surface and cartoon models of GDIR1 (PDBID: 1hh4 [ 88 ]). Pink represents arginine residues and bluehighlights the residues involved in fragmentation of the 9+precursor ion, panel (b). The two disordered regions, one atthe N-terminus [A2-A7] and the other near the flexible arm[V59-P65], are not included in the crystal structure, and thusnot shown. The fragmentation maps of GDIR1 obtained usingnTDMS for a low (b) and intermediate (c) charge state precursorion compared with fragmentation of a high (d) charge stateprecursor ion obtained using dTDMS [ 4 ]. For panels (b)–(d),blue flags represent matched fragment ions with mass toler-ance of ≤15 ppm. In panel (b), arginine residues are colored pinkto match the arginine residues highlighted in panel (a)

N. A. Haverland et al.: Gas-Phase Fragmentation with Native MS 1211

occurring at or near the solvent accessible surface area of theprotein and proximal to several arginine residues (Fig. 5a). Thistrend was also observed in other precursor ions representingboth low and intermediate charge states across a wide range ofmasses (Fig. 6).

With some appreciation for the relationships betweencharge state, tertiary structure of the analogous crystal struc-ture, and preferred fragmentation patterns in nTDMS, we re-visit the mobile proton model. In contrast to dESI, whichimparts charge to a variety of residues with largest protonaffinities (arginine, lysine, histidine, and the N-terminal α-amino group [37]), nESI results in protonation primarily atarginine residues [85]. The difference in protonation of theprecursor is important when considering proton mobility andsequestration. A proton is consideredmobile if the total numberof charges is greater than the number of basic residues presentin the ion, which is often the case with dTDMS experiments.When a low level of activation energy is applied, the protonreadily mobilizes from the side chain of the residue to thebackbone amide bond to induce charge-directed fragmentation.The specificity or preference for specific fragmentation path-ways is diminished with mobile protons, which can result infragmentation that occurs between a greater variety of residuepairs [31, 35, 56], and was consistent with the dTDMS datasetreported here (Fig. 2c) and elsewhere [35]. In contrast, a protonis sequestered if the total number of charges is equal to or lessthan the number of basic residues present in the ion, which ismore reflective of the conditions in nESI. In this case, frag-mentation will occur via pathways that do not require intramo-lecular proton transfer (e.g., charge-remote fragmentation) or ifthe activation energy is increased to promote mobilization ofthe sequestered proton [23, 28–30, 34]. However, if the protonis sequestered and an acidic residue is present, the acidichydrogen in the side chain of the residue can act as the protonsource needed to induce fragmentation [28, 30, 34]. This isobserved as an increased preference for fragmentation occur-ring C-terminal to glutamic acid and aspartic acid, and isconsistent with the nTDMS data presented here (Fig. 2b). Infact, of the 20 possible residue pairs with fragmentation occur-ring C-terminal to aspartic acid, the propensities for 17 of thesepairs were increased by over 50% in nTDMS compared withdTDMS (Fig. 3). Based on these observations, the canonicalview of the mobile proton model as it relates to sequesteredprotons corresponds well with the observations and trendsreported here for both nTDMS and dTDMS.

The Bproline effect^ is the observation that upon collisionalion activation there is an increased fragmentation propensityoccurring N-terminal to proline residues [86], which is likelyinfluenced by the rigid structure and substituted amine in the y-ion’s leaving group of this cyclized residue. Loo and colleaguesreport that as the mass of a precursor ion increases (e.g., peptideto protein to large protein), the likelihood of producing aninterior cleavage event decreases; however, this trend is re-versed when an interior proline residue(s) is present [86]. TheWysocki group expanded on this observation, noting that theenergy demands for fragmentation is increased for precursor

(a)P63244: RACK1; p-score = 4.5 × 10−49

z/zR

= 0.82

z/zR

= 0.83

z/zR

= 0.88

z/zR

= 1.09

12+

35 kDa

69 kDa

(b)P09960: LKH4A; p-score = 5.2 × 10−9

17+

(d)P37802: TAGL2; p-score = 1.2 × 10−20

12+

20 kDa

(c)Q16658: FCSN1; p-score = 2 × 10−11

16+

55 kDa

Figure 6. The cartoon (left) and surface (right) models of (a)RACK1 (PDB ID: 4aow [89]), (b) LKH4A (PDB ID: 4rvb), (c)TAGL2 (PDB ID: 1wym), and (d) FCSN1 (PDB ID: 3llp [90]). Pinkrepresents arginine residues and blue highlights the residuesinvolved in fragmentation. Panels (a) and (b) represent exam-ples of precursor ions with low charge state designations(<0.86), whereas panels (c) and (d) represent examples of pre-cursor ions with intermediate charge state designations (0.86 <z/zR < 1.43)

1212 N. A. Haverland et al.: Gas-Phase Fragmentation with Native MS

ions containing internal prolines, likely a consequence of theunique structure of the residue [23]. Further, an overall increasein fragmentation occurs N-terminal to proline when the residuepairs include valine, histidine, aspartic acid, isoleucine, andleucine; however, fragmentation is less likely when the resi-dues N-terminal to the cleavage site are proline and glycine [26,32]. Additionally, there is a relationship between increasedproton mobility and increased fragmentation occurring N-terminal to proline residues in peptides [26, 31, 32] and pro-teins [35], despite the proton not necessarily being localized tothe proline residue itself [87]. Given that the Bproline effect^ isheavily reliant on the sequence of the precursor ion, it would beexpected that the residue fragmentation propensity occurringN-terminal to proline in both nTDMS and dTDMS would berelatively equivalent. However, the fragmentation propensitiesfor several residue pairs were altered by over 2-fold when usingnTDMS compared with dTDMS, including alanine|proline,tyrosine|proline, and arginine|proline (Fig. 3).

Additional investigations will be required to better de-fine the interplay between gas-phase ion structure, proton-ation, and preferred fragmentation pathways as they relateto the analysis of intact proteins with native-like confor-mations. Likewise, defining the extent of internal fragmen-tation that occurs during nTDMS would also be beneficial,as it could provide a unique insight into how the structureof precursor ions in the gas phase can influence fragmen-tation pathways, especially in comparison to its denaturedcounterpart [35, 73].

ConclusionsHere we have provided the first extended analysis of res-idue fragmentation propensities for proteoforms analyzedusing nTDMS with beam-type HCD fragmentation. Addi-tionally, we established that a great number of fragmenta-tion pathways are significantly altered when using nTDMScompared with dTDMS. Overall, we report a data-drivenassertion that nTDMS funnels fragmentation into a fewernumber of more highly preferred sites than dTDMS. Ad-ditionally, it was unexpected that the already high propen-sities for fragmentation Bhot spots^ in dTDMS (e.g., oc-curring C-terminal to aspartic acid) would be doubledwhen using nTDMS. It also seems that the mobile protonmodel and the proline effect can be used to explain manyof the fragmentation pathways that are enhanced duringnTDMS. The nTDMS results also show a strong correla-tion to surface exposed residues as viewed on the tertiarystructures of selected examples.

Thus, it stands to reason that as nTDMS becomes moreroutine for the analysis of proteoforms and their complexes,expert scoring systems should be updated and informed withthese trends. The residue fragmentation propensities presentedhere will be expanded and become an essential piece of infor-mation when adopting and adapting Bayesian scoring metrics,such as the C-score for proteoforms [40], the MPC-score for

multi-proteoform complexes [58], and other scores generatedby the growing community of developers and practitioners ofTDMS.

AcknowledgementsThe authors acknowledge the W. M. Keck foundation forgenerous support and funding (DT061512). In addition, thiswork was partially supported by R01GM067193 (NIGMS).O.S.S. was supported by an NSF graduate research fellowship(2014171659).

References

1. Catherman, A.D., Skinner, O.S., Kelleher, N.L.: Top-down proteomics:facts and perspectives. Biochem. Biophys. Res. Commun. 445, 683–693(2014)

2. Toby, T.K., Fornelli, L., Kelleher, N.L.: Progress in top-down proteomicsand the analysis of proteoforms. Annu. Rev. Anal. Chem. 9, 499–519(2016)

3. Tran, J.C., Zamdborg, L., Ahlf, D.R., Lee, J.E., Catherman, A.D., Durbin,K.R., Tipton, J.D., Vellaichamy, A., Kellie, J.F., Li, M., Wu, C., Sweet,S.M., Early, B.P., Siuti, N., LeDuc, R.D., Compton, P.D., Thomas, P.M.,Kelleher, N.L.: Mapping intact protein isoforms in discovery mode usingtop-down proteomics. Nature 480, 254–258 (2011)

4. Catherman, A.D., Durbin, K.R., Ahlf, D.R., Early, B.P., Fellers, R.T.,Tran, J.C., Thomas, P.M., Kelleher, N.L.: Large-scale top-down proteo-mics of the human proteome: membrane proteins, mitochondria, andsenescence. Mol. Cell. Proteom. 12, 3465–3473 (2013)

5. Durbin, K.R., Fornelli, L., Fellers, R.T., Doubleday, P.F., Narita, M.,Kelleher, N.L.: Quantitation and identification of thousands of humanproteoforms below 30 kDa. J. Proteome Res. 15, 976–982 (2016)

6. Ansong, C.,Wu, S., Meng, D., Liu, X., Brewer, H.M., DeatherageKaiser,B.L., Nakayasu, E.S., Cort, J.R., Pevzner, P., Smith, R.D., Heffron, F.,Adkins, J.N., Pasa-Tolic, L.: Top-down proteomics reveals a uniqueprotein S-thiolation switch in Salmonella typhimurium in response toinfection-like conditions. Proc. Natl. Acad. Sci. U. S. A. 110, 10153–10158 (2013)

7. Peng, Y., Gregorich, Z.R., Valeja, S.G., Zhang, H., Cai, W., Chen, Y.C.,Guner, H., Chen, A.J., Schwahn, D.J., Hacker, T.A., Liu, X., Ge, Y.: Top-down proteomics reveals concerted reductions in myofilament and Z-discprotein phosphorylation after acute myocardial infarction. Mol. Cell.Proteom. 13, 2752–2764 (2014)

8. Smith, L.M., Kelleher, N.L.: Consortium for top down, P.: proteoform: asingle term describing protein complexity. Nat. Methods 10, 186–187(2013)

9. Loo, J.A., Edmonds, C.G., Smith, R.D.: Primary sequence informationfrom intact proteins by electrospray ionization tandemmass spectrometry.Science 248, 201–204 (1990)

10. Sickmann, A., Reinders, J., Wagner, Y., Joppich, C., Zahedi, R., Meyer,H.E., Schonfisch, B., Perschil, I., Chacinska, A., Guiard, B., Rehling, P.,Pfanner, N., Meisinger, C.: The proteome of Saccharomyces cerevisiaemitochondria. Proc. Natl. Acad. Sci. U. S. A. 100, 13207–13212 (2003)

11. Gygi, S.P., Rist, B., Gerber, S.A., Turecek, F., Gelb, M.H., Aebersold, R.:Quantitative analysis of complex protein mixtures using isotope-codedaffinity tags. Nat. Biotechnol. 17, 994–999 (1999)

12. Haverland, N.A., Fox, H.S., Ciborowski, P.: Quantitative proteomics bySWATH-MS reveals altered expression of nucleic acid binding andregulatory proteins in HIV-1-infected macrophages. J. Proteome Res.13, 2109–2119 (2014)

13. Xie, Y., Zhang, J., Yin, S., Loo, J.A.: Top-down ESI-ECD-FT-ICR massspectrometry localizes noncovalent protein-ligand binding sites. J. Am.Chem. Soc. 128, 14432–14433 (2006)

14. Schennach, M., Schneeberger, E.M., Breuker, K.: Unfolding and foldingof the three-helix bundle protein KIX in the absence of solvent. J. Am.Soc. Mass Spectrom. 27, 1079–1088 (2016)

15. Sleno, L., Volmer, D.A.: Ion activation methods for tandem mass spec-trometry. J. Mass Spectrom. 39, 1091–1112 (2004)

N. A. Haverland et al.: Gas-Phase Fragmentation with Native MS 1213

16. Brodbelt, J.S.: Ion activation methods for peptides and proteins. Anal.Chem. 88, 30–51 (2016)

17. Roepstorff, P., Fohlman, J.: Proposal for a common nomenclature forsequence ions in mass spectra of peptides. Biomed. Mass Spectrom. 11,601 (1984)

18. Frese, C.K., Altelaar, A.F., Hennrich, M.L., Nolting, D., Zeller, M.,Griep-Raming, J., Heck, A.J., Mohammed, S.: Improved peptide identi-fication by targeted fragmentation using CID, HCD, and ETD on anLTQ-Orbitrap Velos. J. Proteome Res. 10, 2377–2388 (2011)

19. Xia, Y., Liang, X., McLuckey, S.A.: Ion trap versus low-energy beam-type collision-induced dissociation of protonated ubiquitin ions. Anal.Chem. 78, 1218–1227 (2006)

20. Zubarev, R.A., Kelleher, N.L., McLafferty, F.W.: Electron capture dis-sociation of multiply charged protein cations. a nonergodic process. J.Am. Chem. Soc. 120, 3265–3266 (1998)

21. Syka, J.E., Coon, J.J., Schroeder, M.J., Shabanowitz, J., Hunt, D.F.:Peptide and protein sequence analysis by electron transfer dissociationmass spectrometry. Proc. Natl. Acad. Sci. U. S. A. 101, 9528–9533(2004)

22. Shaw, J.B., Li, W., Holden, D.D., Zhang, Y., Griep-Raming, J., Fellers,R.T., Early, B.P., Thomas, P.M., Kelleher, N.L., Brodbelt, J.S.: Completeprotein characterization using top-down mass spectrometry and ultravio-let photodissociation. J. Am. Chem. Soc. 135, 12646–12651 (2013)

23. Dongré, A.R., Jones, J.L., Somogyi, Á., Wysocki, V.H.: Influence ofpeptide composition, gas-phase basicity, and chemical modification onfragmentation efficiency: evidence for the mobile proton model. J. Am.Chem. Soc. 118, 8365–8374 (1996)

24. Wysocki, V.H., Tsaprailis, G., Smith, L.L., Breci, L.A.: Mobile andlocalized protons: a framework for understanding peptide dissociation.J. Mass Spectrom. 35, 1399–1406 (2000)

25. Paizs, B., Suhai, S.: Fragmentation pathways of protonated peptides.Mass Spectrom. Rev. 24, 508–548 (2005)

26. Breci, L.A., Tabb, D.L., Yates, J.R.,Wysocki, V.H.: Cleavage N-terminalto proline: analysis of a database of peptide tandem mass spectra. Anal.Chem. 75, 1963–1971 (2003)

27. Gu, C., Tsaprailis, G., Breci, L., Wysocki, V.H.: Selective gas-phasecleavage at the peptide bond C-Terminal to aspartic acid in fixed-chargederivatives of Asp-containing peptides. Anal. Chem. 72, 5804–5813(2000)

28. Gu, C., Somogyi, Á., Wysocki, V.H., Medzihradszky, K.F.: Fragmenta-tion of protonated oligopeptides XLDVLQ (X = L, H, K, or R) by surfaceinduced dissociation: additional evidence for the ‘mobile proton’ model.Anal. Chim. Acta 397, 247–256 (1999)

29. Summerfield, S.G., Gaskell, S.J.: Fragmentation efficiencies of peptideions following low energy collisional activation. Int. J. Mass Spectrom.Ion Processes 165/166, 509–521 (1997)

30. Tsaprailis, G., Nair, H., Somogyi, Á., Wysocki, V.H., Zhong, W., Futrell,J.H., Summerfield, S.G., Gaskell, S.J.: Influence of secondary structureon the fragmentation of protonated peptides. J. Am. Chem. Soc. 121,5142–5154 (1999)

31. Huang, Y., Triscari, J.M., Tseng, G.C., Pasa-Tolic, L., Lipton, M.S.,Smith, R.D., Wysocki, V.H.: Statistical characterization of the chargestate and residue dependence of low-energy CID peptide dissociationpatterns. Anal. Chem. 77, 5800–5813 (2005)

32. Huang, Y., Tseng, G.C., Yuan, S., Pasa-Tolic, L., Lipton, M.S., Smith,R.D., Wysocki, V.H.: A Data-mining scheme for identifying peptidestructural motifs responsible for different MS/MS fragmentation intensitypatterns. J. Proteome Res. 7, 70–79 (2008)

33. Bojesen, G.: The order of proton affinities of the 20 common L-α-aminoacids. J. Am. Chem. Soc. 109, 5557–5558 (1987)

34. Tsaprailis, G., Somogyi, Á., Nikolaev, E.N.,Wysocki, V.H.: Refining themodel for selective cleavage at acidic residues in arginine-containingprotonated peptides2. Int. J. Mass Spectrom. 195/196, 467–479 (2000)

35. Cobb, J.S., Easterling, M.L., Agar, J.N.: Structural characterization ofintact proteins is enhanced by prevalent fragmentation pathways rarelyobserved for peptides. J. Am. Soc. Mass Spectrom. 21, 949–959 (2010)

36. Chowdhury, S.K., Katta, V., Chait, B.T.: Probing conformational changesin proteins by mass spectrometry. J. Am. Chem. Soc. 112, 9012–9013(1990)

37. Loo, J.A., Loo, R.R.O., Udseth, H.R., Edmonds, C.G., Smith, R.D.:Solvent-induced conformational changes of polypeptides probed byelectrospray-ionization mass spectrometry. Rapid Commun. MassSpectrom. 5, 101–105 (1991)

38. Katta, V., Chait, B.T., Carr, S.: Conformational changes in proteinsprobed by hydrogen-exchange electrospray-ionization mass spectrome-try. Rapid Commun. Mass Spectrom. 5, 214–217 (1991)

39. Meng, F., Cargile, B.J., Miller, L.M., Forbes, A.J., Johnson, J.R.,Kelleher, N.L.: Informatics and multiplexing of intact protein identifica-tion in bacteria and the archaea. Nat. Biotechnol. 19, 952–957 (2001)

40. LeDuc, R.D., Fellers, R.T., Early, B.P., Greer, J.B., Thomas, P.M.,Kelleher, N.L.: The C-score: a Bayesian framework to sharply improveproteoform scoring in high-throughput top down proteomics. J. ProteomeRes. 13, 3231–3240 (2014)

41. Siuzdak, G., Bothner, B., Yeager, M., Brugidou, C., Fauquet, C.M.,Hoey, K., Change, C.-M.: Mass spectrometry and viral analysis. Chem.Biol. 3, 45–48 (1996)

42. Robinson, C.V., Grosz, M., Eyles, S.J., Ewbank, J.J., Mayhew, M., Hartl,F.U., Dobson, C.M., Radford, S.E.: Conformation of GroEL-bound α-lactalbumin probed by mass spectrometry. Nature 372, 646–651 (1994)

43. Rostom, A.A., Sunde, M., Richardson, S.J., Schreiber, G., Jarvis, S.,Bateman, R., Dobson, C.M., Robinson, C.V.: Dissection of multi-protein complexes using mass spectrometry: subunit interactions intransthyretin and retinol-binding protein complexes. Proteins 33, 3–11(1998)

44. Gervasoni, P., Staudenmann, W., James, P., Gehrig, P., Plückthun, A.: β-Lactamase binds to GroEL in a conformation highly protected againsthydrogen/deuterium exchange. Proc. Natl. Acad. Sci. U. S. A. 93, 12189–12194 (1996)

45. Ganem, B., Li, Y.T., Henion, J.D.: Detection of noncovalent receptor–ligand complexes by mass spectrometry. J. Am. Chem. Soc. 113, 6294–6296 (1991)

46. Light-Wahl, K.J., Schwartz, B.L., Smith, R.D.: Observation of thenoncovalent quaternary associations of proteins by electrospray ioniza-tion mass spectrometry. J. Am. Chem. Soc. 116, 5271–5278 (1994)

47. Laganowsky, A., Reading, E., Allison, T.M., Ulmschneider, M.B.,Degiacomi, M.T., Baldwin, A.J., Robinson, C.V.: Membrane proteinsbind lipids selectively to modulate their structure and function. Nature510, 172–175 (2014)

48. Gologan, B., Takáts, Z., Alvarez, J., Wiseman, J.M., Talaty, N., Ouyang,Z., Cooks, R.G.: Ion soft-landing into liquids: protein identification,separation, and purification with retention of biological activity. J. Am.Soc. Mass Spectrom. 15, 1874–1884 (2004)

49. Winston, R.L., Fitzgerald, M.C.: Mass spectrometry as a readout ofprotein structure and function. Mass Spectrom. Rev. 16, 165–179 (1997)

50. Sharon, M., Robinson, C.V.: The role of mass spectrometry in structureelucidation of dynamic protein complexes. Annu. Rev. Biochem. 76,167–193 (2007)

51. Breuker, K., McLafferty, F.W.: Stepwise evolution of protein nativestructure with electrospray into the gas phase, 10−12 to 102 s. Proc. Natl.Acad. Sci. U. S. A. 105, 18145–18152 (2008)

52. Compton, P.D., Zamdborg, L., Thomas, P.M., Kelleher, N.L.: On thescalability and requirements of whole protein mass spectrometry. Anal.Chem. 83, 6868–6874 (2011)

53. Muneeruddin, K., Thomas, J.J., Salinas, P.A., Kaltashov, I.A.: Charac-terization of small protein aggregates and oligomers using size exclusionchromatography with online detection by native electrospray ionizationmass spectrometry. Anal. Chem. 86, 10692–10699 (2014)

54. Muneeruddin, K., Nazzaro,M., Kaltashov, I.A.: Characterization of intactprotein conjugates and biopharmaceuticals using ion-exchange chroma-tography with online detection by native electrospray ionization massspectrometry and top-down tandem mass spectrometry. Anal. Chem. 87,10138–10145 (2015)

55. Chen, B., Peng, Y., Valeja, S.G., Xiu, L., Alpert, A.J., Ge, Y.: Onlinehydrophobic interaction chromatography-mass spectrometry for top-down proteomics. Anal. Chem. 88, 1885–1891 (2016)

56. Reid, G.E., Wu, J., Chrisman, P.A., Wells, J.M., McLuckey, S.A.:Charge-state-dependent sequence analysis of protonated ubiquitin ionsvia ion trap tandem mass spectrometry. Anal. Chem. 73, 3274–3281(2001)

57. Skinner, O.S., Do Vale, L.H., Catherman, A.D., Havugimana, P.C., deSousa, M.V., Compton, P.D., Kelleher, N.L.: Native GELFrEE: a newseparation technique for biomolecular assemblies. Anal. Chem. 87, 3032–3038 (2015)

58. Skinner, O.S., Havugimana, P.C., Haverland, N.A., Fornelli, L., Early,B.P., Greer, J.B., Fellers, R.T., Durbin, K.R., Do Vale, L.H., Melani,R.D., Seckler, H.S., Nelp, M.T., Belov, M.E., Horning, S.R., Makarov,A.A., LeDuc, R.D., Bandarian, V., Compton, P.D., Kelleher, N.L.: An

1214 N. A. Haverland et al.: Gas-Phase Fragmentation with Native MS

informatic framework for decoding protein complexes by top-down massspectrometry. Nat. Methods 13, 237–240 (2016)

59. Wojcik, R., Dada, O.O., Sadilek, M., Dovichi, N.J.: Simplified capillaryelectrophoresis nanospray sheath-flow interface for high efficiency andsensitive peptide analysis. Rapid Commun. Mass Spectrom. 24, 2554–2560 (2010)

60. Belov, M.E., Damoc, E., Denisov, E., Compton, P.D., Horning, S.,Makarov, A.A., Kelleher, N.L.: From protein complexes to subunitbackbone fragments: a multi-stage approach to native mass spectrometry.Anal. Chem. 85, 11163–11173 (2013)

61. Olsen, J.V., Macek, B., Lange, O., Makarov, A., Horning, S., Mann, M.:Higher-energy C-trap dissociation for peptide modification analysis. Nat.Methods 4, 709–712 (2007)

62. Fellers, R.T., Greer, J.B., Early, B.P., Yu, X., LeDuc, R.D., Kelleher,N.L., Thomas, P.M.: ProSight Lite: graphical software to analyze top-down mass spectrometry data. Proteomics 15, 1235–1238 (2015)

63. Schaaff, T.G., Cargile, B.J., Stephenson, J.L., McLuckey, S.A.: Ion trapcollisional activation of the (M + 2H)2+ − (M + 17H)17+ ions of humanhemoglobin β-chain. Anal. Chem. 72, 899–907 (2000)

64. Compton, P.D., Fornelli, L., Kelleher, N.L., Skinner, O.S.: Probingasymmetric charge partitioning of protein oligomers during tandem massspectrometry. Int. J. Mass Spectrom. 390, 132–136 (2015)

65. Schwartz, B.L., Bruce, J.E., Anderson, G.A., Hofstadler, S.A.,Rockwood, A.L., Smith, R.D., Chilkoti, A., Stayton, P.S.: Dissociationof tetrameric ions of noncovalent streptavidin complexes formed byelectrospray ionization. J. Am. Soc. Mass Spectrom. 6, 459–465 (1995)

66. Jurchen, J.C.,Williams, E.R.: Origin of asymmetric charge partitioning inthe dissociation of gas-phase protein homodimers. J. Am. Chem. Soc.125, 2817–2826 (2003)

67. Jurchen, J.C., Garcia, D.E., Williams, E.R.: Further studies on the originsof asymmetric charge partitioning in protein homodimers. J. Am. Soc.Mass Spectrom. 15, 1408–1415 (2004)

68. Loo, R.R.O., Loo, J.A.: Salt bridge rearrangement (SaBRe) explains thedissociation behavior of noncovalent complexes. J. Am. Soc. MassSpectrom. 27, 975–990 (2016)

69. Newton, K.A., Pitteri, S.J., Laskowski, M., McLuckey, S.A.: Effects ofsingle amino acid substitution on the collision-induced dissociation ofintact protein ions: Turkey ovomucoid third domain. J. Proteome Res. 3,1033–1041 (2004)

70. Engel, B.J., Pan, P., Reid, G.E., Wells, J.M., McLuckey, S.A.: Chargestate dependent fragmentation of gaseous protein ions in a quadrupole iontrap: bovine ferri-, ferro-, and apo-cytochrome c. Int. J. Mass Spectrom.219, 171–187 (2002)

71. Hogan, J.M., McLuckey, S.A.: Charge state dependent collision-induceddissociation of native and reduced porcine elastase. J. Mass Spectrom. 38,245–256 (2003)

72. Pitteri, S.J., Reid, G.E., McLuckey, S.A.: Affecting proton mobility inactivated peptide and whole protein ions via lysine guanidination. J.Proteome Res. 3, 46–54 (2004)

73. Durbin, K.R., Skinner, O.S., Fellers, R.T., Kelleher, N.L.: Analyzinginternal fragmentation of electrosprayed ubiquitin ions during beam-typecollisional dissociation. J. Am. Soc. Mass Spectrom. 26, 782–787 (2015)

74. Wilm, M.: Principles of electrospray ionization. Mol. Cell. Proteom. 10,(2011)

75. Fernandez de la Mora, J.: Electrospray ionization of large multiplycharged species proceeds via Dole’s charged residue mechanism. Anal.Chim. Acta 406, 93–104 (2000)

76. Smith, R.D., Loo, J.A., Loo, R.R.O., Busman, M., Udseth, H.R.:Principles and practice of electrospray ionization—mass spectrometryfor large polypeptides and proteins. Mass Spectrom. Rev. 10, 359–452(1991)

77. Loo, J.A., Udseth, H.R., Smith, R.D.: Peptide and protein analysis byelectrospray ionization-mass spectrometry and capillary electrophoresis-mass spectrometry. Anal. Biochem. 179, 404–412 (1989)

78. Tolić, L.P., Anderson, G.A., Smith, R.D., Brothers Ii, H.M., Spindler, R.,Tomalia, D.A.: Electrospray ionization Fourier transform ion cyclotronresonance mass spectrometric characterization of high molecular massStarburst dendrimers. Int. J. Mass Spectrom. Ion Processes. 165/166,405–418 (1997)

79. Loo, R.R.O., Lakshmanan, R., Loo, J.A.: What protein charging (andsupercharging) reveal about the mechanism of electrospray ionization. J.Am. Soc. Mass Spectrom. 25, 1675–1693 (2014)

80. Kruger, N.A., Zubarev, R.A., Carpenter, B.K., Kelleher, N.L., Horn,D.M., McLafferty, F.W.: Electron capture versus energetic dissociationof protein ions. Int. J. Mass Spectrom. 182/183, 1–5 (1999)

81. Uversky, V.N., Gillespie, J.R., Fink, A.L.: Why are Bnatively unfolded^proteins unstructured under physiologic conditions? Proteins 41, 415–427(2000)

82. Tompa, P.: Intrinsically unstructured proteins. Trends Biochem. Sci. 27,527–533 (2002)

83. Uversky, V.N.: Natively unfolded proteins: a point where biology waitsfor physics. Protein Sci. 11, 739–756 (2002)

84. Frantz, C., Barreiro, G., Dominguez, L., Chen, X., Eddy, R., Condeelis,J., Kelly, M.J.S., Jacobson, M.P., Barber, D.L.: Cofilin is a pH sensor foractin free barbed end formation: role of phosphoinositide binding. J. CellBiol. 183, 865 (2008)

85. Carbeck, J.D., Severs, J.C., Gao, J., Wu, Q., Smith, R.D., Whitesides,G.M.: Correlation between the charge of proteins in solution and in thegas phase investigated by protein charge ladders, capillary electrophore-sis, and electrospray ionization mass spectrometry. J. Phys. Chem. B 102,10596–10601 (1998)

86. Loo, J.A., Edmonds, C.G., Smith, R.D.: Tandem mass spectrometry ofvery large molecules. 2. Dissociation of multiply charged proline-containing proteins from electrospray ionization. Anal. Chem. 65, 425–438 (1993)

87. Vaisar, T., Urban, J.: Probing proline effect in CID of protonated peptides.J. Mass Spectrom. 31, 1185–1187 (1996)

88. Grizot, S., Fauré, J., Fieschi, F., Vignais, P.V., Dagher, M.C., Pebay-Peyroula, E.: Crystal structure of the Rac1−RhoGDI complex involved inNADPH oxidase activation. Biochemistry 40, 10007–10013 (2001)

89. Ruiz Carrillo, D., Chandrasekaran, R., Nilsson, M., Cornvik, T., Liew,C.W., Tan, S.M., Lescar, J.: Structure of human Rack1 protein at aresolution of 2.45 Å. Acta Crystallogr. Sect. F: Struct. Biol. Cryst.Commun. 68, 867–872 (2012)

90. Chen, L., Yang, S., Jakoncic, J., Zhang, J.J., Huang, X.-Y.: Migrastatinanalogues target fascin to block tumour metastasis. Nature 464, 1062–1066 (2010)

N. A. Haverland et al.: Gas-Phase Fragmentation with Native MS 1215